{"repo": "angular", "sha": "c88ddde1c9f56e3b5c5cf264e87217b86a036a58", "date": "2026-07-14T18:18:35+02:00", "subject": "fix(compiler-cli): re-tag SourceFiles after TsCreateProgramDriver.updateFiles()", "body": "TypeScript reuses SourceFile objects between old and new programs, so untagging the old program also untags shared files in the new program. Re-apply shim tags on the new program to prevent getSemanticDiagnostics() crashes with TS 5.5+.\n\n(cherry picked from commit ae0ec7315c6fbc45ce40581a08a8243a008a91ec)", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "ae0ec7315c6fbc45ce40581a08a8243a008a91ec", "date": "2026-07-14T18:18:35+02:00", "subject": "fix(compiler-cli): re-tag SourceFiles after TsCreateProgramDriver.updateFiles()", "body": "TypeScript reuses SourceFile objects between old and new programs, so untagging the old program also untags shared files in the new program. Re-apply shim tags on the new program to prevent getSemanticDiagnostics() crashes with TS 5.5+.", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "openssl", "sha": "a44ba221abc62a8ca61107749cc776eb72e16969", "date": "2026-07-14T18:26:40+02:00", "subject": "apps: test rsa -text option", "body": "The -text option of the rsa app was not exercised by any test. Add a\nsubtest that prints both a private and a public key in text form and,\nafter stripping the colon-separated hex formatting, verifies the printed\nmodulus and private exponent match the committed testrsa.pem keypair\nrather than merely checking that the labels are present.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Daniel Kubec \nReviewed-by: Paul Dale \n(Merged from https://github.com/openssl/openssl/pull/31950)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "49057cbe91eef40fcc1e4973feeaef8be60b9434", "date": "2026-07-14T18:38:57+02:00", "subject": "apps: test pkey -encopt option", "body": "The -encopt option of the pkey app was not exercised by any test; the\nexisting ML-DSA codec tests only used genpkey -encopt and pkey with\n-provparam. Re-encode the seed-priv key into each supported PKCS#8\noutput format via 'pkey -encopt output_formats:
' and check the\nresult matches the reference for that form. A control compares against\nthe default (no -encopt) encoding so the match is attributed to -encopt\nrather than the default behaviour.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Daniel Kubec \nReviewed-by: Tim Hudson \nMergeDate: Fri Jul 17 08:51:40 2026\n(Merged from https://github.com/openssl/openssl/pull/31951)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "axios", "sha": "b72d3ef190b599bf95e89bf79c46311e8ff7e1b0", "date": "2026-07-14T18:56:24+02:00", "subject": "docs: update sponsors", "body": "Updated sponsorship link and added information about JBO Thailand.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "cb9a9021752b0c3499fccbb965e1e525b357a0f0", "date": "2026-07-14T13:00:59-04:00", "subject": "Fix dsaparams decoding from DER files", "body": "The tests addded in commit d8a7e8e uncovered an odd error case.\n\nhttps://github.com/openssl/openssl/actions/runs/29303845665/job/86993085732\n\nIs failing when attempting to read in a der file converting from a\ncorresponding PEM file containing DSA parameters.\n\nInterestingly The problem was only occuring when:\n1) The input was a DER file\nand\n2) Blake2 was not configured\n\nDoing some tracing of the decoder operation showed that this is occuring\nbecause the OSSL_STORE lookup used to find the proper decoder uses a\n\"first successful decode wins\" approach, after which the loading code\nchecks to see if the decoded type matches the expected key type.\n\nWhen decoding PEM, this isn't a problem, as the PEM armoring gives the\ndecoder a hint as to why type of data the input file is.\n\nBut with DER, there is no such hint, and we're at the mercy of whichever\ndecoder happens to decode the data correctly first. Normally it works\njust fine, but when features are disabled or enabled, the order in which\nthe decoders are attempted may change, affecting the outcome. In this\nparticular case, disabling blake2 caused the DHX decoder to be attempted\nfirst, which decodes the input der file without issue. That in turn\ncaused the subsequent EVP_PKEY_is_a check to fail (as we were expecting\na DSA key), and so the test fails.\n\nFortunately, the code that the dsaparam applet uses to do this decode\nprovides a keytype hint, which we can use to guide the decode process.\nkeep the old store lookup method around in case anyone doesn't pass a\nuri that is a file path or provide a keytype, but if we do both those\nthings, we can use OSSL_DECODER_CTX_new_for_pkey to specifically tell\nthe decoder that we want to decode the input data as the expected type\n(in this case a DSA key).\n\nFixes #31944\n\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Tim Hudson \nMergeDate: Thu Jul 16 15:28:32 2026\n(Merged from https://github.com/openssl/openssl", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4ef7e3b3cdc483cebb55b240770a2ee6e705b8db", "date": "2026-07-14T19:01:24+02:00", "subject": "apps: test ec and ecparam -text options", "body": "The -text option was not exercised for the ec or ecparam apps. Add a\nsubtest to 15-test_ec.t that prints a private and a public EC key and,\nafter stripping the colon-separated hex formatting, verifies the printed\nprivate and public values match the committed testec-p256.pem keypair as\nwell as the curve identification. Add a subtest to 15-test_ecparam.t\nthat prints named and explicit parameters, checking the named form emits\nthe expected curve OID and NIST name while the explicit form emits the\nfield parameters and no OID.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Daniel Kubec \nReviewed-by: Tim Hudson \nMergeDate: Fri Jul 17 08:47:12 2026\n(Merged from https://github.com/openssl/openssl/pull/31952)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "d52c7e78a3531f975591eb1bb1250185a9fd14c2", "date": "2026-07-14T19:06:37+02:00", "subject": "idn: restore `MultiByteToWideChar()` `MB_ERR_INVALID_CHARS` flag", "body": "Also:\n- curlx: pass this flag to the actual conversion calls, for consistency\n and robustness. (It's not stricly necessary because the initial call\n to determine size, with this flag passed, fails already on bad input.)\n- schannel: unfold `MultiByteToWideChar()` line (formatting).\n\nRef: https://learn.microsoft.com/windows/win32/api/stringapiset/nf-stringapiset-multibytetowidechar\n\nFollow-up to 6694a42aa0e820a6fe1e59d85ff8597b6d768d8d #19798\n\nCloses #22326", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "67cb3ed0ac9388e81dafecfc30a24b5a37bf1d7b", "date": "2026-07-14T19:08:20+02:00", "subject": "apps: test pkeyutl -peerform option", "body": "The -peerform option of the pkeyutl app was not exercised by any test.\nExtend the -derive peer key subtest to convert the peer public key to\nDER and derive against it with -peerform DER, checking the shared secret\nmatches the one derived from the PEM peer key, and that reading the DER\npeer key as PEM fails.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Paul Dale \nReviewed-by: Andrew Dinh \nReviewed-by: Daniel Kubec \nMergeDate: Mon Jul 27 09:10:00 2026\n(Merged from https://github.com/openssl/openssl/pull/31953)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7c3c77807023bbf1aef20150dfccce58785e6a30", "date": "2026-07-14T19:18:12+02:00", "subject": "OSSL_FN: Add ModSqrt stanza test in fntest", "body": "Mirrors file_modsqrt() in test/bntest.c. Both square roots are\naccepted for positive cases (ret and p - ret), since a square root has\ntwo roots.\n\nThe operand a is reduced into [0, p) with BN_nnmod() at the boundary\nbefore being handed to OSSL_FN_mod_sqrt(): BN_mod_sqrt() does that\nreduction internally, but OSSL_FN_mod_sqrt() is unsigned and sees only\nthe magnitude, so feeding a raw negative a (e.g. a = -5, p = 7, which\nreduces to 2) would present the wrong residue.\n\nRelated-to: doc/designs/fixed-size-large-numbers.md\nIssue: https://github.com/openssl/project/issues/2026\nAssisted-by: Pi:z-ai/glm-5.2\nSigned-off-by: Richard Levitte \n\nReviewed-by: Jakub Zelenka \nReviewed-by: Igor Ustinov \nReviewed-by: Matt Caswell \nMergeDate: Fri Jul 31 09:39:06 2026\n(Merged from https://github.com/openssl/openssl/pull/31963)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "d705471aa5aa5a54f6f1749d27424ae0d137ccdf", "date": "2026-07-14T18:18:52+01:00", "subject": "Regression tests for display-panes changes.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "38d6c2d5fd6c721f40d800c7bec7a2dee2696593", "date": "2026-07-14T19:20:21+02:00", "subject": "Add value_barrier in MOD_EXP_CTIME_COPY_FROM_PREBUF", "body": "Without it, compilers can figure out the possible values and optimize\nbased on it.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nMergeDate: Wed Jul 29 17:04:46 2026\n(Merged from https://github.com/openssl/openssl/pull/31957)\n\n(cherry picked from commit f8c26111bab30399d0c9262269089639ff8e4746)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f8c26111bab30399d0c9262269089639ff8e4746", "date": "2026-07-14T19:20:21+02:00", "subject": "Add value_barrier in MOD_EXP_CTIME_COPY_FROM_PREBUF", "body": "Without it, compilers can figure out the possible values and optimize\nbased on it.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nMergeDate: Wed Jul 29 17:04:46 2026\n(Merged from https://github.com/openssl/openssl/pull/31957)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ae6309aede393e2891036c5da50baa55e2c5e5a5", "date": "2026-07-14T19:20:43+02:00", "subject": "apps: test pkcs8 -nocrypt reading a PKCS#8 PEM", "body": "Reading an unencrypted PKCS#8 (PrivateKeyInfo) in PEM form with -nocrypt\nwas not exercised by any test; the existing round-trip test only read the\nDER form via -inform DER, hitting d2i_PKCS8_PRIV_KEY_INFO_bio(). The PEM\nbranch in apps/pkcs8.c (the informat == FORMAT_PEM || FORMAT_UNDEF case\ncalling PEM_read_bio_PKCS8_PRIV_KEY_INFO()) was left uncovered.\n\nAdd a subtest that writes an unencrypted PKCS#8 PEM and reads it back\nwith -nocrypt from PEM, exercising PEM_read_bio_PKCS8_PRIV_KEY_INFO(),\nand checks the recovered key matches the original.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Paul Dale \nReviewed-by: Andrew Dinh \nReviewed-by: Matt Caswell \nMergeDate: Mon Jul 27 09:10:45 2026\n(Merged from https://github.com/openssl/openssl/pull/31955)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0c306f0ad0a825c5765dad280c4cacc3b12b39f0", "date": "2026-07-14T19:40:42+02:00", "subject": "Add value_barrier in BN_consttime_swap", "body": "Without it, compilers can figure out the possible values and optimize\nbased on it.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nMergeDate: Wed Jul 29 17:04:47 2026\n(Merged from https://github.com/openssl/openssl/pull/31957)\n\n(cherry picked from commit 1d7a8dd73ff9e0fc197611433b9b60b4b5f5465c)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1d7a8dd73ff9e0fc197611433b9b60b4b5f5465c", "date": "2026-07-14T19:40:42+02:00", "subject": "Add value_barrier in BN_consttime_swap", "body": "Without it, compilers can figure out the possible values and optimize\nbased on it.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nMergeDate: Wed Jul 29 17:04:47 2026\n(Merged from https://github.com/openssl/openssl/pull/31957)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f7ec8c5d5cfb7c375737f48b154d50a83733e1e1", "date": "2026-07-14T19:41:58+02:00", "subject": "apps: test pkcs12 -clcerts/-cacerts and -name options", "body": "The -clcerts and -cacerts certificate filtering and the -name friendly\nname option of the pkcs12 app were not exercised by any test. Add a\nsubtest that exports a PKCS#12 holding a client cert (with a localKeyID),\na CA cert and a friendly name, then checks the friendly name is present,\nthat -clcerts outputs only the client certificate and -cacerts only the\nCA certificate.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Paul Dale \nReviewed-by: Andrew Dinh \nReviewed-by: Daniel Kubec \nMergeDate: Mon Jul 27 09:09:09 2026\n(Merged from https://github.com/openssl/openssl/pull/31956)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "ae1d81c8c5cd81d9c8feeee22f7981d2a167e2b9", "date": "2026-07-14T21:56:41+03:00", "subject": "feat(docs-infra): greet visitors with Angie in the cookie popup", "body": "The cookie consent popup showed only text and two buttons. Add the\nAngie greeting pose to it so the first thing a new visitor sees carries\nthe same mascot treatment already used across the docs (the 404 page,\nempty search results, the preview error state).\n\nThe message and mascot sit in a flex header row; the two action buttons\nnow share the row equally so they fill the card instead of leaving a\ntrailing gap. Layout only, the consent behaviour is unchanged.\n\n(cherry picked from commit 6a3487f7b7b6d859cd91337fa859fbbd84d9495a)", "style": "conventional", "type": "feat", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "6a3487f7b7b6d859cd91337fa859fbbd84d9495a", "date": "2026-07-14T21:56:41+03:00", "subject": "feat(docs-infra): greet visitors with Angie in the cookie popup", "body": "The cookie consent popup showed only text and two buttons. Add the\nAngie greeting pose to it so the first thing a new visitor sees carries\nthe same mascot treatment already used across the docs (the 404 page,\nempty search results, the preview error state).\n\nThe message and mascot sit in a flex header row; the two action buttons\nnow share the row equally so they fill the card instead of leaving a\ntrailing gap. Layout only, the consent behaviour is unchanged.", "style": "conventional", "type": "feat", "scope": "docs-infra", "breaking": false} {"repo": "tmux", "sha": "ad6832e6972663c3aafad890ce983eb213401b7f", "date": "2026-07-14T19:07:03Z", "subject": "Switch menus also to be attached to the window instead of an overlay and drawn as part of the client's scene. This means removing the popup menu.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "e54785ddd6629788e5fa8ebd9f843c784f60aa5e", "date": "2026-07-14T20:07:07+01:00", "subject": "Add menu tests.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "4176aba5e4871a2f1c7c120dd76568f80f5d5ddc", "date": "2026-07-14T22:49:38+02:00", "subject": "tool_cb_prg: avoid integer overflows", "body": "Verify in test 2093\n\nFixes #22316\nReported-by: xmoezzz on github\nCloses #22328", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "11496010c027a06262ecd9bfaa776a0109e7daee", "date": "2026-07-14T22:15:08+01:00", "subject": "Update test outputs.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "4b69d3fd2db74517168385fed0b676f1ab329a05", "date": "2026-07-14T16:37:28-05:00", "subject": "test: adjust exit waits for slow CI", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "160ef2751b524cbc797e59642543790c4a8132ec", "date": "2026-07-15T00:46:56+03:00", "subject": "Fix JSON_TABLE PLAN deparse to keep parentheses around nested joins", "body": "get_json_table_plan() parenthesized the child of a parent/child\n(OUTER/INNER) plan only when that child was a sibling (UNION/CROSS)\njoin, not when it was itself a parent/child join. A plan such as\nPLAN (p0 OUTER (p1 INNER p11)) was therefore deparsed as\nPLAN (p0 OUTER p1 INNER p11), which does not parse back to the same\nplan tree -- a dump/restore hazard. Parenthesize the child whenever it\nis not a bare path name, matching the logic already used for the\noperands of sibling joins.\n\nReported-by: Thom Brown \nDiscussion: https://postgr.es/m/CAA-aLv7aZGSExnbjJRw8eKkoXbu34TdoKLLA2gPye3aHjO5OSA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "cd5fd08be7ec1cef1a4c34791ce743954c2b2293", "date": "2026-07-15T00:47:16+03:00", "subject": "Make JSON_TABLE generated path names avoid collisions", "body": "generateJsonTablePathName() produced names of the form\n\"json_table_path_N\" in the same namespace as user-supplied path and\ncolumn names, without checking whether the name was already in use.\nWhen an unnamed NESTED path's generated name happened to match a\nuser-supplied path name that a PLAN clause referenced, two sibling paths\nmatched the same plan entry and one of them, together with its columns,\nwas silently dropped from the output.\n\nBump the counter until the generated name is unused, so a generated name\ncan no longer coincide with a user-supplied one. The still-uncovered\npath is then correctly reported as not found in the plan.\n\nThe row pattern (root) path is named before the user-supplied column and\npath names are collected, so when it is left unnamed its generated name\ncould not avoid them either, and a user column or path named like a\ngenerated name, e.g.\n\n SELECT * FROM JSON_TABLE(jsonb '1', '$'\n COLUMNS (json_table_path_0 int PATH '$')) jt;\n\nwas rejected with a bogus \"duplicate JSON_TABLE column or path name\"\nerror. Collect the user-supplied names first and generate the row\npattern path's name afterwards, so that it avoids all of them. An\nexplicit root path name is still seeded into the namespace, so a column\nduplicating it is still correctly rejected.\n\nReported-by: Thom Brown \nDiscussion: https://postgr.es/m/CAA-aLv7aZGSExnbjJRw8eKkoXbu34TdoKLLA2gPye3aHjO5OSA@mail.gmail.com\nDiscussion: https://postgr.es/m/CAA-aLv5U94KD4C%2BLhAPYcCeGvs1xBMngcS5oEkZHN9YWwXUHsA%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "4a82912bb2a2bc757142eca48f0fded1817a9a7c", "date": "2026-07-15T00:47:33+03:00", "subject": "Fix and polish JSON_TABLE documentation", "body": "Correct the JSON_TABLE synopsis to place the ON ERROR clause after the\nPLAN clause, matching the grammar (the previous order could not be\ntyped). Replace a dangling reference to json_path_specification with\npath_expression, the term the synopsis actually defines. Mark up the\nPLAN DEFAULT keywords with and fix a couple of wording issues.\nAlso list OUTER before INNER consistently, including in the parent/child\njoin description, where OUTER is the default.\n\nAuthor: Thom Brown \nDiscussion: https://postgr.es/m/CAA-aLv5PGAFmAEpbhKQz8wppoOOTHfo-=LJb2sAB3974-9QtOw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "6d8fd5a88183a65e77308aff374ebc239c44c6af", "date": "2026-07-15T00:47:54+03:00", "subject": "Avoid redundant re-evaluation of JSON_TABLE nested paths", "body": "86ab7f4c721d makes JSON_TABLE with a NESTED PATH re-ran the nested path's\njsonpath expression several times for each parent row. That makes such\nqueries significantly slower as the nested arrays grew, even without a PLAN\nclause.\n\nTwo leftovers from the plan/join executor rework were responsible.\nJsonTablePlanScanNextRow() still reset and advanced the nested plan\nitself, although JsonTablePlanNextRow() now does that; and\nJsonTableResetNestedPlan() eagerly called JsonTableResetRowPattern()\n(which evaluates the path) in addition to setting the reset flag that\nmakes JsonTablePlanNextRow() evaluate it again. Together these caused\nthe nested path to be evaluated multiple times per parent row.\n\nReduce JsonTablePlanScanNextRow() to advancing its own row pattern\niterator, and have JsonTableResetNestedPlan() only reset the transient\nscan state (so a not-yet-advanced sibling still reads as NULL) while\ndeferring the actual path evaluation to the reset flag. The nested path\nis now evaluated exactly once per parent row, as before the PLAN clause\nfeature; results are unchanged and are covered by the existing tests.\n\nReported-by: Thom Brown \nDiscussion: https://postgr.es/m/CAA-aLv5U94KD4C%2BLhAPYcCeGvs1xBMngcS5oEkZHN9YWwXUHsA%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9ee5241d02442911ffec9a0ce35327a0fea4a5c1", "date": "2026-07-15T00:48:07+03:00", "subject": "Remove unreachable error check in JSON_TABLE plan transform", "body": "transformJsonTableNestedColumns() looked up the nested column named by a\nJSON_TABLE PLAN node and raised \"invalid JSON_TABLE plan clause / PATH\nname was %s not found in nested columns list\" if none was found. That\nlookup cannot fail: validateJsonTableChildPlan() runs first, at every\nplan level, and already matches the plan's sibling path names one-to-one\nagainst the nested columns (reporting any uncovered nested path or any\nextra or duplicate sibling node). The check has been unreachable since\nthe PLAN clause code was first written.\n\nReplace it with an Assert documenting the invariant, which also removes\na user-facing message that could never be emitted.\n\nReported-by: Thom Brown \nDiscussion: https://postgr.es/m/CAA-aLv5_9%3DzgA_Y7aoFp-%2BQSeh0kx4dfbAas9Wx%3DyrweQSqa6Q%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2349b106b6661c084dc1171654146c730ecb048f", "date": "2026-07-15T01:39:33+03:00", "subject": "postgres_fdw: don't push down non-relabeling ArrayCoerceExpr", "body": "Commit 62c3b4cd9ddc taught postgres_fdw to push down ArrayCoerceExpr, but\nforeign_expr_walker() only recursed into the input array expression and\nnever examined elemexpr, the per-element conversion that gives the\ncoercion its semantics. deparseArrayCoerceExpr() then shipped a bare\n\"arg::resulttype\" cast, or nothing at all for an implicit-format\ncoercion, leaving the remote server to re-resolve the element conversion\nagainst its own catalogs and session state.\n\nThis produced wrong results or remote errors whenever the element\nconversion was not a plain relabeling, and it was inconsistent with how\npostgres_fdw treats the equivalent scalar coercions. An ArrayCoerceExpr\nwas shipped even when its elemexpr was a cast function (whose\nshippability was never checked), a CoerceViaIO (e.g. float8out or\nbyteaout, which depend on extra_float_digits / bytea_output that\npostgres_fdw sets differently on the remote session), or a\nCoerceToDomain (which pushes domain enforcement to the remote catalog).\nBy contrast, a scalar CoerceViaIO is never shipped, and a scalar cast\nfunction is shipped only when it is shippable.\n\nRestrict pushdown to element coercions that are a plain relabeling, that\nis, elemexpr is a RelabelType or a bare CaseTestExpr. Any other element\ncoercion is now evaluated locally. This keeps the common\nbinary-coercible case pushed down, including \"col = ANY($1)\" with a\nvarchar[]-to-text[] relabeling, which is the case 62c3b4cd9ddc set out to\noptimize.\n\nPushing down shippable element cast functions, to reach parity with the\nscalar case, is left out here for simplicity.\n\nReported-by: Noah Misch \nDiscussion: https://postgr.es/m/20260711024234.43.noahmisch%40microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "ca96fd5190c3910f633ad8fe9f66380c28cd91ad", "date": "2026-07-15T02:18:30+02:00", "subject": "scorecard: fix `max_upload` init value in `ul_parallel()`", "body": "\"In `ul_parallel`, `max_parallel` is computed using\n`self._download_parallel` instead of `self._upload_parallel`. This\ncauses the upload parallelism to incorrectly follow the download\nparallel setting. It should use `self._upload_parallel` to be consistent\nwith how `uploads()` computes `max_parallel`.\"\n\nReported by GitHub Code Quality\n\nFollow-up to 30ef79ed937ca0fc7592ff73d162398773c6a5aa #17295\n\nCloses #22421", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "18105e6db5e5314a575fdb23a99cf4809d8ef062", "date": "2026-07-15T09:22:58+09:00", "subject": "Strip removed-relation references from PHVs in join clauses", "body": "Commit 9a60f295b stripped the stale PlaceHolderVars left behind by\nleft-join removal from the surviving rels' baserestrictinfo and from\nEquivalenceClass member expressions, but it overlooked join clauses.\nA PlaceHolderVar embedded in a join clause can likewise retain the\nremoved rel and join in its phrels, since remove_rel_from_query()\nfixes up the RestrictInfo's own relid sets but not the PHVs inside its\nexpression.\n\nAs before, this is normally harmless, because later processing\nconsults those relid sets rather than the embedded PHVs. However, a\nrestriction clause derived from such an OR join clause inherits the\nstale PlaceHolderVar, and when the derived clause is translated for an\nappendrel child, pull_varnos() recomputes its relids and folds the\nremoved relation back in. The rebuilt clause then references a\nno-longer-existent relation, tripping an assertion during path\ngeneration.\n\nFix by also stripping the removed relation from the PlaceHolderVars in\nthe surviving rels' join clauses, including the sub-clauses of any OR\nclause.\n\nLike 9a60f295b, this is only reachable on v18 and later, where\nmatch_index_to_operand() began ignoring PlaceHolderVars.\n\nAuthor: Arne Roland \nReviewed-by: Tender Wang \nReviewed-by: Richard Guo \nDiscussion: https://postgr.es/m/27a44087-3d65-473e-8d88-7c12228e0d7e@malkut.net\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "387b4c5e4c3fe654187c49db743efd0be41febaf", "date": "2026-07-15T02:28:09+02:00", "subject": "dnsd: fix bounds check in `read_https_alpn_part()`", "body": "\"The check `i > 256` permits `i == 256` to pass through. When `i` is\nthen cast to `uint8_t` in `blob_add(b, (uint8_t)i)`, the value wraps to\n0, silently encoding a zero-length ALPN entry instead of rejecting it.\nThe condition should be `i > 255` (or equivalently `i >= 256`) to\ncorrectly reject any length that does not fit in a single byte.\"\n\nReported by GitHub Code Quality\n\nFollow-up to 86f1e5b3f6c90e453368fdf23c366c1d4a8c953b #21299\n\nCloses #22420", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "99eb806ab48f77d06b9e902742d0b91f289b1ffa", "date": "2026-07-15T10:03:46+09:00", "subject": "Include check on polpermissive relcache for policies", "body": "equalPolicy() is used in the relation cache to check if two policy\ndefinitions are equivalent, but missed to check for polpermissive.\n\nALTER POLICY cannot switch a policy to be PERMISSIVE or RESTRICTIVE, so\nthis would need a dropped and then re-created policy, which would\ntrigger a relcache invalidation. Anyway, there is no harm in being\nconsistent in the check, and if one decides to add an ALTER POLICY to\nswitch PERMISSIVE or RESTRICTIVE, we would be silently in trouble.\n\nAuthor: Andreas Lind \nReviewed-by: Laurenz Albe \nDiscussion: https://postgr.es/m/CAMxA3rv1CS6R7JR5ojz-3CmCEnZEFrqu+XXTnGbLRWrjJRH7sA@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0892319c65e7d7458ee8fb902618491aaa0e4734", "date": "2026-07-15T10:35:19+09:00", "subject": "Rework pgstat_write_statsfile() in combination with to_serialized_data", "body": "Contrary to the from_serialized_data callback used by the pgstats reads\nat startup, the to_serialized_data callback used for the pgstats writes\nmatched with pgstat_write_statsfile(), by not returning a boolean\nstatus, expecting a ferror() failure to deal with the discard of the\nstats file should an error happen while writing the stats. This was\nslightly confusing designed this way.\n\nThings are changed in this commit with:\n- to_serialized_data now returns a boolean status on a write failure.\npgstat_write_statsfile() detects that and switches to failure mode\ninstead of continuing to process the entries to write, speeding up the\nshutdown.\n- pgstat_write_statsfile() now uses STATS_DISCARD if a failure happens,\nto let the registered callbacks directly know that something is wrong,\nand that things need to be cleaned up. This gives a better error path\ndetection for custom stats kinds. For example, they do not have to rely\nsolely on the expectation of an ferror() for an auxiliary file.\n\nThis new set of behaviors matches with what is already done in\npgstat_read_statsfile() for the finish() callback (DISCARD on failure,\nREAD on success) and the from_serialized_data with a status returned.\n\nAuthor: Sami Imseih \nDiscussion: https://postgr.es/m/CAA5RZ0sMgOvuhpb2P=KSJOjgjC6AfUu+GYcu9mHar-y_Xtd=Pg@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "78064046b558db44057306dd3e9b71aad443c096", "date": "2026-07-15T09:35:47+08:00", "subject": "fix(InputNumber): display suffix with form feedback", "body": "", "style": "conventional", "type": "fix", "scope": "InputNumber", "breaking": false} {"repo": "ant-design", "sha": "1ed06602f1f8d6aa1517f801a4e6f66984526f82", "date": "2026-07-15T11:18:42+08:00", "subject": "fix(motion): prevent invalid nested pseudo-element selectors", "body": "", "style": "conventional", "type": "fix", "scope": "motion", "breaking": false} {"repo": "ant-design", "sha": "bf0000b9cf6dff3cb751d2c8fec6e15d25c58df9", "date": "2026-07-15T11:24:18+08:00", "subject": "fix(Splitter): correct percentage-based ARIA value ranges", "body": "", "style": "conventional", "type": "fix", "scope": "Splitter", "breaking": false} {"repo": "postgres", "sha": "7d45a6dc19743d999f5c83c95ad144a0f2eb6745", "date": "2026-07-15T07:59:42+02:00", "subject": "Clean up secure_read()/secure_write() return type", "body": "The return type is ssize_t, not int, but some callers didn't handle\nthis properly.\n\nThe BIO callbacks are constrained by the OpenSSL API, so they take int\nfor the length and return int. This is safe, since the return value\ncan't be greater than the input length. To make it more clear that\nthis is intentional, cast the result of the\nsecure_read()/secure_write() call to int explicitly.\n\nReviewed-by: Heikki Linnakangas \nDiscussion: https://www.postgresql.org/message-id/flat/f9aab072-0078-49e4-ab93-3b08086a4406@eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9ea90ec35a7c9e7ed8ff2293f8aa99704808335a", "date": "2026-07-15T13:59:45+08:00", "subject": "fix(floatwin): don't use same window as altwin", "body": "Fixes a regression from #34486\nFix #40745\n\n(cherry picked from commit 39bc59a8f1c0eacea60e47b3a8b0422bbbbb8cd6)", "style": "conventional", "type": "fix", "scope": "floatwin", "breaking": false} {"repo": "neovim", "sha": "39bc59a8f1c0eacea60e47b3a8b0422bbbbb8cd6", "date": "2026-07-15T13:59:45+08:00", "subject": "fix(floatwin): don't use same window as altwin", "body": "Fixes a regression from #34486\nFix #40745", "style": "conventional", "type": "fix", "scope": "floatwin", "breaking": false} {"repo": "nestjs", "sha": "e270ef54eb5ac30f652088ef2594a0913ee8f0ef", "date": "2026-07-15T08:20:49+02:00", "subject": "chore: update samples deps", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "32bcf024d513b59cd4c9e4985814409419d265b1", "date": "2026-07-15T14:30:55+08:00", "subject": "chore: bump utoopack to v1.4.25", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a20fa498281912c5a34d6d14d2cfb0538fd836cf", "date": "2026-07-15T09:41:06+02:00", "subject": "test/p_ossltest: fix uninitialised bytes written in TLS1-AAD GCM mode", "body": "The test provider's fake AES-128-GCM cipher works by running the real\nsub-cipher for its side effects and then copying the memdup'd input\nback over the output buffer, so the \"encrypted\" record is really the\nplaintext.\n\nIn a TLS record the buffer handed to the cipher reserves an\nuninitialised 8-byte explicit IV at the front and a 16-byte auth tag\nat the end for the cipher to fill. OPENSSL_memdup() duplicated those\nuninitialised IV/tag regions, and the subsequent memcpy(out, inbuf, inl)\ncopied them into out. That output was then written to the socket,\ntriggering valgrind's \"Syscall param write(buf) points to uninitialised\nbyte(s)\":\n\n at sock_write (bss_sock.c:155)\n by bwrite_conv (bio_meth.c:79)\n by BIO_write (bio_lib.c:397)\n by statem_flush (statem.c:963)\n by ossl_statem_client_post_work (statem_clnt.c:881)\n ...\n\nTrack whether EVP_CTRL_AEAD_TLS1_AAD was set (recorded in\nset_ctx_params) and, when encrypting a TLS record, copy only the\nplaintext payload back, leaving the explicit IV and tag that the real\nsub-cipher actually produced intact.\n\nSigned-off-by: Nikola Pajkovsky \n\nReviewed-by: Tim Hudson \nReviewed-by: Paul Dale \nReviewed-by: Neil Horman \nMergeDate: Mon Jul 20 09:07:08 2026\n(Merged from https://github.com/openssl/openssl/pull/31961)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ca326e903df4b2efcc7b9090abc4d1a9c27c3088", "date": "2026-07-15T09:43:03+02:00", "subject": "Clean up read() return type", "body": "and analogously for pg_pread() and FileRead()\n\nBe sure to store the return value in a variable of type ssize_t, not\nint.\n\nAlso make the error messages for short reads consistent. They should\nalways be like \"read %zd of %zu\". Appearance of other placeholders\nindicates the types are probably wrong (although in some cases some\ncasts are added to make macros have the right type and keep the\nstrings consistent, and it some cases it's left as \"%zu of %zu\", which\nis close enough).\n\nIn several cases, the input length is derived from struct stat\nst_size, which has type off_t, which is neither size_t nor ssize_t.\nTo keep the type handling clearer, this introduces intermediate\nvariables in these cases.\n\nIn SendTimeLineHistory() in walsender.c, we need to adjust the logic a\nbit to over underflow wrap if we end up reading more from the file\nthan expected. This is believed to be a theoretical problem only.\nAlternatively, we could treat this as an error. Note that the\nprevious code would have processed the extra data but only up to a\nfull block, which seems wrong in any case.\n\nReviewed-by: Heikki Linnakangas \nDiscussion: https://www.postgresql.org/message-id/flat/f9aab072-0078-49e4-ab93-3b08086a4406@eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "534e5b265077a25ae91d5f8d28c058fd8c8152e8", "date": "2026-07-15T15:44:11+08:00", "subject": "docs(anchor): correct offsetTop default value in index.zh-CN.md", "body": "", "style": "conventional", "type": "docs", "scope": "anchor", "breaking": false} {"repo": "openssl", "sha": "d00cf645c1d631299843d7dda3a21f24ef5848f8", "date": "2026-07-15T09:51:51+02:00", "subject": "test: include PID in valgrind log file names", "body": "Append the %p (process ID) placeholder to valgrind's --log-file in\nboth app() and test(). When a test spawns multiple processes sharing\nthe same result index, they previously wrote to the same log file and\nclobbered each other's output. Using %p gives each process its own\nlog, so no valgrind findings are lost.\n\nSigned-off-by: Nikola Pajkovsky \n\nReviewed-by: Tim Hudson \nReviewed-by: Paul Dale \nReviewed-by: Neil Horman \nMergeDate: Mon Jul 20 09:07:07 2026\n(Merged from https://github.com/openssl/openssl/pull/31961)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f17dfc9ab3bfae39ea987e771bbeb5e03c4374a8", "date": "2026-07-15T15:55:01+08:00", "subject": "test: fix typo", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "9612c861ce0e1a30d4e0293747a554b7fec8f6e9", "date": "2026-07-15T16:58:52+09:00", "subject": "feat(core): support sse comments", "body": "", "style": "conventional", "type": "feat", "scope": "core", "breaking": false} {"repo": "ant-design", "sha": "1b6226c67099d3e993c057c6e1aa1af667da45eb", "date": "2026-07-15T16:01:19+08:00", "subject": "chore: remove utoopack lock", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "b10ab142d7cc7578bad69adfd773f5df2f5979a2", "date": "2026-07-15T16:07:01+08:00", "subject": "chore: enable concatenateModules", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vitest", "sha": "3923cab3d98cd0063377ffd8e4c98b73d2f28db9", "date": "2026-07-15T04:32:43-04:00", "subject": "fix(pool): vm pools to respect cgroupsv2 memory limit", "body": "", "style": "conventional", "type": "fix", "scope": "pool", "breaking": false} {"repo": "postgres", "sha": "01f78a84bf5aca4c00fe1168381830aca6f1b873", "date": "2026-07-15T10:58:13+02:00", "subject": "Add assertion about ssize_t narrowing in AIO code", "body": "The result from pg_preadv() or pg_pwritev(), which is of type ssize_t,\nis assigned to PgAioHandle.result, which is of type int. This should\nbe ok because the maximum result is limited by PG_IOV_MAX times\nBLCKSZ. Add an assertion and a code comment to explain and check\nthis.\n\nReviewed-by: Heikki Linnakangas \nDiscussion: https://www.postgresql.org/message-id/flat/f9aab072-0078-49e4-ab93-3b08086a4406@eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1f8c504e3086cf70a1538038b63b045de5df6a9b", "date": "2026-07-15T10:58:13+02:00", "subject": "Clean up write() return type", "body": "and analogously for pg_pwrite() and FileWrite()\n\nBe sure to store the return value in a variable of type ssize_t, not\nint.\n\nSome callers of FileWrite() did not have a separate error message for\na short write. This is okay in practice because FileWriteV() sets\nENOSPC for all non-error returns, so you'll get a reasonable error\nmessage either way. But callers handled this inconsistently, and this\nbehavior isn't really prominently documented and commit 871fe4917e1\nseems to frown upon it, so it seems better to make all callers handle\nthis consistently by adding the separate error message.\n\nReviewed-by: Heikki Linnakangas \nDiscussion: https://www.postgresql.org/message-id/flat/f9aab072-0078-49e4-ab93-3b08086a4406@eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "29616c15e476eab4a81a6af20259d7d1809d4e03", "date": "2026-07-15T11:05:35+02:00", "subject": "test: clean up thread-local key in QUIC radix tests", "body": "The QUIC radix test creates a new thread-local key for each script but\nonly clears the associated value during teardown. The key itself\nis never deleted.\n\nDelete the thread-local key after all child threads have joined and the\nmain-thread value has been cleared.\n\n 1,536 bytes in 3 blocks are still reachable in loss record 873 of 915\n at 0x488D0B0: calloc (in /usr/libexec/valgrind/vgpreload_memcheck-arm64-linux.so)\n by 0x49561CF: pthread_setspecific@@GLIBC_2.34 (in /usr/lib64/libc.so.6)\n by 0x5C0FD7: CRYPTO_THREAD_set_local (threads_pthread.c:1030)\n by 0x41132F: radix_thread_init (quic_bindings.c:570)\n by 0x4114F7: bindings_process_init (quic_bindings.c:604)\n by 0x422563: test_script (main.c:28)\n by 0x5152DF: run_tests (driver.c:518)\n by 0x517447: main (main.c:52)\n\nSigned-off-by: Nikola Pajkovsky \n\nReviewed-by: Tim Hudson \nReviewed-by: Paul Dale \nReviewed-by: Neil Horman \nMergeDate: Mon Jul 20 09:07:10 2026\n(Merged from https://github.com/openssl/openssl/pull/31961)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "49d250b3d6884a693eacaefb3e08799e7c25f7f0", "date": "2026-07-15T05:13:39-04:00", "subject": "doc PG 19 relnotes: remove duplicate word", "body": "Backpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1865b55846700f48e31b93d6df7b7f95ea07a602", "date": "2026-07-15T11:23:09+02:00", "subject": "test: update provider Valgrind suppression", "body": "The internal provider test intentionally retains its configured provider\nwhen OPENSSL_cleanup() is skipped under OSSL_USE_VALGRIND. The existing\nsuppression accounts for this state but no longer matches the provider\nactivation stack because provider_init() appears between\nOSSL_provider_init() and provider_activate().\n\nAdd the missing frame so the provider context and its diagnostic strings\nare correctly suppressed.\n\nSigned-off-by: Nikola Pajkovsky \n\nReviewed-by: Tim Hudson \nReviewed-by: Paul Dale \nReviewed-by: Neil Horman \nMergeDate: Mon Jul 20 09:07:11 2026\n(Merged from https://github.com/openssl/openssl/pull/31961)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "aa9787e325df8a864cd29dac2fcfdaaf02b6b8ed", "date": "2026-07-15T05:36:00-04:00", "subject": "doc PG 19 relnotes: \"OAUTH validators to supply failure item\"", "body": "Moved to source code section.\n\nReported-by: Andreas Karlsson\n\nDiscussion: https://postgr.es/m/f5f129c3-88fa-43dc-a23a-75155fa10be3@proxel.se\n\nBackpatch-through: 19 only", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "8f2d25024a00daedfe1ff095aaef261539404ddc", "date": "2026-07-15T10:39:46+01:00", "subject": "Remove README.ja which has not been updated since 2018 and is out of date.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "1c7ed8d9adb4274cd5a9554e23d587eed5b10ab2", "date": "2026-07-15T10:40:47+01:00", "subject": "Remove README.ja from EXTRA_DIST.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "70500e4067e74e5b5f79a15774b5d2f01524f397", "date": "2026-07-15T11:49:37+02:00", "subject": "fix(core): preserve explicit input transform write type", "body": "If a directive has an input declared as `dismissible = input(true, {transform: booleanAttribute});` then the following templates were not compiling:\n\n```\n
\n
\n```\n\nThis commit fixes the issue, without breaking contravariant consumers.\n\n(cherry picked from commit 0ae6d81ed2a83b1ebbb75dabc53af77b8db86c41)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "0ae6d81ed2a83b1ebbb75dabc53af77b8db86c41", "date": "2026-07-15T11:49:37+02:00", "subject": "fix(core): preserve explicit input transform write type", "body": "If a directive has an input declared as `dismissible = input(true, {transform: booleanAttribute});` then the following templates were not compiling:\n\n```\n
\n
\n```\n\nThis commit fixes the issue, without breaking contravariant consumers.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "0546889b1b9fa80f0374ddd708c39cb63ed83518", "date": "2026-07-15T05:51:12-04:00", "subject": "doc PG 19 relnotes: swap OAUTH hook and validator items", "body": "Reported-by: Andreas Karlsson\n\nDiscussion: https://postgr.es/m/f5f129c3-88fa-43dc-a23a-75155fa10be3@proxel.se\n\nBackpatch-through: 19 only", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "671e9d4913a3260fb987f93a9f8cff4fc49a5a3d", "date": "2026-07-15T05:55:51-04:00", "subject": "feat(marks): :marks completion, list extmarks #40743", "body": "Problem:\n- No UI for listing extmarks.\n- `:marks` doesn't support cmdline-completion.\n\nSolution:\n- Add completion and extmarks support to `:marks`.\n- Update `get_buf_local_marks` to include the \":\" (prompt-buffer) mark.", "style": "conventional", "type": "feat", "scope": "marks", "breaking": false} {"repo": "axios", "sha": "3077e62097726d22ba30f1cb847d7a07050e339a", "date": "2026-07-15T11:58:53+02:00", "subject": "feat(types): Allow the Params property to be typed, instead of `any`", "body": "", "style": "conventional", "type": "feat", "scope": "types", "breaking": false} {"repo": "tmux", "sha": "5d2b6542b62a81ca0a53c977ad5c33b9c6f0ea5d", "date": "2026-07-15T11:05:15+01:00", "subject": "Fix control mode error and update regression tests.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "b2570e685255891d2e0f68da6eb665c4d3e6ea17", "date": "2026-07-15T05:10:09-05:00", "subject": "fix(showcmd): avoid stale %S contents #40747", "body": "", "style": "conventional", "type": "fix", "scope": "showcmd", "breaking": false} {"repo": "postgres", "sha": "928ed9b653e968ddb559a3fde7ff9a84b08f1196", "date": "2026-07-15T15:47:35+05:30", "subject": "Reject concurrent sequence refreshes.", "body": "'ALTER SUBSCRIPTION ... REFRESH SEQUENCES' can race with an already\nrunning sequence synchronization worker. If a second refresh request\nresets the synchronization state while the worker has already fetched\nsequence values from the publisher but has not yet applied them to the\nsubscriber, the worker can overwrite the subscriber with stale values\nand mark the synchronization as complete.\n\nAvoid this race by rejecting 'ALTER SUBSCRIPTION ... REFRESH SEQUENCES'\nwhen a sequence synchronization worker is already running for the\nsubscription. The command reports an error asking the user to rerun it\nafter the current synchronization completes.\n\nAlso add a wait for the re-added 'regress_s4' sequence to finish\nsynchronizing in 036_sequences.pl, so the subsequent test does not race\nagainst its sequencesync worker.\n\nReported-by: Noah Misch \nAuthor: vignesh C \nReviewed-by: Shveta Malik \nReviewed-by: Hayato Kuroda \nReviewed-by: Amit Kapila \nBackpatch-through: 19, where it was introduced\nDiscussion: https://postgr.es/m/20260710045217.f0.noahmisch@microsoft.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "429ea9a2f006511b9e2cbd7e59472a33571883f7", "date": "2026-07-15T12:20:03+02:00", "subject": "cms: fix AuthenticatedData authAttrs and unauthAttrs element type", "body": "The CMS_AuthenticatedData ASN.1 template declared the authAttrs and\nunauthAttrs fields as X509_ALGOR, whereas per RFC 5652 section 9.1 they\nare [2] and [3] IMPLICIT SET OF Attribute and the CMS_AuthenticatedData\nstructure already declares them as STACK_OF(X509_ATTRIBUTE). The implicit\ntags [2] and [3] were already correct, so only the element type was wrong.\n\nAn Attribute (SEQUENCE { type, SET OF value }) is structurally accepted as\nan AlgorithmIdentifier (SEQUENCE { algorithm, ANY OPTIONAL }), so parsing\ndid not fail; the attributes were merely misinterpreted, e.g. cms -cmsout\n-print rendered them as algorithm/parameter instead of decoding them as\nattributes. Use X509_ATTRIBUTE with the existing tags so the template\nmatches the structure.\n\nAdd a parse test using a BouncyCastle-generated AuthenticatedData message\ncarrying an authenticated and an unauthenticated attribute, asserting both\nare decoded as SET OF Attribute.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 10:52:08 2026\n(Merged from https://github.com/openssl/openssl/pull/31960)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d71c60e98e92701d4a488798235ef80a331e8970", "date": "2026-07-15T12:20:03+02:00", "subject": "cms: fix AuthenticatedData authAttrs and unauthAttrs element type", "body": "The CMS_AuthenticatedData ASN.1 template declared the authAttrs and\nunauthAttrs fields as X509_ALGOR, whereas per RFC 5652 section 9.1 they\nare [2] and [3] IMPLICIT SET OF Attribute and the CMS_AuthenticatedData\nstructure already declares them as STACK_OF(X509_ATTRIBUTE). The implicit\ntags [2] and [3] were already correct, so only the element type was wrong.\n\nAn Attribute (SEQUENCE { type, SET OF value }) is structurally accepted as\nan AlgorithmIdentifier (SEQUENCE { algorithm, ANY OPTIONAL }), so parsing\ndid not fail; the attributes were merely misinterpreted, e.g. cms -cmsout\n-print rendered them as algorithm/parameter instead of decoding them as\nattributes. Use X509_ATTRIBUTE with the existing tags so the template\nmatches the structure.\n\nAdd a parse test using a BouncyCastle-generated AuthenticatedData message\ncarrying an authenticated and an unauthenticated attribute, asserting both\nare decoded as SET OF Attribute.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 10:51:03 2026\n(Merged from https://github.com/openssl/openssl/pull/31960)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "cc56e924626c81ad9a2410fa06a286df97f5cf0f", "date": "2026-07-15T12:20:03+02:00", "subject": "cms: fix AuthenticatedData authAttrs and unauthAttrs element type", "body": "The CMS_AuthenticatedData ASN.1 template declared the authAttrs and\nunauthAttrs fields as X509_ALGOR, whereas per RFC 5652 section 9.1 they\nare [2] and [3] IMPLICIT SET OF Attribute and the CMS_AuthenticatedData\nstructure already declares them as STACK_OF(X509_ATTRIBUTE). The implicit\ntags [2] and [3] were already correct, so only the element type was wrong.\n\nAn Attribute (SEQUENCE { type, SET OF value }) is structurally accepted as\nan AlgorithmIdentifier (SEQUENCE { algorithm, ANY OPTIONAL }), so parsing\ndid not fail; the attributes were merely misinterpreted, e.g. cms -cmsout\n-print rendered them as algorithm/parameter instead of decoding them as\nattributes. Use X509_ATTRIBUTE with the existing tags so the template\nmatches the structure.\n\nAdd a parse test using a BouncyCastle-generated AuthenticatedData message\ncarrying an authenticated and an unauthenticated attribute, asserting both\nare decoded as SET OF Attribute.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 10:49:45 2026\n(Merged from https://github.com/openssl/openssl/pull/31960)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "19a2b69fd27e01e389339e7461ab6b135f906535", "date": "2026-07-15T12:20:03+02:00", "subject": "cms: fix AuthenticatedData authAttrs and unauthAttrs element type", "body": "The CMS_AuthenticatedData ASN.1 template declared the authAttrs and\nunauthAttrs fields as X509_ALGOR, whereas per RFC 5652 section 9.1 they\nare [2] and [3] IMPLICIT SET OF Attribute and the CMS_AuthenticatedData\nstructure already declares them as STACK_OF(X509_ATTRIBUTE). The implicit\ntags [2] and [3] were already correct, so only the element type was wrong.\n\nAn Attribute (SEQUENCE { type, SET OF value }) is structurally accepted as\nan AlgorithmIdentifier (SEQUENCE { algorithm, ANY OPTIONAL }), so parsing\ndid not fail; the attributes were merely misinterpreted, e.g. cms -cmsout\n-print rendered them as algorithm/parameter instead of decoding them as\nattributes. Use X509_ATTRIBUTE with the existing tags so the template\nmatches the structure.\n\nAdd a parse test using a BouncyCastle-generated AuthenticatedData message\ncarrying an authenticated and an unauthenticated attribute, asserting both\nare decoded as SET OF Attribute.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 10:48:20 2026\n(Merged from https://github.com/openssl/openssl/pull/31960)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "45bd99a9ce3dbc16a8572be33f08dd559a70af6f", "date": "2026-07-15T12:20:03+02:00", "subject": "cms: fix AuthenticatedData authAttrs and unauthAttrs element type", "body": "The CMS_AuthenticatedData ASN.1 template declared the authAttrs and\nunauthAttrs fields as X509_ALGOR, whereas per RFC 5652 section 9.1 they\nare [2] and [3] IMPLICIT SET OF Attribute and the CMS_AuthenticatedData\nstructure already declares them as STACK_OF(X509_ATTRIBUTE). The implicit\ntags [2] and [3] were already correct, so only the element type was wrong.\n\nAn Attribute (SEQUENCE { type, SET OF value }) is structurally accepted as\nan AlgorithmIdentifier (SEQUENCE { algorithm, ANY OPTIONAL }), so parsing\ndid not fail; the attributes were merely misinterpreted, e.g. cms -cmsout\n-print rendered them as algorithm/parameter instead of decoding them as\nattributes. Use X509_ATTRIBUTE with the existing tags so the template\nmatches the structure.\n\nAdd a parse test using a BouncyCastle-generated AuthenticatedData message\ncarrying an authenticated and an unauthenticated attribute, asserting both\nare decoded as SET OF Attribute.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 10:02:43 2026\n(Merged from https://github.com/openssl/openssl/pull/31960)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3f51c9f7ac3eddccba407b82c94d3ef19a9aa16a", "date": "2026-07-15T12:20:03+02:00", "subject": "cms: fix AuthenticatedData authAttrs and unauthAttrs element type", "body": "The CMS_AuthenticatedData ASN.1 template declared the authAttrs and\nunauthAttrs fields as X509_ALGOR, whereas per RFC 5652 section 9.1 they\nare [2] and [3] IMPLICIT SET OF Attribute and the CMS_AuthenticatedData\nstructure already declares them as STACK_OF(X509_ATTRIBUTE). The implicit\ntags [2] and [3] were already correct, so only the element type was wrong.\n\nAn Attribute (SEQUENCE { type, SET OF value }) is structurally accepted as\nan AlgorithmIdentifier (SEQUENCE { algorithm, ANY OPTIONAL }), so parsing\ndid not fail; the attributes were merely misinterpreted, e.g. cms -cmsout\n-print rendered them as algorithm/parameter instead of decoding them as\nattributes. Use X509_ATTRIBUTE with the existing tags so the template\nmatches the structure.\n\nAdd a parse test using a BouncyCastle-generated AuthenticatedData message\ncarrying an authenticated and an unauthenticated attribute, asserting both\nare decoded as SET OF Attribute.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 10:02:08 2026\n(Merged from https://github.com/openssl/openssl/pull/31960)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "cbb97b961c62ff31fe746204e203a4000afac42e", "date": "2026-07-15T10:38:31Z", "subject": "Add a mouse binding for empty areas of the window so C-Drag works on them as well.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d2b217b043d0d61862201ff3f4a308ec9a3697ad", "date": "2026-07-15T06:49:45-04:00", "subject": "fix(lsp): improve inlayHint.resolveSupport in client capabilities #40741", "body": "Problem\nThe current list of properties for inlay hint resolve support includes a\nnon-existent \"location\". This field should be \"label.location\" according\nto the specification. Tooltips and commands for inlay hint \"parts\"\naren't advertised as being resolvable.\n\nSolution\nRemove \"location\" and add \"label.location\", \"label.tooltip\", and\n\"label.command\".\n\nCloses #40740\n\n(cherry picked from commit 4bc30d6e1024788c6acf54cf3e4a98618123482e)", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "neovim", "sha": "4bc30d6e1024788c6acf54cf3e4a98618123482e", "date": "2026-07-15T06:49:45-04:00", "subject": "fix(lsp): improve inlayHint.resolveSupport in client capabilities #40741", "body": "Problem\nThe current list of properties for inlay hint resolve support includes a\nnon-existent \"location\". This field should be \"label.location\" according\nto the specification. Tooltips and commands for inlay hint \"parts\"\naren't advertised as being resolvable.\n\nSolution\nRemove \"location\" and add \"label.location\", \"label.tooltip\", and\n\"label.command\".\n\nCloses #40740", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "neovim", "sha": "44d5593afd5e16f8c673ceeed2813967550b3415", "date": "2026-07-15T06:07:17-05:00", "subject": "refactor(lsp): unify LspNotify, decor provider for capabilities #40691", "body": "Problem: Every LSP capability that sends a request on a document change\nto update its state was using its own buffer-local autocmd to do so.\nThat means there is a separate autocmd per buffer per active\ncapability/feature, and the actual code inside the autocmd callback was\nvirtually identical. The same problem occurred for capabilities that had\ncomponents to draw on the screen that had their own decoration providers.\n\nSolution: Introduce a central `LspNotify` autocmd and decoration\nprovider in the capability module itself. New base class methods\n`on_close` and `on_change` that take a client_id have been provided, for\n`didClose` and `didOpen`/`didChange` notifications respectively. New\nbase class method `on_win` that takes topline and botline has been\nprovided to add extmarks or perform other work when lines of a buffer\nare being drawn.\n\nThe autocmd callback loops through each active capability instance for\nthe buffer and calls a corresponding method with the triggering\nclient_id if it is currently attached to the buffer that triggered the\nautocmd. The decoration provider just calls on_win for all active\ncapabilities on the buffer.\n\nThis slightly tweaks folding range to make use of these new client-id\nspecific callbacks.", "style": "conventional", "type": "refactor", "scope": "lsp", "breaking": false} {"repo": "vitest", "sha": "2e1a4eee35460bb2433634a7812865b1262efe86", "date": "2026-07-15T13:26:00+02:00", "subject": "feat: pointer API", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "neovim", "sha": "3a0e6239e9b859fecb5fb543e8af013d2df63a64", "date": "2026-07-15T19:48:34+08:00", "subject": "fix(window): remove float grid from compositor layers before free #40731", "body": "Problem: Closing a floating window from a non-current tab frees its grid\nwithout removing it from the compositor's `layers` table, so the next\n`ui_comp_put_grid()` walks a dangling pointer (UAF).\n\nSolution: Call `ui_comp_remove_grid()` (and `ui_call_win_close()` for\nmultigrid UIs) before `win_free_mem()`, matching `win_close()` since\nPR #21551.\n\n(cherry picked from commit 6f370f34f2c24aed0d2acebab23c4bcc86a06d69)", "style": "conventional", "type": "fix", "scope": "window", "breaking": false} {"repo": "neovim", "sha": "6f370f34f2c24aed0d2acebab23c4bcc86a06d69", "date": "2026-07-15T19:48:34+08:00", "subject": "fix(window): remove float grid from compositor layers before free #40731", "body": "Problem: Closing a floating window from a non-current tab frees its grid\nwithout removing it from the compositor's `layers` table, so the next\n`ui_comp_put_grid()` walks a dangling pointer (UAF).\n\nSolution: Call `ui_comp_remove_grid()` (and `ui_call_win_close()` for\nmultigrid UIs) before `win_free_mem()`, matching `win_close()` since\nPR #21551.", "style": "conventional", "type": "fix", "scope": "window", "breaking": false} {"repo": "curl", "sha": "f369c7ba437962591c8aa9d0519c7b256e7ea5bf", "date": "2026-07-15T12:52:21+01:00", "subject": "quiche: set the max field section size", "body": "quiche 0.29.3 started enforcing a default\n`SETTINGS_MAX_FIELD_SECTION_SIZE` of 32 KiB, so larger response headers\nnow shut down the whole connection with `CURLE_HTTP3`. curl accepts up\nto 300 KiB of response headers with every other backend and HTTP\nversion. Tell quiche to allow what curl itself allows. This keeps\ntest_01_11 passing with quiche 0.29.3 and it also advertises our real\nlimit to servers, which the RFC encourages.\n\nRef: 899e2edec95b3df01db63ad219d3c5547a3e3ae5 #22333\nRef: #22329\nRef: #22325\nRef: https://github.com/cloudflare/quiche/commit/9be0e4fa18594a5cca00120f6ae392e217e837e5\nRef: https://github.com/cloudflare/quiche/releases/tag/0.29.3\n\nCloses #22331", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "1ac3975fcd0a44c0735dbb380d32d5689c160e35", "date": "2026-07-15T14:07:35+02:00", "subject": "add exceptions to the --data* options", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "5f2a70abe3f780b6a97ac7039b7862bd828302cd", "date": "2026-07-15T14:08:18+02:00", "subject": "tidy-up: `TEXT()` vs `_TEXT()` vs `_T()` use (Windows)", "body": "Use `_TEXT()` when interacting with CRT functions (also prefer over\nsynonym `_T()`), `TEXT()` for Win32 functions.\n\nWithin curl, they mean the same because CRT/Win32 Unicode mode are\nalways enabled in sync.\n\nRef: https://devblogs.microsoft.com/oldnewthing/20040212-00/?p=40643/\n\nCloses #22334", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "c89f71a74ce312f9f6522796a106eac48bd33ebb", "date": "2026-07-15T14:10:45+02:00", "subject": "fix(core): ignore processing instruction syntax in templates", "body": "Updates the template parser to detect and ignore processing instruction syntax (e.g. `` or ``). Currently it is being printed out as text.\n\nFixes #34371.\n\n(cherry picked from commit 68ac20407466f0d3e6f5c5016f1273672b5d2b34)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "68ac20407466f0d3e6f5c5016f1273672b5d2b34", "date": "2026-07-15T14:10:45+02:00", "subject": "fix(core): ignore processing instruction syntax in templates", "body": "Updates the template parser to detect and ignore processing instruction syntax (e.g. `` or ``). Currently it is being printed out as text.\n\nFixes #34371.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "ae8c4bd558c5d55c5a357e910d496eea8ff33647", "date": "2026-07-15T21:22:38+09:00", "subject": "Fix argument names in pg_clear_attribute_stats() errors", "body": "pg_clear_attribute_stats() checks its required arguments manually\nbecause the function is not strict. Previously, when schemaname or\nrelname was passed as NULL, the error incorrectly reported the\nargument name as \"relation\" in both cases:\n\n ERROR: argument \"relation\" must not be null\n\nThis was misleading, especially for schemaname, and inconsistent with\nthe function's SQL-visible argument names.\n\nThe cause is that cleararginfo[] in attribute_stats.c used\n\"relation\" for both the schema-name and relation-name arguments.\n\nThis commit fixes the issue by using \"schemaname\" and \"relname\" instead,\nmatching the function's declared argument names so that the error reports\nthe correct argument name.\n\nBackpatch to v18, where pg_clear_attribute_stats() was introduced.\n\nAuthor: Ilia Evdokimov \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/4bf66c5e-8dd7-4ef3-8691-db67ecff6f16@tantorlabs.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "75c874c55b4acae548706a8235f7a7ddc4a83cda", "date": "2026-07-15T12:45:39Z", "subject": "Clear the \"waiting for editor\" area before writing so stray characters aren't visible from underneath.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "b21d6cdd58966d4913daf24df94ce4c57ff5f8f9", "date": "2026-07-15T15:41:01+02:00", "subject": "docs: update \"browser set\" links to avoid bad redirect", "body": "\"Supported browsers\" page got moved from\nhttps://web-platform-dx.github.io/web-features/supported-browsers to https://web-platform-dx.github.io/supported-browsers/\n\nThere's a redirect in place, but it drops the query parameter with the date, so all versions link to the current browser set.\n\n(cherry picked from commit ea83f13388470ceb981b1130fdb2cc6b79e299fc)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "ea83f13388470ceb981b1130fdb2cc6b79e299fc", "date": "2026-07-15T15:41:01+02:00", "subject": "docs: update \"browser set\" links to avoid bad redirect", "body": "\"Supported browsers\" page got moved from\r\nhttps://web-platform-dx.github.io/web-features/supported-browsers to https://web-platform-dx.github.io/supported-browsers/\r\n\r\nThere's a redirect in place, but it drops the query parameter with the date, so all versions link to the current browser set.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a50aa72aba6c3fcaf85faa93513e2798e25efa2c", "date": "2026-07-15T15:44:38+02:00", "subject": "Apply suggestion from @andrewkdinh", "body": "Co-authored-by: Andrew Dinh \n\nReviewed-by: Andrew Dinh \nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nMergeDate: Wed Jul 29 17:04:49 2026\n(Merged from https://github.com/openssl/openssl/pull/31957)\n\n(cherry picked from commit 33482fbddcf137dde9bd84ae12f2f02cbc358c78)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "33482fbddcf137dde9bd84ae12f2f02cbc358c78", "date": "2026-07-15T15:44:38+02:00", "subject": "Apply suggestion from @andrewkdinh", "body": "Co-authored-by: Andrew Dinh \n\nReviewed-by: Andrew Dinh \nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nMergeDate: Wed Jul 29 17:04:49 2026\n(Merged from https://github.com/openssl/openssl/pull/31957)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "3d42cdf380b23da000de0ff9d6c3de5d77e0b0d1", "date": "2026-07-15T22:05:46+08:00", "subject": "fix(runtime-vapor): remove unsafe slot dry runs from vdom interop", "body": "revert #14793\nVapor cannot distinguish between slots.default() calls used for VNode\ninspection and calls used for actual rendering. Eager slot collection can\ntherefore execute user code or claim hydration nodes before rendering.", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "tmux", "sha": "edefce8eb56fae365ee0bfac56e9e8f714efcb01", "date": "2026-07-15T14:14:50Z", "subject": "Small section missed from previous.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "72857780bc71ac83f960a51637d7b0cb4f09971c", "date": "2026-07-15T16:19:26+02:00", "subject": "chore: update multerExceptions constant", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "899e2edec95b3df01db63ad219d3c5547a3e3ae5", "date": "2026-07-15T17:02:21+02:00", "subject": "INTERNALS.md: require quiche 0.20.0+", "body": "For function `quiche_h3_config_set_max_field_section_size()`.\n\nRef: #22331\nRef: https://github.com/curl/curl/pull/22331#issuecomment-4981647189\nRef: https://github.com/cloudflare/quiche/releases/tag/0.20.0\n\nCloses #22333", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "966e5eb2311ced0cc23759369bbb3fa867cbea52", "date": "2026-07-15T18:10:50+03:00", "subject": "feat(devtools): show Angie on the empty-state screens", "body": "The DevTools state screens (Angular not detected, unsupported version,\nproduction build) showed a plain Material icon above the message. Show\nthe Angie mascot instead, matching the treatment already used across\nangular.dev (the 404 page, empty search, the preview error state).\n\nEach screen uses a pose that fits its message: the dizzy `error` pose for\n\"application not detected\", `coding-01` for the version-upgrade prompt,\nand `sad` for the production build notice.\n\n(cherry picked from commit d4f2313e4d6700011ec42979f4b2ca4756bc7e56)", "style": "conventional", "type": "feat", "scope": "devtools", "breaking": false} {"repo": "angular", "sha": "d4f2313e4d6700011ec42979f4b2ca4756bc7e56", "date": "2026-07-15T18:10:50+03:00", "subject": "feat(devtools): show Angie on the empty-state screens", "body": "The DevTools state screens (Angular not detected, unsupported version,\nproduction build) showed a plain Material icon above the message. Show\nthe Angie mascot instead, matching the treatment already used across\nangular.dev (the 404 page, empty search, the preview error state).\n\nEach screen uses a pose that fits its message: the dizzy `error` pose for\n\"application not detected\", `coding-01` for the version-upgrade prompt,\nand `sad` for the production build notice.", "style": "conventional", "type": "feat", "scope": "devtools", "breaking": false} {"repo": "curl", "sha": "7ad33fe56f25424fb51a10405b849de81ff74b98", "date": "2026-07-15T17:21:04+02:00", "subject": "cf-dns: resolve on filter demand", "body": "Use separate dns cache entries for addresses (A+AAAA) and HTTPS-RR\nresults. That makes also \"negative\" results independent of each other.\nDns cache entries, once in use, can no longer be modified safely, as\nconcurrent use would require each access then to be done under lock. By\nusing separate entries, we can update a HTTPS entry without needing to\nduplicate an existing address entry for the same host+port.\n\nConnection filters can now ask for DNS resolves. This works at any time\nduring connection setup and while connect is going on.\n`Curl_conn_dns_add_addr_resolve()` and\n`Curl_conn_dns_add_https_resolve()` are used for that. They check if the\nresolve is already ongoing, can be added to a resolve not started yet\nor, as last resort, create a new filter instance and add it to the\nconnection (it's easier to add more filters than making the same filter\nhandle multiple resolves. Since DNS filters are removed once the\nconnection is established, there is no later penalty).\n\nHTTPS-RR queries are added by the `HTTPS-CONNECT`, `SSL` and `QUIC`\nfilters. The latter will only do that when ECH is configured and\nsupported. That means we trigger HTTPS-RR queries only when the results\nmatter.\n\nAdd test_22_06 for ALPN influenced via HTTPS-RR when tunneling through a\nproxy. This did not work before.\n\nAdjust test2100 to use https: as the previous http: URL no longer\ntriggers HTTPS-RR resolves.\n\nCloses #22216", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "1386135d1c65a3d32c2887a60e66b41afef341ec", "date": "2026-07-15T18:04:59+02:00", "subject": "schannel: fix error check logic in `get_client_cert()` file reader", "body": "Reported by GitHub Code Quality\nFollow-up to 0fdf96512613574591f501d63fe49495ba40e1d5 #5193\n\nCloses #22415", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "7c7be0506e512eb202ceb2e30815b16a32bdcbef", "date": "2026-07-16T00:46:55+08:00", "subject": "Fix initialization of nested for loop head declarations", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "2aecfb0f3829699c573d87bb4ada98d949ecf3b0", "date": "2026-07-15T19:55:40+03:00", "subject": "refactor(devtools): format code", "body": "Format DevTools TypeScript files.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "postgres", "sha": "0c5516a0df97e1e707ea5e9e08712e29ce4ee047", "date": "2026-07-15T13:04:32-04:00", "subject": "Add additional sanity checks when reading a blkreftable.", "body": "Code elsewhere in the system assumes that fork numbers and chunk sizes\nare within bounds, so the code that reads those quantities from disk\nshould validate that they are. Without these additional checks, a\ncorrupted file can cause us to index off the end of fork number or chunk\nentry arrays, potentially resulting in a crash.\n\nReported-by: oxsignal (chunk sizes)\nReported-by: Robert Haas (fork numbers)\nReviewed-by: Daniel Gustafsson \nDiscussion: http://postgr.es/m/CA+TgmoYP8RKoBGosS7C6Fdr-GNCfyz_W1zmK=Tx1Fe0ZvzGh0g@mail.gmail.com\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "3d3b8872814362e061802703bb5d92fe050fc602", "date": "2026-07-15T13:24:00-04:00", "subject": "Make max_pinned_buffers be > 0, don't just Assert that it is.", "body": "It was possible to trigger read_stream_begin_relation's\nassertion that it calculated nonzero max_pinned_buffers,\nin scenarios with unreasonably small numbers of buffers.\nThere seems to be no visible problem in non-Assert builds.\n\nThe bug occurs only in v17 (the first branch containing\nthis logic), because commit 92fc6856c replaced that Assert\nwith a runtime clamp. I won't risk back-patching that entire\ncommit, but let's copy just this minimal fix.\n\nBug: #19551\nReported-by: Roman Zharkov \nAuthor: Thomas Munro \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/19551-d4a55b8fe81db4f2@postgresql.org\nBackpatch-through: 17 only", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "axios", "sha": "58b16c88f0bddf1fadb321aed58ba3f49a90481b", "date": "2026-07-16T01:59:35+08:00", "subject": "refactor(helpers): extract duplicated setFormDataHeaders into a shared helper", "body": "", "style": "conventional", "type": "refactor", "scope": "helpers", "breaking": false} {"repo": "axios", "sha": "3041b8fd1daf17404d1bad1f9d94026ea5ab400b", "date": "2026-07-16T03:32:20+09:00", "subject": "feat(HttpStatusCode): add missing 520 status code", "body": "", "style": "conventional", "type": "feat", "scope": "HttpStatusCode", "breaking": false} {"repo": "angular", "sha": "fa6a1d26bda971a0cd1994e67dd001262dbdb833", "date": "2026-07-15T21:33:36+03:00", "subject": "test(elements): disambiguate the setTimeout spy type", "body": "Bind spyOn explicitly to Window so Jasmine uses the DOM setTimeout signature that returns a number instead of the Node.js Timeout type. Remove the unsafe any cast and the obsolete TODO.\n\n(cherry picked from commit 24a3c63976390f3370d85b147379c66fc0e0802d)", "style": "conventional", "type": "test", "scope": "elements", "breaking": false} {"repo": "angular", "sha": "24a3c63976390f3370d85b147379c66fc0e0802d", "date": "2026-07-15T21:33:36+03:00", "subject": "test(elements): disambiguate the setTimeout spy type", "body": "Bind spyOn explicitly to Window so Jasmine uses the DOM setTimeout signature that returns a number instead of the Node.js Timeout type. Remove the unsafe any cast and the obsolete TODO.", "style": "conventional", "type": "test", "scope": "elements", "breaking": false} {"repo": "angular", "sha": "98115e813718d30b3ae561e3dd5343b4879a9574", "date": "2026-07-15T21:59:04+03:00", "subject": "refactor(devtools): revamp highlighter", "body": "Revamp the highlighter by introducing a completely new mechanism.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "angular", "sha": "f034c02e67d25018ca3ff952904d9bcaf43175d4", "date": "2026-07-15T19:22:52Z", "subject": "build: update cross-repo angular dependencies to v22.1.0-next.4", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "829d98e9ae08b12a2c4a624a64f87660d6dfde26", "date": "2026-07-15T19:25:40Z", "subject": "build: update cross-repo angular dependencies to v22.0.7", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "94e1b104f8db88837e5ef13a6692dad746e3690f", "date": "2026-07-15T12:34:20-07:00", "subject": "Fix like_fixed_prefix_ci() selectivity.", "body": "A wrong calculation introduced by 9c8de15969 could cause trailing\ncharacters from the prefix to be passed to like_selectivity() rather\nthan just the \"rest\".\n\nDiscussion: https://postgr.es/m/c7334a7a44243d2e4ec5e83747589908b3787491.camel@j-davis.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9f8231027bcc9a895537fcab4e0a2e65afc9befc", "date": "2026-07-16T05:36:42+10:00", "subject": "CHANGES.md updates", "body": "Reviewed-by: Bob Beck \nReviewed-by: Norbert Pocs \nMergeDate: Tue Jul 21 14:12:07 2026\n(Merged from https://github.com/openssl/openssl/pull/31964)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "79abbc90a86e930a9ad69c918f2ce1c25549290e", "date": "2026-07-15T21:40:09+02:00", "subject": "pgbench: Fix incorrect parameter name in error message", "body": "Commit 6f164e6d17616 accidentally mistyped --client as --clients in\nthe error message. Backpatch down to v15 where the it was introduced.\n\nAuthor: Semih Doğan \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CALOtZ7tuWisV=v0cUY_q6PLHJ-fOiQ7ZN476JwmM0PyV0t5i7Q@mail.gmail.com\nBackpatch-through: 15", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0fd5595aa3e71e54a297b18bc104f487c8555550", "date": "2026-07-15T15:49:59-04:00", "subject": "Include last block in FSM vacuum of bulk extended relation", "body": "When bulk-extending a relation, we add the newly-added blocks that we\nwon't immediately use to the free space map and then call\nFreeSpaceMapVacuumRange() to propagate that free space up the FSM tree,\nso other backends can find and reuse it.\n\nHowever, the end block argument to FreeSpaceMapVacuumRange() is\nexclusive, and we passed the number of the last added block (since\n00d1e02be24). If that block was the first one covered by a new FSM page,\nits free space wasn't propagated up the tree and was therefore invisible\nto FSM searches until the next FSM vacuum.\n\nFix by passing the block number one past the last added block, so the\nfull range is vacuumed.\n\nAuthor: Jingtang Zhang \nReviewed-by: Melanie Plageman \nDiscussion: https://postgr.es/m/flat/CAPsk3_Bx_vdybN%3D-DZu8HLStf%2BXnuFUBkLwxouONSMkWuO9oug%40mail.gmail.com\nBackpatch-through: 16", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ed62d26cacac96ca5c6b4e6fcc5308b11a8371c8", "date": "2026-07-15T17:24:49-04:00", "subject": "Fix VM clear WAL logging by registering VM blocks", "body": "Heap WAL records that clear bits on the visibility map (like inserts and\ndeletes) did not register the visibility map blocks they modified.\nBecause the WAL summarizer only records registered blocks, an\nincremental backup taken over such operations would omit the changed VM\npages. On restore, the VM would retain stale all-visible/all-frozen\nbits, which can cause wrong results from index-only scans and incorrect\nrelfrozenxid advancement due to vacuum page skipping.\n\nNot registering the VM buffer also meant we never emitted FPIs of VM\npages when clearing bits. A torn VM page won't raise an error because\nthe VM is read with ZERO_ON_ERROR; with checksums on, it would be\ndetected and zeroed, but with checksums off, it is accepted as-is and\ncan lead to data corruption.\n\nFix this by registering the VM buffer in the WAL record when clearing VM\nbits. The VM buffer must now be locked throughout the critical section\nthat modifies the VM and heap pages and emits the WAL record. This can\nslow down operations that clear the VM, since the VM lock is held longer\nand VM FPIs may be emitted, but it is required for correctness.\n\nNote that this fix does not repair existing incremental backups.\n\nBumps XLOG_PAGE_MAGIC.\n\nAuthor: Melanie Plageman \nAuthor: Andres Freund \nReviewed-by: Robert Haas \nReviewed-by: Andrey Borodin \nDiscussion: https://postgr.es/m/flat/CAAKRu_bn%2Be7F4yPFBgFbnP%2BsyJRKyNK092bjD2LKvZW7O4Svag>\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2340fb8f02b28d2fe7f0023d84c3055a38cd5b9e", "date": "2026-07-15T17:24:49-04:00", "subject": "Make VM clear take a RelFileLocator and not fake relcache", "body": "This brings it in line with visibilitymap_set(). An upcoming commit will\nstart reading visibility map buffers with XLogReadBuffer*() functions,\nso we no longer require a relation object to pin the visibility map,\nallowing us to remove this fake relcache business altogether.\n\nWhile we're at it, remove a spurious const qualifier from the\nRelFileLocator parameter to visibilitymap_set().\n\nAuthor: Melanie Plageman \nDiscussion: https://postgr.es/m/flat/CAAKRu_bn%2Be7F4yPFBgFbnP%2BsyJRKyNK092bjD2LKvZW7O4Svag%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b01c31eef9c3a83d0bd9f30656cedaa6722890ee", "date": "2026-07-15T17:32:17-04:00", "subject": "Fix VM clear WAL logging by registering VM blocks", "body": "Heap WAL records that clear bits on the visibility map (like inserts and\ndeletes) did not register the visibility map blocks they modified.\nBecause the WAL summarizer only records registered blocks, an\nincremental backup taken over such operations would omit the changed VM\npages. On restore, the VM would retain stale all-visible/all-frozen\nbits, which can cause wrong results from index-only scans and incorrect\nrelfrozenxid advancement due to vacuum page skipping.\n\nNot registering the VM buffer also meant we never emitted FPIs of VM\npages when clearing bits. A torn VM page won't raise an error because\nthe VM is read with ZERO_ON_ERROR; with checksums on, it would be\ndetected and zeroed, but with checksums off, it is accepted as-is and\ncan lead to data corruption.\n\nFix this by registering the VM buffer in the WAL record when clearing VM\nbits. The VM buffer must now be locked throughout the critical section\nthat modifies the VM and heap pages and emits the WAL record. This can\nslow down operations that clear the VM, since the VM lock is held longer\nand VM FPIs may be emitted, but it is required for correctness.\n\nNote that this fix does not repair existing incremental backups.\n\nBumps XLOG_PAGE_MAGIC. Though it is late in the cycle (post-beta 2) to\nbe doing so, that seemed better than maintaining the backwards\ncompatability code in yet another branch.\n\nAuthor: Melanie Plageman \nAuthor: Andres Freund \nReviewed-by: Robert Haas \nReviewed-by: Andrey Borodin \nDiscussion: https://postgr.es/m/flat/CAAKRu_bn%2Be7F4yPFBgFbnP%2BsyJRKyNK092bjD2LKvZW7O4Svag>\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0672134309afb0f378cf31bf4eedb8e2aab57456", "date": "2026-07-15T17:43:38-04:00", "subject": "Test that VM clear registers VM buffers", "body": "The WAL summarizer only tracks registered buffers, so unregistered VM\nclears are ommitted from incremental backups, corrupting the restored\nvisibility map. Test those cases are now fixed.\n\nAuthor: Melanie Plageman \nReviewed-by: Andrey Borodin \nDiscussion: https://postgr.es/m/oqcsevg35xjan2327x5kdfth6q4fgeqboxfo3v3imeyih2uiny%406sez5dzxl6nt\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c0d9864f5ce47902dd258bb61dd89ada6d020b55", "date": "2026-07-15T17:43:38-04:00", "subject": "Fix VM clear WAL logging by registering VM blocks", "body": "Heap WAL records that clear bits on the visibility map (like inserts and\ndeletes) did not register the visibility map blocks they modified.\nBecause the WAL summarizer only records registered blocks, an\nincremental backup taken over such operations would omit the changed VM\npages. On restore, the VM would retain stale all-visible/all-frozen\nbits, which can cause wrong results from index-only scans and incorrect\nrelfrozenxid advancement due to vacuum page skipping.\n\nNot registering the VM buffer also meant we never emitted FPIs of VM\npages when clearing bits. A torn VM page won't raise an error because\nthe VM is read with ZERO_ON_ERROR; with checksums on, it would be\ndetected and zeroed, but with checksums off, it is accepted as-is and\ncan lead to data corruption.\n\nFix this by registering the VM buffer in the WAL record when clearing VM\nbits. The VM buffer must now be locked throughout the critical section\nthat modifies the VM and heap pages and emits the WAL record. This can\nslow down operations that clear the VM, since the VM lock is held longer\nand VM FPIs may be emitted, but it is required for correctness.\n\nNote that this fix does not repair existing incremental backups.\n\nAuthor: Melanie Plageman \nAuthor: Andres Freund \nReviewed-by: Robert Haas \nReviewed-by: Andrey Borodin \nDiscussion: https://postgr.es/m/flat/CAAKRu_bn%2Be7F4yPFBgFbnP%2BsyJRKyNK092bjD2LKvZW7O4Svag>\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7edec8b5784130b9c175cd8f5ab4d029d4502453", "date": "2026-07-15T17:43:38-04:00", "subject": "Introduce macros for WAL block reference IDs of some heap record types", "body": "When registering a buffer with the WAL machinery, the caller assigns it a\nblock reference ID, and replay must read each block back by that same ID.\nToday these IDs are bare integers assigned by convention (0, 1, 2, ...),\nwhich is easy to follow when a record registers a single block, or when the\nblocks are handled during replay in their registration order.\n\nAn upcoming bug fix registers up to two visibility map blocks in addition\nto the heap block(s) when clearing the VM, and these are not handled during\nreplay in a straightforward 1:1, in-registration-order fashion. Relying on\nbare integers for the block IDs in that case is error-prone.\n\nIntroduce macros naming the block reference IDs for the heap record types\nthat the upcoming commit extends to register visibility map blocks, so the\nregistration and replay sites refer to the same block by a meaningful name.\n\nAuthor: Melanie Plageman \nReviewed-by: Robert Haas \nDiscussion: https://postgr.es/m/66mqpfyti3qhfttcsv6r2lbvqqd32rrmpn6i47ovrsnvguts46%40gou54xc>\nBackpatch through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7c59d28b1060509af998cd6e33152d48465e94fe", "date": "2026-07-15T23:59:08+02:00", "subject": "apps: cover the req -subj modification of an existing request", "body": "Add a subtest exercising the code path in the req app that rewrites the\nsubject of an existing certificate request loaded via -in, i.e. when\n-subj is given without -new/-newkey/-x509.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Matt Caswell \nReviewed-by: Ryan Hooper \nMergeDate: Mon Jul 27 10:19:06 2026\n(Merged from https://github.com/openssl/openssl/pull/31965)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "64d6d47a0c67709ce887a90666fc214c40a283c4", "date": "2026-07-15T17:56:48-05:00", "subject": "fix(forms): preserve intermediate number values in signal forms", "body": "Preserve raw native input text while editing so parsed model values are not written back on every keystroke.\n\n(cherry picked from commit 2e32872720ee95f24029d20e24e91edbc2054b1e)", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "2e32872720ee95f24029d20e24e91edbc2054b1e", "date": "2026-07-15T17:56:48-05:00", "subject": "fix(forms): preserve intermediate number values in signal forms", "body": "Preserve raw native input text while editing so parsed model values are not written back on every keystroke.", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "neovim", "sha": "a1985068154d9e137bb36d46721838c8c35bf596", "date": "2026-07-15T18:05:46-05:00", "subject": "fix(restart): keep :restart error message concise #40746", "body": "(cherry picked from commit 4f1ef2dec1f867c934a42728b69806ffdfecb4db)", "style": "conventional", "type": "fix", "scope": "restart", "breaking": false} {"repo": "neovim", "sha": "4f1ef2dec1f867c934a42728b69806ffdfecb4db", "date": "2026-07-15T18:05:46-05:00", "subject": "fix(restart): keep :restart error message concise #40746", "body": "", "style": "conventional", "type": "fix", "scope": "restart", "breaking": false} {"repo": "neovim", "sha": "5ce9e74f133b86f06ba42d07337660cc7c5785d5", "date": "2026-07-16T07:08:37+08:00", "subject": "feat(completion): commitCharacters #38417", "body": "Problem: LSP completion commitCharacters are not handled. Typing a\ncommit character (e.g. `.`, `(`, `;`) while a completion item is\nselected does not accept the item first.\n\nSolution: Store commit characters as a flat string in complete-items.\nCheck it before completion stops, accept the match and let\nthe character be inserted normally.", "style": "conventional", "type": "feat", "scope": "completion", "breaking": false} {"repo": "neovim", "sha": "fa3a0a7feb248fc953844ef1ef12015ea5c7a9f3", "date": "2026-07-16T00:09:54+01:00", "subject": "fix(window): yeet grid later when closing float in other tabpage #40758", "body": "Problem: #40731 may still crash if close_buffer autocmds reinsert the float's\ngrid. Plus removing the grid (and posting win_close) is unneeded if\nwin_close_othertab refuses to close the window later, which is possible.\n\nSolution: do the stuff before freeing the window, like win_close.\n(cherry picked from commit 1741da8412134fc5a2f83e509da925c4e2d65034)", "style": "conventional", "type": "fix", "scope": "window", "breaking": false} {"repo": "neovim", "sha": "1741da8412134fc5a2f83e509da925c4e2d65034", "date": "2026-07-16T00:09:54+01:00", "subject": "fix(window): yeet grid later when closing float in other tabpage #40758", "body": "Problem: #40731 may still crash if close_buffer autocmds reinsert the float's\ngrid. Plus removing the grid (and posting win_close) is unneeded if\nwin_close_othertab refuses to close the window later, which is possible.\n\nSolution: do the stuff before freeing the window, like win_close.", "style": "conventional", "type": "fix", "scope": "window", "breaking": false} {"repo": "vue-core", "sha": "eb89e935a771ab7fc3f8e67a82bc8fc7ea28bdaa", "date": "2026-07-16T08:16:35+08:00", "subject": "fix(runtime-dom): respect current select model type", "body": "close #15009", "style": "conventional", "type": "fix", "scope": "runtime-dom", "breaking": false} {"repo": "vue-core", "sha": "164460a52988c44dc34a541219c88120078372a0", "date": "2026-07-16T02:16:57+02:00", "subject": "fix(types): don't constrain component $el type to Element", "body": "`$el`'s type param (`TypeEl`) lives in `@vue/runtime-core`, which is\nrenderer-agnostic, yet it's constrained to the DOM `Element` interface.\nCustom renderers (TUI, canvas, native, …) have host nodes that aren't DOM\n`Element`s, so they can't type `$el` without laundering through `& Element`.\n\nRelax the `TypeEl` bound to unconstrained (default stays `any`): DOM\nelements still qualify, non-DOM host nodes now do too. Backward compatible —\nit only widens accepted inputs.", "style": "conventional", "type": "fix", "scope": "types", "breaking": false} {"repo": "vue-core", "sha": "bd962bbaca746d8f65775a07723420406ebad0da", "date": "2026-07-16T08:17:19+08:00", "subject": "fix(shared): prevent SSR comment escaping from creating closing delimiters", "body": "Refs GHSA-9rff-vwfp-83hv.", "style": "conventional", "type": "fix", "scope": "shared", "breaking": false} {"repo": "vue-core", "sha": "4d35ecadb6f625b0f7407ccee0e357c2dd8084c7", "date": "2026-07-16T09:18:35+09:00", "subject": "fix(server-renderer): remove package dependency cycle", "body": "", "style": "conventional", "type": "fix", "scope": "server-renderer", "breaking": false} {"repo": "vue-core", "sha": "ddc132d7c054ecd2029ddcc95bcba770cd44f022", "date": "2026-07-16T02:18:56+02:00", "subject": "fix(runtime-core): unwind dangling blocks when slot content throws", "body": "fix #15070", "style": "conventional", "type": "fix", "scope": "runtime-core", "breaking": false} {"repo": "vue-core", "sha": "474907c0f01d021c564545137051a2b56eb4a769", "date": "2026-07-16T08:19:18+08:00", "subject": "fix(server-renderer): handle errors in optimized component renders", "body": "close #12575", "style": "conventional", "type": "fix", "scope": "server-renderer", "breaking": false} {"repo": "vue-core", "sha": "e0d27234ddd2061383cb5dcc475d6b9558ed8e4f", "date": "2026-07-16T01:19:40+01:00", "subject": "fix(hydration): pass namespace when patching dynamic props", "body": "close #15081\nclose #15050", "style": "conventional", "type": "fix", "scope": "hydration", "breaking": false} {"repo": "vue-core", "sha": "378f9783248a4a4f20781d665d4f7089f39cae6e", "date": "2026-07-16T05:20:06+05:00", "subject": "fix(reactivity): handle effect removal during scope stop", "body": "close #15083", "style": "conventional", "type": "fix", "scope": "reactivity", "breaking": false} {"repo": "vue-core", "sha": "7262e6398e380956cd4640e3cec6987959cd7b8a", "date": "2026-07-16T01:21:06+01:00", "subject": "chore(ci): improve workflows", "body": "", "style": "conventional", "type": "chore", "scope": "ci", "breaking": false} {"repo": "vue-core", "sha": "8801e51ece2bf0c115f3ba31279e0230010d1114", "date": "2026-07-16T08:34:02+08:00", "subject": "test: add expect warn", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "97f3525551ad06103083eb2cbf9858c5e57cd58e", "date": "2026-07-16T08:58:30+08:00", "subject": "fix(runtime-core): skip lazy hydration for detached roots", "body": "close #15091", "style": "conventional", "type": "fix", "scope": "runtime-core", "breaking": false} {"repo": "ant-design", "sha": "4f27ed9be1c9eb6f7254006a30d408474bb33a2c", "date": "2026-07-16T10:01:04+08:00", "subject": "docs: add Button loading animation investigation", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "20c9d263d3bfafdfbcc1921f259fd8f711e18659", "date": "2026-07-16T11:01:05+08:00", "subject": "fix(compiler-core): avoid leaking slot branch keys", "body": "close #15048", "style": "conventional", "type": "fix", "scope": "compiler-core", "breaking": false} {"repo": "ant-design", "sha": "5bcf1489acc9b629ce7d7133fed2d2e47eaa4c6c", "date": "2026-07-16T11:42:18+08:00", "subject": "style(Button): align loading icon styles", "body": "* style(Button): align loading icon styles\n\n* chore: retrigger CI", "style": "conventional", "type": "style", "scope": "Button", "breaking": false} {"repo": "vite", "sha": "c88c236186d8e2384653e382dd0960ca1ff5b24c", "date": "2026-07-16T04:42:33+01:00", "subject": "docs(build): fix incorrect `@default` for build.cssMinify", "body": "", "style": "conventional", "type": "docs", "scope": "build", "breaking": false} {"repo": "postgres", "sha": "e6e8277f420f3b348ad5e414761196c5d02acb82", "date": "2026-07-16T13:35:36+09:00", "subject": "doc: Fix log_parameter_max_length docs to reference log_min_duration_statement", "body": "The documentation for log_parameter_max_length said it affects messages\ngenerated by log_duration. However, log_duration alone does not log bind\nparameter values, so this is misleading.\n\nThis commit updates the documentation to reference log_min_duration_statement,\nwhich can log bind parameters, to better reflect actual behavior.\n\nBackpatch to all supported versions.\n\nAuthor: Fujii Masao \nReviewed-by: Surya Poondla \nDiscussion: https://postgr.es/m/CAHGQGwGnCVMVz8-LU9F8Sh57bkQX3jMZzx7age7M0LFEz5=Fog@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "8a84ddd8c63289e14e1e1cc2c8a4a9c41f652e1a", "date": "2026-07-16T13:37:28+09:00", "subject": "doc: Mention REPACK in MAINTAIN privilege descriptions", "body": "REPACK requires the MAINTAIN privilege, but it was omitted from the\nlists of commands covered by that privilege in ddl.sgml and the\ndescription of the predefined pg_maintain role in user-manag.sgml.\nThis was an oversight in commit ac58465e061, which introduced\nREPACK.\n\nAdd REPACK to both documentation lists, and update the corresponding\ncomment in aclchk.c.\n\nAuthor: Shinya Kato \nReviewed-by: Ewan Young \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAOzEurRJOVokiB2J8nrF569nX-ZMb0oRSB0C=yZQ17mZxd4_BQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2f4f193fdbe057029b87ae45822e22e7a8b85693", "date": "2026-07-16T13:38:34+09:00", "subject": "Check CREATE_REPLICATION_SLOT response shape in libpqwalreceiver", "body": "Previously, libpqrcv_create_slot() checked only that\nCREATE_REPLICATION_SLOT returned PGRES_TUPLES_OK before reading\nvalues from the first row. If the server unexpectedly returned an\ninvalid result, such as zero rows, PQgetvalue() could return NULL,\nleading to a crash while parsing the LSN.\n\nOther replication commands, such as IDENTIFY_SYSTEM, already validate\nthe response shape before accessing result values, but\nCREATE_REPLICATION_SLOT did not.\n\nFix this by verifying that CREATE_REPLICATION_SLOT response contains\nexactly one row with four fields, and report a protocol violation otherwise.\n\nBackpatch to all supported versions.\n\nBug: #19547\nReported-by: Yuelin Wang <1217816127@qq.com>\nAuthor: Kenny Chen \nReviewed-by: Hayato Kuroda \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/19547-f7986f668f71e788@postgresql.org\nDiscussion: https://postgr.es/m/CAPXstDtW2iqe+DJAOTQTX+rRziJp2UhZSo1+HRj1COAtbu+nKw@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "810032097079be1a7da0e2b09ec9d92dd07ec13f", "date": "2026-07-16T13:39:26+09:00", "subject": "fix(bundled-dev): avoid duplicated `buildEnd`", "body": "", "style": "conventional", "type": "fix", "scope": "bundled-dev", "breaking": false} {"repo": "vite", "sha": "c4df6ef9c3c0d796fb5f7968422ed14d528e3388", "date": "2026-07-16T13:39:35+09:00", "subject": "fix(legacy): don't use newer syntax when minifying polyfill chunks", "body": "", "style": "conventional", "type": "fix", "scope": "legacy", "breaking": false} {"repo": "vite", "sha": "f8b38e316bcefbf29f762f90ee49c88cd52c43b5", "date": "2026-07-16T06:52:24+02:00", "subject": "fix(module-runner): don't crash stack-trace source mapping when globalThis.Buffer is absent", "body": "", "style": "conventional", "type": "fix", "scope": "module-runner", "breaking": false} {"repo": "vite", "sha": "0f2e815d14820ad1715022d9c9bf37a089742696", "date": "2026-07-16T12:53:59+08:00", "subject": "docs(static-deploy): fix links for Vercel deployment environments", "body": "", "style": "conventional", "type": "docs", "scope": "static-deploy", "breaking": false} {"repo": "vite", "sha": "5a72b8780705b575026617e86b0b92dea63a56a5", "date": "2026-07-16T13:08:15+08:00", "subject": "fix(client): overlay error message format align rolldown", "body": "", "style": "conventional", "type": "fix", "scope": "client", "breaking": false} {"repo": "vite", "sha": "6c08c39ac4fb5868d080a51ff976a44693fc56ab", "date": "2026-07-16T14:15:23+09:00", "subject": "fix(optimizer): respect importer module format for dynamic import interop with CJS deps", "body": "", "style": "conventional", "type": "fix", "scope": "optimizer", "breaking": false} {"repo": "angular", "sha": "43ea889d3ce3503af2c5a4598c77163a70eeeb01", "date": "2026-07-16T08:36:13+03:00", "subject": "docs: document the inlineStyleLanguage workspace config option", "body": "The `inlineStyleLanguage` build option has existed since v12 but was\nnever listed in the workspace configuration reference. Add it to the\n\"Extra build and test options\" table alongside the other style options,\ndocumenting its accepted values (css, less, sass, scss) and default.\n\nFixes #69801\n\n(cherry picked from commit f032f5ac1ff9f9ce771dd700f6306ab3c2c1d37d)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "f032f5ac1ff9f9ce771dd700f6306ab3c2c1d37d", "date": "2026-07-16T08:36:13+03:00", "subject": "docs: document the inlineStyleLanguage workspace config option", "body": "The `inlineStyleLanguage` build option has existed since v12 but was\nnever listed in the workspace configuration reference. Add it to the\n\"Extra build and test options\" table alongside the other style options,\ndocumenting its accepted values (css, less, sass, scss) and default.\n\nFixes #69801", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5268e1902ec1ddb7efe5dae0e556b3fb18a8e2a8", "date": "2026-07-16T11:25:13+05:30", "subject": "Update tsget man page", "body": "Assisted-by: Claude:claude-sonnet-4-6\nSigned-off-by: Shreenidhi Shedi \n\nReviewed-by: Andrew Dinh \nReviewed-by: Matt Caswell \nMergeDate: Fri Aug 7 13:29:42 2026\n(Merged from https://github.com/openssl/openssl/pull/31445)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "a636f51975121fb39418471429e152695f1832cf", "date": "2026-07-16T14:19:26+08:00", "subject": "fix(Grid): support zero flex values", "body": "", "style": "conventional", "type": "fix", "scope": "Grid", "breaking": false} {"repo": "angular", "sha": "79c81b62b9af18427e265b5f76c549cf6d60e47d", "date": "2026-07-16T06:47:58Z", "subject": "build: update actions/setup-node action to v7", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "716593bd1162b8b5c9179757bf1e63748822203e", "date": "2026-07-16T09:39:20+02:00", "subject": "apps: cover the x509 -serial, name-hash and -fingerprint print options", "body": "Add a subtest exercising the informational print options of the x509 app\nthat operate on any certificate: -serial, -next_serial, -subject_hash\n(and its -hash alias), -issuer_hash, -subject_hash_old, -issuer_hash_old\nand -fingerprint (default SHA-1 and -sha256). These branches of the print\nloop were previously not exercised.\n\nThe expected values are those of the committed test/certs/ca-cert.pem;\nthe fingerprints are the SHA-1 and SHA-256 digests of its DER encoding.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Daniel Kubec \nReviewed-by: Matt Caswell \nMergeDate: Mon Jul 27 10:14:18 2026\n(Merged from https://github.com/openssl/openssl/pull/31967)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "fc1d726545ffae0f5b0e7344eb75457bf8d43657", "date": "2026-07-16T15:51:48+08:00", "subject": "feat(dev): support `hotUpdate` hook", "body": "", "style": "conventional", "type": "feat", "scope": "dev", "breaking": false} {"repo": "openssl", "sha": "47540a76e38317690bff49bf61ecd0242c453280", "date": "2026-07-16T10:02:21+02:00", "subject": "apps: drop invalid trust setting in check_cert_might_be_valid()", "body": "check_cert_might_be_valid() called X509_VERIFY_PARAM_set_trust() with\nX509_TRUST_OK_ANY_EKU, which is a trust_flags bit (1 << 4), not a valid\ntrust id (1..8). The call therefore failed with X509_R_INVALID_TRUST and\nleft that error on the error stack, which was then printed via\nERR_print_errors() for every certificate loaded by the x509 app, even on\nsuccess.\n\nThe intended liberal behavior is already provided by X509_PURPOSE_ANY and\nby suppressing X509_V_ERR_CERT_REJECTED, so just remove the bogus call.\n\nAdd a test checking that printing certificate info emits nothing on stderr.\n\nAssisted-by: Claude:claude-opus-4-8\nSigned-off-by: Norbert Pocs \n\nReviewed-by: Matt Caswell \nReviewed-by: Ryan Hooper \nMergeDate: Mon Jul 27 11:50:39 2026\n(Merged from https://github.com/openssl/openssl/pull/31968)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "1a2cafaa23342565f5d474b7e87ae03025127d9c", "date": "2026-07-16T16:14:38+08:00", "subject": "doc: support watch.nodeModulesRegexes", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ffa7eb92109a8f95faf83482086f9d4e33d86a6a", "date": "2026-07-16T10:20:04+02:00", "subject": "apps: cover the crl -crlnumber and -issuer print options", "body": "Add a subtest exercising the -crlnumber and -issuer print options of the\ncrl app: with a CRL that carries a CRL Number extension (printed as a hex\nvalue), with a CRL that has none (printed as ), and checking the\nprinted issuer DN in both cases.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Daniel Kubec \nReviewed-by: Matt Caswell \nMergeDate: Mon Jul 27 10:15:03 2026\n(Merged from https://github.com/openssl/openssl/pull/31969)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "2562246429658917aa2e6fceb0c5b54552068baf", "date": "2026-07-16T08:21:57Z", "subject": "docs: add v23 release and change to yearly release cycle", "body": "## Summary\n\nThis PR adds v22.x and v23 release dates. It also changes Angular's release cadence to a yearly cycle.\n\n## Why we are making this change\n\nThe community has long requested less frequent major releases due to the impact of breaking changes and upgrades for their projects as well as for enterprise customers. Additionally, a longer release cycle provides increased API stability for developers using agentic workflows, while still delivering a reasonable cadence of API upgrades and migrations.\n\n(cherry picked from commit 25dbe795079d4b57ba7f334013a399aff640f5e2)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "25dbe795079d4b57ba7f334013a399aff640f5e2", "date": "2026-07-16T08:21:57Z", "subject": "docs: add v23 release and change to yearly release cycle", "body": "## Summary\n\nThis PR adds v22.x and v23 release dates. It also changes Angular's release cadence to a yearly cycle.\n\n## Why we are making this change\n\nThe community has long requested less frequent major releases due to the impact of breaking changes and upgrades for their projects as well as for enterprise customers. Additionally, a longer release cycle provides increased API stability for developers using agentic workflows, while still delivering a reasonable cadence of API upgrades and migrations.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "babel", "sha": "d957b38f09fb9d3ca49da024885a5790f2368d31", "date": "2026-07-16T16:26:59+08:00", "subject": "Add types for `@babel/standalone`", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "809f50b440310d2746955846be1327513bba76f6", "date": "2026-07-16T16:31:29+08:00", "subject": "fix(Tree): restore rootStyle compatibility", "body": "* fix(Tree): restore deprecated rootStyle compatibility\n\n* fix(Tree): constrain semantic style types", "style": "conventional", "type": "fix", "scope": "Tree", "breaking": false} {"repo": "openssl", "sha": "94fcd62565f0ccce31193096fd1d7deab20e0fe7", "date": "2026-07-16T02:47:24-06:00", "subject": "Extract the documentation for reference id setters", "body": "I need to make further changes, and to do so, need the documentation\nof the current functions to be a bit better organized.\n\nThis extracts the documentation for the verify parameter reference id\nsetters to a dedicated page, and cleans it up to prepare for later\nchanges in the area. It should not have any funcitonal change, it\nonly documents them as they exist today.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMergeDate: Tue Aug 18 15:17:02 2026\n(Merged from https://github.com/openssl/openssl/pull/31972)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e8a742176a23b8c9b827bdb613d6e1c74130bebb", "date": "2026-07-16T10:53:53+02:00", "subject": "rand: fix jitter seed macro logic", "body": "The seeding macro logic was mixed up and incompletely applied. The\nmacro logic was also cleaned up.\n\nCo-authored-by: Paul Dale \n\nReviewed-by: Tim Hudson \nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nMergeDate: Fri Jul 17 04:47:07 2026\n(Merged from https://github.com/openssl/openssl/pull/31970)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "3cc659376ccb45d0c188aab4cef3fb6b185b570d", "date": "2026-07-16T17:16:57+08:00", "subject": "fix(Tag): prevent navigation when closing link tags", "body": "", "style": "conventional", "type": "fix", "scope": "Tag", "breaking": false} {"repo": "curl", "sha": "b41c28e70af47e1a49606db9c28f622f2a0052d0", "date": "2026-07-16T11:26:33+02:00", "subject": "lib: client writer tweaks", "body": "- keep pause state at request\n- keep presense of content decoding writers at request\n- only add cw-pause writer when content decoding happens\n- define macro for passing data down the writer chain\n\nscorecard: add option to test with h2o server\n\nCloses #22335", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "4854b2373c738a3e2fe5a46c2667d908f7c7e13a", "date": "2026-07-16T11:47:41+02:00", "subject": "Require ICU 55 or later", "body": "Support for older versions is removed. Since we no longer support\nRHEL 7, we don't need to support these old versions anymore. This\nallows a fair amount of code cleanup, including some code blocks that\nspecifically catered to old versions that probably received very\nlittle actual testing and use.\n\nReviewed-by: Jelte Fennema-Nio \nDiscussion: https://www.postgresql.org/message-id/flat/CA+hUKGL7trhWiJ4qxpksBztMMTWDyPnP1QN+Lq341V7QL775DA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b6d3ccf2f5f1a179e25e277bae3f432cc084ce58", "date": "2026-07-16T11:47:41+02:00", "subject": "Make PL/Tcl require Tcl 8.6 or later", "body": "Support for Tcl 8.5 and 8.4 is removed. Since we no longer support\nRHEL 7, we don't need to support these old versions anymore. This\nallows some small amount of code cleanup.\n\nReviewed-by: Jelte Fennema-Nio \nDiscussion: https://www.postgresql.org/message-id/flat/CA+hUKGL7trhWiJ4qxpksBztMMTWDyPnP1QN+Lq341V7QL775DA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "da9451830455a402fd6e29403a6512ebb2b41d97", "date": "2026-07-16T11:47:41+02:00", "subject": "Drop support for _MSC_VER less than 1933", "body": "Commit aa7c8685234 added a workaround for _MSC_VER less than 1933,\nwhich is Visual Studio 2022 version 17.3. Since we dropped support\nfor Visual Studio before 2022, and the versions up to 17.3 are long\nEOL, we can drop this extra code.\n\nReviewed-by: Jelte Fennema-Nio \nDiscussion: https://www.postgresql.org/message-id/flat/CA+hUKGL7trhWiJ4qxpksBztMMTWDyPnP1QN+Lq341V7QL775DA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "6319d6372c8f28cfb30bd2b7b0893f28e44491e6", "date": "2026-07-16T11:47:41+02:00", "subject": "Raise requirement to Visual Studio 2022", "body": "The use of _Generic from commit \"Replace __builtin_types_compatible_p\nwith _Generic\" causes compiler errors from VS 2019. Apparently, that\ncompiler is just broken for that (even though it appears to support\n_Generic in general, for example in commit 59c2f03d1ec).\n\nPer discussion, just drop support for VS 2019 and require at least VS\n2022. This just updates the documentation about that. In passing,\nsome information about VS 2026 is added.\n\nReviewed-by: Jelte Fennema-Nio \nDiscussion: https://www.postgresql.org/message-id/flat/CA+hUKGL7trhWiJ4qxpksBztMMTWDyPnP1QN+Lq341V7QL775DA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c27da9bde25a7d8295dd73ca48ced0abf475e629", "date": "2026-07-16T11:55:37+02:00", "subject": "Separate out the host flag documentation and tidy set_flags.pod", "body": "Move X509_VERIFY_PARAM_set_hostflags() and _get_hostflags() into a\ndedicated X509_VERIFY_PARAM_set_hostflags.pod, then reorganize and\ncorrect what remains on the set_flags page (flag sections, the\nmissing SUITEB flags, and the trust/time/policy/depth descriptions).\n\nDocumentation only; no behavioural change.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMergeDate: Tue Aug 18 15:17:09 2026\n(Merged from https://github.com/openssl/openssl/pull/31972)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ed9ea9710dfd148473535f554031c63ac6d90b77", "date": "2026-07-16T12:01:16+02:00", "subject": "OSSL_FN: Add lshift1 round-trip checks to fntest", "body": "file_lshift1() now mirrors bntest.c's file_lshift1() where OSSL_FN\nprimitives allow: it cross-checks LShift1 against A + A and A * 2,\nthen round-trips LShift1 >> 1 back to A with OSSL_FN_rshift1.\n\nThe LShift1 / 2, LShift1 % 2, and set-LSB rshift1-flooring checks\nfrom bntest.c are deferred to a future commit: they need\nOSSL_FN_div() and OSSL_FN_set_bit(), neither of which exists yet.\n\nIssue: https://github.com/openssl/project/issues/2015\nAssisted-by: Pi:z-ai/glm-5.2\nSigned-off-by: Richard Levitte \n\nReviewed-by: Igor Ustinov \nReviewed-by: Tomas Mraz \nMergeDate: Thu Jul 23 14:15:42 2026\n(Merged from https://github.com/openssl/openssl/pull/31830)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "fc2668e0b33f02a076bc7e356ad6c754f7b39e93", "date": "2026-07-16T12:47:12+02:00", "subject": "Clarify the X509_VERIFY_PARAM_set_trust() documentation", "body": "Spell out that the X509_TRUST_* selectors do not check the extKeyUsage\nextension they appear to name: they match only auxiliary trust\ninformation attached to the in-memory certificate from outside, so absent\nthat they have no effect. Mark the selector as a legacy mechanism and\npoint callers at X509_VERIFY_PARAM_set_purpose() instead.\n\nDocumentation only; no behavioural change.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMergeDate: Tue Aug 18 15:17:16 2026\n(Merged from https://github.com/openssl/openssl/pull/31972)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "824a07290f853a97219ee2624a46c0aada246efb", "date": "2026-07-16T10:52:32Z", "subject": "Do not redraw overlays unless actually asked to, GitHub issue 5398 from Noam Stolero.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "f531e6c8835ec8f6b7dc000ccf95ea7cc693198d", "date": "2026-07-16T20:35:12+09:00", "subject": "docs: clarify which hook is per-environment", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "7c091348af0ea3e0f6e05b5f387a972b418bfd9b", "date": "2026-07-16T19:36:13+08:00", "subject": "test(api/ui_spec): fix flaky test", "body": "The poke_eventloop() and screen:expect() sometimes don't run the test\nrunner's event loop long enough for the data to arrive at the reading\nend of the pipe. Use a screen:expect_unchanged() so that the event loop\ncan run a bit longer.\n\nFAILED ...ovim/build/Xtest_xdg_api/test/functional/api/ui_spec.lua @ 105: nvim_ui_send works with stdout_tty\nExpected values to be equal.\nExpected:\n\"\\27]11;?\\7Hello world\"\nActual:\n\"\\27]11;?\\7\"\nstack traceback:\n\t...ovim/build/Xtest_xdg_api/test/functional/api/ui_spec.lua:138: in function <...ovim/build/Xtest_xdg_api/test/functional/api/ui_spec.lua:105>", "style": "conventional", "type": "test", "scope": "api/ui_spec", "breaking": false} {"repo": "vite", "sha": "14fcb66c7513957beea28de9657d15936ad48728", "date": "2026-07-16T20:36:15+09:00", "subject": "perf(optimizer): check popular package manager lockfiles first", "body": "", "style": "conventional", "type": "perf", "scope": "optimizer", "breaking": false} {"repo": "vite", "sha": "302c755a8125b9a26214e3b413922b5513e41981", "date": "2026-07-16T19:42:14+08:00", "subject": "feat(css): export PostCSS config type for type-safe configs", "body": "", "style": "conventional", "type": "feat", "scope": "css", "breaking": false} {"repo": "vite", "sha": "6319827116c5be2a19c1b91c84ba3d38ad26a41c", "date": "2026-07-16T13:43:05+02:00", "subject": "feat(optimizer): support aube lockfile", "body": "", "style": "conventional", "type": "feat", "scope": "optimizer", "breaking": false} {"repo": "vite", "sha": "9e79b51579457a9af4fa623b68a0bfabbf38010b", "date": "2026-07-16T04:44:50-07:00", "subject": "feat(wasm): expand test suite, unwrap WebAssembly.Global and enable js-string builtins", "body": "", "style": "conventional", "type": "feat", "scope": "wasm", "breaking": false} {"repo": "vite", "sha": "65d3604f6fdbfcf6e86244d7fe3c1ca86acae701", "date": "2026-07-16T04:59:08-07:00", "subject": "feat(optimizer): support nub lockfile", "body": "", "style": "conventional", "type": "feat", "scope": "optimizer", "breaking": false} {"repo": "tmux", "sha": "381eaaa3a384f87c68673d20746b1c25a3dfdf97", "date": "2026-07-16T13:04:29+01:00", "subject": "No more popups and overlays.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "c7a6a735bc42a491efc8cb5c399d58b45c86aa18", "date": "2026-07-16T13:10:51+01:00", "subject": "Another popup leftover.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "63afe639141faf49a62611808651b2b80030c64b", "date": "2026-07-16T14:15:56+02:00", "subject": "quic: report SRTM entry match status via an output argument", "body": "Coverity (CID 1696969, CHECKED_RETURN) flagged the ignored return value of\nossl_quic_srtm_remove() in ch_enqueue_retire_conn_id(). The return value\ncould not be usefully checked because 0 was overloaded to mean both \"no\nmatching entry\" and a genuine internal failure (alloc_failed or an\nlhash consistency error).\n\nGive the return value a single meaning (1 on success, 0 on internal\nerror) and report whether a matching entry was found through a new\nuint8_t *match output argument, which may be NULL if the caller does not\nneed it.\n\nResolved: https://scan5.scan.coverity.com/#/project-view/65138/10222?selectedIssue=1696969\nSigned-off-by: Nikola Pajkovsky \n\nReviewed-by: Bob Beck \nReviewed-by: Paul Dale \nReviewed-by: Neil Horman \nMergeDate: Tue Jul 21 09:24:45 2026\n(Merged from https://github.com/openssl/openssl/pull/31974)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "fba51526af88c338abae31f97182951c4c9f16aa", "date": "2026-07-16T13:16:29+01:00", "subject": "Modal pane regression test.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "5bb7d7a4f27ca16eb12a02cdd43f40493a51272b", "date": "2026-07-16T13:29:44+01:00", "subject": "http: stop dropping large custom headers", "body": "Closes #22336", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "1d0d22df223d6801d827b2ec5027288eed2d0e23", "date": "2026-07-16T12:36:58Z", "subject": "Fix uninitialized var for sscanf %n.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "d5e8b1ef7a02c84d4fd70a6b4d748ead9ff815bf", "date": "2026-07-16T08:19:17-05:00", "subject": "feat(forms): allow permanent hidden fields in signal forms", "body": "Allow the hidden utility function to be called without a configuration object to make fields permanently hidden.", "style": "conventional", "type": "feat", "scope": "forms", "breaking": false} {"repo": "openssl", "sha": "0ffaa24c6148514fa77f76a48ae48852df7be9e7", "date": "2026-07-16T15:30:27+02:00", "subject": "Rework v3nametest for SAN-first matching, add flag and long-name coverage", "body": "Drive the match matrix primarily through SAN dNSName/rfc822Name rather\nthan subject CN/emailAddress (subject setters cut to one each, decoy\nmulti-SAN variants added). Add explicit coverage of ALWAYS/NEVER_CHECK_\nSUBJECT and the non-matching-SAN-plus-matching-subject case, and a\nlong-name test for the 64/68-byte dNSName boundary and truncation guards.\n\nTest only; no behavioural change.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMergeDate: Thu Jul 30 09:11:28 2026\n(Merged from https://github.com/openssl/openssl/pull/31981)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "30c0442eb308cdb9a0d1b8551758775ec4d0e75d", "date": "2026-07-16T16:05:36+02:00", "subject": "doc: Fix link text for data checksums", "body": "Commit 67846550dc6d removed the xreflabels for initdb options, which\nturned the sentence \"The second field contains the page checksum if\ndata checksums are enabled\" into \"The second field contains the page\nchecksum if -k are enabled\", as well \"Only has effect if data checksums\nare enabled\" into \"Only has effect if -k are enabled\".\n\nFix by setting an explicit link text, and while there also change the\nlink to point to the data checksum page which has more information\nthan just the initdb option.\n\nThe original report was for one instance, further inspection turned\nup quite a few more cases. Also redirect the link in the amcheck\ndocs which albeit was reading right, but will be more helpful if\nlinking to the main page on data checksums. Backpatch to v18 where\nthe xreflabels were removed.\n\nAuthor: Daniel Gustafsson \nReported-by: y.saburov@gmail.com\nReviewed-by: Laurenz Albe \nDiscussion: https://postgr.es/m/178350739237.73862.4549076173872335741@wrigleys.postgresql.org\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "00565bb35cbaddad3fdd9c2a00d6ae360b7161f5", "date": "2026-07-16T16:13:38+02:00", "subject": "Stop consulting the subject DN by default in X509 name checks", "body": "do_x509_check() now matches the subject commonName (host checks) and\nemailAddress (email checks) only when X509_CHECK_FLAG_ALWAYS_CHECK_SUBJECT\nis set, rather than whenever no matching subjectAltName is present.\n\nUpdate v3nametest to request ALWAYS_CHECK_SUBJECT for its subject-DN cases.\n\nFixes: openssl/project#1897\n\nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMergeDate: Wed Aug 26 15:21:48 2026\n(Merged from https://github.com/openssl/openssl/pull/31982)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a297985d05dafe92260d6733ba30f737b1997099", "date": "2026-07-16T17:23:27+03:00", "subject": "fix(messages): show hint in E77 error #40765", "body": "", "style": "conventional", "type": "fix", "scope": "messages", "breaking": false} {"repo": "angular", "sha": "0650966fb164d2b17d23106351e3f1c157ba90c3", "date": "2026-07-16T11:24:16-04:00", "subject": "docs: modernize directives guides", "body": "Co-authored-by: Matthieu Riegler \n\nPR Close #69822", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "f0f84d1251f84ee1d421c2faf419a64a76d28f6c", "date": "2026-07-16T16:30:19+01:00", "subject": "http: avoid length underflow in Curl_compareheader", "body": "Closes #22338", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a00e43ac32d4f7367567791a68b2cf815f8c4d3f", "date": "2026-07-17T00:49:26+09:00", "subject": "postgres_fdw: stabilize terminated-connection regression tests", "body": "The regression test for postgres_fdw_get_connections(true) assumed that\na terminated remote connection would still remain visible in the FDW\nconnection cache long enough to be reported as closed with a nonzero\nremote_backend_pid.\n\nThat assumption is not always valid. postgres_fdw_get_connections()\nreports only entries that are still present in ConnectionHash, while\npgfdw_inval_callback() may immediately discard an idle cached connection\n(xact_depth == 0) when a relevant invalidation arrives. In CI, that can\nhappen between terminating the remote backend and querying\npostgres_fdw_get_connections(true), causing the function to return no\nrows.\n\nAdjust the idle-connection test to accept either outcome: if the cache\nentry is still present, verify that it reports the expected server name,\nclosed status, and nonzero remote backend PID; otherwise treat zero rows\nas a legitimate result.\n\nTo preserve coverage of the terminated-backend reporting path, add a\nseparate check inside an explicit transaction. In that case, concurrent\ninvalidation may mark the connection invalid but cannot discard it\nbefore transaction end, so postgres_fdw_get_connections(true) should\nstill report the terminated connection as in-use, closed, and associated\nwith a nonzero remote backend PID.\n\nBackpatch to v18, where the affected postgres_fdw_get_connections(true)\ntest was introduced.\n\nReported-by: Robert Haas \nAuthor: Fujii Masao \nReviewed-by: Robert Haas \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CA+Tgmoax3cHXHsm9OidN4F-xiu16y8q2W8T5dTNFic1Zoo2cOw@mail.gmail.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "24a2b541bd450a2db9d8f3ddefc92ace58ef9a42", "date": "2026-07-17T00:50:54+09:00", "subject": "Handle concurrent sequence drops during synchronization", "body": "Commit d4a657b0a4d added a call to has_sequence_privilege() while\nfetching sequence information from the publisher, so that\npublisher-side permission failures could be distinguished from missing\nsequences. It also assumed that has_sequence_privilege() could never\nreturn NULL, and asserted accordingly.\n\nHowever, that assumption was incorrect. If a sequence is dropped after\nthe synchronization worker collects its metadata but while fetching the\nsequence information, has_sequence_privilege() can return NULL.\nThis can trigger the assertion failure. This was also reported in\na buildfarm failure on member culicidae.\n\nFix this by treating a NULL result from has_sequence_privilege() as\nindicating that the sequence was dropped concurrently, and report it as\na missing sequence instead of asserting that the result is never NULL.\n\nReported-by: Noah Misch \nAuthor: Vignesh C \nReviewed-by: Hayato Kuroda \nReviewed-by: Zsolt Parragi \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/20260710045217.f0.noahmisch@microsoft.com\nDiscussion: https://postgr.es/m/CALDaNm2fHGLeiQKj0r6OG7N9QeayxSmpLrWYJRyt4dL_m3VRWw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "637aa273e6335c2e4d196f510d6787210403ed24", "date": "2026-07-16T11:56:01-04:00", "subject": "In transformIndexStmt, transform index expressions in INCLUDING.", "body": "Process these just like the adjacent loop for expressions in regular\nindex columns. This logic was originally omitted because there was\nno intention of supporting index expressions in INCLUDING.\nThere's still no near-term intention of that, but without this change\nthe ChooseIndexExpressionName code added by 181b6185c spits up on\nexpressions in INCLUDING: that expects to handle parse-transformed\nexpressions, and it runs before we reach the place that is currently\nsupposed to throw the \"not supported\" error. We could fix this\nproblem in other ways, but this way avoids contorting the logic,\nand it results in less code to be revised not more if we ever get\naround to supporting INCLUDING expressions.\n\nReported-by: Maaz Syed Adeeb \nAuthor: Tom Lane \nTest-authored-by: Maaz Syed Adeeb \nDiscussion: https://postgr.es/m/CAG+FJqOxYj=sVyHcys64h9DbvzP6EUPGHJ7oKj-PW=Qp5Ebk_g@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "2ba2fe354026aa4d60ab06fbc9b6ac1e56d1b92f", "date": "2026-07-16T22:03:35+05:30", "subject": "ldap: base64-encode LDIF values beginning with colon or less-than", "body": "A value whose first byte is ':' or '<' is not a SAFE-INIT-CHAR per RFC\n2849 and must be base64-encoded, but ldap_value_needs_base64() and the\ninline check in oldap_recv() only encoded on a control byte or\nleading/trailing blank. Such a value was emitted verbatim, producing\nLDIF that strict parsers reject.\n\nCloses #22339", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "f7459dca248ca001777d99a774164d14112d9071", "date": "2026-07-17T00:53:15+08:00", "subject": "refactor(demo): migrate styles from markdown to CSS-in-JS", "body": "* refactor(demo): migrate styles from markdown to CSS-in-JS\n\n* update snap\n\n* update snap\n\n* update snap\n\n* update snap\n\n* update snap\n\n* update", "style": "conventional", "type": "refactor", "scope": "demo", "breaking": false} {"repo": "neovim", "sha": "c15ac3d64cb41fb9da03d520a4d2bdbf8ba01c52", "date": "2026-07-16T12:08:47-05:00", "subject": "fix(lsp): capability cleanup #40763", "body": "", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "axios", "sha": "a092bae50d1884782151b2fcea12974d6da6e376", "date": "2026-07-16T23:46:54+05:30", "subject": "fix(core): synchronous interceptors swallow errors and proceed with request", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "neovim", "sha": "9cabba671d99ca63d6d28e8aa6a9f270b51038b4", "date": "2026-07-16T21:17:54+03:00", "subject": "feat(wasm-emcc): add browser support for Neovim", "body": "", "style": "conventional", "type": "feat", "scope": "wasm-emcc", "breaking": false} {"repo": "neovim", "sha": "9edf57e665820c5dff20425236fb309c4fb48ff6", "date": "2026-07-16T13:26:01-05:00", "subject": "fix(dir): custom 3P directory-browser #40725", "body": "Problem:\nThird-party directory browsers do not have a documented way to take over local\ndirectory buffers from nvim.dir.\n\nSolution:\nDocument deleting the nvim.dir augroup and handling FileType directory as the\nafter-startup handoff, and avoid installing the built-in buffer-local \"-\" mapping\nwhen \"-\" is already mapped.", "style": "conventional", "type": "fix", "scope": "dir", "breaking": false} {"repo": "postgres", "sha": "2a933deaa23bff4245c327bc8e9f4676095d0f6c", "date": "2026-07-16T11:50:16-07:00", "subject": "Reject infinite and out-of-range interval shifts in uuidv7().", "body": "uuidv7(interval) shifts the current time by the given interval before\nencoding it into the 48-bit Unix-millisecond timestamp field of the\ngenerated UUID. Two cases were mishandled:\n\nAn infinite interval ('infinity' or '-infinity') produced an infinite\ntimestamp, which overflowed during the conversion to Unix-epoch\nmicroseconds and yielded a garbage UUID. Reject infinite intervals up\nfront, before any timestamp arithmetic.\n\nA shift that moved the timestamp outside the range representable by\nthe 48-bit field was silently accepted. Timestamps before the Unix\nepoch wrapped when cast to unsigned, and timestamps beyond\napproximately year 10889 overflowed the field; both produced UUIDs\nwith bogus timestamps that break sort ordering. Reject any shifted\ntimestamp outside the supported range.\n\nAlso document that infinite intervals and out-of-range shifts are\nrejected.\n\nAlthough raising a new error changes behavior in a stable branch, this\nis back-patched to 18 (where uuidv7(interval) was introduced) because\nthe previous behavior can silently corrupt data. Failing loudly is far\nsafer than silently accepting the wraparound; otherwise users may not\ndiscover that their UUIDv7 values are no longer sortable until years\nlater, when recovery is painful. It also matches how PostgreSQL\nalready handles timestamp + interval overflow, which raises an\nerror. The change only affects applications passing an interval large\nenough to push the result outside the representable range.\n\nBackpatch to 18, where uuidv7(interval) was introduced.\n\nReported-by: Christophe Pettus \nAuthor: Baji Shaik \nReviewed-by: Masahiko Sawada \nReviewed-by: Zsolt Parragi \nReviewed-by: Tristan Partin \nReviewed-by: Kyotaro Horiguchi \nDiscussion: https://www.postgresql.org/message-id/799A70FA-6E5C-4118-99EB-2FBBE1CBAC54@thebuild.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "baf0ec4652c2c42c76cf108bb95860b61c1eff16", "date": "2026-07-16T12:13:16-07:00", "subject": "Correct logical decoding status at end of recovery with minimal WAL level.", "body": "Crash recovery running with wal_level='minimal' can replay an\nXLOG_LOGICAL_DECODING_STATUS_CHANGE record that activates logical\ndecoding, if the server previously ran with a higher wal_level and\ncrashed after the last logical slot was dropped but before the\ncheckpointer deactivated logical decoding. Replaying such a record is\ncorrect since it reflects the status at the time it was\nwritten. However, UpdateLogicalDecodingStatusEndOfRecovery() asserted\nthat logical decoding is never active with wal_level='minimal',\ncausing an assertion failure at the end of recovery. In production\nbuilds, logical decoding would remain active while running with\nwal_level='minimal'.\n\nInstead of special-casing wal_level='minimal', recompute the status at\nthe end of recovery as usual: no logical slot can exist with\nwal_level='minimal' as RestoreSlotFromDisk() would have rejected it,\nso the recomputation always deactivates logical decoding in this case,\nalso writing the corresponding status change record.\n\nOversight in 67c20979ce7.\n\nReviewed-by: Guoqing Yang \nReviewed-by: Hayato Kuroda \nDiscussion: https://postgr.es/m/CAD21AoAnPAugUnDic+ESvrfXjXHk2bss9eHAD7zP0-Chy2UabA@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "8e93e0c72b8aa225b0a66d4afdfb69e38d8b2424", "date": "2026-07-16T12:20:45-07:00", "subject": "pg_controldata: Show logical decoding status.", "body": "The logical decoding status is stored in checkpoint records and used\nto restore the status at server startup, but pg_controldata did not\nshow it. This information is useful for diagnosing issues around the\ndynamic activation and deactivation of logical decoding.\n\nOversight in 67c20979ce7.\n\nReviewed-by: Guoqing Yang \nReviewed-by: Hayato Kuroda \nDiscussion: https://postgr.es/m/CAD21AoAnPAugUnDic+ESvrfXjXHk2bss9eHAD7zP0-Chy2UabA@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "81926f04b449ed875fac7423a5bb93538a4232f0", "date": "2026-07-16T21:34:41+02:00", "subject": "refactor(search): SearchState", "body": "", "style": "conventional", "type": "refactor", "scope": "search", "breaking": false} {"repo": "eslint", "sha": "d584e31098f262563b97dc5d5f23a63a187879a0", "date": "2026-07-16T21:54:42+02:00", "subject": "chore: fix failing ecosystem test for `eslint-plugin-unicorn`", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "neovim", "sha": "4d90288c694e0956214c1c54f3c1b2e19eee507a", "date": "2026-07-16T17:25:28-04:00", "subject": "test: unreliable \"bumps 'busy' on each watched buffer\" #40775", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "angular", "sha": "fec5977df4dda3a10d5ce2923e3e06d86ba11ee7", "date": "2026-07-16T16:53:40-05:00", "subject": "fix(http): match header values exactly when deleting", "body": "Normalize value-specific HttpHeaders deletions before filtering. The string overload previously used String#indexOf and removed shorter values contained within the requested deletion value, potentially widening outgoing request metadata.\n\nPreserve delete-all behavior only when no value is supplied, and cover string, array, and empty-string deletion.", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "angular", "sha": "39e362eea5603fa36bda436c9f03d9304960c1ea", "date": "2026-07-16T16:53:40-05:00", "subject": "fix(http): match header values exactly when deleting", "body": "Normalize value-specific HttpHeaders deletions before filtering. The string overload previously used String#indexOf and removed shorter values contained within the requested deletion value, potentially widening outgoing request metadata.\n\nPreserve delete-all behavior only when no value is supplied, and cover string, array, and empty-string deletion.\n\n(cherry picked from commit f33ee95045ac3ab1e96baa41f44dd130ef3724ab)", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "postgres", "sha": "cea74f48d36bc1073da0adab755946ad94aa6902", "date": "2026-07-16T18:08:43-04:00", "subject": "Use fake LSNs consistently in hash index AM.", "body": "Defensively make sure that all hash index atomic actions use a fake LSN\nwith an unlogged relation.\n\nOversight in commit e5836f7b, which added fake LSN support to the hash\nindex AM, but missed log_split_page.\n\nAuthor: Peter Geoghegan \nReviewed-by: Ewan Young \nDiscussion: https://postgr.es/m/CAH2-WzkC-opX8iS6X=a470DDC31er_x5rzPw=HjRxha9N8brZw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7af998936d304b676d4e8fb2f2780560cf8443ca", "date": "2026-07-17T07:40:10+09:00", "subject": "Remove SQL function getpgusername()", "body": "Since 457ac0331cd3, current_user() is the recommended way to get access\nto this information, and getpgusername() was marked as deprecated since,\nwithout being documented.\n\nBump catalog version.\n\nReported-by: Thom Brown \nDiscussion: https://postgr.es/m/CAA-aLv44sPoKhHZqBvhqRdrFUEKwatGEmXBtqC360UuD5Lt7Nw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "323530e00d279f267265ef32c7e6b40a04c94105", "date": "2026-07-16T18:55:36-04:00", "subject": "Fix wrong variable offset sanity check.", "body": "Commit c7aeb775 rewrote the HOT-chain offset sanity checks in three\nplaces, but in heap_get_root_tuples it accidentally tested offnum -- the\nouter loop variable, which is already bounded by the loop condition --\ninstead of nextoffnum, the offset actually passed to PageGetItemId. The\npre-c7aeb775 check tested nextoffnum.\n\nWith the check ineffective, a stale t_ctid could make PageGetItemId read\npast the end of the line pointer array (which is data corruption that we\nexpect to be able to catch here).\n\nAuthor: Peter Geoghegan \nReported-by: Konstantin Knizhnik \nDiscussion: https://postgr.es/m/87c7d8a4-3a82-4334-bee6-e8c2ad3f3293@garret.ru\nBackpatch-through: 15", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d36d05cbd59a3accc4bc41f561b81187a3071d93", "date": "2026-07-16T20:38:44-04:00", "subject": "refactor(globals): group Visual-mode globals in VisualState #40777", "body": "", "style": "conventional", "type": "refactor", "scope": "globals", "breaking": false} {"repo": "postgres", "sha": "03480907e9ff5d9bb3296b56c9b49db3df756e0f", "date": "2026-07-17T09:47:33+05:30", "subject": "Doc: Clarify DROP SUBSCRIPTION behavior after SET (slot_name = NONE).", "body": "The previous text claimed that once the slot is disassociated with\nALTER SUBSCRIPTION ... SET (slot_name = NONE), DROP SUBSCRIPTION \"will no\nlonger attempt any actions on a remote host\". That is inaccurate:\nDROP SUBSCRIPTION may still connect to the publisher to drop\ninternally-created table synchronization slots when some table\nsynchronization is left unfinished. Reword to describe this, and note\nthat if the publisher is unreachable those slots (and the main slot, if\nit still exists) must be dropped manually to avoid indefinitely reserving\nWAL.\n\nReported-by: Jeff Davis \nAuthor: Amit Kapila \nBackpatch-through: 14\nDiscussion: https://postgr.es/m/CAA4eK1+tyYSpPxMBy1974kjivuGeR7YY=yopwRGrK3+vCTysdg@mail.gmail.com\nDiscussion: https://postgr.es/m/D908370F-2695-4231-851D-17179A6A6F2A@gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "b2c608c014a396800a24a89343265d0616bee2d8", "date": "2026-07-17T07:07:25+02:00", "subject": "fix: NewExpression with parenthesized callee in `preserve-caught-error`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vite", "sha": "05302b07267f6b4f9dbeac5b1d73fcc3dc06d730", "date": "2026-07-17T15:10:19+09:00", "subject": "feat(config): warn features incompatible with native loader in bundle loader", "body": "", "style": "conventional", "type": "feat", "scope": "config", "breaking": false} {"repo": "angular", "sha": "3854174ec4a25e21abaef4ddd9d968a5b46fc832", "date": "2026-07-17T09:17:34+03:00", "subject": "refactor(compiler): enforce exhaustive defer trigger handling", "body": "Store the trigger kind before each switch and assign the value to `never` in the fallback branch.\n\nThis removes the `any` casts and makes the switches exhaustive. Adding a new `DeferTriggerKind` without handling it in either phase now produces a TypeScript compilation error.\n\nRuntime behavior and error messages remain unchanged.\n\n(cherry picked from commit 8201cebc49c2d43307731191479749fb4dc3e854)", "style": "conventional", "type": "refactor", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "8201cebc49c2d43307731191479749fb4dc3e854", "date": "2026-07-17T09:17:34+03:00", "subject": "refactor(compiler): enforce exhaustive defer trigger handling", "body": "Store the trigger kind before each switch and assign the value to `never` in the fallback branch.\n\nThis removes the `any` casts and makes the switches exhaustive. Adding a new `DeferTriggerKind` without handling it in either phase now produces a TypeScript compilation error.\n\nRuntime behavior and error messages remain unchanged.", "style": "conventional", "type": "refactor", "scope": "compiler", "breaking": false} {"repo": "openssl", "sha": "ab42643c66dc64933bd93c12506a69e10f8b850f", "date": "2026-07-17T02:37:59-04:00", "subject": "[test] bntest: add ModInv file tests", "body": "Add file_modinv() to bntest.c, driven by a new bnmodinv.txt corpus and\nwired into 10-test_bn.t.\n- Positive tests assert equalBN and a mult sanity check\n- Negative tests assert NULL return plus the BN_R_NO_INVERSE error\nCorpus covers:\n- in-range A\n- out-of-range A\n- positive and negative A for both of the above\n- degenerate moduli M = 0 and M = 1\n- RFC 4754 sourced KATs\n- RFC 7091 sourced KATs\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Tim Hudson \nReviewed-by: Richard Levitte \nMergeDate: Thu Jul 23 14:20:08 2026\n(Merged from https://github.com/openssl/openssl/pull/31985)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "78accc42a5b8887d9df624f7d4a934d3ead677d1", "date": "2026-07-17T14:01:23+07:00", "subject": "feat(dev): label network URLs with their interface name", "body": "", "style": "conventional", "type": "feat", "scope": "dev", "breaking": false} {"repo": "vitest", "sha": "51acd5778fa1e47b41ed7e4c0e9675880274626d", "date": "2026-07-17T00:08:04-07:00", "subject": "fix(browser): inlined awaited import should not trigger syntax error after transform", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "vitest", "sha": "22d353a80c230dbc7c69a22b390070f7aa9afabe", "date": "2026-07-17T09:08:39+02:00", "subject": "fix: prevent node builtins double prefix (fix #10630)", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "babel", "sha": "79b62de6147f349945c4f7d05cfeb776eaa34773", "date": "2026-07-17T09:19:04+02:00", "subject": "Up source map packages", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "b5666dc0324ffd9b3bc8ac9f86e96833b687b931", "date": "2026-07-17T15:44:09+08:00", "subject": "refactor(Typography): simplify ellipsis Tooltip control", "body": "* fix(Typography): restore ellipsis Tooltip popup interaction\n\n* fix(Typography): delay Tooltip restore after actions hover", "style": "conventional", "type": "refactor", "scope": "Typography", "breaking": false} {"repo": "vite", "sha": "9beae37d7221b25463a011feb40b0303ca328d87", "date": "2026-07-17T16:50:31+09:00", "subject": "feat: add `input` option", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "babel", "sha": "6ce28e10297ed8cd3445a04199d4cf2e1c8cb135", "date": "2026-07-17T10:06:37+02:00", "subject": "Add whitespace in source-map-visual tests", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "d5b1f3ad890259f6b91309f72dafcdeb4612bbcb", "date": "2026-07-17T08:13:23Z", "subject": "Clear cfg_client if the client is destroyed to prevent UAF found by Ilya Grigoriev.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "43092f8d541908009e79d1b51ebfe9cbee225f11", "date": "2026-07-17T09:13:33+01:00", "subject": "Test for client lost during config file.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "5cd43626ef66b9854625d860f3b68c0c47e362d3", "date": "2026-07-17T09:29:40+01:00", "subject": "cmd_template_replace test.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "2711780dffe2a5348f8d0e02095acd34a532a5bb", "date": "2026-07-17T08:37:29Z", "subject": "Move per-client window sizes into control.c since the only user is for control mode.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "b46684427ffbc1c20686731ef01beeaec0ba53f1", "date": "2026-07-17T09:37:32+01:00", "subject": "Extend control client size test.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "f7aee46a5b05245ce74dd57a2efdae0050aeefe8", "date": "2026-07-17T17:44:40+08:00", "subject": "docs(Image): add wheel prop for mouse wheel zoom control", "body": "- Add wheel prop to PreviewType for controlling mouse wheel zoom\n- Add wheel prop to PreviewGroupType for preview group\n- Set default value to true to keep backward compatibility", "style": "conventional", "type": "docs", "scope": "Image", "breaking": false} {"repo": "neovim", "sha": "58fe621cfcd94a3572c5934b100d801ba648bb8a", "date": "2026-07-17T06:09:04-05:00", "subject": "feat(health): rearrange vim.ui.img checks #40782", "body": "", "style": "conventional", "type": "feat", "scope": "health", "breaking": false} {"repo": "vite", "sha": "55bba7bbd9de40d031360e4408fe91bff5b29ec9", "date": "2026-07-17T20:12:16+09:00", "subject": "fix: resolve root to real path", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "neovim", "sha": "0d92bf376170dd6c311f1e7dfefb323e336ff8eb", "date": "2026-07-17T07:16:31-04:00", "subject": "build(vim-patch): '#ifdef FEAT_' tweaking is N/A #40785", "body": "Filter out `#ifdef` line changes when analyzing C files (*.c, *.h)\nvia git \"-I\" flag.\n\n```\n-I\n--ignore-matching-lines=\nIgnore changes whose all lines match . This option may be specified more than once.\n```\n\n- https://git-scm.com/docs/diff-options#Documentation/diff-options.txt--Iregex", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "postgres", "sha": "3aa54433b0cdce48facb610a5b720208cc760654", "date": "2026-07-17T20:16:34+09:00", "subject": "Restrict pg_stat_io entries for data checksum processes", "body": "The data checksums launcher and workers were exposed in pg_stat_io\nwith the same broad set of object/context combinations as general\nbackground workers. However, several of those entries can never\naccumulate I/O statistics for these processes, such as bulkwrite,\nrelation init, temporary relation, and launcher vacuum entries.\n\nTeach pgstat_tracks_io_object() and pgstat_tracks_io_op() about the\nactual I/O performed by the data checksum processes. Keep the entries\nneeded for catalog scans, including bulkread catalog scans, worker\nrelation processing with a vacuum access strategy, and WAL writes and\ninitialization, while excluding WAL reads and other object/context\ncombinations that can never be used.\n\nAuthor: Fujii Masao \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CAHGQGwHz_-nt+YkHDMRZNBZrnoHro8cMOgSwuXEmSYT6vxgQ=w@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c5fd5eb55a58f0f370c69d56fd0dabfe76035474", "date": "2026-07-17T14:10:03+02:00", "subject": "uint-spbset: reused empty chunks", "body": "Fix chunk allocations by reusing existing empty chunks when a new offset\nis needed. Before this fix, spbsets would only ever grow with added\nnumbers outside the range of existing chunks.\n\nCloses #22340", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "1ce000062af83bb1cb0aaef9ce6c5e80e67c3d58", "date": "2026-07-17T12:42:51Z", "subject": "Remove the active-pane flag for now, there are some gaps in how this works and I don't like it. May come back in a different form (maybe just for windows).", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "e16ff3a1199293ac9cdfa6132c08fdea162215f3", "date": "2026-07-17T18:38:27+05:30", "subject": "fix(build): map CSS chunks in chunk import maps (fix #22946)", "body": "", "style": "conventional", "type": "fix", "scope": "build", "breaking": false} {"repo": "postgres", "sha": "1f414035135cca34b606d4c0d3db0eb57045e00b", "date": "2026-07-17T15:40:16+02:00", "subject": "Fix truncation rules for base64 encoding", "body": "Commit e1d917182 added support for base64url encoding, a base64 variant\nintended to be safe for usage in URLs and filenames. The padding rules\nfor base64url and base64 differ in that base64url require no extra '='\npadding, but the commit unintentionally relaxed this requirement for\nbase64 as well. Fix by making sure that the truncation logic check\nfor the encoding and add a test to make sure.\n\nBackpatch down to v19 where support for base64url was introduced.\n\nAuthor: Daniel Gustafsson \nReviewed-by: David E. Wheeler \nDiscussion: https://postgr.es/m/3258FC72-F5E1-40B9-B5D7-64478CAF7728@yesql.se\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2521f9e75e910cec13591c1af76a69ea7f96109e", "date": "2026-07-17T15:45:30+02:00", "subject": "doc: Fix SQL quoting in example", "body": "The identifier \"order\" needs to be quoted, just like in the example a\nlittle bit earlier.\n\nAuthor: Thom Brown \nDiscussion: https://www.postgresql.org/message-id/CAA-aLv4xyAaxm8vq5LEhznh-SGphs4wUAGC6Vpas%2B1hHZzzD7A%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "82b214d9bc273d75efaffade17472ffab2520f35", "date": "2026-07-17T16:30:40+02:00", "subject": "OSSL_FN: Make RAND top selector count-based", "body": "Renumber OSSL_FN_RAND_TOP_ANY / _ONE / _TWO from -1 / 0 / 1 to\n0 / 1 / 2, so each value is the number of high bits the result is\nforced to set. This matches the already-count-based\nOSSL_FN_RAND_BOTTOM_* (0 = unconstrained, 1 = force the low bit),\nmaking the whole selector family read the same way.\n\nThis is a deliberate departure from the BN counterpart: BN_RAND_TOP_*\nstays -1 / 0 / 1, but OSSL_FN_RAND_TOP_* has no legacy tier and no\nneed for a sentinel, so the values are the counts themselves.\nCallers that use the symbolic macros are unaffected; only raw-literal\ncallers would notice, and there are none in tree.\n\nAssisted-by: Pi:z-ai/glm-5.2\nSigned-off-by: Richard Levitte \n\nReviewed-by: Igor Ustinov \nReviewed-by: Tomas Mraz \nMergeDate: Thu Jul 23 14:17:18 2026\n(Merged from https://github.com/openssl/openssl/pull/31959)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "81f752bfc8580e21218e840e06f9bdf88b35edf5", "date": "2026-07-17T10:33:54-04:00", "subject": "fix(register): do not inherit parent execArgv", "body": "* fix(register): do not inherit parent execArgv\n\nPer https://nodejs.org/api/worker_threads.html#launching-worker-threads-from-preload-scripts, new Worker threads will inherit command line flags, including `-e` and `-r`.\n\nWhen they are combined with the `--input-type`, Node.js somehow hangs and the worker does not start.\n\n* feat: resolve token in process tests\n\n* add tests\n\n* fix: ensure args are always string\n\n* replace rootDir by file:// url\n\n* test: use for --import tests", "style": "conventional", "type": "fix", "scope": "register", "breaking": false} {"repo": "babel", "sha": "b13121b462805508ab30ff425a8d1c67b093ef96", "date": "2026-07-17T16:35:36+02:00", "subject": "Remove cycles support from tsconfig generator", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "38808509d597ef1f92a32846a7042ce9a3da586c", "date": "2026-07-17T15:48:41+01:00", "subject": "Add support for local windows for a client.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "225e69341e18b707055347b474b24fe1b5dbc80d", "date": "2026-07-17T16:55:44+02:00", "subject": "build(reactivity): restore published esm-browser bundles", "body": "The dev + prod esm-browser builds are useful for direct browser/CDN\nusage of @vue/reactivity, so publish them again instead of building\nthe prod bundle on demand for benchmarks only.", "style": "conventional", "type": "build", "scope": "reactivity", "breaking": false} {"repo": "vue-core", "sha": "05b9255c9cdbd1317414a26ee0a3afdfa384df05", "date": "2026-07-17T16:57:31+02:00", "subject": "build: drop global/iife bundles for all packages except vue", "body": "- compiler-dom, runtime-dom: esm-only published output; remove the\n unpkg/jsdelivr fields since there is no browser-loadable bundle left\n- reactivity: point unpkg/jsdelivr at the esm-browser build instead\n- size-global: build each target separately so runtime-dom and\n compiler-dom global prod sizes stay tracked via the on-demand\n single-target format override in build.js (bundles are no longer\n published but remain the size-regression signal per package)\n- buildOptions.name kept: scripts/dev.js still uses it for local\n global-format dev builds", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "b26c1abe0773a5bd64997614cea26262d864a6cb", "date": "2026-07-17T18:11:44+03:00", "subject": "feat(devtools): show Angie on tab empty states", "body": "The Transfer State and signal-graph tabs render bare-text placeholders\nwhen there's nothing to show. Add the Angie mascot to both, matching the\npattern already shipped on the top-level DevTools screens (#69786):\n\n- Transfer State \"isn't using Transfer State\" card: replace the\n swap_horiz icon with the teaching pose.\n- Signal graph \"No signals in this component\": add the orthos-back pose\n above the message.", "style": "conventional", "type": "feat", "scope": "devtools", "breaking": false} {"repo": "vue-core", "sha": "33fe534fd916f68ec987a4e2784caf1981c29be1", "date": "2026-07-17T17:13:32+02:00", "subject": "refactor: simplify pkg exports + types for better compat", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "postgres", "sha": "42413e702913135efafa9180d0f101c455101ac5", "date": "2026-07-17T11:23:18-04:00", "subject": "meson: Fix ccache issues when using precompiled headers with gcc", "body": "Unfortunately the combination of gcc, precompiled headers, ccache and meson\ncurrently is not safe without further options. The dependencies emitted by gcc\nare insufficient to trigger rebuilds when headers \"below\" the precompiled\nheaders are changed. Whether that's a ccache, gcc or meson bug is\ndebatable. Luckily gcc's -fpch-deps option fixes the issue.\n\nThis problem occasionally leads to build failures, e.g. if only c.h,\npostgres.h or pg_config_manual.h change. That's e.g. the case when creating a\nnew major version branch.\n\nReviewed-by: Nazir Bilal Yavuz \nReviewed-by: Jelte Fennema-Nio \nDiscussion: https://postgr.es/m/CAN55FZ0tqR6Xz%3DiVFLc1BBoLOEHU775ARhcGYwggHA3XLA%3DoQg%40mail.gmail.com\nDiscussion: https://postgr.es/m/CA+hUKG+s7Yvt0PUnSQUEjCjysV-7-51n9B1h468Le3VJi0x4ZQ@mail.gmail.com\nDiscussion: https://postgr.es/m/phsrssp75npoyalqsolcd7fmnmlbzbmquc2p7w7mqjlw7432jk@bzskz3luyjvb\nDiscussion: https://github.com/ccache/ccache/issues/1686\nBackpatch-through: 16, where meson support was added", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "377a8101683ac808ed53e18a5beb9d2a45564443", "date": "2026-07-17T15:24:30Z", "subject": "Add some missing calls to invalidate scene.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "f9bc7d0cf8b4be1d07a8f9200b755c5c40d079c7", "date": "2026-07-17T17:24:32+02:00", "subject": "build: remove global size check for runtime-dom and compiler-dom", "body": "", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "tmux", "sha": "fde879c347b5c9f52af641f406534a73685c4a2b", "date": "2026-07-17T16:30:02+01:00", "subject": "Add local panes as well.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "2710b182472cd1a95e4aff6186cb2c4fbadb4ee0", "date": "2026-07-18T00:35:49+09:00", "subject": "ci: Add explicit permissions to rebuild-docs-sites workflow", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "postgres", "sha": "3601be26b3fa076e1aa03e9f1e0ad0e6077ce759", "date": "2026-07-17T11:43:52-04:00", "subject": "ci: Generate crashlogs on Windows", "body": "This configures cdb.exe to log all crashes to\n\"\\crashlogs\\crashlog-.txt\" (as it was previously set up\nfor cirrus-ci based CI). The upload logs step already collects these logs.\n\nThe logic is copied from the generation of Postgres CI Windows images for\ncirrus-ci [1]. Since this would be too long to include inline in pg-ci.yml,\nit is implemented as 'src/tools/ci/gha_setup_windows_debugger.ps1' script.\n\n[1] https://github.com/anarazel/pg-vm-images/blob/main/scripts/windows_install_dbg.ps1\n\nAuthor: Nazir Bilal Yavuz \nReviewed-by: Andres Freund \nDiscussion: https://postgr.es/m/CAN55FZ1BgsXSTzOpehnMa4NzWL8Aivsxx-di7-VT6bZ3j2Omow%40mail.gmail.com\nDiscussion: https://postgr.es/m/iggjozfshwbqpv33x5jqwtju5k5zrkyu3257dlifxkhtpg7eoq@k2ccyrdi5dtu\nBackpatch-through: 19, where GHA CI was added", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "postgres", "sha": "cce2bbbfd3cb8df4844013d20c3e08cef6d29084", "date": "2026-07-17T11:43:52-04:00", "subject": "ci: Use optimized build for mingw", "body": "The test runtime dominates over the compile time on GHA. Note that we just\nneed to remove options, as postgres's default is debugoptimized.\n\nReviewed-by: Nazir Bilal Yavuz \nDiscussion: https://postgr.es/m/a2ejn7lfqolutzz7kozalbhy3bixdrujb4buc3pgbtlk4am2ba@wbv6v7riia33\nBackpatch-through: 19, where GHA CI was added", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "6c083c206633d67868997f7db39b5e2f4903ed87", "date": "2026-07-17T23:45:39+08:00", "subject": "refactor(demo): migrate styles from markdown to CSS-in-JS", "body": "* refactor(demo): migrate styles from markdown to CSS-in-JS\n\n* update", "style": "conventional", "type": "refactor", "scope": "demo", "breaking": false} {"repo": "eslint", "sha": "dcc9312c1081931de0fe1b555fbb6aa82fe696b9", "date": "2026-07-18T00:45:40+09:00", "subject": "test: add error locations to `eqeqeq`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "eslint", "sha": "3d36589a4917326500f0707bb41745160f387d00", "date": "2026-07-18T00:50:41+09:00", "subject": "chore: add missing backticks to `serialization.js`", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "tmux", "sha": "496bfdce3879ccf5d057edabdc388c854a1466a4", "date": "2026-07-17T16:03:15Z", "subject": "Prevent display-panes from shifting the active pane scroll position by copying the screen from before zoom, also do not let resize consume blank lines.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "481d48b53bb7d4ff0dae525e1aea39e627c1317e", "date": "2026-07-17T17:44:43+01:00", "subject": "3.8 changes.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "eac8ed4825213a80644bb5ea15234a59daa57352", "date": "2026-07-17T13:04:29-04:00", "subject": "Add return check to OSSL_DECODER_up_ref", "body": "Coverity flagged a location where we didn't check this return code, fix\nit up.\n\nFixes https://scan5.scan.coverity.com/#/project-view/60762/10222?selectedIssue=1696759\n\nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMergeDate: Thu Jul 23 08:24:22 2026\n(Merged from https://github.com/openssl/openssl/pull/31992)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1261c82f1e8a961294f9b119c258b1952bdcfbe3", "date": "2026-07-17T19:20:38+02:00", "subject": "Add BIO bss_file test", "body": "Add a cmocka based unit test for the file BIO covering BIO_new_file,\nBIO_new_fp and the file method callbacks, with stdio and openssl_fopen.\n\nAdditionally add a Windows-only Detours based side test for the\nWindows specific branches.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Andrew Dinh \nMergeDate: Sat Aug 8 15:26:34 2026\n(Merged from https://github.com/openssl/openssl/pull/31991)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "02d0f65c9f8579cd1625a8d084ee8479093beb1b", "date": "2026-07-17T13:35:56-04:00", "subject": "CMS/PKCS7: use EVP_CIPHER_fetch() for SMIMECapabilities", "body": "Fixes #25919\n\nReplace EVP_get_cipherbynid()/EVP_get_digestbynid() with\nEVP_CIPHER_fetch()/EVP_MD_fetch() when building SMIMECapabilities\nattributes, so that only algorithms available in the active providers\nare advertised. This prevents RC2, DES, and GOST from appearing by\ndefault when only the default provider is loaded.\n\nAdd CMS_add_standard_smimecap_ex() as a provider-aware replacement\nfor the newly-deprecated CMS_add_standard_smimecap().\n\nAdds missing documention and new unit-tests validating the new behavior.\n\nSigned-off-by: Todd Short \n\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Jakub Zelenka \nMergeDate: Mon Aug 3 07:06:44 2026\n(Merged from https://github.com/openssl/openssl/pull/31990)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "f4e95f65759fc686592bd3ceda130359ef0c93ae", "date": "2026-07-18T02:39:56+09:00", "subject": "fix(microservices): correct pub/sub client type in redis listeners", "body": "", "style": "conventional", "type": "fix", "scope": "microservices", "breaking": false} {"repo": "openssl", "sha": "4bd8a62af12729493247d7b9085595cefd135df2", "date": "2026-07-17T19:43:18+02:00", "subject": "OSSL_FN: Add mod_inverse stanza tests to fntest", "body": "Add a file_modinv() stanza handler to test/fntest.c, exercising\nOSSL_FN_mod_inverse() through the BIGNUM/OSSL_FN bridge. It reuses\nthe bntest ModInv corpus (test/recipes/10-test_bn_data/bnmodinv.txt),\nso 11-test_fn.t now feeds that file to fntest.\n\nThe handler mirrors bntest's file_modinv(): a ModInv value of -1 marks\na negative test that expects OSSL_FN_mod_inverse() to fail with\nOSSL_FN_R_NO_INVERSE (covering a == 0, non-coprime pairs, and the\ndegenerate moduli M = 0 and M = 1), while a non-negative ModInv is\nchecked against the computed inverse and verified by A * inv(A) == 1\n(mod M).\n\nOSSL_FN is unsigned, so OSSL_FN_mod_inverse() computes the inverse of\n|A|; for negative A the result is adjusted to M - r before the\ncomparison, matching the sign convention of the corpus.\n\nAssisted-by: Pi:z-ai/glm-5.2\n\nReviewed-by: Igor Ustinov \nReviewed-by: Tomas Mraz \nMergeDate: Fri Jul 31 09:32:20 2026\n(Merged from https://github.com/openssl/openssl/pull/32027)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "19ddd861289f8951707f81ff6db9efcb7a34e639", "date": "2026-07-17T12:52:56-05:00", "subject": "fix(restart): messages EXX error codes #40759", "body": "(cherry picked from commit ed5dcc0edf2ebea02c4c5d667a612cd6fb23cf81)", "style": "conventional", "type": "fix", "scope": "restart", "breaking": false} {"repo": "neovim", "sha": "ed5dcc0edf2ebea02c4c5d667a612cd6fb23cf81", "date": "2026-07-17T12:52:56-05:00", "subject": "fix(restart): messages EXX error codes #40759", "body": "", "style": "conventional", "type": "fix", "scope": "restart", "breaking": false} {"repo": "neovim", "sha": "feec7b848f864e45f16081dfea3b19855fbc1fb5", "date": "2026-07-17T12:59:45-05:00", "subject": "docs: remove incorrect mention of `ZR` in the docs for `v:startreason`", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9d3edf345f8e25f1a7976efff54818a426661802", "date": "2026-07-17T19:21:29+01:00", "subject": "fix(path): avoid shell fallback for unknown users #40757", "body": "Problem:\nExpanding an unknown ~user path falls back to the shell. In a fast event,\nthis re-enters the event loop and aborts Nvim.\n\nSolution:\nLeave unknown users unexpanded when the system account lookup fails.", "style": "conventional", "type": "fix", "scope": "path", "breaking": false} {"repo": "openssl", "sha": "9375418a8758b9229182da5d5922342c08bc7c74", "date": "2026-07-17T14:22:50-04:00", "subject": "use safe_math on do_name_ex", "body": "Coverity caught some potential overflows in do_name_ex. Seems unlikely\nto occur, but lets just do all the adds under safe_math to ensure we\ndon't overflow our buffer length.\n\nFixes https://scan5.scan.coverity.com/#/project-view/60762/10222?selectedIssue=1695278\n\nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMergeDate: Thu Jul 23 08:24:23 2026\n(Merged from https://github.com/openssl/openssl/pull/31992)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "eccc68d42564e46bc9020d19ec52d2988f3b7bfa", "date": "2026-07-18T03:52:27+09:00", "subject": "docs: correct --suppressions-location option description", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "85e0559d46c700e810ef605d39f9b376ad497ea1", "date": "2026-07-18T05:23:34+10:00", "subject": "feat(ui): \":%detach!\" detaches all other UIs #39216", "body": "Problem:\nNot easy for one UI to kick the others off a shared Nvim server.\n\nSolution:\nTreat `:%detach` as detach-other-UIs.", "style": "conventional", "type": "feat", "scope": "ui", "breaking": false} {"repo": "postgres", "sha": "c9ea37231a2e23120bafe0c4dd542642d4b87aad", "date": "2026-07-17T15:53:10-04:00", "subject": "Fix GiST index-only scan column alignment issue.", "body": "An index-only scan filled its result slot from the HeapTuple an index AM\nreturns in scan->xs_hitup by deforming it with the virtual slot's own\ntuple descriptor (during GiST and SP-GiST index-only scans). But index\nAMs form that heap tuple using their own descriptor, scan->xs_hitupdesc.\nThe AM's descriptor may disagree with the IoS virtual slot's descriptor\nabout each column's precise alignment, leading to \"can't happen\" errors\nin certain rare edge cases. Hard crashes were possible but much less\nlikely.\n\nTo fix, deform the tuple with the descriptor it was formed with. This\nis simpler, and makes xs_hitup handling (used by GiST and SP-GiST)\nuniform with the nearby existing xs_itup handling (used by nbtree).\n\nIn practice this issue was very unlikely to be hit (it was found during\ntesting of a patch that will change the table AM API used during index\nscans). The only currently affected core opclass is GiST's range_ops.\nIt was only possible for the datum to be accessed at an incorrectly\naligned offset when reading the second or subsequent column from a\nmulticolumn GiST index. This couldn't happen in the common case where\nthe datum used an unaligned short varlena header. Moreover, an earlier\ncolumn had to leave the range datum at an offset where the two\nalignments actually disagree (e.g., an odd-length varlena datum).\n\nAuthor: Peter Geoghegan \nReviewed-by: Tomas Vondra \nDiscussion: https://postgr.es/m/CAH2-WzkGXa2SKnebdW29RT1hCcQBo_p03v3iqif2u9bjzLB-aQ@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "3b5a3cfb903eca6dfc87319e0ac29e3e23430e08", "date": "2026-07-17T15:56:54-04:00", "subject": "doc PG 19 relnote: move Oauth items", "body": "Reported-by: Jacob Champion\n\nAuthor: Jacob Champion\n\nDiscussion: https://postgr.es/m/CAOYmi+k5h_w5p6HsVJF5k+bNGmChMAhpGU_1gjJMxHg080XibA@mail.gmail.com\n\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e0e7f65b74f7e0e901b415bbb62f6ce52a13b7cb", "date": "2026-07-17T16:37:45-04:00", "subject": "refactor(detach): eliminate `nvim__ui_detach` #40793", "body": "Problem:\n`nvim__ui_detach` was added in 85e0559d46c7 in order to implement\n`detach_others` from Lua. Using Lua in this case is doing more harm than\ngood.\n\nSolution:\nExtract `ui_detach_channel`, which also allows it to be used for the\n\"self detach\" path from `ex_detach`.\n\nBonus: the old MSWIN path always called `os_swap_to_hidden_console()`;\nnow `ui_detach_channel` only does so for a stdio channel. For the common\nstdio TUI this is identical; for a socket UI it's more correct (no\nparent console to swap).", "style": "conventional", "type": "refactor", "scope": "detach", "breaking": false} {"repo": "neovim", "sha": "a373466cd984eeec7109c07c09e7d3d0978e5ff2", "date": "2026-07-17T16:45:11-04:00", "subject": "test: unreliable \"bumps 'busy' on each watched buffer\" #40794", "body": "Problem:\nTest still unreliable.\n\nSolution:\nDon't require both 'busy' pulses to coincide.\n\nfollowup to #40775", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "3180ce3d7a836e7fbcf8946a0e0be9f263353481", "date": "2026-07-17T16:47:33-04:00", "subject": "Turn visibilitymap_clear() Assert back into an error", "body": "Commit ed62d26caca fixed a bug in clearing the visibility map and, while\ndoing so, made some incidental changes to visibilitymap_clear(). One of\nthem replaced the error thrown when the wrong buffer is passed to\nvisibilitymap_clear() with an Assert().\n\nWhile anyone adding a new visibilitymap_clear() caller should be running\nassert-enabled builds, visibilitymap_set() still reports the same\nwrong-buffer condition with an elog(ERROR), so visibilitymap_clear()\nshould also do so for consistency. This change was also unrelated to\nthe bug fix and is better made as a separate commit. Restore the error.\n\nReported-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAHGQGwF_PzOv5y7ucFh7Fqqqa8ar83zYwvWugqarRD6%2B7GCtEQ%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "e02a0debcaa312ef602e97e027a3f8599d2934c2", "date": "2026-07-18T00:10:19+03:00", "subject": "refactor(devtools): add alt text to the empty-state screen illustrations", "body": "The Angie illustrations on the DevTools empty-state screens (added in\n#69786) shipped with an empty `alt` and `aria-hidden=\"true\"`. Give them\ndescriptive alt text and drop `aria-hidden` to improve accessibility.\n\n(cherry picked from commit 00699f3e66bf5b1a48590410c2390f95a7988950)", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "angular", "sha": "00699f3e66bf5b1a48590410c2390f95a7988950", "date": "2026-07-18T00:10:19+03:00", "subject": "refactor(devtools): add alt text to the empty-state screen illustrations", "body": "The Angie illustrations on the DevTools empty-state screens (added in\n#69786) shipped with an empty `alt` and `aria-hidden=\"true\"`. Give them\ndescriptive alt text and drop `aria-hidden` to improve accessibility.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "openssl", "sha": "7810f3341ab34ea17592c6f3c7a1f967a61bf401", "date": "2026-07-17T17:34:11-04:00", "subject": "errors: fix stale string overrides in openssl.txt for renamed SSL error codes", "body": "Commit b3d26e438a (\"merge feature/removesslv3\", PR #29338) renamed SSL3\nerror codes to TLS equivalents and updated the string descriptions in\ncrypto/ssl_err.c (e.g. from \"ssl/tls alert bad certificate\" to\n\"tls alert bad certificate\"), but did not update the corresponding\nstrings in crypto/err/openssl.txt.\n\nThis caused a latent divergence: running mkerr.pl -rebuild would\nregenerate ssl_err.c with the old openssl.txt strings, undoing the\nintentional name cleanup. Similarly, adding any new SSL error code\ntriggers a full SSL regeneration that would revert the strings and\nreorder entries to their correct alphabetical positions, producing\nspurious changes unrelated to the new code.\n\nUpdate openssl.txt to match the strings already in ssl_err.c, which\nalso happen to be the auto-generated strings derived from the reason\ncode names. Regenerate ssl_err.c and sslerr.h with the correct\nalphabetical ordering so that future make update runs are idempotent:\nadding and removing a new SSL error code now produces only the expected\nchanges to the generated files, with no side-effect reordering or\nstring reversions.\n\nFixes: b3d26e438a (\"Rename SSL3 error codes to TLS equivalents\")\nSigned-off-by: Todd Short \n\nReviewed-by: Tim Hudson \nReviewed-by: Paul Dale \nMergeDate: Tue Jul 21 14:28:22 2026\n(Merged from https://github.com/openssl/openssl/pull/31993)\n\n(cherry picked from commit ba2f3b557a660cbe2a0fc0e99b1830cb930a090f)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ba2f3b557a660cbe2a0fc0e99b1830cb930a090f", "date": "2026-07-17T17:34:11-04:00", "subject": "errors: fix stale string overrides in openssl.txt for renamed SSL error codes", "body": "Commit b3d26e438a (\"merge feature/removesslv3\", PR #29338) renamed SSL3\nerror codes to TLS equivalents and updated the string descriptions in\ncrypto/ssl_err.c (e.g. from \"ssl/tls alert bad certificate\" to\n\"tls alert bad certificate\"), but did not update the corresponding\nstrings in crypto/err/openssl.txt.\n\nThis caused a latent divergence: running mkerr.pl -rebuild would\nregenerate ssl_err.c with the old openssl.txt strings, undoing the\nintentional name cleanup. Similarly, adding any new SSL error code\ntriggers a full SSL regeneration that would revert the strings and\nreorder entries to their correct alphabetical positions, producing\nspurious changes unrelated to the new code.\n\nUpdate openssl.txt to match the strings already in ssl_err.c, which\nalso happen to be the auto-generated strings derived from the reason\ncode names. Regenerate ssl_err.c and sslerr.h with the correct\nalphabetical ordering so that future make update runs are idempotent:\nadding and removing a new SSL error code now produces only the expected\nchanges to the generated files, with no side-effect reordering or\nstring reversions.\n\nFixes: b3d26e438a (\"Rename SSL3 error codes to TLS equivalents\")\nSigned-off-by: Todd Short \n\nReviewed-by: Tim Hudson \nReviewed-by: Paul Dale \nMergeDate: Tue Jul 21 14:28:22 2026\n(Merged from https://github.com/openssl/openssl/pull/31993)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b5f0da4c9412202bc546555f0e7f459ee427a791", "date": "2026-07-17T23:44:14+02:00", "subject": "test: assert the specific alert in DTLS sslrecords tests", "body": "The DTLS unknown-record-type tests in 70-test_sslrecords.t checked only\nthat s_client exited non-zero, which does not distinguish a genuine\nrejection of the record from an unrelated failure (or a crash).\n\nCapture s_client's stdout+stderr in TLSProxy and parse the alerts it\nlogs via the -msg message callback, exposing them through new accessors:\nclientoutput, client_alerts and client_sent_fatal_alert(). The check\nreflects the alert s_client generated locally and so does not depend on\nthe peer receiving it, keeping the test deterministic.\n\nIn the DTLS cases add -msg to the client flags and assert that s_client\ngenerated the fatal unexpected_message alert and exited cleanly with a\nfailure, using a new client_failed() helper that decodes the child\nstatus so a signal death is not mistaken for a rejection. On failure\nthe captured client output is dumped for diagnostics.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Tomas Mraz \nReviewed-by: Ryan Hooper \nReviewed-by: Neil Horman \nMergeDate: Thu Sep 10 21:33:34 2026\n(Merged from https://github.com/openssl/openssl/pull/32135)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5ed9cee74c6cf23729bd5cc4dee3b0ebbd4aa7aa", "date": "2026-07-17T18:08:23-04:00", "subject": "Fix yet another portability problem in new NLS test.", "body": "Álvaro reported offlist that his machine was passing the new-in-v19\nnls.sql test in \"make check\" but not in \"make installcheck\". On\ninvestigation, the cause turned out to be that he has LANGUAGE set in\nhis environment, and with (at least recent versions of) glibc that\noverrides LC_MESSAGES and friends, as per previous research by Bryan\nGreen. \"make check\" works because pg_regress unsets LANGUAGE before\nstarting the postmaster, but in installcheck mode we're exposed to\nthe prevailing value and we lose.\n\nWe're already hacking the value of LANGUAGE in this test for Solaris,\nso let's just extend that to unsetting LANGUAGE on every other platform.\n\nReported-by: Álvaro Herrera \nDiagnosed-by: Andrew Dunstan \nAuthor: Tom Lane \nDiscussion: https://postgr.es/m/a337896e-5bff-490b-afc9-c545f06c014c@gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f931f44ea553c26fa69180a72bddd77e091aa4dc", "date": "2026-07-18T00:09:30+02:00", "subject": "apps: decode DTLSv1.2 records in s_client/s_server -msg output", "body": "The msg_cb message callback used by the -msg option only recognised\nDTLSv1.0 among the DTLS versions, so DTLSv1.2 records were logged as\n\"Not TLS data or unknown version\" instead of being decoded. Add\nDTLS1_2_VERSION to the recognised version check and to the ssl_versions\nlookup table, and wrap the now-overlong condition.\n\nAdd a test that runs s_client against s_server over TLSv1.2, TLSv1.3 and\nDTLSv1.2, logging the protocol messages via -msg, and checks that every\nrecord is decoded (no \"Not TLS data or unknown version\" lines).\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Tim Hudson \nReviewed-by: Paul Dale \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Mon Jul 20 09:09:23 2026\n(Merged from https://github.com/openssl/openssl/pull/31994)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e0bf61917aa5e9f75c0ee36d0809620823fe818f", "date": "2026-07-18T00:09:30+02:00", "subject": "apps: decode DTLSv1.2 records in s_client/s_server -msg output", "body": "The msg_cb message callback used by the -msg option only recognised\nDTLSv1.0 among the DTLS versions, so DTLSv1.2 records were logged as\n\"Not TLS data or unknown version\" instead of being decoded. Add\nDTLS1_2_VERSION to the recognised version check and to the ssl_versions\nlookup table, and wrap the now-overlong condition.\n\nAdd a test that runs s_client against s_server over TLSv1.2, TLSv1.3 and\nDTLSv1.2, logging the protocol messages via -msg, and checks that every\nrecord is decoded (no \"Not TLS data or unknown version\" lines).\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Tim Hudson \nReviewed-by: Paul Dale \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Mon Jul 20 09:09:07 2026\n(Merged from https://github.com/openssl/openssl/pull/31994)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2cf212db5286baa7d7e80449300fe472d8c4ba9f", "date": "2026-07-17T16:41:14-07:00", "subject": "Generate unicode_limits.h.", "body": "To ensure we do not overflow a size_t while case mapping on a 32-bit\nplatform, we need to know the maximum amount a UTF8 string can expand.\n\nCalculate that maximum while generating unicode tables as a part of\nthe update-unicode target, and output into a new header\nunicode_limits.h. Minor refactoring along the way.\n\nAdd a StaticAssertDecl to check that a MaxAllocSize text value\nexpanded by that amount would still have a length that fits in\nsize_t. (We couldn't actually create a new text value out of that, but\nwe still need to avoid overflow.)\n\nReported-by: Tom Lane \nDiscussion: https://postgr.es/m/3213927.1783950167@sss.pgh.pa.us", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "2fff7c4025dc29de9965cc0f0dd9fac2709bbd18", "date": "2026-07-17T19:52:19-04:00", "subject": "chore: fix @babel/eslint-plugin peer deps", "body": "* chore: fix @babel/eslint-plugin peer deps\n\n* chore: update lock file", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "1027d07704836c8d642abbed34ab98535de47433", "date": "2026-07-18T02:42:39+02:00", "subject": "timeval: make `Curl_freq` variable static (Windows)", "body": "Follow-up to 64d4f5c4b660be112a02be104e7314e8bf264caa #20852\n\nCloses #22346", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9fae7234f661b0c701d4a18c13b429b517be5d25", "date": "2026-07-18T09:07:01+08:00", "subject": "feat(completion): support completionList.applyKind #40769", "body": "Problem: values from completionList.itemDefaults always replace the\nitem's own commitCharacters and data. Since 3.18 a server can ask for\nthem to be merged instead.\n\nSolution: Handle CompletionList.applyKind and advertise applyKindSupport.\nDrop itemDefaults once folded into the items: get_items() runs twice on\nthe same result and a Merge isn't idempotent.\n\nhttps://microsoft.github.io/language-server-protocol/specifications/lsp/3.18/specification/#completionItemApplyKinds", "style": "conventional", "type": "feat", "scope": "completion", "breaking": false} {"repo": "neovim", "sha": "9f3f326acd05055f324cffb15764266a40291737", "date": "2026-07-17T20:07:58-05:00", "subject": "fix(tui): urxvt reset cursor style #40800", "body": "Problem:\nurxvt doesn't support resetting to default cursor style like we assumed.\nhttps://github.com/neovim/neovim/issues/38987#issuecomment-5004636437\n\nSolution:\nInstead of the reset sequence, always set it to steady block. This will\nstill be wrong for users who configured their initial cursor to\nunderline, but it's about the best we can do.\n\n(cherry picked from commit 28818702e94de0a606afb9d3fe39a76c1a37d026)", "style": "conventional", "type": "fix", "scope": "tui", "breaking": false} {"repo": "neovim", "sha": "28818702e94de0a606afb9d3fe39a76c1a37d026", "date": "2026-07-17T20:07:58-05:00", "subject": "fix(tui): urxvt reset cursor style #40800", "body": "Problem:\nurxvt doesn't support resetting to default cursor style like we assumed.\nhttps://github.com/neovim/neovim/issues/38987#issuecomment-5004636437\n\nSolution:\nInstead of the reset sequence, always set it to steady block. This will\nstill be wrong for users who configured their initial cursor to\nunderline, but it's about the best we can do.", "style": "conventional", "type": "fix", "scope": "tui", "breaking": false} {"repo": "curl", "sha": "1128177a6f725d1f8c4cba4f7ff6532969109e7a", "date": "2026-07-18T03:18:11+02:00", "subject": "timeval: rename timer frequency variable", "body": "It's no longer a global variable.\n\nFollow-up to 1027d07704836c8d642abbed34ab98535de47433 #22346\n\nCloses #22347", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "a8a3b55fcc24e05151c03fce89d88a7f43672ae2", "date": "2026-07-18T03:50:47+02:00", "subject": "feat(common): add override mimetype option to file type validator", "body": "", "style": "conventional", "type": "feat", "scope": "common", "breaking": false} {"repo": "openssl", "sha": "b80761945b62bda5d78110bb4bb39f0fedbef17a", "date": "2026-07-17T22:02:19-04:00", "subject": "Fix capitalization of 'Swiss Army knife' in README", "body": "CLA: trivial\n\nReviewed-by: Andrew Dinh \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Paul Dale \nMergeDate: Thu Jul 23 08:30:57 2026\n(Merged from https://github.com/openssl/openssl/pull/31871)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3a5a104da6e7b890b70f188f83dad1e7b633c01f", "date": "2026-07-18T11:11:26+09:00", "subject": "docs: clarify EVP_PKEY context workflow and operation descriptions", "body": "Document that a generated key is returned separately from its generation\ncontext and that a new key-bound context is needed for subsequent\noperations. Correct the encrypt and decrypt initialization descriptions\nand examples accordingly.\n\nCorrect two copy-and-paste errors in the verify-recover documentation:\ndescribe its initializer as starting a verify-recover operation and refer\nto EVP_PKEY_verify_recover() when discussing its input.\n\nFixes #31989\n\nAssisted-by: Grok:Grok-4.5\n\nReviewed-by: Tomas Mraz \nReviewed-by: Dmitry Belyavskiy \nMergeDate: Thu Jul 23 08:29:41 2026\n(Merged from https://github.com/openssl/openssl/pull/31995)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f7112c925c0f3f6239bcab218cb32fa3e52bf621", "date": "2026-07-18T11:11:26+09:00", "subject": "docs: clarify EVP_PKEY context workflow and operation descriptions", "body": "Document that a generated key is returned separately from its generation\ncontext and that a new key-bound context is needed for subsequent\noperations. Correct the encrypt and decrypt initialization descriptions\nand examples accordingly.\n\nCorrect two copy-and-paste errors in the verify-recover documentation:\ndescribe its initializer as starting a verify-recover operation and refer\nto EVP_PKEY_verify_recover() when discussing its input.\n\nFixes #31989\n\nAssisted-by: Grok:Grok-4.5\n\nReviewed-by: Tomas Mraz \nReviewed-by: Dmitry Belyavskiy \nMergeDate: Thu Jul 23 08:29:24 2026\n(Merged from https://github.com/openssl/openssl/pull/31995)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "6e5b5a784c5447110caca4d19b97b1a51685ec39", "date": "2026-07-18T04:40:36+02:00", "subject": "test(common): add override mimetype validation coverage", "body": "", "style": "conventional", "type": "test", "scope": "common", "breaking": false} {"repo": "neovim", "sha": "371e60dceb22cd383beda74c30c8b7d63ddd2ccf", "date": "2026-07-18T03:07:14-05:00", "subject": "fix(restart): distinguish v:exitreason for :restart! #40801", "body": "(cherry picked from commit 12e8db38471eacd12da28d42103d67619ea9af0f)", "style": "conventional", "type": "fix", "scope": "restart", "breaking": false} {"repo": "neovim", "sha": "12e8db38471eacd12da28d42103d67619ea9af0f", "date": "2026-07-18T03:07:14-05:00", "subject": "fix(restart): distinguish v:exitreason for :restart! #40801", "body": "", "style": "conventional", "type": "fix", "scope": "restart", "breaking": false} {"repo": "neovim", "sha": "446b9d8c5589f931a71904cd5bf384fec56ebf2b", "date": "2026-07-18T04:29:51-04:00", "subject": "docs: misc, remove some old help tags #40214", "body": "- Remove old help tags: they add noise to cmdline completion.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9abe01c7be4cc616eed2b88f9140f4b004a5a043", "date": "2026-07-18T05:01:41-04:00", "subject": "test: unreliable \"bumps 'busy' on each watched buffer\" #40804", "body": "Problem:\nSampling a transient busy=1 pulse is racy under load.\n\nSolution:\nSet a long debounce, assert the stable state, then use a short debounce\nto let the reload clear it.", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "eslint", "sha": "8b6f6c0b33411f34485512456d94f221daf7321f", "date": "2026-07-18T12:17:17+03:00", "subject": "fix: apply ignore configs to computed methods in class-methods-use-this", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "neovim", "sha": "5e2ad295c07fa66c6e8ec05b1142b2964eabba9b", "date": "2026-07-18T05:39:36-04:00", "subject": "test: unreliable \":terminal buffer … partial mappings\" #40806", "body": "Problem:\n\n FAILED …/terminal/buffer_spec.lua @ 348: :terminal buffer is refreshed with partial mappings in Normal mode\n …/terminal/buffer_spec.lua:348: Row 2 did not match.\n Expected:\n |tty ready |\n |*^jjjjjjjjjj |\n | |\n | |\n | |\n | |\n | |\n Actual:\n |tty ready |\n |*^jjjjjjjjj |\n | |\n | |\n | |\n | |\n | |\n\nSolution:\nIncrease sleep time.", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "b2bb60d70a449b5ce8041ed27df7c1248fea17d8", "date": "2026-07-18T06:33:38-05:00", "subject": "fix(ui2): avoid ModeChanged in message window #40788", "body": "", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "neovim", "sha": "99a16121d4aa0affabbd37697bccc388d48d0229", "date": "2026-07-18T07:48:36-04:00", "subject": "test: unreliable \"put command . register special … ring the bell\" #40812", "body": "Problem:\n\n FAILED …/put_spec.lua:894: …/put_spec.lua @ 898: put command . register special tests should ring the bell when deleting if not appropriate\n …/put_spec.lua:894: …/put_spec.lua:898: Bell was not rung after action\n\n Snapshot:\n screen:expect([[\n ^ine of words 1 |\n Line of words 2 |\n {1:~ }|*4\n {3:[No Name] [+] }|\n |\n {1:~ }|*4\n {UNKNOWN_HL_ID(4):[No Name] }|\n {UNKNOWN_HL_ID(1): }|\n ]])\n stack traceback:\n …/ui/screen.lua:917: in function '_wait'\n …/ui/screen.lua:540: in function 'expect'\n …/editor/put_spec.lua:894: in function 'bell_test'\n …/editor/put_spec.lua:940: in function <…/editor/put_spec.lua:932>\n\nAnalysis:\nThe bell was asserted via the UI 'bell' event (screen.bell), but\nvim_beep() rate-limits that to 3 per 500ms. These tests share one\nlong-lived session, so beeps from earlier tests could exhaust the window\nand drop the asserted beep.\n\nSolution:\nUse assert_beeps()/assert_nobeep(), which check called_vim_beep (set\nbefore the rate-limit gate).", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "27999c571c79b5c4d4949000f64bdd562ce5632a", "date": "2026-07-18T14:01:25+02:00", "subject": "test(tui): failures if build lacks terminfo", "body": "", "style": "conventional", "type": "test", "scope": "tui", "breaking": false} {"repo": "openssl", "sha": "78669ae813f552ffb13608b92866f6f18dae671c", "date": "2026-07-18T21:14:23+09:00", "subject": "doc: clarify raw RSA migration", "body": "RSA_private_encrypt() and RSA_public_decrypt() support both PKCS#1 v1.5\nand no padding, but the migration guide currently describes their EVP\nreplacements as using no padding only.\n\nDocument direct mappings to EVP_PKEY_sign() and\nEVP_PKEY_verify_recover(), and explain that compatibility requires no\nsignature digest and the same legacy padding mode, RSA_PKCS1_PADDING or\nRSA_NO_PADDING.\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Paul Dale \nMergeDate: Mon Jul 27 07:54:00 2026\n(Merged from https://github.com/openssl/openssl/pull/31996)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "cef11407852e369888ef9ac58cd4f2de70f2edbc", "date": "2026-07-18T14:56:27+02:00", "subject": "test(tui): avoid \"Screen test succeeded immediately\" warning", "body": "", "style": "conventional", "type": "test", "scope": "tui", "breaking": false} {"repo": "neovim", "sha": "9129bb04cac1b1d7dc0d6dca7bca77d2ee6c5187", "date": "2026-07-18T14:59:32+02:00", "subject": "test(pack): \"vim.pack\" tests fail on local machine", "body": "", "style": "conventional", "type": "test", "scope": "pack", "breaking": false} {"repo": "eslint", "sha": "20b5ad052360a443786a202e94624a3f81846511", "date": "2026-07-18T15:37:13+02:00", "subject": "fix: quadratic-time regex in `prefer-template`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "6b0c9b21914132f71d9d312b0c3c80d28062c193", "date": "2026-07-18T19:57:36+05:30", "subject": "fix(core): drop the g flag from middleware overlap regexes", "body": "removeOverlappedRoutes builds one regex per parameterized route and\nreuses it while filtering every route. The regexes were created with the\ng flag, and a global regex advances lastIndex on a successful test(), so\nthe next call resumed mid-string and failed the ^ anchor.\n\nOverlap detection therefore depended on iteration order: with cats/:id,\ncats/1, cats/2 and cats/3, the filter kept ['cats/:id', 'cats/2'] and\nthe middleware ran twice for /cats/2.\n\nThe pattern is anchored and only ever used with test(), so the flag\nbought nothing. Without it the same input collapses to ['cats/:id'].\n\nFixes #17334", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "neovim", "sha": "a99f8989f7f32eb052b2dfcc5d18b7c6705c93ed", "date": "2026-07-18T10:35:18-04:00", "subject": "build(vim-patch): alloc.c N/A functions", "body": "Note this doesn't find new N/A patches by itself\nand this is to align with ':h dev-vimpatch'.\nI use the 'vim_na_*' files to debug the filtered diff of files\nthat seem N/A but contain relevant code.\nI can't blacklist the entire file.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "4cef487f2a5615e59d919e6fd4094a02038950ff", "date": "2026-07-18T16:49:06+02:00", "subject": "Guard against NULL data in empty ASN1_STRINGs", "body": "Since 28179061bf a zero-length ASN1_STRING has data == NULL in\nfuzzing builds instead of a 1-byte allocation. Several call sites\ndid pointer arithmetic or memcpy on the data pointer before any\nlength check, which is undefined behaviour for NULL even with a\nzero offset and aborts the fuzz targets under UBSan:\n\n- asn1_string_canon: skip canonicalisation of empty values\n- do_buf, do_hex_dump: return early on an empty buffer\n- i2d_ocsp_nonce: skip the memcpy for an empty nonce\n\nThe loops at these sites were already no-ops for zero length, so\nthere is no behaviour change outside sanitizer builds.\n\nAssisted-By: Claude:claude-fable-5\n\nSigned-off-by: Norbert Pocs \n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nMergeDate: Tue Jul 21 12:54:48 2026\n(Merged from https://github.com/openssl/openssl/pull/31998)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a2dfa195b2b9db3f6c442a131e1c9268989fa70d", "date": "2026-07-18T18:20:41+02:00", "subject": "feat(lua): vim.keycode() gets structured parse result #38636", "body": "Problems:\n1. Can't get individual parts of a key-chord separately: modifiers, key.\n2. Can't separate a key-combo into individual key-chords.\n\nSolution:\nEnhance `vim.keycode()` to optionally return a structured parse result\nas a list of key-chords:\n- `key_raw` the key-chord (problem 2)\n- `mod` the modifiers of `key_raw` (problem 1)\n- `key_orig` the key part of the key-chord, only here if differing from `key`\n (this doesn't solve any of the above mentioned problems, but it may provide\n useful and it's (in terms of code) free)\n- `key` a normalized version of `key_orig` (solving problem 1), example(the\n first is `key_orig` and second is `key`): `lt` and `<`, `Bar` and `|` (in\n ``)", "style": "conventional", "type": "feat", "scope": "lua", "breaking": false} {"repo": "postgres", "sha": "e727c6f2ef5c75a282fbef9cb8adc0c1c6c7978e", "date": "2026-07-18T09:22:40-07:00", "subject": "Fix REASSIGN OWNED for subscriptions in other databases.", "body": "Subscription objects are conceptually database-local objects, but\npg_subscription is a shared catalog so that the launcher process can\nscan it.\n\nCheck readers of pg_subscription to ensure that, unless it's the\nlauncher process, it filters by MyDatabaseId. Most readers were\nalready doing so, but this commit fixes REASSIGN OWNED and adds guards\nto catch other problems in the future. Also, clarify documentation.\n\nAuthor: Dilip Kumar \nReported-by: Noah Misch \nReviewed-by: shveta malik \nDiscussion: https://postgr.es/m/20260710192533.4f.noahmisch@microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "54f6db2f99d2db4d362c8e9ab7c056c548447bc0", "date": "2026-07-18T19:44:43+03:00", "subject": "fix(common): Correct JSDoc in two http exceptions", "body": "", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "nestjs", "sha": "15f30d304920eec249b6d84f5a083022419c91b2", "date": "2026-07-18T19:45:49+03:00", "subject": "fix(common): Add @publicApi to ParseDatePipe, rename ParseArrayOptions", "body": "", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "postgres", "sha": "8d2beee027af41a6b66c670b6a595895f24a1bb0", "date": "2026-07-18T12:57:42-04:00", "subject": "Restore the ability to use | and -> as prefix operators.", "body": "Commit 2f094e7ac changed the parser to treat these as built-in\noperator names, where before they were just generic Op. While\nit correctly gave them the same precedence as Op and added new\nproductions to allow them to still be used as infix operators,\nit missed allowing them to still be used as prefix operators.\nAt least one extension expects to be able to do that, so add\nthe necessary productions.\n\nBug: #19558\nReported-by: Pierre Senellart \nAuthor: Pierre Forstmann \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/19558-ad1fca59a3a471a0@postgresql.org\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "5f2618ac605b9cab7aeb5ef87ef32a190fae507a", "date": "2026-07-18T20:30:02+03:00", "subject": "fix(common): Reject hex and non-numeric strings in parse-float pipe", "body": "", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "postgres", "sha": "3e1fe25e686d788a32c926e2aa79bc8d9312841b", "date": "2026-07-18T14:09:10-04:00", "subject": "Fix edge case in remove_useless_result_rtes() with outer joins.", "body": "find_dependent_phvs() and find_dependent_phvs_in_jointree() decide\nwhether a PlaceHolderVar depends on the RTE_RESULT rel we're\nconsidering removing by comparing the PHV's phrels to a singleton set\ncontaining that rel's RT index, reasoning that if phrels contains any\nother relid bits then those define an appropriate place where we can\nevaluate the PHV. But since this code was originally written, we've\nredefined phrels to include outer-join relids, and that breaks this\nlogic, potentially allowing us to remove an RTE_RESULT that leaves no\nvalid place to evaluate the PHV. The planner doesn't throw an error\nwhen that happens, but it does produce an incorrect plan that will not\nreplace the PHV's value with NULL when needed.\n\nIn the known test case for this bug, the \"extra\" OJ relid is one that\nwe've actually decided to remove but haven't yet cleaned out of the\nquery's PHVs. It's not entirely clear though that that would always\nbe the case. Let's restore this code to the way it was designed to\nwork, by considering only base relids within the PHV's phrels.\n\nBug: #19553\nReported-by: Viktor Leis \nAuthor: Matheus Alcantara \nCo-authored-by: Richard Guo \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/19553-4561747f93f368a7@postgresql.org\nBackpatch-through: 16", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "6f5fae3f8cf8b4f8b9baa8a8eaca9deffcf929e5", "date": "2026-07-18T14:31:36-04:00", "subject": "fix(lua): vim.keycode cleanup #40817", "body": "", "style": "conventional", "type": "fix", "scope": "lua", "breaking": false} {"repo": "postgres", "sha": "5951bf2ba56cd4806955fb6299529bde026de1ac", "date": "2026-07-19T03:50:01+09:00", "subject": "Further cleanup for commit 54cd6fc83.", "body": "Commit 54cd6fc83 set the version argument for the stats-import functions\nintroduced by that commit, which is of type int, using UInt32GetDatum,\nnot Int32GetDatum. This would be completely harmless as it's positive\nand currently ignored in the functions, but let's fix that code to use\nInt32GetDatum for consistency.\n\nAuthor: Etsuro Fujita \nDiscussion: https://postgr.es/m/CAPmGK14aremJGrPezVwFqWt7dnrMhD3KF1DgzsRygAUPETBU7w%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "1ecd9c31e5d3ab0301507ad221b499ce3007a5de", "date": "2026-07-18T21:52:03+03:00", "subject": "docs: fix broken link to Angular AI Skills page", "body": "The MCP server setup guide linked to /ai/skills, which 404s. The\npage lives at /ai/agent-skills. Update the link to point to the\ncorrect route.\n\nFixes #69838\n\n(cherry picked from commit d6cad90b3a865cbedc9575a1966192720c4014ea)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "d6cad90b3a865cbedc9575a1966192720c4014ea", "date": "2026-07-18T21:52:03+03:00", "subject": "docs: fix broken link to Angular AI Skills page", "body": "The MCP server setup guide linked to /ai/skills, which 404s. The\npage lives at /ai/agent-skills. Update the link to point to the\ncorrect route.\n\nFixes #69838", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "3ff610077744d749849eaeeaaac34f9cf6aee217", "date": "2026-07-18T21:57:49+03:00", "subject": "fix(docs-infra): only claim a search on the 404 page when results exist", "body": "The \"Page Not Found\" page always rendered the sentence \"We have\ninitiated a search for the term extracted from the URL\", but the\nNotFound component only runs a search when a term can be extracted,\nand only renders results when there is at least one hit. On a 404\nwith no extractable term or no matching results, the page asserted\na search it never backed up.\n\nGate the sentence behind the same searchResults() check that guards\nthe results list, so the empty state simply reads \"We couldn't find\nwhat you were looking for.\"\n\n(cherry picked from commit 79fe1c3a058a2380aabcf7b0d0cf5401425cb6bc)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "79fe1c3a058a2380aabcf7b0d0cf5401425cb6bc", "date": "2026-07-18T21:57:49+03:00", "subject": "fix(docs-infra): only claim a search on the 404 page when results exist", "body": "The \"Page Not Found\" page always rendered the sentence \"We have\ninitiated a search for the term extracted from the URL\", but the\nNotFound component only runs a search when a term can be extracted,\nand only renders results when there is at least one hit. On a 404\nwith no extractable term or no matching results, the page asserted\na search it never backed up.\n\nGate the sentence behind the same searchResults() check that guards\nthe results list, so the empty state simply reads \"We couldn't find\nwhat you were looking for.\"", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "neovim", "sha": "277cb1b431b3d1a6fc18132dd3d879747914379f", "date": "2026-07-18T15:09:55-04:00", "subject": "build(vim-patch): \"vim_\" char/string N/A funcs", "body": "Note this doesn't find new N/A patches by itself\nand this is to align with ':h dev-vimpatch'.\n\nNeovim refactored Vim's \"char/string\" functions to be either safer\n(ie. xstrlcat) or target Unicode.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "neovim", "sha": "37da2446859d27cdd6f8454c591eefd9accc318f", "date": "2026-07-18T15:19:34-04:00", "subject": "backport: test: unreliable \"put command . register special … ring the bell\"", "body": "test: unreliable \"put command . register special … ring the bell\"\n\nProblem:\n\n FAILED …/put_spec.lua:894: …/put_spec.lua @ 898: put command . register special tests should ring the bell when deleting if not appropriate\n …/put_spec.lua:894: …/put_spec.lua:898: Bell was not rung after action\n\n Snapshot:\n screen:expect([[\n ^ine of words 1 |\n Line of words 2 |\n {1:~ }|*4\n {3:[No Name] [+] }|\n |\n {1:~ }|*4\n {UNKNOWN_HL_ID(4):[No Name] }|\n {UNKNOWN_HL_ID(1): }|\n ]])\n stack traceback:\n …/ui/screen.lua:917: in function '_wait'\n …/ui/screen.lua:540: in function 'expect'\n …/editor/put_spec.lua:894: in function 'bell_test'\n …/editor/put_spec.lua:940: in function <…/editor/put_spec.lua:932>\n\nAnalysis:\nThe bell was asserted via the UI 'bell' event (screen.bell), but\nvim_beep() rate-limits that to 3 per 500ms. These tests share one\nlong-lived session, so beeps from earlier tests could exhaust the window\nand drop the asserted beep.\n\nSolution:\nUse assert_beeps()/assert_nobeep(), which check called_vim_beep (set\nbefore the rate-limit gate).", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "6cb9334aa0ab291c6f04a0058305a212b1c7207b", "date": "2026-07-18T19:46:25Z", "subject": "backport: fix(path): avoid shell fallback for unknown users", "body": "fix(path): avoid shell fallback for unknown users\n\nProblem:\nExpanding an unknown ~user path falls back to the shell. In a fast event,\nthis re-enters the event loop and aborts Nvim.\n\nSolution:\nLeave unknown users unexpanded when the system account lookup fails.\n\n(cherry picked from commit 9d3edf345f8e25f1a7976efff54818a426661802)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9536d35099f48bc5eeda63fe0be8bd9c8bd6b4ad", "date": "2026-07-18T16:05:38-04:00", "subject": "build(vim-patch): ignore amount of whitespace change", "body": "Patch 9.1.1363 is N/A after ignoring whitespace.\n\n-----\n\nUnrelated N/A patches\nvim-patch:8.1.1985: code for dealing with paths is spread out\n\nhttps://github.com/neovim/neovim/pull/36125#issuecomment-5012670153", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "eslint", "sha": "0174428dd7543d118cf67a823b98dd97032fe68c", "date": "2026-07-19T05:10:18+09:00", "subject": "chore: add missing backticks to `translate-cli-options.js`", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f7bc98156e1170c60a03dfbeee7871e59888de82", "date": "2026-07-18T20:43:24-04:00", "subject": "doc PG 19 relnotes: \\crosstab to \\crosstabview", "body": "Backpatch-through: 19 only", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "84de9caccdda655e0308874aa12de34fde44b88e", "date": "2026-07-19T10:10:14+09:00", "subject": "Restore empty raw PKCS#1 verify-recover behavior", "body": "EVP_PKEY_verify_recover() with RSA PKCS#1 v1.5 padding and no configured\nsignature digest rejected a valid signature whose recovered payload is\nempty, instead of returning success with a recovered length of zero.\n\nRSA_public_decrypt() returns -1 on error and otherwise the number of\nrecovered bytes, which may legitimately be zero for a raw PKCS#1 v1.5\nsignature that encodes an empty payload. Pull request #28306 (\"Make\nerror checks on RSA_public_decrypt() consistent\") changed the error\ncheck in the raw, no-digest verify-recover path from \"ret < 0\" to\n\"ret <= 0\", so a valid zero recovered length was mistaken for an error.\n\nRestore the \"< 0\" check in that branch only, leaving the digest-aware\nPKCS#1, X9.31, PSS and ordinary verification paths untouched.\n\nAdd test_RSA_verify_recover_empty_payload, which performs a complete\nzero-length raw PKCS#1 v1.5 sign-and-recover round trip through EVP and\nfails on the unmodified implementation specifically at the\nEVP_PKEY_verify_recover() call. OpenSSL 3.0.13 recovers the same\nsignature and reports a zero-byte result.\n\nThe behavior regressed as a result of #28306.\n\nFixes #32000\n\nAssisted-by: OpenCode:GLM-5.2\n\nReviewed-by: Paul Dale \nReviewed-by: Jakub Zelenka \nMergeDate: Fri Jul 24 13:06:58 2026\n(Merged from https://github.com/openssl/openssl/pull/32001)\n\n(cherry picked from commit 5f48317d6d847e555bf788f8b71d51ac032a6344)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5f48317d6d847e555bf788f8b71d51ac032a6344", "date": "2026-07-19T10:10:14+09:00", "subject": "Restore empty raw PKCS#1 verify-recover behavior", "body": "EVP_PKEY_verify_recover() with RSA PKCS#1 v1.5 padding and no configured\nsignature digest rejected a valid signature whose recovered payload is\nempty, instead of returning success with a recovered length of zero.\n\nRSA_public_decrypt() returns -1 on error and otherwise the number of\nrecovered bytes, which may legitimately be zero for a raw PKCS#1 v1.5\nsignature that encodes an empty payload. Pull request #28306 (\"Make\nerror checks on RSA_public_decrypt() consistent\") changed the error\ncheck in the raw, no-digest verify-recover path from \"ret < 0\" to\n\"ret <= 0\", so a valid zero recovered length was mistaken for an error.\n\nRestore the \"< 0\" check in that branch only, leaving the digest-aware\nPKCS#1, X9.31, PSS and ordinary verification paths untouched.\n\nAdd test_RSA_verify_recover_empty_payload, which performs a complete\nzero-length raw PKCS#1 v1.5 sign-and-recover round trip through EVP and\nfails on the unmodified implementation specifically at the\nEVP_PKEY_verify_recover() call. OpenSSL 3.0.13 recovers the same\nsignature and reports a zero-byte result.\n\nThe behavior regressed as a result of #28306.\n\nFixes #32000\n\nAssisted-by: OpenCode:GLM-5.2\n\nReviewed-by: Paul Dale \nReviewed-by: Jakub Zelenka \nMergeDate: Fri Jul 24 13:06:58 2026\n(Merged from https://github.com/openssl/openssl/pull/32001)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "829bacaeb323e4811eb804b5e4e29f643a0c619d", "date": "2026-07-19T10:47:29+09:00", "subject": "Remove redundant null-treatment check in window function dedup.", "body": "Commit 25a30bbd423 (IGNORE NULLS / RESPECT NULLS for window functions)\nmade ExecInitWindowAgg() treat two otherwise-equal window functions as\nduplicates only when their ignore_nulls settings also matched:\n\n if (i <= wfuncno && wfunc->ignore_nulls == perfunc[i].ignore_nulls)\n\nThat extra term reads WindowStatePerFuncData.ignore_nulls, but the field\nwas never populated when a per-function entry was filled in, so it stayed\nzero from palloc0_array(). Consequently a duplicate call carrying\nIGNORE NULLS or an explicit RESPECT NULLS never matched an identical\nearlier entry and was needlessly given its own per-function slot and\nevaluated twice. (Results stayed correct; this was a missed sharing, not\na wrong answer.)\n\nThe extra term is in fact redundant. WindowFunc.ignore_nulls is a plain\nscalar field with no pg_node_attr, so _equalWindowFunc() already compares\nit; the preceding equal() call therefore never matches two WindowFuncs\nthat differ only in null treatment. If equal() matches, ignore_nulls\nnecessarily matched too, so the term can never change the outcome, and\nWindowStatePerFuncData.ignore_nulls existed only to feed it.\n\nRather than populate the shadow field, drop the redundant term and the\nfield (and adjust the now-stale comment) and let equal() do the work.\nThat fixes the same bug while removing the hand-maintained duplicate\nstate that caused it, so it cannot silently drift again.\n\nAuthor: Chao Li \nCo-authored-by: Ewan Young \nReviewed-by: Tatsuo Ishii \nDiscussion: https://postgr.es/m/5D2C9081-5DFE-4E27-AB14-7358238EA1BC%40gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "e44ce90727eb756bba516b79e105b8dc2ee2de7c", "date": "2026-07-19T12:24:50+08:00", "subject": "refactor(type): improve type inference for semantic functions", "body": "", "style": "conventional", "type": "refactor", "scope": "type", "breaking": false} {"repo": "babel", "sha": "57585c01be14c0e4ce3b330ff4b24239763d6d51", "date": "2026-07-19T14:49:43+08:00", "subject": "fix: Avoid bundling `package/sub`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "eslint", "sha": "d28f697e0d32d7446c825fd89c362db028450134", "date": "2026-07-19T15:53:17+09:00", "subject": "docs: replace Code Climate CLI links with Qlty CLI links", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "eslint", "sha": "c47e8dc9bbfba797430db45ae08cb6f7392cfc9d", "date": "2026-07-19T16:38:32+09:00", "subject": "chore: add missing backticks to `languages/js/index.js`", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "eslint", "sha": "68eb4a57572409dbdcfeb3bc5c92ddbba46bc770", "date": "2026-07-19T17:19:22+09:00", "subject": "docs: fix broken Specify Globals anchor links in rule pages", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "94e530fd748572a80afc829f82cc61a5958fa1a9", "date": "2026-07-19T16:48:04+08:00", "subject": "refactor(demo): migrate styles from markdown to CSS-in-JS", "body": "* refactor(demo): migrate styles from markdown to CSS-in-JS\n\n* update\n\n* update\n\n* update\n\n* update", "style": "conventional", "type": "refactor", "scope": "demo", "breaking": false} {"repo": "eslint", "sha": "a99906ffb8342dde03336b5c6372b5658f21d5f2", "date": "2026-07-19T17:52:54+09:00", "subject": "test: Add error location details to `no-delete-var` rule", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "eslint", "sha": "13ab0ec447650c079d8a5e5d67222f79e69c741e", "date": "2026-07-19T18:05:04+09:00", "subject": "test: add error locations to `no-label-var`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "eslint", "sha": "8dde64570e240f3ef8af873d59a752db6d8519aa", "date": "2026-07-19T18:20:01+09:00", "subject": "test: add error locations to `no-ex-assign`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "503ee34d91d5a73d60c160661a2ff2d42565f49d", "date": "2026-07-19T18:58:17+08:00", "subject": "fix(demo): preserve search text in remote users demo", "body": "", "style": "conventional", "type": "fix", "scope": "demo", "breaking": false} {"repo": "axios", "sha": "878bb29de570765b0d4c0970e30400d9ea9399f7", "date": "2026-07-19T16:07:40+05:00", "subject": "fix(sandbox): resolve TypeError on constant variable path assignment", "body": "", "style": "conventional", "type": "fix", "scope": "sandbox", "breaking": false} {"repo": "ant-design", "sha": "de885b2e53a7e4b4f440401f73083b606a321dd6", "date": "2026-07-19T20:08:21+08:00", "subject": "chore(demo): set transition to none when type=\"editable-card\"", "body": "* chore(demo): set transition to none when type=editable-card\n\n* update", "style": "conventional", "type": "chore", "scope": "demo", "breaking": false} {"repo": "eslint", "sha": "92bb13f13065db72d62454b9830d3ae52db15c76", "date": "2026-07-19T21:19:34+09:00", "subject": "docs: replace quake link", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "eslint", "sha": "e23fafe8d4b15355adef6cdebef414c3a2019454", "date": "2026-07-19T18:09:00+05:30", "subject": "fix: prefer-object-spread add semicolon when adding parenthesis", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "axios", "sha": "c44f8d0a910df99486da9175584b99f56a94a73b", "date": "2026-07-19T07:53:02-05:00", "subject": "ci: use bundled npm for v1 publish", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "axios", "sha": "a930f390d8ecd262bb91c25bf22840f6ed6ec204", "date": "2026-07-19T08:22:26-05:00", "subject": "ci: use bundled npm for v0 publish", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "d302c7ab833c0bd3bba951135e76e0c48273b3d7", "date": "2026-07-19T15:41:08+02:00", "subject": "fix(forms): ensure `pending` status propagates to the root form in signal forms", "body": "Previously, the `pending()` status on a field's `ValidationState` only checked if the field itself or its immediate children had a pending asynchronous validator by directly inspecting `asyncErrors()`. This meant that a pending asynchronous validator deep within a nested form (e.g. on a grand-child) would not correctly bubble the `pending` state up to the root form.\n\nfixes #69840\n\n(cherry picked from commit 3497c9b943748d0f21587fcdc1482215f857fa28)", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "3497c9b943748d0f21587fcdc1482215f857fa28", "date": "2026-07-19T15:41:08+02:00", "subject": "fix(forms): ensure `pending` status propagates to the root form in signal forms", "body": "Previously, the `pending()` status on a field's `ValidationState` only checked if the field itself or its immediate children had a pending asynchronous validator by directly inspecting `asyncErrors()`. This meant that a pending asynchronous validator deep within a nested form (e.g. on a grand-child) would not correctly bubble the `pending` state up to the root form.\n\nfixes #69840", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "neovim", "sha": "239125b8c8ae699beaf0753fd4042fdeb2da0ff5", "date": "2026-07-19T10:03:53-04:00", "subject": "feat(eval): declare more \"fast\" functions #40834", "body": "", "style": "conventional", "type": "feat", "scope": "eval", "breaking": false} {"repo": "neovim", "sha": "851656c6288362bbce702319137db60bc6a01200", "date": "2026-07-19T10:07:18-04:00", "subject": "fix(coverity): STRING_OVERFLOW #40835", "body": "CID 651340: Security best practices violations (STRING_OVERFLOW)\n /src/nvim/keycodes.c: 383 in get_special_key()\n 377 data->key = *s;\n 378 data->key_alt = (String){ NULL, 0 };\n 379 }\n 380 }\n 381\n 382 if ((int)s->size + idx + 2 <= MAX_KEY_NAME_LEN) {\n >>> CID 651340: Security best practices violations (STRING_OVERFLOW)\n >>> You might overrun the 33-character fixed-size string \"string + idx\" by copying \"s->data\" without checking the length.\n 383 STRCPY(string + idx, s->data);\n 384 idx += (int)s->size;\n 385 }\n 386 }\n 387 string[idx++] = '>';\n 388 string[idx] = NUL;", "style": "conventional", "type": "fix", "scope": "coverity", "breaking": false} {"repo": "neovim", "sha": "746f5682a7d7c8e5cb122da5d68eacfd55d58887", "date": "2026-07-19T15:03:46Z", "subject": "fix(ui): combine float border with window background #40796", "body": "Problem:\nFloat border highlight groups (FloatBorder, FloatTitle and FloatFooter)\nfall back to Normal background highlight if no background color set.\n\nSolution:\nCombine border colors with window-local Normal highlight.\n\nFix #38330", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "ant-design", "sha": "01cef0398bbc47e0162f559342cb6d44fae1e6e5", "date": "2026-07-19T23:34:03+08:00", "subject": "refactor(demo): replace hardcoded styles with design tokens", "body": "* refactor(demo): replace hardcoded styles with design tokens\n\n* update\n\n* update snap", "style": "conventional", "type": "refactor", "scope": "demo", "breaking": false} {"repo": "neovim", "sha": "d1b3a9924d6d0e77cdd1196efbbf2b57e9946ea0", "date": "2026-07-19T11:38:14-04:00", "subject": "feat(api): mark nvim_create_autocmd as api-fast #40836", "body": "Problem:\nnvim_create_autocmd() isn't |api-fast|, so modules that create autocmds\n(e.g. vim.treesitter.query) can't be require()d in a fast event context.\n\nSolution:\nMark it |api-fast|. Compile autocmd patterns with RE_NOBREAK so\naucmd_next() won't os_breakcheck() mid-iteration, where a fast\nnvim_create_autocmd() could realloc the autocmds vector and dangle the\ncaller's AutoPat/AutoCmd.\n\nRE_NOBREAK is low-risk because:\n- aucmd_next()'s loop checks CTRL-C: `(for (… i < apc->ausize && !got_int; …)`.\n- `line_breakcheck()` (`autocmd.c:1912`) runs once per matched autocmd.\n- Each autocmd _execution_ runs through `do_cmdline`, which has its own\n breakchecks.\n\nHowever this does admit risk of a pathological case:\na catastrophic-backtracking glob matched against a very long `User`\nevent-pattern.", "style": "conventional", "type": "feat", "scope": "api", "breaking": false} {"repo": "eslint", "sha": "7ec733a3ba5acb053a73ac656030a3879aa48fda", "date": "2026-07-20T00:42:31+09:00", "subject": "docs: Fix typos and grammar in glossary", "body": "* docs: Fix typos and grammar in glossary\n\n* docs: Include TypeScript config file extensions in glossary", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "eslint", "sha": "a746ec6ea4a2249b1eb3f512264da5f4d5f7c886", "date": "2026-07-20T01:02:57+09:00", "subject": "test: add error locations to `no-new-wrappers`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "eslint", "sha": "9715887ec94a5ff936447d7b680d039fbe2f0541", "date": "2026-07-20T01:05:27+09:00", "subject": "test: Add error location details to `no-div-regex`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "40ab941f7948a2b7cc9bf053c9e792dcd7742bfa", "date": "2026-07-19T12:45:40-04:00", "subject": "test: improve Nvim EOF handling (for Windows CI) #40841", "body": "Problem:\nMysterious Windows CI failure points to resource exhaustion, but we have\nno visibility/instrumentation:\n\n RUN T610 nvim_set_keymap, nvim_del_keymap can set mappings with special characters, lhs: , rhs: : 93.70 ms OK\n RUN T611 nvim_set_keymap, nvim_del_keymap can set mappings with special characters, lhs: , rhs: : 49.48 ms FAIL\n …\\testnvim.lua:144: EOF was received from Nvim. Likely the Nvim process crashed.\n stack traceback:\n D:/a/neovim/neovim/test/functional/testnvim.lua:144: in function 'nvim_set_keymap'\n …/api/keymap_spec.lua:769: in function <…/api/keymap_spec.lua:768>\n RUN T612 nvim_set_keymap, nvim_del_keymap can set mappings with special characters, lhs: , rhs: : 18.49 ms FAIL\n …\\testnvim.lua:144: EOF was received from Nvim. Likely the Nvim process crashed.\n stack traceback:\n D:/a/neovim/neovim/test/functional/testnvim.lua:144: in function 'nvim_set_keymap'\n …/api/keymap_spec.lua:769: in function <.../build/Xtest_xdg_api/test/functional/api/keymap_spec.lua:768>\n\nSolution:\nAppend the child exit-code and signal to the \"EOF … crashed\" message.\n\nExample:\n\n Nvim EOF (crash?) exit code: 42 (0x0000002A)\n\nOn the next failing Windows run, the first crash line will classify the failure:\n- 0xC0000005 → access violation (nvim bug)\n- 0xC0000374 → heap corruption (nvim bug)\n- spawn/resource error or \"clean\" exit-code → system resource exhaustion", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "61abef356441c0914b5e6a0392c69f2b48d521fe", "date": "2026-07-19T12:52:49-04:00", "subject": "ci: combine PR-labeler jobs #40842", "body": "Problem:\nWe have 5 \"labeler\" jobs, this is kind of nuts, it bloats the CI report,\nand also introduces race conditions...\n\nSolution:\nMerge 3 jobs into 1 `label` job with multiple sequential steps.\n- Eliminates the `needs` chain\n- Drops unused `contents:write` permission.\n- GITHUB_TOKEN is scoped to the single `gh` step.\n - Note: is the automatic job token (`contents:read` + `pull-requests:write`,\n not a PAT), so worst case is PR/label mischief.\n\nThe `ai-assisted` and `request-reviewer` jobs stay separate, bc they are\ntriggered on different events (not only \"opened\").", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "tmux", "sha": "6ef280fb0f138b3830fbfedbdff4736b2da07707", "date": "2026-07-19T17:25:38Z", "subject": "Guard x as well as y to avoid overflow, GitHub issue 5409 from Park Seongjae.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "7ba955fe079d4aa2554fea8e7235651fafd40efb", "date": "2026-07-19T19:29:57+02:00", "subject": "fix(statusline): prevent integer overflow with huge item groups #40837", "body": "Problem: when PR #40621 removed a 50 cell restriction to item group\nminwid, a possible integer overflow was introduced when the added\npadding is multiplied with the number of fillchar bytes.\n\nSolution: when there is not enough buffer capacity left, reduce the\nnumber of added fillchars before calculating the number of added bytes.", "style": "conventional", "type": "fix", "scope": "statusline", "breaking": false} {"repo": "tmux", "sha": "2a2bce3dd413274894c42caaaa59bd817a0c1ed2", "date": "2026-07-19T18:34:18+01:00", "subject": "Fix tty-keys.sh test.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "c766669b3b01d24b2fd60e6a2c4b316910abff77", "date": "2026-07-19T17:36:38Z", "subject": "Add unzoomed width and height formats and use them for the sizes shown in panes mode, reported by tb@.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "3dbe73a9ef9fff2cae616ff67b8fee488445b88d", "date": "2026-07-19T20:38:46+02:00", "subject": "feat(benchmark): add pluggable benchmark provider API", "body": "Introduce a `benchmark.provider` config option that lets a plugin own\nbenchmark execution end to end. A provider receives the expanded\nregistrations of a single test and returns the results Vitest reports;\nthe returned array is the sole source of results, matched to\nregistrations by name.\n\nThe built-in tinybench execution moves behind a default provider,\nselected when `benchmark.provider` is `'default'`. A custom provider is\na module path whose default export exposes `getProvider()`; it is\nresolved to an absolute path during config resolution and imported once\nper worker. `createBench` now runs every path — `bench().run()`,\n`bench.compare()`, `writeResult` and comparison storage — through the\nprovider instead of reading results back off the tinybench instance.\n\nThe benchmark name used for reporting and the getter-overuse warning now\ncomes from `options.name` when set, falling back to the full test name,\ndropping the previous per-run index suffix.", "style": "conventional", "type": "feat", "scope": "benchmark", "breaking": false} {"repo": "tmux", "sha": "8024fab1f2365b86a52f7cc2881790033d33344c", "date": "2026-07-19T19:09:30Z", "subject": "Copy the home directory if it comes from getpwuid.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "689b45dda32d3cdb1827ed3c3e50d66bfb02a627", "date": "2026-07-20T03:42:59+08:00", "subject": "fix(write): show E17 on all platforms #40595", "body": "Problems:\nOn Windows, `:w somedir` (where somedir is an existing directory\nin pwd) shows the misleading `E13: File exists (add ! to override)`\ninstead of `E17: \"somedir\" is a directory`.\n\nSolution:\ndrop unnecessary `#ifdef` guard.", "style": "conventional", "type": "fix", "scope": "write", "breaking": false} {"repo": "tmux", "sha": "aea209b1423d2cd6d6b2cf958c517ad1c13ff9f1", "date": "2026-07-19T19:53:11Z", "subject": "Remove bits to disable removing empty lines since actually we want to do that - zoom added them in the first place.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "c750c014b2054c059d20024d8fa2ad474c0caa15", "date": "2026-07-19T20:58:20+01:00", "subject": "Extend test.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f87375bfba5cee2dcea4c19879ee990463be9636", "date": "2026-07-19T23:27:52+02:00", "subject": "Fix parsing of underscores in pg_plan_advice occurrence numbers", "body": "The pg_plan_advice scanner recognizes underscores as digit separators\njust like the core parser, but used strtoint() to convert occurrence\nnumbers which does not support underscores. Consequently, advice such\nas SEQ_SCAN(x#1_0) failed to parse. Fix by using pg_strtoint32_safe()\nlike the core scanner, and also add regression test coverage.\n\nThis bug was independently found and reported by Lukas Fittl and Chao\nLi. Backpatch down to v19 where pg_plan_advice was introduced.\n\nAuthor: Chao Li \nCo-authored-by: Daniel Gustafsson \nReported-by: Lukas Fittl \nReported-by: Chao Li \nReviewed-by: Daniel Gustafsson \nReviewed-by: Lukas Fittl \nDiscussion: https://postgr.es/m/22E2ECE0-B768-43D5-8575-61C3EBC2E4E8@gmail.com\nDiscussion: https://postgr.es/m/CAP53PkzKeD=t90OfeMsniYrcRe2THQbUx3g6wV17Y=ZtiwmWTQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "01c4d37959227ccb7a7fce1a8d63616ef2ed7479", "date": "2026-07-19T20:16:56-04:00", "subject": "fix(lsp): preview window converted to normal window is auto-closed", "body": "Problem:\nA floating preview window (hover, signature help) converted to a normal\nwindow (e.g. with CTRL-W_H or nvim_win_set_config()) is still closed by\nthe open_floating_preview() autocommands, and a later preview may focus\nor close it as if it were still a float.\n\nSolution:\nWhen the preview window is no longer floating, remove its auto-close\nautocommands and stop tracking it as the buffer's floating preview.\n\nCloses #36659", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "vue-core", "sha": "411ec5c35225016169a67854d8723e8b8ade95ad", "date": "2026-07-20T09:21:40+08:00", "subject": "docs: remove outdated agent guidance", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "458bfef110dc0f1d2bbe4eb93840f32c51b9d50d", "date": "2026-07-20T09:05:59+07:00", "subject": "Fix clean target find/-prune precedence bug", "body": "The clean target's find command intended -prune to apply to the whole\n-o chain of excluded submodule paths, but -prune only binds to the\nimmediately preceding -path, so only the last path (wycheproof) was\nactually protected from recursion. Earlier\nsubmodules (cloudflare-quiche, pkcs11-provider, etc.) were still\ndescended into, letting make clean delete vendored symlinks inside\nthem. Move -prune outside the parenthesized path list so it applies\nwhenever any excluded path matches.\n\nAssisted-by: Claude:claude-sonnet-5\nFixes: 70741ee62a50 \"Configurations/unix-Makefile.tmpl: make cleanup faster\"\n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Bob Beck \nReviewed-by: Milan Broz \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Tue Jul 21 09:21:37 2026\n(Merged from https://github.com/openssl/openssl/pull/32006)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b087a88c40e45453682bf6d24c42eff33f401662", "date": "2026-07-19T22:15:37-04:00", "subject": "Run nbtree test module tests under autoconf builds", "body": "Commit 1e4e5783e added the src/test/modules/nbtree test module, but only\nregistered it in the meson build, not in the module list in\nsrc/test/modules/Makefile. As a result, autoconf builds never ran the\nmodule's tests.\n\nTo fix, add the module to the Makefile's lists of\ninjection-point-dependent modules.\n\nOversight in commit 1e4e5783e.\n\nAuthor: Peter Geoghegan \nReviewed-by: Michael Paquiër \nDiscussion: https://postgr.es/m/CAH2-Wz=JchiD5ksiT35p8Ar02gaNv8_y6w2wBAST+Zzen-eNjw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "46c1ea141468f1b87cdabee6c10231948663075b", "date": "2026-07-20T10:24:57+08:00", "subject": "fix(Typography): respect border design tokens in reset styles", "body": "", "style": "conventional", "type": "fix", "scope": "Typography", "breaking": false} {"repo": "ant-design", "sha": "3b34326d7b35b0575674c85a53f33e2d7a3fe7bd", "date": "2026-07-20T10:25:27+08:00", "subject": "fix(Tree): respect border design tokens in tree styles", "body": "", "style": "conventional", "type": "fix", "scope": "Tree", "breaking": false} {"repo": "ant-design", "sha": "f3bfef9c3ef0e919d3eafb421751c8e13def5167", "date": "2026-07-20T10:25:42+08:00", "subject": "fix(Collapse): respect border design tokens in borderless styles", "body": "", "style": "conventional", "type": "fix", "scope": "Collapse", "breaking": false} {"repo": "ant-design", "sha": "80b230f5c92e9dfa1d84aef91dc43a26d4a3db41", "date": "2026-07-20T10:26:06+08:00", "subject": "fix(Layout): respect border design tokens in sider styles", "body": "", "style": "conventional", "type": "fix", "scope": "Layout", "breaking": false} {"repo": "ant-design", "sha": "78c3d8461978044558287f83aff911d9c4973eb9", "date": "2026-07-20T10:26:47+08:00", "subject": "fix(Input.Search): sync custom button disabled and loading states", "body": "", "style": "conventional", "type": "fix", "scope": "Input.Search", "breaking": false} {"repo": "openssl", "sha": "f8da3aa370b1c267a2edaa335e74e0073bd658ce", "date": "2026-07-19T22:44:18-04:00", "subject": "Update SSL_CTX_set1_curves.pod for OpenSSL 3.5 additions", "body": "Added documentation of the addition of group tuples in openSSL 3.5\nin HISTORY section of SSL_CTX_set1_curves.pod.\n\nCLA: trivial\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Nikola Pajkovsky \nMergeDate: Fri Jul 24 13:13:26 2026\n(Merged from https://github.com/openssl/openssl/pull/32007)\n\n(cherry picked from commit 971b8d060e52499d6ffd2f9ca697fe23f72a629a)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "971b8d060e52499d6ffd2f9ca697fe23f72a629a", "date": "2026-07-19T22:44:18-04:00", "subject": "Update SSL_CTX_set1_curves.pod for OpenSSL 3.5 additions", "body": "Added documentation of the addition of group tuples in openSSL 3.5\nin HISTORY section of SSL_CTX_set1_curves.pod.\n\nCLA: trivial\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Nikola Pajkovsky \nMergeDate: Fri Jul 24 13:13:26 2026\n(Merged from https://github.com/openssl/openssl/pull/32007)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "ce842bc14024ecdb1191ea43e9551358a60fa75b", "date": "2026-07-20T11:02:50+08:00", "subject": "fix(Tree): respect margin-inline-start design tokens", "body": "", "style": "conventional", "type": "fix", "scope": "Tree", "breaking": false} {"repo": "postgres", "sha": "d610d8e8be5678f828e07aa99c16c1dcc2e4ef63", "date": "2026-07-20T12:13:11+09:00", "subject": "Skip unnecessary get_relids_in_jointree() when there are no PHVs", "body": "Commit 1df9e8d96 made remove_useless_result_rtes() compute the set of\nbaserels in the jointree, to pass down to the find_dependent_phvs()\nchecks. But those checks are no-ops when the query contains no PHVs,\nsince find_dependent_phvs() and find_dependent_phvs_in_jointree() both\nreturn early in that case. So we can avoid the\nget_relids_in_jointree() scan altogether when root->glob->lastPHId is\nzero, leaving baserels as NULL.\n\nAuthor: Richard Guo \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/CAMbWs49H275KzgZr3Cd1Hy+6Lmwp35bZ+5PrVc62k3HDLj6hNQ@mail.gmail.com\nBackpatch-through: 16", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "a686662492f00c4cb2552cf9ffcc39fc13e0730f", "date": "2026-07-20T11:21:11+08:00", "subject": "fix(runtime-vapor): validate props for vdom interop", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "eslint", "sha": "5abd878740fe417fd8a910cac3741f2a0317b365", "date": "2026-07-20T12:27:25+09:00", "subject": "test: add error locations to `no-proto`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "45cf7b1e5bf923ca48dfd9aa5001bdd0630d11c3", "date": "2026-07-20T11:11:45+05:30", "subject": "Handle concurrent sequence refreshes.", "body": "'ALTER SUBSCRIPTION ... REFRESH SEQUENCES' can race with a running\nsequence synchronization worker. If the worker has fetched a sequence's\nvalue from the publisher but not yet marked it READY, a concurrent refresh\nthat resets the sequence to INIT can be overwritten by the worker's stale\nvalue, silently losing the refresh request.\n\nHandle this by stopping any running sequence sync worker before resetting\nthe sequences to INIT. This is race-free because AlterSubscription()\nalready holds AccessExclusiveLock on the subscription object. That lock\nblocks a running worker's UpdateSubscriptionRelState(), which takes\nAccessShareLock on the object, and also any worker the apply worker\nre-launches, because a new worker takes AccessShareLock on the object in\nInitializeLogRepWorker() before it reads pg_subscription_rel. Such a\nworker cannot act on the sequence states until the refresh commits, by\nwhich time they are reset to INIT and it will synchronize the latest\npublisher values.\n\nReported-by: Noah Misch \nAuthor: Amit Kapila \nReviewed-by: vignesh C \nReviewed-by: Shveta Malik \nReviewed-by: Hayato Kuroda \nBackpatch-through: 19\nDiscussion: https://postgr.es/m/20260710045217.f0.noahmisch@microsoft.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "b707ef3cc968648251744618ab21b09af0bf4ee3", "date": "2026-07-20T14:13:13+08:00", "subject": "docs(Alert): fix spacing and typos", "body": "* docs(alert): fix wording in demos\n\n* docs(alert): fix spacing\n\n* docs(alert): fix typo\n\n* docs(Alert): improve wording", "style": "conventional", "type": "docs", "scope": "Alert", "breaking": false} {"repo": "postgres", "sha": "20fe90bbb93df755e00dc88ace5988fda0b1d7d3", "date": "2026-07-20T08:38:08+02:00", "subject": "Fix RLS checks for FOR PORTION OF leftover rows", "body": "UPDATE/DELETE FOR PORTION OF may insert leftover rows to preserve the\nparts of the old row that are outside the target range. Those inserts\ngo through ExecInsert(), which checks RLS policies using\nWCO_RLS_INSERT_CHECK.\n\nHowever, the rewriter only added RLS WITH CHECK options for the\noriginal statement command. For UPDATE, that meant only\nWCO_RLS_UPDATE_CHECK options were available, so ExecInsert() skipped\nthem. For DELETE, no RLS WITH CHECK options were added at all. As a\nresult, leftover rows could be inserted even when they violated INSERT\nRLS policies.\n\nFix this by adding INSERT RLS WITH CHECK options for UPDATE/DELETE FOR\nPORTION OF target relations. Also add regression coverage for both\nUPDATE and DELETE, including cases where allowed leftovers still\nsucceed and disallowed leftovers are rejected.\n\nAuthor: Chao Li \nCo-authored-by: Paul A Jungwirth \nReviewed-by: Paul A Jungwirth \nReviewed-by: Ayush Tiwari \nReviewed-by: Dean Rasheed \nDiscussion: https://www.postgresql.org/message-id/flat/6C34A987-AC50-4477-BD71-2D4AFEE1A589%40gmail.com\nDiscussion: https://www.postgresql.org/message-id/flat/CAJTYsWWdeBkoH5g8D-k9LDw9ciqsMxb21EJSiFXAzP4J%3DXyxOQ%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "c475e4225941330e1ce14bc26ca857fe3beec5a5", "date": "2026-07-20T09:48:46+03:00", "subject": "fix(Tag): align and space bare `` icons", "body": "", "style": "conventional", "type": "fix", "scope": "Tag", "breaking": false} {"repo": "postgres", "sha": "16ebc196d8926823bc8d64931557b679cecd67aa", "date": "2026-07-20T15:52:18+09:00", "subject": "Improve pgstat_get_entry_ref_cached() behavior on out-of-memory errors", "body": "A failure in allocating a new cache entry in the backend-level hash\ntable holding references to shared stats entries would leave the table\nin an inconsistent state, crash or FATAL at session exit, depending on\nif there are pending stats.\n\nRather than leaving things in an inconsistent state on OOM, the code is\nswitched to use MemoryContextAllocExtended(MCXT_ALLOC_NO_OOM), so as an\nallocation failure leads to a cleanup of the hash table before issuing\nthe allocation error.\n\nThe problem is unlikely going to show up in practice, so no backpatch is\ndone. Note that shared memory is not impacted, only a backend-level\nhash table.\n\nReported-by: Alex Masterov \nDiscussion: https://postgr.es/m/CA+8z=zumV9sscgK=j1Es+-564maVoO9CMDdB9CsW9=FCziCj3w@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0c0c96c2a37f57c92129a8aa7fc46b7a4703983a", "date": "2026-07-20T10:30:18+03:30", "subject": "EVP_CTRL_AEAD_SET_IV_FIXED: reject arg values below -1", "body": "EVP_CIPHER_CTX_ctrl() casts the signed int arg to size_t before\ndispatching. For EVP_CTRL_AEAD_SET_IV_FIXED the value -1 is an\nintentional sentinel (\"copy full IV\"), so it must be preserved, but\nany value below -1 wraps to a near-SIZE_MAX size_t. The two guards\ninside gcm_tls_iv_set_fixed() both fail to catch this:\n\n * the sentinel check tests for SIZE_MAX (i.e. (size_t)-1), not\n SIZE_MAX-1 or lower;\n * the bounds check back-casts len to int, undoing the wrap and\n making the comparison evaluate to false.\n\nThe result is memcpy(ctx->iv, iv, SIZE_MAX-1) — a heap overflow.\n\nFix the root cause in EVP_CIPHER_CTX_ctrl() by rejecting any arg\nbelow -1 for EVP_CTRL_AEAD_SET_IV_FIXED, consistent with the guards\nalready present for EVP_CTRL_SET_KEY_LENGTH, EVP_CTRL_AEAD_SET_IVLEN,\nEVP_CTRL_GCM_SET_IV_INV and others.\n\nAs a defense-in-depth measure, also replace the (int) back-cast in\ngcm_tls_iv_set_fixed() with a safe size_t upper-bound check so that\nthe provider itself rejects any out-of-range length independently of\nthe caller.\n\nCo-authored-by: Tomáš Mráz (comment fix)\nReviewed-by: Jakub Zelenka \nReviewed-by: Paul Dale \nReviewed-by: Richard Levitte \nMerge-date: Tue Aug 18 07:41:57 2026\nMerged-from: https://github.com/openssl/openssl/pull/31879", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "e9cb7c86109f860c873470751a7cc653f9f31f43", "date": "2026-07-20T15:20:43+08:00", "subject": "fix(compiler-vapor): avoid helper and cache variable name collisions", "body": "", "style": "conventional", "type": "fix", "scope": "compiler-vapor", "breaking": false} {"repo": "vue-core", "sha": "1b423b5b23f6ecf99a7d0666ab28748bd018870c", "date": "2026-07-20T15:34:22+08:00", "subject": "refactor(runtime-vapor): reduce TransitionGroup bundle size", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "tmux", "sha": "efcfc555356b9d02c4ee6899554831e316e900d9", "date": "2026-07-20T07:42:13Z", "subject": "Correctly include status line in pane height.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ee72a8e124123e759de95684684ac6f66c2391ef", "date": "2026-07-20T09:44:53+02:00", "subject": "Add ossl_list_TYPE_join(head, tail) function", "body": "The function appends list tail to list head. List tail becomes\nempty after the function returns.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMergeDate: Thu Aug 6 11:59:15 2026\n(Merged from https://github.com/openssl/openssl/pull/32031)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2a2c5643eb9a5af91c2b826a817e2422292d29fd", "date": "2026-07-20T09:44:53+02:00", "subject": "Add ossl_list_TYPE_join(head, tail) function", "body": "The function appends list tail to list head. List tail becomes\nempty after the function returns.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMergeDate: Thu Aug 6 11:58:58 2026\n(Merged from https://github.com/openssl/openssl/pull/32031)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "fd4e510b1b2d4078759c27213901f566db59b0c5", "date": "2026-07-20T09:44:53+02:00", "subject": "Add ossl_list_TYPE_join(head, tail) function", "body": "The function appends list tail to list head. List tail becomes\nempty after the function returns.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMergeDate: Thu Aug 6 11:58:45 2026\n(Merged from https://github.com/openssl/openssl/pull/32031)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "da7d774e0f3072a6951c05687f7507f2163bb828", "date": "2026-07-20T09:44:53+02:00", "subject": "Add ossl_list_TYPE_join(head, tail) function", "body": "The function appends list tail to list head. List tail becomes\nempty after the function returns.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMergeDate: Thu Aug 6 11:58:31 2026\n(Merged from https://github.com/openssl/openssl/pull/32031)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d7191d9d35ce320fc1e81c95200078f261d411ce", "date": "2026-07-20T09:44:53+02:00", "subject": "Add ossl_list_TYPE_join(head, tail) function", "body": "The function appends list tail to list head. List tail becomes\nempty after the function returns.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMergeDate: Thu Aug 6 11:57:39 2026\n(Merged from https://github.com/openssl/openssl/pull/32031)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "37d7b11bc15ff2560b590994dfeee7a0d953ecde", "date": "2026-07-20T02:48:36-05:00", "subject": "fix(window): avoid UAF in `close_windows`", "body": "(cherry picked from commit a787a314900ecf98c759a82030ac28992896cc88)", "style": "conventional", "type": "fix", "scope": "window", "breaking": false} {"repo": "neovim", "sha": "a787a314900ecf98c759a82030ac28992896cc88", "date": "2026-07-20T02:48:36-05:00", "subject": "fix(window): avoid UAF in `close_windows`", "body": "", "style": "conventional", "type": "fix", "scope": "window", "breaking": false} {"repo": "vitest", "sha": "78ed73afd841d74a4638b39e371021fdf1784374", "date": "2026-07-20T09:04:04+01:00", "subject": "feat: promote fsModuleCache to a top-level option", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vitest", "sha": "088c0173d347497eb20a3742b36848d5c0dc7acb", "date": "2026-07-20T09:11:26+01:00", "subject": "docs: update latitude logo [backport to v4]", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "e30ccfe1c52f3a632336c9701a7b20adb44671d3", "date": "2026-07-20T09:11:29+01:00", "subject": "docs: update latitude logo", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "6bff85be68785e8dce7131279775688af75b0339", "date": "2026-07-20T10:29:41+02:00", "subject": "tool_xattr: add support for Windows alternate data stream", "body": "Requires a supported filesystem, i.e. NTFS.\n\nExample:\n```console\n> curl.exe https://curl.se/index.html --output test.txt --xattr --referer https://curl.se/\n> cat < test.txt:Zone.Identifier\n\n[ZoneTransfer]\nHostUrl=https://curl.se/index.html\nReferrerUrl=https://curl.se/\n```\n\nWhere newlines are CLRF, `ReferrerUrl` is set when using `--referer`\noption.\n\nAlso:\n- test688: make it test `--referer`.\n- document `user.xdg.referrer.url` in `--xattr` man page.\n\nRefs:\nhttps://en.wikipedia.org/wiki/Mark_of_the_Web\nhttps://en.wikipedia.org/wiki/NTFS#Alternate_data_stream_(ADS)\nhttps://learn.microsoft.com/en-us/windows/win32/fileio/file-streams\nhttps://learn.microsoft.com/en-us/openspecs/windows_protocols/ms-fscc/c54dec26-1551-4d3a-a0ea-4fa40f848eb3\n\nRef: #22345\n\nCloses #22354", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "075ed054800c03b05707ae6d448aba1242d8ead8", "date": "2026-07-20T11:46:50+03:00", "subject": "docs(devtools): add a settings doc", "body": "Add a doc describing the settings mechanism and the process of updating the settings options.", "style": "conventional", "type": "docs", "scope": "devtools", "breaking": false} {"repo": "neovim", "sha": "4121f2cf25c45ec340135847e094e530fc2ff083", "date": "2026-07-20T03:53:21-05:00", "subject": "test(api): normalize context comparisons #40854", "body": "", "style": "conventional", "type": "test", "scope": "api", "breaking": false} {"repo": "neovim", "sha": "09a4cc8c50285fa5c7c094650f505de11f90dccc", "date": "2026-07-20T02:02:17-07:00", "subject": "fix(docs): multi-level list dedent indentation #40850", "body": "Problem:\nThe list-item indentation logic in `gen_help_html.lua` tracks `opt.indent` that\nonly decrements by one level when a list item is less indented than its sibling.\nDedenting across multiple nesting levels at once produces the wrong left margin.\n\nSolution:\nTrack the exact leading whitespace for each indent level in a `opt.indent_ws` stack:\n- On reset or a top-level item initialize the stack with the current item's\n whitespace.\n- When indenting deeper push the current whitespace at the new level.\n- When dedenting pop every level whose tracked whitespace is deeper than the\n current item (reaching the correct ancestor level).", "style": "conventional", "type": "fix", "scope": "docs", "breaking": false} {"repo": "ant-design", "sha": "4ab41429a4c8d067655f6d318d4852dc27398b94", "date": "2026-07-20T17:18:31+08:00", "subject": "fix(Transfer): preserve custom action disabled state", "body": "* fix(Transfer): preserve custom action disabled state\n\n* fix(Transfer): prevent disabled custom action clicks\n\n* fix(Transfer): prevent disabled custom action navigation", "style": "conventional", "type": "fix", "scope": "Transfer", "breaking": false} {"repo": "eslint", "sha": "044a627fa3e28ee1410d515acc5378eb4b49f8ba", "date": "2026-07-20T18:36:57+09:00", "subject": "chore: update minimatch to ^10.2.5", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0345927a022020ab578f55b10b195f90aaec61b0", "date": "2026-07-20T12:02:17+02:00", "subject": "test_ocsp: bump test number", "body": "Signed-off-by: Norbert Pocs \n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMergeDate: Mon Jul 20 10:19:08 2026\n(Merged from https://github.com/openssl/openssl/pull/32014)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "9d40f8aefef9dcd893db5d01bc58d3e65e1cb4c2", "date": "2026-07-20T15:36:01+05:30", "subject": "fix(http): prevent transfer cache key collisions", "body": "`makeCacheKey` joined the request fields with `|` before hashing. The url\nand the serialized body can contain `|` themselves, so a shifted field\nboundary (url `/items/a` + body `b|c` vs url `/items/a|b` + body `c`)\nproduced the same joined string and the same key, letting two distinct\nrequests share a transfer cache slot.\n\nJoin with `\\0` instead, which cannot occur in a valid url or in encoded\nparams, so the field boundaries cannot be forged by field content.\n\n(cherry picked from commit 3192dccaa3e1488f8bf351864ea48e44a21246aa)", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "angular", "sha": "3192dccaa3e1488f8bf351864ea48e44a21246aa", "date": "2026-07-20T15:36:01+05:30", "subject": "fix(http): prevent transfer cache key collisions", "body": "`makeCacheKey` joined the request fields with `|` before hashing. The url\nand the serialized body can contain `|` themselves, so a shifted field\nboundary (url `/items/a` + body `b|c` vs url `/items/a|b` + body `c`)\nproduced the same joined string and the same key, letting two distinct\nrequests share a transfer cache slot.\n\nJoin with `\\0` instead, which cannot occur in a valid url or in encoded\nparams, so the field boundaries cannot be forged by field content.", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "postgres", "sha": "69fe2514fdd829cc4aebcd084f6f300f60e84f57", "date": "2026-07-20T12:12:13+02:00", "subject": "Move code to get_tables_to_repack_partitioned", "body": "Some of its code was pointlessly in its caller. This makes it better\ncontained and clearer.\n\nBackpatch to 19, to avoid having two different copies in case we have to\nmodify it again later.\n\nAuthor: Álvaro Herrera \nReviewed-by: Bharath Rupireddy \nReviewed-by: ChangAo Chen \nDiscussion: https://postgr.es/m/alD9l-XlCuu3eUEe@alvherre.pgsql", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "457b6cbefd91361e2c973feae93dd0f5c7c21106", "date": "2026-07-20T11:24:33+01:00", "subject": "Put OSC 133 data into grid_line.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "22b5350ff486ea2031148af9be542395c393a6b9", "date": "2026-07-20T18:24:52+08:00", "subject": "chore: add ES2020 syntax check for dist files", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "d6a2656a8ef1a89d89f49486f2062b0304725f22", "date": "2026-07-20T18:46:03+08:00", "subject": "docs(Alert): revert icon description", "body": "", "style": "conventional", "type": "docs", "scope": "Alert", "breaking": false} {"repo": "angular", "sha": "8616ba9db6240f3fbf8b905342d07326e096302b", "date": "2026-07-20T10:54:54Z", "subject": "fix(core): ensure SVG animation attributeName is checked case-insensitively", "body": "Currently, the SVG sanitizer checks a static set of candidate attribute\nnames (`attributeName` and `attributename`). This approach misses other case\nvariations (such as `attributenAme` or others), which could potentially\nbypass sanitization when binding sensitive attributes like `href` on\n`` or `` elements.\n\nThis change retrieves all attribute names of the SVG element, performs a\ncase-insensitive comparison with `'attributename'`, and sanitizes the value\nif a match is found.\n\n(cherry picked from commit 3499a1321bfe4801c38744ddb500715486680366)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "3499a1321bfe4801c38744ddb500715486680366", "date": "2026-07-20T10:54:54Z", "subject": "fix(core): ensure SVG animation attributeName is checked case-insensitively", "body": "Currently, the SVG sanitizer checks a static set of candidate attribute\nnames (`attributeName` and `attributename`). This approach misses other case\nvariations (such as `attributenAme` or others), which could potentially\nbypass sanitization when binding sensitive attributes like `href` on\n`` or `` elements.\n\nThis change retrieves all attribute names of the SVG element, performs a\ncase-insensitive comparison with `'attributename'`, and sanitizes the value\nif a match is found.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "neovim", "sha": "84fd9214cc71860e8bcbcdbdb4bd45bf00cabf50", "date": "2026-07-20T18:59:21+08:00", "subject": "fix(diagnostic): skip out-of-range lnum in underline handler #40839", "body": "Problem:\n`vim.diagnostic.handlers.underline.show` throws `E565: Index out of bounds` when\nit tries to underline a diagnostic whose `lnum` is past the end of the buffer.\nThis happens with stale diagnostics set on an unloaded buffer (e.g. via\n`bufadd()`) that are drawn only after the buffer is loaded: by then the file on\ndisk can be shorter than the `lnum` the diagnostic carried. File pickers\n(snacks.nvim) that open files via `bufadd` + `:buffer` hit this whenever an LSP\nserver has already emitted diagnostics for that URI.\n\nSolution:\nSkip diagnostics with an out-of-range `lnum` in the underline handler.\n\nOther handlers `M.virtual_text.show()`, `M.signs.show()`, have a similar condition.", "style": "conventional", "type": "fix", "scope": "diagnostic", "breaking": false} {"repo": "tmux", "sha": "d13b69ece4eda6fba458711f514db9dd4dfe326c", "date": "2026-07-20T11:16:33Z", "subject": "Change cellused/size to 16 bits and time to 32 bits in grid_line and add the OSC 133 positions (size stays the same).", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d06bac614c2bc09db62e439b2e42ac7101653333", "date": "2026-07-20T19:21:53+08:00", "subject": "refactor(window): extract w_kind for window role #40830", "body": "Problem:\nw_float_is_info conflates a window's role with its floating layout,\nwhich are orthogonal properties.\n\nSolution:\nReplace w_float_is_info with a w_kind enum for the window's role.", "style": "conventional", "type": "refactor", "scope": "window", "breaking": false} {"repo": "vitest", "sha": "cb3bd6383c271a961d602c0ed9078e948f4d12a6", "date": "2026-07-20T13:40:30+02:00", "subject": "test: pointer API", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "curl", "sha": "0c4eebb3fa9d18f04b03c2bb2a53ef670bf6e389", "date": "2026-07-20T13:50:34+02:00", "subject": "test: adjust test_06_13 for 0100::/64 being blackholed", "body": "Fixes #22344\nReported-by: Carlos Henrique Lima Melara\nCloses #22355", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "10389a7e15e0ed0214e19c6560edd4dc6bb07a35", "date": "2026-07-20T13:50:45+02:00", "subject": "Fix LSN format in REPACK worker debug message", "body": "Commit 6f6f284c7ee4 introduced use of LSN_FORMAT_ARGS across the whole\ntree to remove use of manual bit-shifting, and commit 2633dae2e487\nchanged the printf format to be %X/%08X; however commit 28d534e2ae0a\nviolated both conventions by reintroducing the old manual-shift style\nwith the deprecated %X/%X format in one debug message. Make that new\nmessage conform to our style.\n\nAuthor: kenny \nBackpatch-through: 19\nDiscussion: https://postgr.es/m/CAPXstDuWD8jg0=C8PXTXGSTTsZcjqJ+u+xKCrMpN99CXsxQzCg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a7e5a6eea8fd3ccca6b6fbba031a5fbf8a3d93b4", "date": "2026-07-20T14:10:47+02:00", "subject": "CMP unexpected sender DN used as format string in ERR_raise_data()", "body": "ossl_cmp_msg_check_update() converts an unexpected CMP response sender DN with\nX509_NAME_oneline() and passes that peer-controlled string directly as the\nformat argument to ERR_raise_data(). Printable percent characters survive the\nDN conversion, so a sender such as CN=%s%n reaches vsnprintf() as active format\nsyntax without matching varargs.\n\nFixes: CVE-2026-63073\n\nOriginal patch by: Filipe Casal of Trail of Bits in collaboration with OpenAI\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Igor Ustinov \nMerge-date: Mon Aug 24 13:02:29 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0cc20b322639919aa423e90799d9a57c3b4b76ca", "date": "2026-07-20T14:10:47+02:00", "subject": "CMP unexpected sender DN used as format string in ERR_raise_data()", "body": "ossl_cmp_msg_check_update() converts an unexpected CMP response sender DN with\nX509_NAME_oneline() and passes that peer-controlled string directly as the\nformat argument to ERR_raise_data(). Printable percent characters survive the\nDN conversion, so a sender such as CN=%s%n reaches vsnprintf() as active format\nsyntax without matching varargs.\n\nFixes: CVE-2026-63073\n\nOriginal patch by: Filipe Casal of Trail of Bits in collaboration with OpenAI\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Igor Ustinov \nMerge-date: Mon Aug 24 13:02:10 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6a0acc072b4d37a7cac1252a29c1ce1f00c5ec29", "date": "2026-07-20T14:10:47+02:00", "subject": "CMP unexpected sender DN used as format string in ERR_raise_data()", "body": "ossl_cmp_msg_check_update() converts an unexpected CMP response sender DN with\nX509_NAME_oneline() and passes that peer-controlled string directly as the\nformat argument to ERR_raise_data(). Printable percent characters survive the\nDN conversion, so a sender such as CN=%s%n reaches vsnprintf() as active format\nsyntax without matching varargs.\n\nFixes: CVE-2026-63073\n\nOriginal patch by: Filipe Casal of Trail of Bits in collaboration with OpenAI\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Igor Ustinov \nMerge-date: Mon Aug 24 13:01:47 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7eb2e3ec9d1d4f35c8022fccd4b03398b3f33e21", "date": "2026-07-20T14:10:47+02:00", "subject": "CMP unexpected sender DN used as format string in ERR_raise_data()", "body": "ossl_cmp_msg_check_update() converts an unexpected CMP response sender DN with\nX509_NAME_oneline() and passes that peer-controlled string directly as the\nformat argument to ERR_raise_data(). Printable percent characters survive the\nDN conversion, so a sender such as CN=%s%n reaches vsnprintf() as active format\nsyntax without matching varargs.\n\nFixes: CVE-2026-63073\n\nOriginal patch by: Filipe Casal of Trail of Bits in collaboration with OpenAI\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Igor Ustinov \nMerge-date: Mon Aug 24 13:00:49 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "958f6f6040385a88a322673dbb465f7259ebf0f2", "date": "2026-07-20T14:10:47+02:00", "subject": "CMP unexpected sender DN used as format string in ERR_raise_data()", "body": "ossl_cmp_msg_check_update() converts an unexpected CMP response sender DN with\nX509_NAME_oneline() and passes that peer-controlled string directly as the\nformat argument to ERR_raise_data(). Printable percent characters survive the\nDN conversion, so a sender such as CN=%s%n reaches vsnprintf() as active format\nsyntax without matching varargs.\n\nFixes: CVE-2026-63073\n\nOriginal patch by: Filipe Casal of Trail of Bits in collaboration with OpenAI\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Andrew Dinh \nReviewed-by: Igor Ustinov \nMerge-date: Mon Aug 24 12:53:37 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "4eb4b0d080723d328c4bf88c3be86faf02a39c5f", "date": "2026-07-20T14:37:13+02:00", "subject": "vquic: use ngtcp2 v1.25.0 new close2 callback", "body": "Forward only the app error code from the receiving side to the h3 layer.\n\nThis only takes effect when building against ngtcp2 v1.25.0 or higher.\n\nFixes #22270\n\nCloses #22356", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1009339b3acaf06b457a8182bf50b504847a643c", "date": "2026-07-20T21:40:31+09:00", "subject": "Allow PostgreSQL::Test::Cluster::start() to pass postmaster options", "body": "Previously, tests that needed extra postmaster command-line options had\nto invoke pg_ctl start directly, because\nPostgreSQL::Test::Cluster::start() provided no way to pass them. That\nbypassed the test framework's postmaster PID tracking, so a postmaster\ncould be left running if the test failed after startup.\n\nAdd an options parameter to\nPostgreSQL::Test::Cluster::start(), which is passed to pg_ctl's\n--options argument. This allows tests to use start() while\npreserving the framework's normal cleanup behavior.\n\nAuthor: Fujii Masao \nReviewed-by: JoongHyuk Shin \nDiscussion: https://postgr.es/m/CAHGQGwEpfE0CDUUODjBt7GO9U4ZF11hqga_Ci3wP8=O49oFKVw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d5751c33cc3e842b20dbe67545fd2c850be8fa59", "date": "2026-07-20T21:41:14+09:00", "subject": "Avoid ERROR in recovery target GUC assign hooks", "body": "Recovery target parameters are postmaster-startup GUCs, but their\nassign hooks previously did more than assign individual parameter\nvalues. They also updated the global recoveryTarget state and raised\nERROR if more than one recovery target appeared to be set.\n\nThis was not a good fit for GUC assign hooks. Assign hooks should not\nthrow ERROR, and deriving cross-parameter state while individual GUCs\nare still being assigned makes the result depend on assignment order\nrather than the final configuration.\n\nFor example, setting one recovery target and then setting another\nrecovery_target_* parameter to an empty string could clear\nrecoveryTarget, causing recovery to proceed with no target even\nthough a valid target remained configured.\n\nFix this by having the assign hooks only store their own parameter\nvalues. The effective recoveryTarget is now derived once from the\nfinal recovery_target* settings in\nvalidateRecoveryParameters(), which also rejects configurations that\nspecify more than one recovery target with FATAL. This preserves the\nexpected behavior for repeated assignments of the same GUC, treats empty\nvalues as \"not set\", and removes cross-GUC validation from the assign\nhooks.\n\nAuthor: JoongHyuk Shin \nReviewed-by: Greg Lamberson \nReviewed-by: Michael Paquier \nReviewed-by: Scott Ray \nReviewed-by: Álvaro Herrera \nReviewed-by: Kyotaro Horiguchi \nReviewed-by: Henson Choi \nReviewed-by: Zsolt Parragi \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CACSdjfPUa4UvKjADgOERXoxNYmCg2mqqiqKkiJk6mX6E4qgVFw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "14a20419506264bdd69b36dacde4ac81c426f937", "date": "2026-07-20T14:51:49+02:00", "subject": "test: failing double click", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "angular", "sha": "b8a0fa00c9dd21d0493bd0562d0d4f00aa12fb1a", "date": "2026-07-20T15:14:39+02:00", "subject": "build: update pnpm-lock.yaml", "body": "This commit updates the pnpm-lock.yaml to reflect differences in node and https-proxy-agent transitive dependencies.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a56b74bde3806b74a2e28424bdce69e351560172", "date": "2026-07-20T20:53:06+07:00", "subject": "feat(lua): Iter:count() #40831", "body": "", "style": "conventional", "type": "feat", "scope": "lua", "breaking": false} {"repo": "neovim", "sha": "ba2ea67a1f45f6d4b5780efefbdbbe273665c724", "date": "2026-07-20T10:03:11-04:00", "subject": "fix(lsp): incorrect rendering of markdown codeblock #40861", "body": "Problem:\nLSP hover erroneously drops blank lines before a 4-space-indented\ncodeblock, which is not valid Markdown. This causes incorrect parsing\nand wrong display.\n\nSolution:\nFix `split_lines` so that it doesn't drop the blank line just before\na 4-space-indented codeblock.\n\nfix https://github.com/neovim/neovim/issues/40860\n\n(cherry picked from commit 317c5ddda6c71d1e7709c7066c91522ecaeb081a)", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "neovim", "sha": "317c5ddda6c71d1e7709c7066c91522ecaeb081a", "date": "2026-07-20T10:03:11-04:00", "subject": "fix(lsp): incorrect rendering of markdown codeblock #40861", "body": "Problem:\nLSP hover erroneously drops blank lines before a 4-space-indented\ncodeblock, which is not valid Markdown. This causes incorrect parsing\nand wrong display.\n\nSolution:\nFix `split_lines` so that it doesn't drop the blank line just before\na 4-space-indented codeblock.\n\nfix https://github.com/neovim/neovim/issues/40860", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "angular", "sha": "3d0971338673a8e793dcf64be77a0d5d18e3f6b3", "date": "2026-07-20T17:26:35+03:00", "subject": "feat(devtools): show Angie on the remaining empty-state screens", "body": "Extend the Angie illustrations to the empty states that were still plain\ntext: the injector providers \"No such providers\" result, the injector tree\nunsupported-version notice, the two profiler timeline placeholders, and the\nrouter tree unsupported-version notice.\n\nAdds the magnifying-glass and question poses as assets; the orthos-back\nand sad poses are already in the tree.", "style": "conventional", "type": "feat", "scope": "devtools", "breaking": false} {"repo": "neovim", "sha": "86fbcd67be123a1894611ad90de73d413e92eda1", "date": "2026-07-20T11:12:53-04:00", "subject": "test: unreliable \"CursorHold not triggered after only K_EVENT on startup\" #40863", "body": "Problem:\n\n FAILED …/cursorhold_spec.lua @ 73: CursorHold is not triggered after only K_EVENT on startup\n Expected values to be equal.\n Expected:\n 1\n Actual:\n 0\n stack traceback:\n …/cursorhold_spec.lua:79: in function <…/cursorhold_spec.lua:73>\n\nSolution:\nRetry instead of hardcoding sleep(50).", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "6d8c8b18d551d0be0490640c00fabb816727ec10", "date": "2026-07-20T17:16:51+02:00", "subject": "test(harness): migrate away from magic globals", "body": "Problem:\nThe magic globals `it`, `describe`, etc., are more trouble than they are\nworth.\n\n- Hooking into `after_each` requires `getfenv()` hacks.\n- They confuse luals/emmylua, because the top-level `.luarc.json` isn't\n merged with `test/.luarc.json` (apparently a luals limitation?)\n- They totally defeat discoverability because the user just has to\n \"know\" about the various magic symbols.\n\nSo they harm DX, which means they serve no purpose at all.\n\nSolution:\n- Expose the test API from `testutil`, so tests can call `t.it()`,\n `t.describe()`, etc., in the conventional way.\n- Drop `getfenv()` hacks.\n- Drop the `setfenv()` injection in `load_chunk`.\n- Drop `test/_meta.lua`.", "style": "conventional", "type": "test", "scope": "harness", "breaking": false} {"repo": "postgres", "sha": "1d6c654c8189cf198bc4059a9a0c9a51502e4b5f", "date": "2026-07-20T17:21:20+02:00", "subject": "Fix restore of partitions with exclusion constraints", "body": "Commit 8c852ba9a4 allowed exclusion constraints to be added to\npartitioned tables, but wasn't careful to verify that pg_restore worked\ncorrectly for them. Fix that by making CompareIndexInfo() more\nselective about what needs to be rejected.\n\nAuthor: Japin Li \nReported-by: Keith Paskett \nDiscussion: https://postgr.es/m/2A40921D-83AB-411E-ADA6-7E509A46F1E4@logansw.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "bb39cda6483de2edab2f8221459b0ed5b73ef221", "date": "2026-07-20T20:55:14+05:30", "subject": "fix(migrations): preserve NgClass import on partial migration", "body": "When only some NgClass usages are migrated (partial migration),\nthe NgClass import should be preserved in the module/component\nimports if it is still used elsewhere.\n\n(cherry picked from commit 223e40279f3f6df979a323f79cb370528366c24a)", "style": "conventional", "type": "fix", "scope": "migrations", "breaking": false} {"repo": "angular", "sha": "223e40279f3f6df979a323f79cb370528366c24a", "date": "2026-07-20T20:55:14+05:30", "subject": "fix(migrations): preserve NgClass import on partial migration", "body": "When only some NgClass usages are migrated (partial migration),\nthe NgClass import should be preserved in the module/component\nimports if it is still used elsewhere.", "style": "conventional", "type": "fix", "scope": "migrations", "breaking": false} {"repo": "tmux", "sha": "d080378464c82de538eb06145bef78caeb5d5fa1", "date": "2026-07-20T16:26:02+01:00", "subject": "Add collapsible lines.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "bd58857201452aba99d78742990296f1a6b72c6a", "date": "2026-07-20T17:34:20+02:00", "subject": "vtls_scache: allocate absolute filename on heap (Windows)", "body": "To allow any full path lengths, beyond the 260 characters allowed by\n`_MAX_PATH`.\n\nFollow-up to fa0ccd9f1fbbbd77bf50b26e3ba231ea6c729474 #15774\n\nCloses #22357", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "d85288d75f3f85dc6b10bf813c805b1b7fed7329", "date": "2026-07-20T20:54:02+05:00", "subject": "fix(migrations): skip tsconfig files of non-Angular projects", "body": "Only touch tsconfig files for targets that use an Angular builder, including community ones like Nx, so non-Angular projects in mixed workspaces are left alone.\n\nFixes #69837\n\n(cherry picked from commit f72600eaddfa6a299526e922eb9ad5356be821a6)", "style": "conventional", "type": "fix", "scope": "migrations", "breaking": false} {"repo": "angular", "sha": "f72600eaddfa6a299526e922eb9ad5356be821a6", "date": "2026-07-20T20:54:02+05:00", "subject": "fix(migrations): skip tsconfig files of non-Angular projects", "body": "Only touch tsconfig files for targets that use an Angular builder, including community ones like Nx, so non-Angular projects in mixed workspaces are left alone.\n\nFixes #69837", "style": "conventional", "type": "fix", "scope": "migrations", "breaking": false} {"repo": "neovim", "sha": "1250821ba817bffe14b25cd95421b122ba91dfc7", "date": "2026-07-20T18:11:52+02:00", "subject": "test: drop `assert.is_true()` and friends", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "angular", "sha": "030779c4f038a41d8ab9699e169cf972bfc251f9", "date": "2026-07-20T18:25:43+02:00", "subject": "build: update dependency magic-string to v1", "body": "Magic-string is now ESM only.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "neovim", "sha": "c7707e2a10dc5551f7362f1dedd8c030f920b37d", "date": "2026-07-20T09:30:29-07:00", "subject": "build: raise `preloadFileSize` in `src/.luarc.json` to match root config", "body": "", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5b05b7da7c8109c8128dee71ce5ed73eb326fc15", "date": "2026-07-20T19:10:23+02:00", "subject": "Rename ASN1_STRING_length_ex() to ASN1_STRING_get_length()", "body": "To make it in line with ASN1_BIT_STRING_get_length() and other (size)\ngetters. The majority of the changes was done by\n\n for i in `git grep -l ASN1_STRING_length_ex`; do \\\n sed -i 's/ASN1_STRING_length_ex/ASN1_STRING_get_length/g' \"$i\"; \\\n done\n\nwith a manual touch-up in util/libcrypto.num.\n\nComplements: 28179061bfcd \"Prepare a now opaque ASN1_STRING for the size_t rapture.\"\nSigned-off-by: Eugene Syromiatnikov \n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Bob Beck \nReviewed-by: Simo Sorce \nMergeDate: Tue Aug 11 13:56:42 2026\n(Merged from https://github.com/openssl/openssl/pull/32015)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "786353f4c56ceadd2f8d8d2351cbd59c89b923f9", "date": "2026-07-20T19:18:59+02:00", "subject": "Change set suffix in ASN1_STRING_set_{data,string} to set1", "body": "To comply with the naming guidelines for the API functions that transfer\nownership of the data. The majority of the changes was done by\n\n for i in `git grep -l ASN1_STRING_set_data`; do \\\n sed -i 's/ASN1_STRING_set_data/ASN1_STRING_set1_data/g' \"$i\"; \\\n done\n for i in `git grep -l ASN1_STRING_set_string`; do \\\n sed -i 's/ASN1_STRING_set_string/ASN1_STRING_set1_string/g' \"$i\"; \\\n done\n\nwith a manual touch-up in util/libcrypto.num.\n\nComplements: 28179061bfcd \"Prepare a now opaque ASN1_STRING for the size_t rapture.\"\nSigned-off-by: Eugene Syromiatnikov \n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Bob Beck \nReviewed-by: Simo Sorce \nMergeDate: Tue Aug 11 13:56:44 2026\n(Merged from https://github.com/openssl/openssl/pull/32015)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "665bf40a132a857f1790de2aebffb14f85da7b6a", "date": "2026-07-20T13:36:39-04:00", "subject": "doc: Granting TRIGGER or REFERENCES on table is dangerous.", "body": "It's always been the case that granting these privileges to users that\nyou don't fully trust was a bad idea, but it hasn't always been\nobvious to people reading the documentation that this is the case.\nTo prevent confusion, and also repeated reports to pgsql-security,\nmention it explicitly.\n\nDiscussion: http://postgr.es/m/CA+TgmobrjCHBuWHrvX3=2vndUCO2thUOdevrCcMDFW86cqCYvw@mail.gmail.com\nReviewed-by: Nathan Bossart \nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d343a1e230bf61bff41a78583c97ae39c5c961f3", "date": "2026-07-20T12:58:30-05:00", "subject": "fix(ui2): position message windows without autocmds", "body": "Problem: Fix applied in b2bb60d7 does not properly replace the normal command.\nSolution: Restore normal command with `noautocmd` to resolve #40780.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "postgres", "sha": "83486b6a65aa48ed18e4e872d5e0fb5b0bfaca41", "date": "2026-07-21T03:24:58+09:00", "subject": "Fix recovery target test waiting on unavailable WAL", "body": "Buildfarm member skink reported a failure in\nrecovery/003_recovery_targets after commit d5751c33cc3. The newly\nadded recovery_target_xid set-then-cleared test could time out while\nwaiting for pg_last_wal_replay_lsn() to reach the expected LSN.\n\nThe test recorded lsn6 after calling pg_switch_wal(). As a result,\nlsn6 pointed into the next WAL segment, but pg_switch_wal() only\narchived the previous one. Since the standby in this test restores WAL\nfrom the archive only, it could not obtain the segment containing\nlsn6 and waited indefinitely.\n\nFix this by recording lsn6 before calling pg_switch_wal(), so the\narchived WAL contains the LSN that the standby is waiting for.\n\nPer buildfarm member skink.\n\nReported-by: Álvaro Herrera \nAuthor: Fujii Masao \nDiscussion: https://postgr.es/m/al5Y2mWffRs1NP34@alvherre.pgsql", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "61bfda575b03e6a32af7877b65651a37eb339426", "date": "2026-07-21T02:37:08+08:00", "subject": "chore: Update prettier v3.9", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "neovim", "sha": "18d6b2309fc2f8824854882933f11ee95d086f49", "date": "2026-07-20T19:25:24Z", "subject": "fix(ui): handle combining chunked float titles", "body": "Problem:\nThe title is combined with window's attributes only if the title is a\nstring (which implies the FloatTitle / FloatFooter highlight groups),\nbut not when the title is text-hl chunks.\n\nSolution:\nCombine specified highlight group with window-local Normal highlight as\nwell.", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "neovim", "sha": "ca07e505f939e8ff6e70571ef130722bdc40b292", "date": "2026-07-20T21:35:16+02:00", "subject": "refactor(ruler)!: replace C implementation with 'rulerformat' expression", "body": "Problem:\nThe default 'ruler' is implemented in C instead of the 'statusline' DSL.\n\nSolution:\nReplace the C implementation with a default 'rulerformat' expression.\nThis is a continuation of #1248 and #33036.\n\nAdvantages:\n- configuration is more discoverable, the default being a useful example\n- users and plugins can augment the default\n- code reuse and less C code to maintain\n- ui2: due to the use of an item group with `minwid`, it can expand\n instead of truncating when the content gets too long, which is\n particularly useful for locales with long translations of Top/Bot/All\n\nImplementation details:\nAs is the case for 'statusline', when trying to set 'rulerformat' to an\nempty string, the default expression is restored instead, mimicking how\npreviously the default C implementation would have been activated.\nJust like before, `:set rulerformat=` and `:set rulerformat&` have the\nsame effect, and the ruler is disabled with `:set noruler`.\n\nThe default expression uses an item group with `%=`, unlike the fallback\nin the previous default statusline `%-14.(%l,%c%V%) %P`, because the\ntotal width and how it is configured is immediately clear without\ndocumentation, it is a more useful pattern in general that works when\nboth sides have flexible width, and it also works for vim, which is\nuseful for configuration sharing/reuse.\n\nA truncation marker `%<` is added at the end to mimic how at small\nscreen widths, the scroll percentage would disappear first, so that the\ncursor position can remain fully visible.\n\nBREAKING CHANGES:\n- `&rulerformat` can no longer be set to an empty string\n- ui2: the default ruler is no longer of fixed width, but can expand\n- at very small screen widths (< 36 columns)\n - ui2: it will no longer try to shrink white-space before truncating\n - it truncates gradually from the right, whereas previously, the\n scroll percentage would disappear all at once\n- l10n can no longer add a space after the comma between line and column\n (this was only done for one ", "style": "conventional", "type": "refactor", "scope": "ruler", "breaking": true} {"repo": "openssl", "sha": "8834b4b1b1ccdd4b450f71825c18f9ad976bfd2e", "date": "2026-07-20T21:35:48+02:00", "subject": "test: add low-level OSSL_ENCODER / OSSL_DECODER API test", "body": "Add a test that exercises the encoder and decoder chain processing in\nencoder_lib.c and decoder_lib.c through a built-in test provider\nimplementing a two-stage encoder chain (\"TEST-KEY\" -> \"inter\" -> \"pem\")\nand the mirrored decoder chain.\n\nThis covers previously untested code paths:\n\n- encoder chaining through abstract objects, including the intermediate\n memory BIO handling in encoder_process()\n- OSSL_ENCODER_to_fp() and OSSL_DECODER_from_fp()\n- OSSL_ENCODER_to_data() edge cases (size query, pre-allocated buffer,\n too small buffer)\n- OSSL_DECODER_CTX_add_decoder() and OSSL_DECODER_export()\n- decoding from a non-seekable BIO (BIO_f_readbuffer wrapping)\n- various error paths and NULL argument checks in both APIs\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Matt Caswell \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 13:45:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32019\n(cherry picked from commit b542f2545cc10f6b481fd70600bcb0b8523e09a9)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b542f2545cc10f6b481fd70600bcb0b8523e09a9", "date": "2026-07-20T21:35:48+02:00", "subject": "test: add low-level OSSL_ENCODER / OSSL_DECODER API test", "body": "Add a test that exercises the encoder and decoder chain processing in\nencoder_lib.c and decoder_lib.c through a built-in test provider\nimplementing a two-stage encoder chain (\"TEST-KEY\" -> \"inter\" -> \"pem\")\nand the mirrored decoder chain.\n\nThis covers previously untested code paths:\n\n- encoder chaining through abstract objects, including the intermediate\n memory BIO handling in encoder_process()\n- OSSL_ENCODER_to_fp() and OSSL_DECODER_from_fp()\n- OSSL_ENCODER_to_data() edge cases (size query, pre-allocated buffer,\n too small buffer)\n- OSSL_DECODER_CTX_add_decoder() and OSSL_DECODER_export()\n- decoding from a non-seekable BIO (BIO_f_readbuffer wrapping)\n- various error paths and NULL argument checks in both APIs\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Matt Caswell \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 13:45:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32019", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "angular", "sha": "f3027367eb87556f2d632753e5aeb1a74136fe17", "date": "2026-07-21T01:43:14+06:00", "subject": "refactor(core): use native `Promise.withResolvers()` in tests", "body": "Replaces the temporary `promiseWithResolvers` polyfill with the\nnative `Promise.withResolvers()` API in test files and Updates the TypeScript configuration to include the `es2024.promise`.\n\n(cherry picked from commit b5ce15c659b76e7c25e8cd98358ffd44a24cdea2)", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "b5ce15c659b76e7c25e8cd98358ffd44a24cdea2", "date": "2026-07-21T01:43:14+06:00", "subject": "refactor(core): use native `Promise.withResolvers()` in tests", "body": "Replaces the temporary `promiseWithResolvers` polyfill with the\nnative `Promise.withResolvers()` API in test files and Updates the TypeScript configuration to include the `es2024.promise`.", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "neovim", "sha": "16c145d24042ebdb0e2a0479697c4c5808d3ebcf", "date": "2026-07-20T14:44:37-05:00", "subject": "fix(ui2): preserve Visual mode messages #40828", "body": "Problem:\nThe existing `showmode` overlay can immediately cover messages emitted while\nVisual mode is active, including the `g CTRL-G` word count.\n\nSolution:\nProtect Visual mode messages with the existing message delay and temporarily\nhide the previous last-line overlay until it is restored.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "angular", "sha": "90e81e43672c4330c0af4f940e8306fc4c499921", "date": "2026-07-20T22:52:13+03:00", "subject": "feat(docs-infra): bookend the tutorials with Angie", "body": "The tutorial intro and \"next steps\" pages were plain headings with no\nsend-off. Bookend each tutorial with Angie: a greeting pose welcoming\nlearners on the intro page, and a superhero pose congratulating them on\nthe completion page, each beside a speech bubble.\n\nThe bubble treatment from the 404 page lived inside the not-found\ncomponent, so it's extracted into a reusable docs-content style\n(docs-tutorial-mascot) usable from tutorial markdown, and applied to the\nintro and completion pages of the Learn Angular, Signals, and Signal\nForms tutorials.\n\n(cherry picked from commit d2c0e1ec77c352e07a13647dde153bf83fca603e)", "style": "conventional", "type": "feat", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "d2c0e1ec77c352e07a13647dde153bf83fca603e", "date": "2026-07-20T22:52:13+03:00", "subject": "feat(docs-infra): bookend the tutorials with Angie", "body": "The tutorial intro and \"next steps\" pages were plain headings with no\nsend-off. Bookend each tutorial with Angie: a greeting pose welcoming\nlearners on the intro page, and a superhero pose congratulating them on\nthe completion page, each beside a speech bubble.\n\nThe bubble treatment from the 404 page lived inside the not-found\ncomponent, so it's extracted into a reusable docs-content style\n(docs-tutorial-mascot) usable from tutorial markdown, and applied to the\nintro and completion pages of the Learn Angular, Signals, and Signal\nForms tutorials.", "style": "conventional", "type": "feat", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "388daea2fc188aad3ab69fb81bd3f893ac3cd846", "date": "2026-07-21T01:23:12+05:30", "subject": "fix(migrations): correctly migrate ngClass with mixed space-separated keys", "body": "Preserve NgClass import on partial migration and increment\nskippedNgClassCount when an unmigrable mixed binding is encountered.\n\n(cherry picked from commit 49672c437b440564e9c20b77a7ff151fbf316706)", "style": "conventional", "type": "fix", "scope": "migrations", "breaking": false} {"repo": "angular", "sha": "49672c437b440564e9c20b77a7ff151fbf316706", "date": "2026-07-21T01:23:12+05:30", "subject": "fix(migrations): correctly migrate ngClass with mixed space-separated keys", "body": "Preserve NgClass import on partial migration and increment\nskippedNgClassCount when an unmigrable mixed binding is encountered.", "style": "conventional", "type": "fix", "scope": "migrations", "breaking": false} {"repo": "angular", "sha": "7eec520725b28ba02c2213c467345f71eba36e9e", "date": "2026-07-20T13:30:07-07:00", "subject": "test(compiler-cli): cover DOM-only vs full instruction set across compilation modes", "body": "Adds a compliance case pinning the instruction-set selection for a\ndirective-free standalone component:\n\n - full compile: the compiler can prove the template has no directive\n dependencies, so it takes the DOM-only fast path\n (`ɵɵdomElementStart`/`ɵɵdomElementEnd`).\n - local compile: the compiler cannot inspect dependencies, so\n `hasDirectiveDependencies` is forced true and the full instruction\n set is emitted (`ɵɵelementStart`/`ɵɵelementEnd`).\n\nThis mode-dependent switch was previously only exercised incidentally by\nthe `foreign_component` case, which couples it with foreign-component\ncompilation. The new case isolates it.\n\n(cherry picked from commit eb3e480360d0f24a725bb453af49bd9072fdc87e)", "style": "conventional", "type": "test", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "eb3e480360d0f24a725bb453af49bd9072fdc87e", "date": "2026-07-20T13:30:07-07:00", "subject": "test(compiler-cli): cover DOM-only vs full instruction set across compilation modes", "body": "Adds a compliance case pinning the instruction-set selection for a\ndirective-free standalone component:\n\n - full compile: the compiler can prove the template has no directive\n dependencies, so it takes the DOM-only fast path\n (`ɵɵdomElementStart`/`ɵɵdomElementEnd`).\n - local compile: the compiler cannot inspect dependencies, so\n `hasDirectiveDependencies` is forced true and the full instruction\n set is emitted (`ɵɵelementStart`/`ɵɵelementEnd`).\n\nThis mode-dependent switch was previously only exercised incidentally by\nthe `foreign_component` case, which couples it with foreign-component\ncompilation. The new case isolates it.", "style": "conventional", "type": "test", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "d45bd2f53d8a3f94452e05025815909893d6c5ce", "date": "2026-07-20T13:30:22-07:00", "subject": "refactor(core): implement toggleWatchSignal for DevTools signal debugging", "body": "Implement \\`toggleWatchSignal(id)\\` in \\`signal_debug.ts\\` to enable toggling reactive watch listeners on individual signal graph nodes.\n- Create reactive \\`Watch\\` instances using \\`createWatch\\` to log signal value/state changes to the console when active.\n- Use \\`WeakRef\\` mapping and \\`FinalizationRegistry\\` (\\`watchCleanupRegistry\\`) so active watches do not retain strong references to signal nodes or prevent garbage collection.\n- Expose \\`watched\\` status on \\`DebugSignalGraphNode\\` and publish \\`toggleWatchSignal\\` onto \\`window.ng\\` in development mode.\n- Add acceptance tests covering watch activation, signal mutation logging, manual disposal, GC cleanup, and safe invalid ID handling.", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "82a767535e522139e6a6d0f2a33b454cb391ef8c", "date": "2026-07-20T13:31:45-07:00", "subject": "feat(devtools): add \"Watch signal\" button to signal-details component", "body": "Add a \"Watch signal\" button to the signal-details component to allow users to \"watch\" changes to the associated signal.", "style": "conventional", "type": "feat", "scope": "devtools", "breaking": false} {"repo": "openssl", "sha": "512483db5ab1d5961f4bec47b0c37f3c84a14899", "date": "2026-07-20T22:59:41+02:00", "subject": "encoder: pass the data type up through the encoder chain", "body": "In encoder_process(), the data type (the name of the encoder\nimplementation that produced the current encoding) is determined at the\ndeepest recursion level, but was not propagated up through the\nrecursion together with the other results of the recursive call. As a\nconsequence, the abstract object handed to a chained encoder\nimplementation contained an OSSL_OBJECT_PARAM_DATA_TYPE parameter with\nNULL data, in violation of provider-object(7), which specifies it as a\nUTF8 string carrying the type of the object content.\n\nThis went unnoticed because no encoders provided by OpenSSL itself are\never chained.\n\nAlso extend endecode_api_test to verify that a chained encoder\nimplementation receives the correct data type.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Matt Caswell \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 13:45:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32019\n(cherry picked from commit 13b0be7cb62e21c23916937999446c00a6cd31f9)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "13b0be7cb62e21c23916937999446c00a6cd31f9", "date": "2026-07-20T22:59:41+02:00", "subject": "encoder: pass the data type up through the encoder chain", "body": "In encoder_process(), the data type (the name of the encoder\nimplementation that produced the current encoding) is determined at the\ndeepest recursion level, but was not propagated up through the\nrecursion together with the other results of the recursive call. As a\nconsequence, the abstract object handed to a chained encoder\nimplementation contained an OSSL_OBJECT_PARAM_DATA_TYPE parameter with\nNULL data, in violation of provider-object(7), which specifies it as a\nUTF8 string carrying the type of the object content.\n\nThis went unnoticed because no encoders provided by OpenSSL itself are\never chained.\n\nAlso extend endecode_api_test to verify that a chained encoder\nimplementation receives the correct data type.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Matt Caswell \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 13:45:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32019", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "90910715011211a20d011d807d398a7005127f35", "date": "2026-07-21T00:20:49+03:00", "subject": "fix: prevent `no-unreachable-loop` crash when all loop types are ignored", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "eslint", "sha": "ee225b60c26f0cc5f4d71641888a5bd2ec5626d6", "date": "2026-07-21T06:22:15+09:00", "subject": "test: Add error location details to `no-eq-null` rule", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "10413cceb66411389bbd65a9e4698dc6a172113c", "date": "2026-07-20T23:34:08+02:00", "subject": "Add OSSL_ECHSTORE_read_echconfiglist fuzzer", "body": "Fuzz OSSL_ECHSTORE_read_echconfiglist with raw\nfuzzer input in stage 1. Then fixup the starting\nbytes to reach deeper code withotu relying only on\ncorpus. Fixup is done for binary encoding only (not base64).\n\nAssisted-by: ChatGPT:gpt-5.5\n\nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMergeDate: Thu Jul 23 13:57:54 2026\n(Merged from https://github.com/openssl/openssl/pull/32021)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "f65253056a1b3d0bc6b27f450a6291ed253f6491", "date": "2026-07-20T23:44:38+02:00", "subject": "cf-quiche.c: minor tidy-ups", "body": "Pointed out by the GitHub AI thing\n\nCloses #22360", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "20da449b76daf8c93fd38cfbfac2a6776217a8f8", "date": "2026-07-20T18:28:19-04:00", "subject": "QUIC: ignore rejected default stream candidates", "body": "Auto-rejected streams remain in the stream map until garbage collection, but\nthey are never linked to the incoming accept queue. Connection-level reads\ncurrently treat such a stream as a default stream candidate and try to remove\nits unlinked accept node, which aborts in debug builds and dereferences NULL\nwhen assertions are disabled.\n\nOnly select peer streams that are still linked to the accept queue, and add\na regression test requiring SSL_ERROR_WANT_READ after the peer stream is\nrejected.\n\nFixes #31685\n\nAssisted-by: Codex:GPT-5\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Jakub Zelenka \nMergeDate: Tue Aug 11 07:15:53 2026\n(Merged from https://github.com/openssl/openssl/pull/32020)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5c00c80b54db33fab6db02c0c7fd2322f09f929c", "date": "2026-07-21T00:33:18+02:00", "subject": "Add a red-black tree implementation", "body": "The code comes from David Gwynne .\nThe same implmentation can be found in OpenBSD.\n\nSeveral minor adjustments have been done to include the code\ninto OpenSSL:\n * the prefix has been changed to OSSL_RBT_,\n * the support augment is removed,\n * _RBT_CHECK()/_RBT_POISON() got removed too,\n * _RB_REMOVE() resets link pointer to NULL in debug build,\n\nThe documentation is based on tree(3) from OpenBSD, originally authored\nby Niels Provos .\n\nCo-Authoerd-by: David Gwynne \nCo-Authored-by: Niels Provos \nCo-Authored-by: Andrew Dinh \n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Bob Beck \nReviewed-by: Norbert Pocs \nMergeDate: Tue Aug 11 19:17:41 2026\n(Merged from https://github.com/openssl/openssl/pull/32033)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e91ef97992fa48d732b0023b2b20ba4bd50ecd27", "date": "2026-07-21T00:33:18+02:00", "subject": "Add a red-black tree implementation", "body": "The code comes from David Gwynne .\nThe same implmentation can be found in OpenBSD.\n\nSeveral minor adjustments have been done to include the code\ninto OpenSSL:\n * the prefix has been changed to OSSL_RBT_,\n * the support augment is removed,\n * _RBT_CHECK()/_RBT_POISON() got removed too,\n * _RB_REMOVE() resets link pointer to NULL in debug build,\n\nThe documentation is based on tree(3) from OpenBSD, originally authored\nby Niels Provos .\n\nCo-Authoerd-by: David Gwynne \nCo-Authored-by: Niels Provos \nCo-Authored-by: Andrew Dinh \n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Bob Beck \nReviewed-by: Norbert Pocs \nMergeDate: Tue Aug 11 19:15:59 2026\n(Merged from https://github.com/openssl/openssl/pull/32033)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "d22e909c5fa6ecb027f9cdb888132417bf44f182", "date": "2026-07-21T06:34:00+08:00", "subject": "Apply suggestions from code review", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "2b412f098e1d6e82e5646265f73779acb1ba6a68", "date": "2026-07-21T01:57:57+03:00", "subject": "feat: initial WASM RPC implementation (RPC handshake incomplete)", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "angular", "sha": "6bfbfcc3a1d5ae8a9a94ce9278d257c1eb2caaa0", "date": "2026-07-20T23:07:57Z", "subject": "refactor(core): expose debuggableFn for effects in signal debug graph", "body": "Exposes the internal fn reference of effect reactive nodes on DebugSignalGraphNode as debuggableFn. This allows Angular DevTools to inspect and set breakpoints on effect callbacks.", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "neovim", "sha": "c6717f1f38c86e61099cfab4be9527f326a0d3b0", "date": "2026-07-20T23:32:50Z", "subject": "perf(ui): blend and combine title default attr once", "body": "", "style": "conventional", "type": "perf", "scope": "ui", "breaking": false} {"repo": "neovim", "sha": "1f045b62ba4d6f74e2afe35a098235ed3a0badaf", "date": "2026-07-20T23:40:59Z", "subject": "perf(ui): defer title attr resolution", "body": "In case of an overflow, the iteration will\nbreak before the grid_line_puts call.", "style": "conventional", "type": "perf", "scope": "ui", "breaking": false} {"repo": "postgres", "sha": "4ca78f292de7b86495cec0a416dc047d94df96c4", "date": "2026-07-20T17:11:21-07:00", "subject": "Add logical decoding status to pg_control_checkpoint().", "body": "Commit 8108765f04b added the logical decoding status to the\npg_controldata output, but overlooked the pg_control_checkpoint() SQL\nfunction, which reports the same checkpoint information. This commit\nadds a logical_decoding column to pg_control_checkpoint(), placed\nafter full_page_writes to match the pg_controldata output order.\n\nOversight in 8108765f04b.\n\nBump catalog version.\n\nReported-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAHGQGwEkp1-1n5iC38+yHSNh955+KshwtCL6DzA0vk_vuUF_Eg@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "ce85e8ad27fb3f4f2763eb8e64e1ea61c59e8cbe", "date": "2026-07-21T01:46:36+01:00", "subject": "fix(runtime-vapor): hydrate vapor components in deferred vdom hydration via interop", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "69211821539136ee51c67a1512c9edf1bdbfe86d", "date": "2026-07-21T02:22:45+01:00", "subject": "fix(runtime-vapor): contain setup errors in prod", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "810a1ef0b3ab1a8b3398213a5784c0207603f775", "date": "2026-07-21T09:40:02+08:00", "subject": "feat(compiler-vapor): make event delegation opt-in", "body": "* feat(compiler-vapor): make event delegation opt-in\n\nAttach Vapor DOM event listeners directly by default and use the\nVapor-only .delegate modifier to opt into document-level delegation.\n\nWarn and fall back to direct listeners for dynamic or unsupported\nevents, and remove the eventDelegation compiler option.\n\nThis prevents ancestor stopPropagation() calls from blocking delegated\nchild handlers while retaining delegation as an explicit optimization.\n\nBREAKING CHANGE: Vapor event listeners are no longer delegated by\ndefault, and the eventDelegation compiler option has been removed.", "style": "conventional", "type": "feat", "scope": "compiler-vapor", "breaking": false} {"repo": "vue-core", "sha": "208f1ed3bbe0079f7ca5d6024fb9e4899e7fdf6f", "date": "2026-07-21T02:54:09+01:00", "subject": "fix(runtime-vapor): restore instance context when handling async setup result", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "42a718673cb09eda9f3b599faa1352fa792b5edb", "date": "2026-07-21T03:33:26+01:00", "subject": "fix(runtime-vapor): create slot anchor when hydrating unwrapped interop slot content", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "openssl", "sha": "420fd332d3ae72b98016271ea1cd030055f772a3", "date": "2026-07-21T12:54:28+10:00", "subject": "param parsing: cleaned generated file includes, flagged things as unused", "body": "Cleaned up the generated files so that they only #include dependencies once\nrather than once per generated parser.\n\nFlagged the generated OSSL_PARAM list and parsing function with ossl_unused\nso that they won't generate warnings if compilation unit doesn't use them.\nThis is necessary if the generated parser is included from different files.\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "567c515bd0f2897269e0b44c1f1c2c5c2fdb12a5", "date": "2026-07-21T04:46:10+01:00", "subject": "fix(runtime-vapor): apply transition hooks before suspense insertion", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "d774576f6f05f65e3a944eb509ef0620ea6b107e", "date": "2026-07-21T09:23:11+05:30", "subject": "Allow logical replication workers to ignore default_transaction_read_only.", "body": "Sequence synchronization updates sequence state via setval(), which\nexplicitly calls PreventCommandIfReadOnly(). If\ndefault_transaction_read_only is enabled on the subscriber, this causes\nsequencesync workers to fail with \"cannot execute setval() in a read-only\ntransaction\". Apply and tablesync workers are not affected, since they\nwrite via direct heap access rather than through these read-only-checked\nfunctions.\n\nRather than special-casing sequencesync, override\ndefault_transaction_read_only to \"off\" for all logical replication workers\nin InitializeLogRepWorker(), the same way session_replication_role and\nsearch_path are already forced there. This keeps the initialization\nuniform.\n\nFor PG-19, we kept the fix narrow by overriding\ndefault_transaction_read_only to \"off\" only for sequencesync workers.\n\nReported-by: Noah Misch \nAuthor: vignesh C \nReviewed-by: Amit Kapila \nBackpatch-through: 19\nDiscussion: https://postgr.es/m/20260710045217.f0.noahmisch@microsoft.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "800e12ac8a3b8f011e29478e153be81c15f22d1f", "date": "2026-07-20T23:59:19-04:00", "subject": "build(vim-patch): n/a runtime/doc/*.txt updates", "body": "- Vim version header (date)\n- N/A Vimscript functions\n - \"listener_\" and \"redraw_listener_\" API\n- N/A line changes\n - \"^.*{.* compiled with .*}$\"", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vite", "sha": "f6547698b1b71387d7337b8ceac92d9b8aa68e1c", "date": "2026-07-21T13:10:29+08:00", "subject": "chore: fix ts erros in test cases", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "f3965519e9b41e1f26c4eadb93e3ab0cbcdb7693", "date": "2026-07-21T14:49:46+09:00", "subject": "fix(fastify): avoid double prefixing inherited middleware", "body": "", "style": "conventional", "type": "fix", "scope": "fastify", "breaking": false} {"repo": "vitest", "sha": "3b8d5257ffa84c275b61293836f872198c41ac2a", "date": "2026-07-21T14:54:54+09:00", "subject": "test: use top-level fsModuleCache option", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vite", "sha": "0a3c1dd20976bdbcaaf9ed46e85c417952b78e95", "date": "2026-07-21T14:00:57+08:00", "subject": "docs: add file extension to import for native config loader compat", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9170c8b71693cf76a4b9a9fbcd16b1fb20b3357c", "date": "2026-07-21T08:33:25+02:00", "subject": "Test what BEFORE UPDATE triggers do to FOR PORTION OF", "body": "If a BEFORE trigger changes NEW.valid_at, what is the interaction with\nFOR PORTION OF? This commit gives a test to capture our current\nbehavior: The trigger's change replaces the value we computed\nautomatically, but it does not change the bounds of the temporal\nleftovers.\n\nThis matches the behavior of MariaDB. On the other hand, DB2 rejects\nchanging the start/end columns of a PERIOD. Since we don't have\nPERIODs, we can't reject the change at trigger definition time as DB2\ndoes, but we could reject it at run time by comparing the values\nbefore and after running triggers.\n\nAuthor: Paul A. Jungwirth \nDiscussion: https://www.postgresql.org/message-id/CA%2BrenyV3Cr9BvWsPeb1t8b%3DPk24apuzyGbubAEs_YsgLUTfXpg%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "abb7108e25ab40753a1aa352b5f03f651bd3619e", "date": "2026-07-21T07:42:20+01:00", "subject": "Add vertical bar next to expanded lines.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "8aaae25e56a3498c13149f5581808b8b65143bad", "date": "2026-07-21T08:00:39+01:00", "subject": "Fix a merge and add test bits.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "418ddada23d290ab02417b62213b402a08f4d9b5", "date": "2026-07-21T07:12:49Z", "subject": "Increase escape-time for tmux's own queries (refresh-client -l and window size) as well, GitHub issue 5388 from Jackson Edmonds.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0e2576a842104886f3e33f0e62928551e54309cb", "date": "2026-07-21T07:20:18Z", "subject": "Use correct cursor position for range end, GitHub issue 5414.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "ec4682737783c1415bdfd834b6ad67f1cabd5d3c", "date": "2026-07-21T15:26:55+08:00", "subject": "chore(sfc-playground): limit displayed pre-release versions", "body": "", "style": "conventional", "type": "chore", "scope": "sfc-playground", "breaking": false} {"repo": "angular", "sha": "4461e8fdbc47647c4b87a9f79589d4af2ac93c64", "date": "2026-07-21T07:42:57Z", "subject": "build: update devinfra commit and correct module file reference in ts version validation", "body": "Update devinfra sha to latest version.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "fc6425fbe7a1fc7458fc27060ec78c5d98757690", "date": "2026-07-21T16:50:18+09:00", "subject": "Improve generate_partition_qual()'s cache handling on out-of-memory errors", "body": "An in-flight failure when trying to set rd_partcheckcxt or\nrd_partcheck, while for example doing an allocation in copyObject(),\nwould leave a backend cache in a corrupted state. The operations are\nnow ordered so as we avoid a leak in the cache memory context and a\nsemi-filled cache state when an allocation failure happens.\n\nThis is unlikely going to be hit in practice. Like the other\nimprovements of this kind, no backpatch is done.\n\nReported-by: Alexander Lakhin \nAuthor: Matthias van de Meent \nDiscussion: https://postgr.es/m/95c64dc2-3abe-4f4e-b285-4c681f565d9f@gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "b5f8518379b77c3b62a7a9d2b52f6c76cda09bd5", "date": "2026-07-21T16:21:40+08:00", "subject": "chore: allow `@parcel/watcher` builds", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vite", "sha": "abb793e18c92592c21fbb8e1f3fc450b5839f04f", "date": "2026-07-21T10:28:56+02:00", "subject": "fix(css): rewrite urls in OnceExit-injected content", "body": "", "style": "conventional", "type": "fix", "scope": "css", "breaking": false} {"repo": "ant-design", "sha": "501aba50e68e7f4394977edfe46e8523bd64d4ff", "date": "2026-07-21T16:29:37+08:00", "subject": "fix(Table): preserve nested table top border inside custom content", "body": "", "style": "conventional", "type": "fix", "scope": "Table", "breaking": false} {"repo": "curl", "sha": "ab8d771d31f015327b555981992ade034603acf0", "date": "2026-07-21T10:30:42+02:00", "subject": "lib: reader/writer fixes", "body": "- clear writer paused bit when destroying writer stack\n- remove the reader paused bit that turned out not to be a good idea\n\nfollow-up to b41c28e70af47e1\n\nCloses #22361", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "22116feb775b0f407eeed9c04a8b58dd303c7ac5", "date": "2026-07-21T10:08:43+01:00", "subject": "fix: set non-zero exit code when teardown throws during close", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ef398036d1ad1acb707dd5fca6c5a02d21516860", "date": "2026-07-21T11:18:53+02:00", "subject": "Add test for CVE-2026-63076", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:11:28 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e77160c3780c10e5364ab08548072e64d4965110", "date": "2026-07-21T11:18:53+02:00", "subject": "Add test for CVE-2026-63076", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:11:12 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7f48cc3d0bff7fc1e78410b4ed6f58724b1f7b1b", "date": "2026-07-21T11:18:53+02:00", "subject": "Add test for CVE-2026-63076", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:10:52 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7b84dec93fc02b8e97a7b620c6195cbaee9e62d9", "date": "2026-07-21T11:18:53+02:00", "subject": "Add test for CVE-2026-63076", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:10:00 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6fa8ce64f35df0bf7f74cdd6f1cedf2637642c5f", "date": "2026-07-21T11:18:53+02:00", "subject": "Add test for CVE-2026-63076", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:12:03 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2647bfe5d7e1868f436cb0d4bcb75d1822744b81", "date": "2026-07-21T11:18:53+02:00", "subject": "Add test for CVE-2026-63076", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:11:44 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "48c9aaf6cc3195d92fa31a3afc39b984778b5700", "date": "2026-07-21T11:19:29+02:00", "subject": "Fix Remote NULL deref in ossl_cmp_calc_protection() via crafted protectionAlg", "body": "ossl_cmp_calc_protection() only checked whether the protectionAlg parameter\n(ppval) was NULL before treating it as a PBMParameter ASN1_STRING.\n\nX509_ALGOR_get0() does not validate the ASN.1 type of the parameter against what\nthe caller expects. For id-PasswordBasedMAC, a crafted message can encode the\nparameter as a BOOLEAN instead of the expected PBMParameter SEQUENCE. Because\nthe ASN1_TYPE value union overlays the boolean int on the pointer field, ppval\ncomes back as a bogus non-NULL pointer (e.g. 0xff).\n\nFixes CVE-2026-63076\n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:11:29 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "49de27169f4cc42619a096c38412e47a3f890803", "date": "2026-07-21T11:19:29+02:00", "subject": "Fix Remote NULL deref in ossl_cmp_calc_protection() via crafted protectionAlg", "body": "ossl_cmp_calc_protection() only checked whether the protectionAlg parameter\n(ppval) was NULL before treating it as a PBMParameter ASN1_STRING.\n\nX509_ALGOR_get0() does not validate the ASN.1 type of the parameter against what\nthe caller expects. For id-PasswordBasedMAC, a crafted message can encode the\nparameter as a BOOLEAN instead of the expected PBMParameter SEQUENCE. Because\nthe ASN1_TYPE value union overlays the boolean int on the pointer field, ppval\ncomes back as a bogus non-NULL pointer (e.g. 0xff).\n\nFixes CVE-2026-63076\n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:11:13 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a17cc8d612ecff6d94a9b7ca8b5283ddf5ff570e", "date": "2026-07-21T11:19:29+02:00", "subject": "Fix Remote NULL deref in ossl_cmp_calc_protection() via crafted protectionAlg", "body": "ossl_cmp_calc_protection() only checked whether the protectionAlg parameter\n(ppval) was NULL before treating it as a PBMParameter ASN1_STRING.\n\nX509_ALGOR_get0() does not validate the ASN.1 type of the parameter against what\nthe caller expects. For id-PasswordBasedMAC, a crafted message can encode the\nparameter as a BOOLEAN instead of the expected PBMParameter SEQUENCE. Because\nthe ASN1_TYPE value union overlays the boolean int on the pointer field, ppval\ncomes back as a bogus non-NULL pointer (e.g. 0xff).\n\nFixes CVE-2026-63076\n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:10:53 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "022a99e2b50cccb94b7d3efb7080077c28276272", "date": "2026-07-21T11:19:29+02:00", "subject": "Fix Remote NULL deref in ossl_cmp_calc_protection() via crafted protectionAlg", "body": "ossl_cmp_calc_protection() only checked whether the protectionAlg parameter\n(ppval) was NULL before treating it as a PBMParameter ASN1_STRING.\n\nX509_ALGOR_get0() does not validate the ASN.1 type of the parameter against what\nthe caller expects. For id-PasswordBasedMAC, a crafted message can encode the\nparameter as a BOOLEAN instead of the expected PBMParameter SEQUENCE. Because\nthe ASN1_TYPE value union overlays the boolean int on the pointer field, ppval\ncomes back as a bogus non-NULL pointer (e.g. 0xff).\n\nFixes CVE-2026-63076\n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:10:03 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "37882aa2e0256e1072442a8f62f7db45b995c45b", "date": "2026-07-21T11:19:29+02:00", "subject": "Fix Remote NULL deref in ossl_cmp_calc_protection() via crafted protectionAlg", "body": "ossl_cmp_calc_protection() only checked whether the protectionAlg parameter\n(ppval) was NULL before treating it as a PBMParameter ASN1_STRING.\n\nX509_ALGOR_get0() does not validate the ASN.1 type of the parameter against what\nthe caller expects. For id-PasswordBasedMAC, a crafted message can encode the\nparameter as a BOOLEAN instead of the expected PBMParameter SEQUENCE. Because\nthe ASN1_TYPE value union overlays the boolean int on the pointer field, ppval\ncomes back as a bogus non-NULL pointer (e.g. 0xff).\n\nFixes CVE-2026-63076\n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:12:04 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a1f348ccb328c3afbd4ba6883f9b7c813c043259", "date": "2026-07-21T11:19:29+02:00", "subject": "Fix Remote NULL deref in ossl_cmp_calc_protection() via crafted protectionAlg", "body": "ossl_cmp_calc_protection() only checked whether the protectionAlg parameter\n(ppval) was NULL before treating it as a PBMParameter ASN1_STRING.\n\nX509_ALGOR_get0() does not validate the ASN.1 type of the parameter against what\nthe caller expects. For id-PasswordBasedMAC, a crafted message can encode the\nparameter as a BOOLEAN instead of the expected PBMParameter SEQUENCE. Because\nthe ASN1_TYPE value union overlays the boolean int on the pointer field, ppval\ncomes back as a bogus non-NULL pointer (e.g. 0xff).\n\nFixes CVE-2026-63076\n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Sat Aug 22 06:11:45 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "61585cc984ba859efc0b18dde277424db5cc88f1", "date": "2026-07-21T11:38:38+02:00", "subject": "docs: fix styling of logos", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "f4d17bb31f80075b26ff6963c06715e712b00dab", "date": "2026-07-21T11:38:48+02:00", "subject": "docs: fix styling of logos, update sponsors list", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "ee9a9ab678f53a78c02307d27521511660afd6b2", "date": "2026-07-21T18:15:19+08:00", "subject": "fix: respect line type design token in component borders", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "eslint", "sha": "2fee9bb7416116cbed4d8c8100b1ae713b6356a1", "date": "2026-07-21T13:34:14+03:00", "subject": "feat: export `ConfigObject` from `eslint/config`", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "angular", "sha": "a79eb0ded36976b1b8f336ba9853e1f18daeb9fa", "date": "2026-07-21T10:45:13Z", "subject": "build: update devinfra digest to 24f9214", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a71ee754ee4bb88fc2d2fbe67972f24c5a6d7883", "date": "2026-07-21T12:53:41+02:00", "subject": "build(api): unused labels in generated Lua bindings #40880", "body": "Problem:\nGenerated API bindings for Lua contain unused jump labels when the API\nfunction has a non-KeyDict (e.g., Array) param after a KeyDict param, causing\nthe build to fail.\n\nSolution:\nDon't output a label when the current parameter is not a KeyDict but the\nprevious one is.", "style": "conventional", "type": "build", "scope": "api", "breaking": false} {"repo": "vitest", "sha": "bd825d961b93597bb00bc0fd1836be2daeceaa83", "date": "2026-07-21T12:59:25+02:00", "subject": "chore: fix right click test", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vitest", "sha": "0f8e63f54759c10bdc17c33aea9af32da6ddda87", "date": "2026-07-21T14:18:56+03:00", "subject": "refactor: `@vitest/browser` to not contain V8 coverage logic", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "postgres", "sha": "02597b42e5ff0df44e6a089801b7f6830ad19d83", "date": "2026-07-21T13:27:15+02:00", "subject": "Unify error messages", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "545cdd4b504310df7d8823a3acf6764b77b4f05b", "date": "2026-07-21T13:28:59+02:00", "subject": "asyn-thrdd: retry link-local ipv6 if missing scope id", "body": "When the threaded resolver gets AAAA results that carry a link-local\naddress without scope-id, it now re-queues a query with AF_UNSPEC and\nstrips ipv4 addresses from that result. Whatever the resulting addresses\nand scope-ids are, this becomes the result of the resolve.\n\nFixes #22330\nReported-by: Bartel Sielski\nCloses #22368", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e31d75713e0772b7264c0d73958ab4dfbb83250c", "date": "2026-07-21T13:47:30+02:00", "subject": "rand: factor out seed source creation and fix store race", "body": "rand_get0_primary() created the seed source ahead of the primary DRBG\nand stored both under the write lock at the end. When two threads\nraced to create the primary DRBG and one of them failed after having\ncreated and stored the seed source, the other thread overwrote the\nstored seed source with its own instance, leaking the previous one.\n\nFactor the get-or-create logic out into rand_get0_seed() which\nrechecks the global under the write lock, so exactly one instance is\nkept and returned to every racing thread, and the seed source is\nstored independently of the primary DRBG creation outcome.\n\nAdd a regression test checking that creating the primary DRBG creates\nand stores the seed source and that later use keeps returning the same\ninstance.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Paul Dale \nReviewed-by: Matt Caswell \nReviewed-by: Ryan Hooper \nMergeDate: Mon Jul 27 09:13:25 2026\n(Merged from https://github.com/openssl/openssl/pull/32029)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1e9b70735ec4d616219f9c0173d277edd6889db6", "date": "2026-07-21T13:47:30+02:00", "subject": "rand: factor out seed source creation and fix store race", "body": "rand_get0_primary() created the seed source ahead of the primary DRBG\nand stored both under the write lock at the end. When two threads\nraced to create the primary DRBG and one of them failed after having\ncreated and stored the seed source, the other thread overwrote the\nstored seed source with its own instance, leaking the previous one.\n\nFactor the get-or-create logic out into rand_get0_seed() which\nrechecks the global under the write lock, so exactly one instance is\nkept and returned to every racing thread, and the seed source is\nstored independently of the primary DRBG creation outcome.\n\nAdd a regression test checking that creating the primary DRBG creates\nand stores the seed source and that later use keeps returning the same\ninstance.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Paul Dale \nReviewed-by: Matt Caswell \nReviewed-by: Ryan Hooper \nMergeDate: Mon Jul 27 09:13:12 2026\n(Merged from https://github.com/openssl/openssl/pull/32029)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "abd2f3a9a57851b68d93dcfc95651671c49dd1a2", "date": "2026-07-21T11:52:13Z", "subject": "Add copy-mode-current-line-style to set style for current line in copy mode, GitHub issue 5391 from Leo Henon.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "9d583cad5af9792c42fea65f968710546facf4bd", "date": "2026-07-21T11:53:41Z", "subject": "If drawing border but not actually on a border, use the default line style.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "3e61e63851f343439f63bfd2bb5ca357dfc60dc0", "date": "2026-07-21T12:56:32+01:00", "subject": "Test for outside panes with no border.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a87e91677c8d25d4db9454f96fae3abf295ffdd5", "date": "2026-07-21T16:58:44+05:00", "subject": "fix(system): use native path for Windows executables #40885", "body": "Problem:\nWhen 'shellslash' is set, exepath() returns forward-slashed paths.\nPassing that path to vim.system() can prevent cmd.exe from launching.\n\nSolution:\nConvert the resolved executable path to native separators before spawning.\nAdd a Windows regression test for cmd.exe with 'shellslash' enabled.", "style": "conventional", "type": "fix", "scope": "system", "breaking": false} {"repo": "tmux", "sha": "8bf2a4e807a092970dabfac89936d6e887389ecd", "date": "2026-07-21T12:28:43Z", "subject": "Add a -C flag to new-pane to have a modal pane close when the mouse is clicked outside it, GitHub issue 5400.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "65df7468dfff120436242e9348c25917f9101743", "date": "2026-07-21T14:31:48+02:00", "subject": "chore: times option", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vite", "sha": "2ced1fe4e4e480ed78cb7aa5c78319e57bfa7783", "date": "2026-07-21T20:56:17+08:00", "subject": "fix(config): exclude virtual modules from native config compat check", "body": "", "style": "conventional", "type": "fix", "scope": "config", "breaking": false} {"repo": "neovim", "sha": "6d5b774362e9d9b586280c723060813ed7ca8757", "date": "2026-07-21T08:56:42-04:00", "subject": "build(luals): drop src/.luarc.json, test/.luarc.json #40887", "body": "Problem:\nemmylua reports `unresolved-require` for `require('test.testutil')` and\nfriends, and a file under `test/` cannot even resolve its own siblings.\n\nThis happens because `test/.emmyrc.json` and `test/.luarc.json`, cause\nvim.lsp to root the workspace at `test/` instead of the repo top-level.\n\nSolution:\nDrop the nested configs:\n\n src/.luarc.json\n test/.luarc.json\n test/.emmyrc.json\n\nOutcomes:\n- No \"libraries\" needed: luv types come from `runtime/lua/uv/_meta.lua`,\n not `${3rd}/luv` (which would only duplicate them).\n- Drop `test/` from the root `.luarc.json` `ignoreDir` so its modules are\n resolvable by LuaLS/Emmylua.\n - Note: this means that LuaLS diagnostics will now be reported for\n `test/`, but that is not a bad thing...", "style": "conventional", "type": "build", "scope": "luals", "breaking": false} {"repo": "tmux", "sha": "c59db2d007351f0da953334b81f5394508204df9", "date": "2026-07-21T13:04:01Z", "subject": "Invalidate scene when window is resized.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "39b1ddf5b9dfec07de8e05d96e8845117c54edc5", "date": "2026-07-21T14:22:39+01:00", "subject": "fix(runtime-vapor): sync transition hooks on interop update", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "c90c9678e533e622a393d3296f8fe4b9e5e317af", "date": "2026-07-21T15:27:02+02:00", "subject": "Message style fixes", "body": "Change DETAIL messages to conform to the style guide by capitalizing\nthe first word of sentences and ending sentences with a period.\n\nAuthor: Peter Smith \nReviewed-by: Chao Li \nReviewed-by: vignesh C \nReviewed-by: Xiaopeng Wang \nReviewed-by: Peter Eisentraut \nDiscussion: https://www.postgresql.org/message-id/flat/CAHut%2BPszSntkUgN%2BQa9matGY6MLEoFGSuVbuKDgnnTdZ7YPRwg%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "a57b0f1282129231d3e0934062991feb11381976", "date": "2026-07-21T13:27:41Z", "subject": "Clip floating panes exactly at window edge, don't let the border creep into the first outside column.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "5ceb20867de5b60e239dd14150bf937d1f9c886b", "date": "2026-07-21T15:31:33+02:00", "subject": "chore: offsets", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "tmux", "sha": "2848a12562c0958d0fef68f9ead7c52fa59f877e", "date": "2026-07-21T14:41:52+01:00", "subject": "Update tests for fix to floating panes at window edges.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "3e41d52536dfe9ccd7440248824cdc0d3d43b2aa", "date": "2026-07-21T14:47:14+01:00", "subject": "fix(runtime-core,runtime-vapor): restore undefined effect scope after `setCurrentInstance`", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-core,runtime-vapor", "breaking": false} {"repo": "vitest", "sha": "fb959b826fa0724305a8e0dacab0c1cc28815249", "date": "2026-07-21T15:51:47+02:00", "subject": "chore: types", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vitest", "sha": "15e0af953fd67e445324ceccda4547526e092cb4", "date": "2026-07-22T01:53:16+12:00", "subject": "feat: support non-ascii characters in `for/each` title placeholders", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "7d63723a0e3784545f5566656c6db58ea7f31a73", "date": "2026-07-21T15:01:08+01:00", "subject": "fix(runtime-vapor): defer mount hooks to owning suspense boundary", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "tmux", "sha": "b49cc2d936e2a142934c6e827c5e7d903782acd3", "date": "2026-07-21T15:03:52+01:00", "subject": "Fix stray L on empty commands.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "af740dadd83dc111f87dab4da86a6dbfb1ed93b1", "date": "2026-07-21T16:14:04+02:00", "subject": "chore: update middle click test", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "angular", "sha": "fb68711f04e26109dc8d5dfa2f4c31023f961089", "date": "2026-07-21T17:22:58+03:00", "subject": "feat(docs-infra): react with Angie in the playground minigame", "body": "The angle-guessing minigame's result popup personifies your score with a\nhand-drawn stick-figure that changes expression by accuracy. Now that Angie\nappears across adev — the 404 page, docs search, the embedded editor, the\ntutorial intros and completions — give the minigame the same treatment so\nits result feels consistent with the rest of the docs.\n\nReplace the result reactor's stick-figure illustration with an Angie pose\nkeyed to accuracy (seven tiers, superhero down to angry). She is revealed\nonce the accuracy counter finishes counting up, popping in beside a speech\nbubble that carries the round's existing quote. The result popup is widened\nso Angie and the bubble sit side by side, and the share link is moved from\ntwitter.com to x.com.\n\nThis removes the previous hand-drawn stick-figure result art. NG the Angle,\nthe interactive character in the play area, is unchanged.\n\n(cherry picked from commit 8422da814d8311d605eaba8b99bf3bc63c21241a)", "style": "conventional", "type": "feat", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "8422da814d8311d605eaba8b99bf3bc63c21241a", "date": "2026-07-21T17:22:58+03:00", "subject": "feat(docs-infra): react with Angie in the playground minigame", "body": "The angle-guessing minigame's result popup personifies your score with a\nhand-drawn stick-figure that changes expression by accuracy. Now that Angie\nappears across adev — the 404 page, docs search, the embedded editor, the\ntutorial intros and completions — give the minigame the same treatment so\nits result feels consistent with the rest of the docs.\n\nReplace the result reactor's stick-figure illustration with an Angie pose\nkeyed to accuracy (seven tiers, superhero down to angry). She is revealed\nonce the accuracy counter finishes counting up, popping in beside a speech\nbubble that carries the round's existing quote. The result popup is widened\nso Angie and the bubble sit side by side, and the share link is moved from\ntwitter.com to x.com.\n\nThis removes the previous hand-drawn stick-figure result art. NG the Angle,\nthe interactive character in the play area, is unchanged.", "style": "conventional", "type": "feat", "scope": "docs-infra", "breaking": false} {"repo": "vitest", "sha": "9a632dbb023e995cea32326d13c3dc1dc152f85e", "date": "2026-07-21T16:35:00+02:00", "subject": "chore: improve drag and drop test", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "postgres", "sha": "078aac2ed649f983e52f039ded9014bf6447963e", "date": "2026-07-21T16:56:37+02:00", "subject": "pg_upgrade: Message wording fix", "body": "For internally consistent terminology", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7d1b39fc3d9aca1175d00b2859d896f6e81ecad4", "date": "2026-07-21T17:18:06+02:00", "subject": "Remove assertion added by commit 7dcea51c2a4d", "body": "We've got no reports of problems. Get rid of it.\n\nAuthor: Álvaro Herrera \nBackpatch-through: 19\nDiscussion: https://postgr.es/m/alewd1f2G0kKeM1i@alvherre.pgsql", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d9d235510e10fb61457744fc7798852c46260f29", "date": "2026-07-21T11:59:37-04:00", "subject": "doc: clarify to_char(\"OF\") HH/MM doesn't represent actual chars", "body": "Change formatting and chars to be less of a match against actual\nformatting characters.\n\nReported-by: Phil\n\nDiscussion: https://postgr.es/m/177801333530.795.16999885814007014333@wrigleys.postgresql.org\n\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9fa107f07f06d8b0281f729931512ee896aa3440", "date": "2026-07-21T12:06:10-04:00", "subject": "doc: clarify how TIMESTAMP WITH TIME ZONE behaves", "body": "Mention \"time zone conversion\" as a way to clarify the time zone is not\nstored in the database.\n\nReported-by: Richard Neill\n\nDiscussion: https://postgr.es/m/ddf41f033a8add84e1f28a095defafae@richardneill.org\n\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f8cfd7f06ae87d47aac04ddce860b302f7820858", "date": "2026-07-21T18:43:55+02:00", "subject": "feat(options): schema, \"dict\" options, messages", "body": "Problem:\nOptions parsing is still painful for dict-style options.\n\nSolution:\nschema-maxxing => better `opt:get()` (will be the basis for `vim.o()`),\nunified (and more-detailed) err msgs.\n\n- Drop bespoke structure-builder in `_core/options.lua`.\n- Define `schema` for all non-primitive options (except 'guicursor' and\n statusline-style options); generate reified keysets `OptKeyDict`).\n - Generate 'fillchars' => `fcs_tab`, 'listchars' => `lcs_tab`.\n- `nvim_set_option_value`:\n - Return the improved structures. Also from `vim.opt.x:get()`.\n - Eliminate api <=> lua roundtrip, centralize option structure\n handling.\n- Improve/unify errors.\n - Bump ERR_BUFLEN 80 → 256 so the \"one of\" list isn't truncated.\n- Eliminate old 'diffopt' order-dependence (`iwhiteall` before `iwhite`)\n\nError samples:\n\n Typed-key path (opt_strings_check → diffopt/mousescroll/breakindentopt):\n E474: Unknown item 'foo'\n E474: 'context' requires a number\n E474: 'ver' number is out of range\n E474: 'algorithm' must be one of: myers, minimal, patience, histogram\n E474: 'filler' does not take a value\n\nRelated:\n\n- #31084\n- #34661\n- #31820\n- #14739\n- #20107\n- fix #18875\n - :get() returns `{ sbr = true, shift = '3' }` (reified-keyset) instead of `{'sbr', 'shift:3'}`\n - Setting via table now works too. `object_as_optval_for` `is_map` now recognizes struct options.\n- fix #30296\n - instead of `E474: Invalid argument`, errors now look like:\n ```\n E474: Invalid value 'x', expected one of: single, double: ambiwidth=x\n E474: Unknown item 'foo': diffopt=foo\n E474: 'context' requires a number: diffopt=context:x\n ```\n\nsimplify `win_float_parse_option` from #26799.", "style": "conventional", "type": "feat", "scope": "options", "breaking": false} {"repo": "vitest", "sha": "e38b94f9db3fdd8b5039ed94a2875dcd39548a34", "date": "2026-07-21T19:25:24+02:00", "subject": "docs: fix latitude logo path", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9180c6e093391969b2751d09b6fa0397dbfc7882", "date": "2026-07-21T20:32:13+03:00", "subject": "fix(lsp): guard linked_editing_range refresh on attach", "body": "Problem:\non_attach() calls refresh(), but there is no guarantee the attached\nbuffer is the current buffer. This can make linked editing request\nhandling assume the wrong window.\n\nSolution:\nCall refresh() only if the current buffer is attached. This keeps the\ninitial highlighting behavior while avoiding making incorrect request.", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "tmux", "sha": "d8ad7cd9c53b4e75fefaa49498952a368d2f3708", "date": "2026-07-21T18:48:16+01:00", "subject": "Add exit status on right.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "5948722d5aaa5875c24b0c911a81ae77b94e3265", "date": "2026-07-21T19:19:31+01:00", "subject": "Add highlighting to the current line.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "381c3d56e332b7decb5392a84ccac0e972b4c56d", "date": "2026-07-21T21:47:33+03:00", "subject": "ngtcp2: clean up after ngtcp2 in `curl_global_cleanup`", "body": "Frees the memory allocated by `ngtcp2_crypto_ossl_init`, thus making\nValgrind happy. Works for versions of ngtcp2 >= 1.24.0.\n\nRef: https://github.com/ngtcp2/ngtcp2/issues/2207\n\nCloses #22363", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "05b241cb094941c7f83ebd94d5516648d82bc2d9", "date": "2026-07-21T14:52:59-04:00", "subject": "Fix missing NULL check in statem_clnt_construct_test", "body": "Fixes https://scan5.scan.coverity.com/#/project-view/60762/10222?selectedIssue=1695905\n\nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMergeDate: Thu Jul 23 08:24:25 2026\n(Merged from https://github.com/openssl/openssl/pull/31992)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "8728a5f6395d6c93384ad84c4c59e58be9602abd", "date": "2026-07-21T22:45:45+03:00", "subject": "Fix top-level enter and exit visitor types", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "1554d587b31460f135adf65812d199f963a34e61", "date": "2026-07-22T03:46:24+08:00", "subject": "chore: Some minor cleanup", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "openssl", "sha": "acddc9b6d02fbe62c31b62c4aeb49736a52bd6ed", "date": "2026-07-21T23:06:43+02:00", "subject": "apps: test dgst -r and -c output formats", "body": "Add coverage for the coreutils style output of `openssl dgst -r`,\nincluding the newline escaping done by newline_escape_filename() when\nthe digested file name contains a newline, and for the colon separated\noutput of `openssl dgst -c`.\n\nThe newline in file name part is skipped where such a file cannot be\ncreated.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Daniel Kubec \nReviewed-by: Ryan Hooper \nMergeDate: Mon Jul 27 10:19:58 2026\n(Merged from https://github.com/openssl/openssl/pull/32035)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "50f990a67a6d116c7e5c2461355b8ddea46bd5d7", "date": "2026-07-21T23:15:11+02:00", "subject": "docs: update attribute testing guide to include canonical example with local component", "body": "(cherry picked from commit 738b8fe9d2a9c164f3601b123cbdbb33bb9a0a04)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "738b8fe9d2a9c164f3601b123cbdbb33bb9a0a04", "date": "2026-07-21T23:15:11+02:00", "subject": "docs: update attribute testing guide to include canonical example with local component", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "eslint", "sha": "eccbe7bc9257a6c99880468e452c8dfafbf47f74", "date": "2026-07-22T06:23:45+09:00", "subject": "test: add error locations to `no-class-assign`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "eslint", "sha": "2d063e224e9bb0ccb0a28c120f7e380a8d62f320", "date": "2026-07-22T06:24:44+09:00", "subject": "chore: update HTTP URLs to HTTPS in JSDoc and comments", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "2c3200f5785fed310d0541d9c428d822248869ed", "date": "2026-07-22T00:05:27+02:00", "subject": "dnscache: avoid double unlock", "body": "Follow-up to 7ad33fe56f25424fb51a1040\n\nPointed out by Zeropath\n\nCloses #22364", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "e62905e536c16cfe7b688de8295f7260cad93a86", "date": "2026-07-22T01:09:27+03:00", "subject": "docs: update stale Twitter reference to X in the URL matcher guide", "body": "Twitter is now X. In the custom-route-matcher guide, point the author-credit\nlink at x.com and refer to an \"X (formerly Twitter) handle\" (clarified once,\nthen \"X handle\"), matching the \"X (formerly Twitter)\" wording already used in\nthe footer, navigation, and update guide.\n\n(cherry picked from commit 6557df5dbec8e205a5cecfc7039d76928ee4000e)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "6557df5dbec8e205a5cecfc7039d76928ee4000e", "date": "2026-07-22T01:09:27+03:00", "subject": "docs: update stale Twitter reference to X in the URL matcher guide", "body": "Twitter is now X. In the custom-route-matcher guide, point the author-credit\nlink at x.com and refer to an \"X (formerly Twitter) handle\" (clarified once,\nthen \"X handle\"), matching the \"X (formerly Twitter)\" wording already used in\nthe footer, navigation, and update guide.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "02047ece21cf8f6126fdf4119a463ef4d950b1cc", "date": "2026-07-22T00:11:18+02:00", "subject": "apps: fail skeyutl -genkey on bad -skeyopt", "body": "app_params_new_from_opts() returns NULL when a key option cannot be\nconverted, either because the key management does not settle the\nparameter or because the opt:value separator is missing. skeyutl\nignored that and generated the key without the requested options, only\nprinting the parameter error to stderr and still exiting successfully.\n\nBail out instead, the way the mac, kdf and enc applications do. A NULL\nreturn is only an error when key options were actually given, as\napp_params_new_from_opts() also returns NULL for an empty option list.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Paul Dale \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Daniel Kubec \nMergeDate: Mon Jul 27 09:06:03 2026\n(Merged from https://github.com/openssl/openssl/pull/32036)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "22f58fd9cce68d32deb8a08ba54fcf2d2eaf0135", "date": "2026-07-22T00:11:24+02:00", "subject": "test: support key options in the fake cipher provider", "body": "The fake cipher key management accepted a key name and raw key bytes on\ngeneration and import, but did not advertise them as settable\nparameters, so applications converting opt:value strings with\nOSSL_PARAM_allocate_from_text() had nothing to look them up in.\n\nAdd the settable parameter table for both generation and import, and\nreport the key name as the key id, so that setting it has an observable\neffect.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Paul Dale \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Daniel Kubec \nMergeDate: Mon Jul 27 09:06:05 2026\n(Merged from https://github.com/openssl/openssl/pull/32036)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "851a95ccfbcaf17eb84b4d63f107c76ec55c513b", "date": "2026-07-22T00:11:29+02:00", "subject": "apps: test skeyutl -skeyopt handling", "body": "Cover the -skeyopt option of skeyutl using the fake cipher provider: a\nkey name passed as a key option shows up in the reported key id, while\nan option unknown to the key management and an option missing the\nopt:value separator both make the command fail.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Paul Dale \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Daniel Kubec \nMergeDate: Mon Jul 27 09:06:06 2026\n(Merged from https://github.com/openssl/openssl/pull/32036)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "1b09130640dd3a74381db9bd21f7ab788d79ba56", "date": "2026-07-21T23:06:16Z", "subject": "feat(devtools): rename DevTools tab to \"Angular & Wiz\" for Googlers", "body": "Conditionally update the browser DevTools tab name to \"Angular & Wiz\" if the user is identified as a Google employee via internal corp network detection.", "style": "conventional", "type": "feat", "scope": "devtools", "breaking": false} {"repo": "postgres", "sha": "e65c331b8fbf8c9632b62c5a9dcb589cbd3046a8", "date": "2026-07-22T08:15:00+09:00", "subject": "Add new pgstats routine to split pending data setup", "body": "This commit adds pgstat_prep_pending_from_entry_ref(), a new pgstats\nroutine that is able to prepare an existing PgStat_EntryRef to receive\npending stats. This split gives a way for callers to obtain first a\nreference via pgstat_get_entry_ref(), then set up pending data as two\nseparate, distinctive, steps.\n\nPreviously, the only way to get an entry reference with pending data\nready was pgstat_prep_pending_entry(), which bundles lookup, creation,\nand pending setup in a single call. Callers that need finer control\nover the entry creation had no way to attach pending data to an\nalready-obtained entry reference. One case where this has shown to\nmatter for a stats kind is where one wants to check some capacity (for\nexample where a GUC bounds the maximum numer of entries allowed) before\ndeciding if a new entry should be created. So this split can help in\nreducing calls to pgstat_get_entry_ref(), meaning less shmem hash table\nlookups.\n\nThe only logical ordering change is that pgStatPendingContext is\ninitialized after calling pgstat_get_entry_ref() in\npgstat_prep_pending_entry(). This does not matter in practice.\n\npgstat_prep_pending_entry() is refactored to use the new function\ninternally. All the existing callers are unchanged. Existing\nout-of-core custom stats kinds should see no impact.\n\nAuthor: Sami Imseih \nReviewed-by: Kyotaro Horiguchi \nReviewed-by: Michael Paquier \nDiscussion: https://postgr.es/m/CAA5RZ0sV+TsLejUMhAM=PJoOm8u-t8ru7B67KvyLCy=19sM87g@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ccfd4b683867a5eded09a82b624992c87396d02d", "date": "2026-07-22T09:52:04+09:00", "subject": "Improve lookup_type_cache() handling on out-of-memory errors", "body": "If an error happens during the initialization of the TYPEOID catcache,\nas part of lookup_type_cache(), the error handling of that lookup would\ncause an assertion failure via finalize_in_progress_typentries(), called\nduring error recovery, the presence of an in-progress type OID causing\na catcache initialization outside of a transaction context.\n\nThe in-progress list is now delayed to happen after the initial entry\nlookup. Alexander Lakhin has found a fancy way to reproduce the\nproblem, with the injection of probabilistic memory allocation failures.\n\nThis problem is unlikely going to show up in practice. Like the other\nchanges of this kind, no backpatch is done.\n\nReported-by: Alexander Lakhin \nAuthor: Matthias van de Meent \nDiscussion: https://postgr.es/m/95c64dc2-3abe-4f4e-b285-4c681f565d9f@gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "429006e3ac914658861ee77cbbcc5edfa7740c4a", "date": "2026-07-22T08:54:12+08:00", "subject": "fix(runtime-vapor): defer post-render effects to vdom suspense boundaries", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "1abb830874fb6fd67c01b05fc2dafd643d30f674", "date": "2026-07-22T09:24:47+08:00", "subject": "fix(type): improve document type detection in utilities", "body": "", "style": "conventional", "type": "fix", "scope": "type", "breaking": false} {"repo": "vue-core", "sha": "d385f8960751dc9866639acd4ac60eb729922d61", "date": "2026-07-22T10:58:48+08:00", "subject": "fix(runtime-vapor): update logical child cache after hydration mutations", "body": "- transfer the last located logical child when mismatch recovery replaces nodes\n- maintain the cache during Transition template unwrapping and fragment cleanup", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vite", "sha": "18535d11cff9ce2a508798146a29d56a8fc5d901", "date": "2026-07-22T12:29:46+08:00", "subject": "perf(config): skip native config compat check when warning is ignored", "body": "", "style": "conventional", "type": "perf", "scope": "config", "breaking": false} {"repo": "vite", "sha": "be9631658f5191ee5c5665e780239d42a330280a", "date": "2026-07-22T13:29:55+09:00", "subject": "fix(hmr): remove hot data after prune", "body": "", "style": "conventional", "type": "fix", "scope": "hmr", "breaking": false} {"repo": "openssl", "sha": "7ef104e779ce00b5e63cbc90586853a28c44284e", "date": "2026-07-22T07:37:25+02:00", "subject": "quic: use @-style Doxygen tags in ossl_quic_srtm_remove doc comment", "body": "Convert the doc comment for ossl_quic_srtm_remove() from backslash-style\nDoxygen tags (\\brief, \\param, \\return) to the @-style equivalents\n(@brief, @param, @returns), matching the STYLE.md convention.\n\nComplements: 63afe639141f \"quic: report SRTM entry match status via an output argument\"\nSigned-off-by: Nikola Pajkovsky \n\nReviewed-by: Saša Nedvědický \nReviewed-by: Jakub Zelenka \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Mon Aug 3 12:03:18 2026\n(Merged from https://github.com/openssl/openssl/pull/32037)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "961643a44c2268b4226a885c89adfdd1d3ace635", "date": "2026-07-22T15:04:00+09:00", "subject": "docs: remove `@experimental` from `resolve.tsconfigPaths` JSDoc", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "6ee41d3351bc0b36d12cee29471484548dee063c", "date": "2026-07-22T10:16:00+03:00", "subject": "docs: outdated section in the `Signal Forms | Custom controls` page", "body": "(cherry picked from commit cb1841d10ba96df1e058b4c3a3778d7a3157a2a2)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "cb1841d10ba96df1e058b4c3a3778d7a3157a2a2", "date": "2026-07-22T10:16:00+03:00", "subject": "docs: outdated section in the `Signal Forms | Custom controls` page", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d01f86f219489d333cb9fafe040c5f9ad13ea969", "date": "2026-07-22T09:13:18+01:00", "subject": "Update CHANGES/NEWS to mention the HollowByte fix", "body": "Add a previously missing CHANGES.md/NEWS.md entry for the Hollowbyte fix.\n\nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Norbert Pocs \nMergeDate: Wed Jul 29 09:19:06 2026\n(Merged from https://github.com/openssl/openssl/pull/32043)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c500fbd8a44600de969452fe533cf97040504707", "date": "2026-07-22T09:13:18+01:00", "subject": "Update CHANGES/NEWS to mention the HollowByte fix", "body": "Add a previously missing CHANGES.md/NEWS.md entry for the Hollowbyte fix.\n\nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Norbert Pocs \nMergeDate: Wed Jul 29 09:18:25 2026\n(Merged from https://github.com/openssl/openssl/pull/32042)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4db90ebab383412fe2b7dc92a5b77bb19d0627ff", "date": "2026-07-22T09:13:18+01:00", "subject": "Update CHANGES/NEWS to mention the HollowByte fix", "body": "Add a previously missing CHANGES.md/NEWS.md entry for the Hollowbyte fix.\n\nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Norbert Pocs \nMergeDate: Wed Jul 29 09:17:49 2026\n(Merged from https://github.com/openssl/openssl/pull/32041)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "659de2429f0735ebc58b905da225ed472141cf55", "date": "2026-07-22T09:13:18+01:00", "subject": "Update CHANGES/NEWS to mention the HollowByte fix", "body": "Add a previously missing CHANGES.md/NEWS.md entry for the Hollowbyte fix.\n\nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Norbert Pocs \nMergeDate: Wed Jul 29 09:17:15 2026\n(Merged from https://github.com/openssl/openssl/pull/32040)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c4c5d2bbd7abe5630d8f0e6a41b23fff4398eb9f", "date": "2026-07-22T09:13:18+01:00", "subject": "Update CHANGES/NEWS to mention the HollowByte fix", "body": "Add a previously missing CHANGES.md/NEWS.md entry for the Hollowbyte fix.\n\nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Norbert Pocs \nMergeDate: Wed Jul 29 09:15:56 2026\n(Merged from https://github.com/openssl/openssl/pull/32039)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b1ad3972091d546198dc7d935fe834e4c5e27fc0", "date": "2026-07-22T09:13:18+01:00", "subject": "Update CHANGES/NEWS to mention the HollowByte fix", "body": "Add a previously missing CHANGES.md/NEWS.md entry for the Hollowbyte fix.\n\nReviewed-by: Paul Dale \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Norbert Pocs \nMergeDate: Wed Jul 29 09:15:32 2026\n(Merged from https://github.com/openssl/openssl/pull/32039)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "bd4ac0299484b0930c0644e48680f3a63ad6a0c7", "date": "2026-07-22T08:19:14Z", "subject": "Add A modifier to cycle through a series of values, GitHub issue 5412 from Fernando Daciuk.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "bbcdb4d97df214a67a43bf583ca7fbc524f40f57", "date": "2026-07-22T10:37:18+02:00", "subject": "refactor(benchmark): simplify provider contract to a default export", "body": "Address review feedback on the pluggable benchmark provider API. The\ninitial design mirrored the coverage provider (a `getProvider` factory\nplus a `BenchmarkProviderModule` wrapper), which was a special case that\nbenchmarks — running only in tests — don't need.\n\n- Drop `BenchmarkProviderModule` and `getProvider`: a custom provider\n module now default-exports a `BenchmarkProvider` directly.\n- Reuse the existing `TestModuleRunner` interface instead of the bespoke\n `BenchmarkProviderLoader`.\n- Remove the `TestRunner.runBenchmarks` override hook, now redundant with\n the provider API, and inline `tinybench.run()` in the default provider.", "style": "conventional", "type": "refactor", "scope": "benchmark", "breaking": false} {"repo": "ant-design", "sha": "afe2c19fc03442ded9675a1260126bbda1108ca8", "date": "2026-07-22T16:55:03+08:00", "subject": "docs(time-picker): sync `status` enum value", "body": "", "style": "conventional", "type": "docs", "scope": "time-picker", "breaking": false} {"repo": "nestjs", "sha": "803fc2c471f4fb986f786a2ccb9dd4bb856d20f7", "date": "2026-07-22T17:57:02+09:00", "subject": "fix(core): serialize SSE retry field when value is zero", "body": "Truthy checks omitted retry: 0, so clients never received an explicit zero reconnection delay. Use isNil like other MessageEvent fields.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "curl", "sha": "474ebb524708ecddf4fe0fdb5493e8d638bdc04d", "date": "2026-07-22T11:06:39+02:00", "subject": "api-guard: check lock on session cache", "body": "Add a property to easy/multi API calls that prohibit calling\nthe function when the involved SSL session cache is under lock\nby the current thread.\n\nChecks are only in effect when pthreads/Windows threads are\navailable.\n\nCloses #22367", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "fb4538ccd77c0d8b0d4b1ec0fd925a438ce0eb05", "date": "2026-07-22T17:23:11+08:00", "subject": "fix(runtime-vapor): preserve pending async component position", "body": "- use a live comment placeholder while async setup is pending\n- avoid duplicate dependency registration during hydration\n- clean up pending placeholders during unmount\n- preserve first-mount and active state for cached pending components", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "angular", "sha": "125a7948916186926a3f8f10a059160be0f86473", "date": "2026-07-22T12:48:47+03:00", "subject": "docs: use https for external links in adev", "body": "Several guides, examples, and the update-guide recommendations linked to\nexternal resources over insecure http. Switch them to https.\n\n(cherry picked from commit 28d59e8d6357b6f5015657f06ecd7b543f00aba0)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "28d59e8d6357b6f5015657f06ecd7b543f00aba0", "date": "2026-07-22T12:48:47+03:00", "subject": "docs: use https for external links in adev", "body": "Several guides, examples, and the update-guide recommendations linked to\nexternal resources over insecure http. Switch them to https.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "d8cd38830251b95fd7dddcd0eee0ce94cc61c2f4", "date": "2026-07-22T19:03:34+09:00", "subject": "test: config CJS module vars in ESM case", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "curl", "sha": "41b4a911234207b7c5271f8147053769570b06e7", "date": "2026-07-22T12:22:18+02:00", "subject": "scripts: replace/extend `--` with `--end-of-options` in git commands", "body": "It makes these scripts require Git 2.24.0+ (2019-11-04).\n\nRefs:\nhttps://git-scm.com/docs/gitcli\nhttps://nesbitt.io/2026/07/21/end-of-options.html\nhttps://github.com/git/git/commit/19e8789b236dfe33667747d5523d6689bb59b5ef\n\nCloses #22369", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "c0368f227846024bb26d3628c59541e870bb36e4", "date": "2026-07-22T05:31:52-05:00", "subject": "fix(common): preserve crossorigin on image preloads", "body": "Propagate the crossorigin attribute from priority NgOptimizedImage hosts to SSR-generated preload links. Keep preload and image requests in the same credentials mode to avoid an anonymous image issuing an earlier credentialed request.\n\n(cherry picked from commit d14696e4308c5cd95cf2184cae3582b52f7c7cf3)", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "angular", "sha": "d14696e4308c5cd95cf2184cae3582b52f7c7cf3", "date": "2026-07-22T05:31:52-05:00", "subject": "fix(common): preserve crossorigin on image preloads", "body": "Propagate the crossorigin attribute from priority NgOptimizedImage hosts to SSR-generated preload links. Keep preload and image requests in the same credentials mode to avoid an anonymous image issuing an earlier credentialed request.", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "tmux", "sha": "eaeb234dd437d36e0b17c4b234be84532928d3b7", "date": "2026-07-22T11:33:46+01:00", "subject": "Remove code to put exit-code on right and go for just simple ! on left of +.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b84ce96f2e6f6ff7d49acacfe308e5a5a7b20c4a", "date": "2026-07-22T12:34:07+02:00", "subject": "DOCS: Fix KEM init function documentation", "body": "Document the missing params argument for OSSL_FUNC_kem_decapsulate_init(),\nremove stale name references, and fix the malformed provider-keymgmt(7) link.\n\nComplements: 78c44b05945b Add HPKE DHKEM provider support for EC, X25519 and X448.\nSigned-off-by: olszomal \n\nReviewed-by: Jakub Zelenka \nReviewed-by: Paul Dale \nMergeDate: Mon Jul 27 07:57:46 2026\n(Merged from https://github.com/openssl/openssl/pull/32045)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3a1c8d428328f628d3ae4842fc237128e6dabc73", "date": "2026-07-22T12:34:07+02:00", "subject": "DOCS: Fix KEM init function documentation", "body": "Document the missing params argument for OSSL_FUNC_kem_decapsulate_init(),\nremove stale name references, and fix the malformed provider-keymgmt(7) link.\n\nComplements: 78c44b05945b Add HPKE DHKEM provider support for EC, X25519 and X448.\nSigned-off-by: olszomal \n\nReviewed-by: Jakub Zelenka \nReviewed-by: Paul Dale \nMergeDate: Mon Jul 27 07:57:36 2026\n(Merged from https://github.com/openssl/openssl/pull/32045)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f603a95b87bc1a5648bed80aa722bbc0bc62fed3", "date": "2026-07-22T12:34:07+02:00", "subject": "DOCS: Fix KEM init function documentation", "body": "Document the missing params argument for OSSL_FUNC_kem_decapsulate_init(),\nremove stale name references, and fix the malformed provider-keymgmt(7) link.\n\nComplements: 78c44b05945b Add HPKE DHKEM provider support for EC, X25519 and X448.\nSigned-off-by: olszomal \n\nReviewed-by: Jakub Zelenka \nReviewed-by: Paul Dale \nMergeDate: Mon Jul 27 07:57:23 2026\n(Merged from https://github.com/openssl/openssl/pull/32045)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "758b1190c2669dc1b0061a4e54faba77c4b982a0", "date": "2026-07-22T12:34:07+02:00", "subject": "DOCS: Fix KEM init function documentation", "body": "Document the missing params argument for OSSL_FUNC_kem_decapsulate_init(),\nremove stale name references, and fix the malformed provider-keymgmt(7) link.\n\nComplements: 78c44b05945b Add HPKE DHKEM provider support for EC, X25519 and X448.\nSigned-off-by: olszomal \n\nReviewed-by: Jakub Zelenka \nReviewed-by: Paul Dale \nMergeDate: Mon Jul 27 07:57:13 2026\n(Merged from https://github.com/openssl/openssl/pull/32045)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0b65b147e241c22df65ba2a240209da327ce69c0", "date": "2026-07-22T12:34:07+02:00", "subject": "DOCS: Fix KEM init function documentation", "body": "Document the missing params argument for OSSL_FUNC_kem_decapsulate_init(),\nremove stale name references, and fix the malformed provider-keymgmt(7) link.\n\nComplements: 78c44b05945b Add HPKE DHKEM provider support for EC, X25519 and X448.\nSigned-off-by: olszomal \n\nReviewed-by: Jakub Zelenka \nReviewed-by: Paul Dale \nMergeDate: Mon Jul 27 07:57:01 2026\n(Merged from https://github.com/openssl/openssl/pull/32045)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "tmux", "sha": "670ea5d083ec4517229ac8b5415d3a19a598275b", "date": "2026-07-22T11:35:28+01:00", "subject": "Add regression test for copy-mode-exit-status.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "4cfd155ffce3fa920d34e6e3f27c7fc908a8d94d", "date": "2026-07-22T11:38:07+01:00", "subject": "Add copy-mode collapse output regression test.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "290f7e69a12eb7f4f4fe804e583912d4d5c2f9d4", "date": "2026-07-22T13:43:05+03:00", "subject": "backport: fix(lsp): guard linked_editing_range refresh on attach", "body": "Problem:\non_attach() calls refresh(), but there is no guarantee the attached\nbuffer is the current buffer. This can make linked editing request\nhandling assume the wrong window.\n\nSolution:\nCall refresh() only if the current buffer is attached. This keeps the\ninitial highlighting behavior while avoiding making incorrect request.\n\n(cherry picked from commit 9180c6e093391969b2751d09b6fa0397dbfc7882)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "526365a8d13fc28ebb52042ccd2b8b29e0054ea7", "date": "2026-07-22T19:26:09+08:00", "subject": "fix(help): handle escape edge cases #40901", "body": "* fix(help): preserve multi-letter CTRL modifiers in :help subject\n\nProblem: After PR 39537 normalized `CTRL-U-default` by excluding\n`-` from the trigger char class, the regex `(CTRL%-[^{])([^-_\\\\])`\nalso fires between `CTRL-S` and the `H` of the `SHIFT` modifier\nin tags like `c_CTRL-SHIFT-V`, splitting it into the non-existent\n`c_CTRL-S_HIFT-V`.\n\nSolution: Exclude uppercase letters from the char class so a\nmulti-letter modifier following `CTRL-X` (e.g. `SHIFT`, `ALT`) is\nno longer split. The `-` exclusion from 39537 is preserved so\n`i_CTRL-U-default` still works.\n\n* fix(help): escape ~ in expr-=~? exception to keep regex valid\n\nProblem: The `expr-=~?` exception in the tag alias table maps to\n`=~?`, but in Vim's regex engine `~` refers to the previous\nsubstitute pattern. With no prior substitute the pattern fails to\ncompile, so `:help expr-=~?` returns E149 even though the tag exists.\n\nSolution: Escape `~` as \\\\\\~` so the alias expands to a valid\nregex (`=\\\\\\~?`) that substring-matches the `expr-=~?` tag. Other\n`expr-X?` entries are unaffected since they don't contain `~`.\n\n* fix(help): don't insert _ before CTRL- when preceded by -\n\nProblem: The rule `([^_])CTRL%- -> %1_CTRL-` inserts an underscore\nbetween any non-underscore char and `CTRL-`, which wrongly rewrites\ntags like `map-CTRL-C` and `telnet-CTRL-]` (where `CTRL-` is\nliteral text in the tag, not a key chord).\n\nSolution: Exclude `-` from the trigger char class so a hyphen\nimmediately before `CTRL-` is preserved. The existing behavior for\n`iCTRL-GCTRL-J` -> `i_CTRL-G_CTRL-J` is unaffected because the\n`G` before `CTRL-J` is not `-`.\n\n* fix(help): only treat ^X as caret notation for valid control chars\n\nProblem: The regex `%^([^_])` converts any `^X` to `CTRL-X`,\nincluding `:set^=` where `^=` is literal text in the tag, not a\ncaret-notation control character.\n\nSolution: Restrict the transformation to caret-notation chars\n(`%a` or one of `?@[\\\\]^{}`). This matches the C reference\nbehavior where `^X` is only conv", "style": "conventional", "type": "fix", "scope": "help", "breaking": false} {"repo": "angular", "sha": "e26b9b42df199cf60c40ca6c986955d58f0dbe3b", "date": "2026-07-22T14:32:56+03:00", "subject": "docs: fix first heading rendering as paragraph on two template guides", "body": "The ng-container and binding template guide files each began with a\nUTF-8 BOM (EF BB BF) before the leading `#`. The docs markdown parser\nonly promotes `#` to an H1 when it is the first character on the line,\nso the BOM demoted the title to paragraph text (`

# ...

`) and the\nstandard docs header (breadcrumbs, page title, edit button) never\nrendered.\n\nStripping the BOM restores `#` as the first character, so both pages\nnow generate the proper `
` block. Verified\nby rebuilding //adev/src/content/guide/templates:templates and\ninspecting the generated HTML.\n\nFixes #69889\n\n(cherry picked from commit f12db896597f246cea7b26dda43c776d8a1e8595)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "f12db896597f246cea7b26dda43c776d8a1e8595", "date": "2026-07-22T14:32:56+03:00", "subject": "docs: fix first heading rendering as paragraph on two template guides", "body": "The ng-container and binding template guide files each began with a\nUTF-8 BOM (EF BB BF) before the leading `#`. The docs markdown parser\nonly promotes `#` to an H1 when it is the first character on the line,\nso the BOM demoted the title to paragraph text (`

# ...

`) and the\nstandard docs header (breadcrumbs, page title, edit button) never\nrendered.\n\nStripping the BOM restores `#` as the first character, so both pages\nnow generate the proper `
` block. Verified\nby rebuilding //adev/src/content/guide/templates:templates and\ninspecting the generated HTML.\n\nFixes #69889", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c812307c9c058ee3bbdec3bf3cedc2049c19972c", "date": "2026-07-22T13:33:05+02:00", "subject": "DOCS: Fix derive_skey provider documentation", "body": "Correct the OSSL_FUNC_keyexch_derive_skey() return type, derive_skey\nreturn-value descriptions, stale function names, and broken links.\n\nSigned-off-by: olszomal \nReviewed-by: Paul Dale \nReviewed-by: Dmitry Belyavskiy \nMerge-date: Fri Aug 28 11:13:32 2026\nMerged-from: https://github.com/openssl/openssl/pull/32046", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7e643ba676787603e06a155333d9c4bdfed43f54", "date": "2026-07-22T13:33:05+02:00", "subject": "DOCS: Fix derive_skey provider documentation", "body": "Correct the OSSL_FUNC_keyexch_derive_skey() return type, derive_skey\nreturn-value descriptions, stale function names, and broken links.\n\nSigned-off-by: olszomal \nReviewed-by: Paul Dale \nReviewed-by: Dmitry Belyavskiy \nMerge-date: Fri Aug 28 11:13:19 2026\nMerged-from: https://github.com/openssl/openssl/pull/32046", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1fbe7151a1d4e056de354053aa5e9f9ce65836c4", "date": "2026-07-22T13:33:05+02:00", "subject": "DOCS: Fix derive_skey provider documentation", "body": "Correct the OSSL_FUNC_keyexch_derive_skey() return type, derive_skey\nreturn-value descriptions, stale function names, and broken links.\n\nSigned-off-by: olszomal \nReviewed-by: Paul Dale \nReviewed-by: Dmitry Belyavskiy \nMerge-date: Fri Aug 28 11:13:06 2026\nMerged-from: https://github.com/openssl/openssl/pull/32046", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "45eef0103e9429a98ea24e10e71c1c49793af847", "date": "2026-07-22T14:49:28+03:00", "subject": "fix(docs-infra): fail the guide build when a markdown file starts with a BOM", "body": "A leading UTF-8 byte order mark (U+FEFF) before the first `#` stops the\nMarkdown parser from recognizing the heading, so the guide renders its\ntitle as a paragraph and drops the standard docs header. The character\nis invisible, so it cannot be caught in review.\n\nAdd a check in the guides generation pipeline that throws when a source\nfile starts with a BOM, failing the build with the offending file name.\nThis sits alongside the existing unknown-anchor check and prevents the\nregression fixed in #69889 from recurring.\n\n(cherry picked from commit ab52df470a3be72d5d08d88af0c026cbf2ce8ee3)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "ab52df470a3be72d5d08d88af0c026cbf2ce8ee3", "date": "2026-07-22T14:49:28+03:00", "subject": "fix(docs-infra): fail the guide build when a markdown file starts with a BOM", "body": "A leading UTF-8 byte order mark (U+FEFF) before the first `#` stops the\nMarkdown parser from recognizing the heading, so the guide renders its\ntitle as a paragraph and drops the standard docs header. The character\nis invisible, so it cannot be caught in review.\n\nAdd a check in the guides generation pipeline that throws when a source\nfile starts with a BOM, failing the build with the offending file name.\nThis sits alongside the existing unknown-anchor check and prevents the\nregression fixed in #69889 from recurring.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "neovim", "sha": "41163b773136638e4f8212752fb87e3f9b1fee1f", "date": "2026-07-22T20:27:54+08:00", "subject": "fix(inccommand): enforce timeout during matching, not just after a line #40892", "body": "Problem: do_sub() only checks the timeout limit after finishing a line.\n A pathological regex will run on a single line input unbounded\n until the compute is completed.\n\nSolution: Pass the timeout limit to `vim_regexec_multi()` so the\n computation on the regex engine is bounded per-line.\n\nSigned-off-by: XiaowenHu96 \n(cherry picked from commit bd73d8c01146695b727849e67fc1f1f209d5884f)", "style": "conventional", "type": "fix", "scope": "inccommand", "breaking": false} {"repo": "neovim", "sha": "bd73d8c01146695b727849e67fc1f1f209d5884f", "date": "2026-07-22T20:27:54+08:00", "subject": "fix(inccommand): enforce timeout during matching, not just after a line #40892", "body": "Problem: do_sub() only checks the timeout limit after finishing a line.\n A pathological regex will run on a single line input unbounded\n until the compute is completed.\n\nSolution: Pass the timeout limit to `vim_regexec_multi()` so the\n computation on the regex engine is bounded per-line.", "style": "conventional", "type": "fix", "scope": "inccommand", "breaking": false} {"repo": "postgres", "sha": "bdcea66f0f3b2f16c51d6c95bb9d0bb317fdfd99", "date": "2026-07-22T08:47:44-04:00", "subject": "walsummarizer: Guard against WAL files whose tail ends are not valid.", "body": "SummarizeWAL documents that maximum_lsn should be passed as \"the switch\npoint when reading a historic timeline, or the most-recently-measured end of\nWAL when reading the current timeline.\" But the caller always passed the\nmost recently measured end-of-WAL even when reading from a historic\ntimeline, due to an oversight on my part. Fix that.\n\nAs far as I can determine, for this to become an issue in practice, it's\nnecessary to have a corrupted WAL file in the archive. SummarizeWAL checks\nthat every record it processes both starts and ends before switch_lsn; so if\nall the WAL files in the archive are valid, SummarizeWAL will still discover\nwhere it should stop summarizing and do the right thing. However, if\nthere's a corrupted file in the WAL archive, and if it is also the case that\nthe end of the current timeline has advanced past the switch point, then the\nincorrect maximum_lsn value can result in trying to read an invalid record\nand erroring out, which leads repeatedly retrying and failing with an error\nevery time.\n\nOne way this could occur is if a new primary is promoted and creates a\n.partial file, and the user manually renames that file to remove the suffix,\nand it is then archived. In that situation, the tail end of the file need\nnot be valid WAL, and that could lead to a stuck WAL summarizer.\n\nReported-by: Fabrice Chapuis \nAnalyzed-by: Thom Brown (using claude)\nDiscussion: http://postgr.es/m/CAA5-nLDdvGMkN6Z-GaHGHG5T7QWEgv4YoHO7XvOJbeD00cghNg@mail.gmail.com\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "96ee499ba4c821ca6ef507b2198094091be95c74", "date": "2026-07-22T15:15:40+02:00", "subject": "chore: cleanup, remove `'default'`", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0868435e81986b53c1d58ab93a2b973d2f311f59", "date": "2026-07-22T14:17:22+01:00", "subject": "Add test for A: modifier, from Fernando Daciuk.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "d14b73d855f7b565a46940c2ba16bcdb6cd46989", "date": "2026-07-22T15:19:16+02:00", "subject": "docs: add a guide for the custom benchmark provider", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "b19e5ece98aba23953149a3e3e0a7cc1238f6cab", "date": "2026-07-22T18:55:58+05:30", "subject": "fix: don't race the typechecker spawn grace period on windows", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vitest", "sha": "86c700ed8b77b7afb0ee9ec677ef8379f34a4bed", "date": "2026-07-22T15:54:33+02:00", "subject": "feat(benchmark): add pluggable benchmark provider API", "body": "", "style": "conventional", "type": "feat", "scope": "benchmark", "breaking": false} {"repo": "angular", "sha": "e428df2b89b2f947825be79a380ed957909b3bf0", "date": "2026-07-22T16:01:17+02:00", "subject": "build: update pnpm-lock.yaml", "body": "Update pnpm lock after it got desychronized in\n791b0646c8 build: update all non-major dependencies", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e397b2c2ca930382b6f56d6882ead563cad31527", "date": "2026-07-22T22:39:29+08:00", "subject": "refactor(path): path_has_wildcard() #40893", "body": "Problem:\nRedundant code.\n\nSolution:\nCombine path_has_wildcard() and path_has_exp_wildcard(). They have similar logic.\nAlso, the latter operates on whatever wildcards remain after shell/env/backtick\nexpansion, so the two felt somewhat related.\n\nAlso drop some unnecessary MB_PTR_ADV calls.", "style": "conventional", "type": "refactor", "scope": "path", "breaking": false} {"repo": "vite", "sha": "428bda3b98886708006f4f2f409095f89d705df1", "date": "2026-07-22T23:44:57+09:00", "subject": "test: add `resolve.tsconfigPaths` + `import.meta.glob` case", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "26476e287da466721a4e353009794ae895d98893", "date": "2026-07-22T22:54:04+08:00", "subject": "fix: avoid Form warning hook compiler memoization", "body": "* fix: avoid memoizing Form warning hook\n\n* test: harden Form compiler regression test\n\n* fix: avoid memoizing Form warning hooks\n\n* fix: expose Form warning helper as hook", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "fd571a66e813713b27a3c5efab04237951d32927", "date": "2026-07-22T23:12:04+08:00", "subject": "docs(Image): sync missing TransformAction `reset` and ToolbarRenderInfoType`total` in English docs", "body": "", "style": "conventional", "type": "docs", "scope": "Image", "breaking": false} {"repo": "ant-design", "sha": "f06ce22f89c04c120b59a275411f41873a212525", "date": "2026-07-22T23:21:24+08:00", "subject": "docs(ConfigProvider): fix link anchor typo", "body": "", "style": "conventional", "type": "docs", "scope": "ConfigProvider", "breaking": false} {"repo": "openssl", "sha": "bd8585e8119c02ff0f4ea0e74ea8320e7076b420", "date": "2026-07-22T11:55:02-04:00", "subject": "X509: keep names modified until canonicalization succeeds", "body": "x509_name_encode() cleared the modified flag before x509_name_canon().\nIf canonicalization then failed, later encodes reused the DER cache while\ncanonical state was absent, allowing a malformed name to compare and hash\nas an empty name.\n\nClear modified only after both cache representations succeed. Validate both\noutput-pass encodes and propagate nested ASN.1 encoding failures so partial\noutput is not reported as success.\n\nAdd malformed-name cache regression coverage and multi-valued RDN\nallocation-failure coverage.\n\nRefs #32010\n\nAssisted-by: OpenAI Codex:gpt-5.6-sol\nReviewed-by: Jakub Zelenka \nReviewed-by: Mounir Idrassi \nMerge-date: Mon Sep 14 14:13:33 2026\nMerged-from: https://github.com/openssl/openssl/pull/32047\n(cherry picked from commit 4fc10a5bfe8658d080a44b08fb6229fa666949ac)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4fc10a5bfe8658d080a44b08fb6229fa666949ac", "date": "2026-07-22T11:55:02-04:00", "subject": "X509: keep names modified until canonicalization succeeds", "body": "x509_name_encode() cleared the modified flag before x509_name_canon().\nIf canonicalization then failed, later encodes reused the DER cache while\ncanonical state was absent, allowing a malformed name to compare and hash\nas an empty name.\n\nClear modified only after both cache representations succeed. Validate both\noutput-pass encodes and propagate nested ASN.1 encoding failures so partial\noutput is not reported as success.\n\nAdd malformed-name cache regression coverage and multi-valued RDN\nallocation-failure coverage.\n\nRefs #32010\n\nAssisted-by: OpenAI Codex:gpt-5.6-sol\nReviewed-by: Jakub Zelenka \nReviewed-by: Mounir Idrassi \nMerge-date: Mon Sep 14 14:13:33 2026\nMerged-from: https://github.com/openssl/openssl/pull/32047", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "fe6c1cb4bd945041b615240c6c5434f9d4774e68", "date": "2026-07-22T09:03:34-07:00", "subject": "fix(treesitter): queries for bundled diff parser #40866", "body": "", "style": "conventional", "type": "fix", "scope": "treesitter", "breaking": false} {"repo": "neovim", "sha": "8b6bdb79e6340b74f2c66c57ebbfac09cafde5c8", "date": "2026-07-23T01:05:37+09:00", "subject": "fix(ui): invalid ruler items in other windows #40891", "body": "Problem:\nWhen buffer is open in multiple windows and its line count changes, any\n'statusline' or 'rulerformat' items depending on it (e.g. %P) will\nbecome invalid in non-current windows because w_redr_status is only set\non first change.\n\nSolution:\nSet w_redr_status on all windows with the buffer when the line count\nchanges.", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "openssl", "sha": "2ed423c6d0239963071c9b8073cfce8e1b85417e", "date": "2026-07-22T18:41:08+02:00", "subject": "apps: test rsa -modulus output", "body": "Add a test recipe subtest checking that 'openssl rsa -modulus' prints the\nexpected modulus for both a private key and a -pubin public key, mirroring\nthe existing dsa -modulus test.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Paul Dale \nReviewed-by: Paul Yang \nReviewed-by: Daniel Kubec \nMergeDate: Mon Jul 27 09:08:13 2026\n(Merged from https://github.com/openssl/openssl/pull/32048)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "ada20074535c253fe31ef5c6a927f9c38367fa8f", "date": "2026-07-22T10:21:39-07:00", "subject": "feat(dir): provider framework #40738", "body": "Problem:\nThe core behaviors of `nvim.dir` cannot be reused by consumers\nsuch as the archive plugin, ssh browser, etc.\n\nSolution:\n- Define a basic \"provider\" shape.\n- Extract the filesystem provider.", "style": "conventional", "type": "feat", "scope": "dir", "breaking": false} {"repo": "angular", "sha": "2f82601662a958b3accd4d22dff7ad4ec7c56d77", "date": "2026-07-22T12:46:52-05:00", "subject": "fix(router): limit protocol-relative URL handling to serialization", "body": "Preserve createUrlTree command semantics, including custom serializer inputs, while keeping the single-leading-slash guarantee at the default serialization boundary.\n\nExpand coverage for command forms, public UrlTree values, secondary outlets, and preserved query parameters and fragments.\n\nFixes #69700\n\n(cherry picked from commit 435f8b2b8b67277fc69888911251d14b32705828)", "style": "conventional", "type": "fix", "scope": "router", "breaking": false} {"repo": "angular", "sha": "435f8b2b8b67277fc69888911251d14b32705828", "date": "2026-07-22T12:46:52-05:00", "subject": "fix(router): limit protocol-relative URL handling to serialization", "body": "Preserve createUrlTree command semantics, including custom serializer inputs, while keeping the single-leading-slash guarantee at the default serialization boundary.\n\nExpand coverage for command forms, public UrlTree values, secondary outlets, and preserved query parameters and fragments.\n\nFixes #69700", "style": "conventional", "type": "fix", "scope": "router", "breaking": false} {"repo": "tmux", "sha": "9625fe3b5f7d23817f4f0f07b24877cea58a7d7f", "date": "2026-07-22T18:58:48Z", "subject": "Show user hooks correctly with show-hooks.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "18fc547b829317a886693827b8c4aa531e0e8143", "date": "2026-07-22T19:58:55+01:00", "subject": "Update show hooks tests.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "2591fa6bd950881a3ee45a59ab2b55dc83dfcf37", "date": "2026-07-22T14:14:49-05:00", "subject": "docs: clarify behavior of credentials option during SSR on Node.js", "body": "(cherry picked from commit 90ac011fa92d28d5b681998e60029e2ee6a2695a)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "90ac011fa92d28d5b681998e60029e2ee6a2695a", "date": "2026-07-22T14:14:49-05:00", "subject": "docs: clarify behavior of credentials option during SSR on Node.js", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "tmux", "sha": "2a5389416a2ef4c65e56a1479b1c34cdf234127a", "date": "2026-07-22T19:23:59Z", "subject": "Add a menu on empty areas to allow a new pane to be created.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "09a2ea190d325523c4eb0dd45b4bc09c09df8f72", "date": "2026-07-22T20:24:21+01:00", "subject": "Update test.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1cc87ac6739f97501d3f2aae2b0522de56a69d3e", "date": "2026-07-22T21:24:57+02:00", "subject": "apps: test pkey -inform and -outform handling", "body": "Add test recipe subtests for the pkey app format options: the public key\nPEM/DER round-trip, format detection when -inform is omitted, rejection of\na mismatching -inform, the restriction of -outform to PEM and DER, and\nreading a private key from a PKCS#12 file with -inform P12.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Daniel Kubec \nReviewed-by: Paul Dale \nMergeDate: Thu Jul 30 08:54:04 2026\n(Merged from https://github.com/openssl/openssl/pull/32049)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "196a4b978499604b2ddb875e967bf3aaa49dd4ef", "date": "2026-07-22T21:34:49+02:00", "subject": "Add test for CVE-2026-63073", "body": "Signed-off-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Igor Ustinov \nMerge-date: Mon Aug 24 13:02:30 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9d838f67fec7ff72cf02322856f1220738a17430", "date": "2026-07-22T21:34:49+02:00", "subject": "Add test for CVE-2026-63073", "body": "Signed-off-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Igor Ustinov \nMerge-date: Mon Aug 24 13:02:11 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1951c3cbd9d5101e027a1eb126f26c3c758f6cc0", "date": "2026-07-22T21:34:49+02:00", "subject": "Add test for CVE-2026-63073", "body": "Signed-off-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Igor Ustinov \nMerge-date: Mon Aug 24 13:01:48 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1f12245e12d3930a11a10e96ef783d8bd1426e61", "date": "2026-07-22T21:34:49+02:00", "subject": "Add test for CVE-2026-63073", "body": "Signed-off-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Igor Ustinov \nMerge-date: Mon Aug 24 13:00:50 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a6286cd01801a50aec5fc5b10e7ac73399d160f0", "date": "2026-07-22T21:34:49+02:00", "subject": "Add test for CVE-2026-63073", "body": "Signed-off-by: Norbert Pocs \nReviewed-by: Andrew Dinh \nReviewed-by: Igor Ustinov \nMerge-date: Mon Aug 24 12:53:37 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "7277712ca8b603d3b11edbb470de50eb1c64bbf1", "date": "2026-07-22T20:12:58Z", "subject": "Switch show-options over to using a format and add -F.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "a81154d6d97a46f81cd7e3c5bca8204f8e473d5b", "date": "2026-07-22T23:40:01+01:00", "subject": "Fix missing ! before commands with no output and non-zero exit status.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "e1450d8fdaf05f27ec75eb15df303fe931755a59", "date": "2026-07-23T01:50:06+02:00", "subject": "tidy-up: use more `static`, `sizeof()`, `char[]`, double-const", "body": "- make `const` data `static`, where missing and possible.\n- replace `strlen()` on literal or const strings with `sizeof()`.\n While the latter is optimized by popular C compiler, e.g. MSVC only\n does it with `/O2`.\n- replace magic numbers with `sizeof()`, where missing.\n- introduce `CURL_CSTRLEN()` macro for `sizeof(char[]) - 1`.\n- use `CURL_CSTRLEN()` macro.\n- move `const` before integer types, where missing.\n- replace `char *var` with `var[]`, where missing and possible.\n- use double const, where missing.\n `static const char *` -> `static const char * const`.\n- lib1514: constify pointers.\n- unit3205: drop redundant cast, avoid another one.\n- unit1666: map `OID()` macro to identical `STRCONST()`.\n\nCloses #22406", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "c3bc9e15cc043b3c32e23952bd500833f939624b", "date": "2026-07-23T08:38:28+08:00", "subject": "fix(runtime-vapor): preserve functional component root bindings", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "openssl", "sha": "2a3dac874c8057c1f0186849bf1ede1ae7b6b756", "date": "2026-07-23T03:04:33+02:00", "subject": "Fix heap buffer overflow (8-byte OOB write) in AES-WRAP-PAD unwrap", "body": "On its integrity-failure paths that primitive writes and cleanses up to inlen\nbytes of the output buffer. Size the buffer for that worst case so a failed\nunwrap cannot write past the allocation.\n\nFixes CVE-2026-63072\n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Mon Aug 24 14:38:35 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9530a5fd1aacaeccdced4478ea2340a480613335", "date": "2026-07-23T03:04:33+02:00", "subject": "Fix heap buffer overflow (8-byte OOB write) in AES-WRAP-PAD unwrap", "body": "On its integrity-failure paths that primitive writes and cleanses up to inlen\nbytes of the output buffer. Size the buffer for that worst case so a failed\nunwrap cannot write past the allocation.\n\nFixes CVE-2026-63072\n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Mon Aug 24 14:38:16 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "13bcff0ab2f5c55696c92cc6b3480d94f80e8e3b", "date": "2026-07-23T03:04:33+02:00", "subject": "Fix heap buffer overflow (8-byte OOB write) in AES-WRAP-PAD unwrap", "body": "On its integrity-failure paths that primitive writes and cleanses up to inlen\nbytes of the output buffer. Size the buffer for that worst case so a failed\nunwrap cannot write past the allocation.\n\nFixes CVE-2026-63072\n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Mon Aug 24 14:37:11 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "8edbf9dac07904435550e36bd6c70c109a0d45b8", "date": "2026-07-22T22:12:13-04:00", "subject": "build(vim-patch): n/a doc for channel,job api", "body": "Vim's code and test for channel,job API are already N/A.\nBlacklisting runtime/doc/channel.txt ensures most patches for it\nwill be detected as N/A.\nVim's API that is too incompatible with Neovim must be handled\noutside vim-patch.sh .", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "ant-design", "sha": "166196df449eba5103e5122fb4f15ddbb520149e", "date": "2026-07-23T10:30:57+08:00", "subject": "docs(Select): correct placeholder type to `ReactNode`", "body": "lonewolfyx:correct-placeholder-type", "style": "conventional", "type": "docs", "scope": "Select", "breaking": false} {"repo": "vite", "sha": "5d18326e488f566771cb73527371bbe7e3e39aff", "date": "2026-07-23T11:34:44+08:00", "subject": "fix(playground): include legacy dir in tsconfig for type checking", "body": "", "style": "conventional", "type": "fix", "scope": "playground", "breaking": false} {"repo": "vite", "sha": "428d8af8135626d783f963e3ba9b2378b64f680b", "date": "2026-07-23T11:59:03+08:00", "subject": "fix(create-vite): ignore --eslint for non-React templates instead of crashing", "body": "", "style": "conventional", "type": "fix", "scope": "create-vite", "breaking": false} {"repo": "vite", "sha": "91f32574fa19544e5ea810c8f536c4dfbccef17c", "date": "2026-07-23T05:59:55+02:00", "subject": "docs: fix browser target grammar", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "989a42ec52b527846bf43a8281fefd0a3aab7358", "date": "2026-07-23T12:15:17+08:00", "subject": "fix(legacy): don't use newer syntax when minifying legacy chunks", "body": "", "style": "conventional", "type": "fix", "scope": "legacy", "breaking": false} {"repo": "vite", "sha": "fb59b469a5e395ff8990af29a0e310c0e2c429fa", "date": "2026-07-23T12:17:26+08:00", "subject": "docs: remove unused imports", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2417e4ffdd09417cab6db8fab0b7bbadf4d185a7", "date": "2026-07-23T07:06:28+02:00", "subject": "test/quictestlib.c: add functions to create QUIC objects on dgram bio", "body": "Add the following helper functions:\n * create_quic_ctx_pair() - creates pair of SSL_CTX (server, client).\n * create_quic_conn_objects() - creates pair of SSL objects, client\n and listener. They both are 'connected' by BIO_dgram_pair.\n * create_quic_client() - creates SSL QUIC client object bound to BIO\n object provided by caller.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Wed Aug 12 15:00:27 2026\n(Merged from https://github.com/openssl/openssl/pull/32052)\n\n(cherry picked from commit f20e51341d3134bed58e6bb18475b80bfd5712ce)\n(cherry picked from commit 50c55eed777b1f9041256399f9cc6b3cc0c91f23)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "50c55eed777b1f9041256399f9cc6b3cc0c91f23", "date": "2026-07-23T07:06:28+02:00", "subject": "test/quictestlib.c: add functions to create QUIC objects on dgram bio", "body": "Add the following helper functions:\n * create_quic_ctx_pair() - creates pair of SSL_CTX (server, client).\n * create_quic_conn_objects() - creates pair of SSL objects, client\n and listener. They both are 'connected' by BIO_dgram_pair.\n * create_quic_client() - creates SSL QUIC client object bound to BIO\n object provided by caller.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Wed Aug 12 15:00:27 2026\n(Merged from https://github.com/openssl/openssl/pull/32052)\n\n(cherry picked from commit f20e51341d3134bed58e6bb18475b80bfd5712ce)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f20e51341d3134bed58e6bb18475b80bfd5712ce", "date": "2026-07-23T07:06:28+02:00", "subject": "test/quictestlib.c: add functions to create QUIC objects on dgram bio", "body": "Add the following helper functions:\n * create_quic_ctx_pair() - creates pair of SSL_CTX (server, client).\n * create_quic_conn_objects() - creates pair of SSL objects, client\n and listener. They both are 'connected' by BIO_dgram_pair.\n * create_quic_client() - creates SSL QUIC client object bound to BIO\n object provided by caller.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Wed Aug 12 15:00:27 2026\n(Merged from https://github.com/openssl/openssl/pull/32052)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1a8c172228db9f8ffd7add086e3b0a3cc4986b1c", "date": "2026-07-23T10:46:42+05:30", "subject": "Reject sequence synchronization against pre-PostgreSQL 19 publishers.", "body": "Sequence synchronization requires the page_lsn field returned by\npg_get_sequence_data(), which was added in PostgreSQL 19. Previously,\nrequesting sequence synchronization against an older publisher (via\nALTER SUBSCRIPTION ... REFRESH SEQUENCES or by running\nALTER SUBSCRIPTION ... CONNECTION on a disabled subscription with\nsequences in the INIT state and subsequently enabling the subscription)\nwould cause the sequence synchronization worker to repeatedly fail with a\nconfusing \"invalid query response\" error.\n\nCheck the publisher's server version up front in both\nAlterSubscription_refresh_seq() and copy_sequences(), and error out\nimmediately when it predates PostgreSQL 19.\n\nAlso document the PostgreSQL 19 publisher requirement for sequence\nreplication in the logical replication documentation and in\nALTER SUBSCRIPTION ... REFRESH SEQUENCES.\n\nReported-by: Noah Misch \nAuthor: vignesh C \nReviewed-by: Shveta Malik \nReviewed-by: Hayato Kuroda \nReviewed-by: Amit Kapila \nBackpatch-through: 19\nDiscussion: https://postgr.es/m/20260710045217.f0.noahmisch@microsoft.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "821a4b4d074cb89d3f85f57d22896a63e9405b6a", "date": "2026-07-23T14:37:48+09:00", "subject": "injection_points: Clear waiter slot on error and exit", "body": "injection_wait() only clears its slot in the waiter array after the\nwait loop finishes. When the waiting query is canceled or the backend\nis terminated (wait look has a CHECK_FOR_INTERRUPS), the slot leaks.\nLater wakeups of the same point then bump the counter of the leaked slot\ninstead of the real waiter, that sleeps forever. Repeated leaks can\nexhaust all the slots.\n\nThe code is changed so as the waiting loop is wrapped with\nPG_ENSURE_ERROR_CLEANUP, so as the injection point slots, that are\nshared resources, can be cleaned up on ERROR as much as a FATAL.\n\nAn isolation test is added: cancel one waiter, terminate another waiter,\nthen check that a later waiter still receives a wakeup. Without the\nfixed code, the test would fail on timeout.\n\nAuthor: Zsolt Parragi \nDiscussion: https://postgr.es/m/CAN4CZFO+KF=cc0-iEg28RhqRBp_fTs6D4b8b7D7DB-pGYP3Ccg@mail.gmail.com\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "cabfa17480a73e79348143f6bb97bd422472b5d7", "date": "2026-07-23T02:19:28-04:00", "subject": "build(vim-patch): logfile.c is N/A", "body": "Vim channel API logging is N/A.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "curl", "sha": "489a4c1b48096db7df4875af0bedf4f7d0973382", "date": "2026-07-23T12:09:04+05:30", "subject": "ctype: exclude control bytes from ISPRINT and ISGRAPH", "body": "Closes #22371", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "38feebb5d2a5b2be475aca56c406fbdf57cee13d", "date": "2026-07-23T08:06:09+01:00", "subject": "Show options tests.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "abbd74ce8738d536e8d99151122b7a650e3b63d5", "date": "2026-07-23T16:09:13+09:00", "subject": "doc: Improve description of pg_stat_activity.backend_type", "body": "The documentation of pg_stat_activity used an incomplete list of values\nfor backend_type. While on it, it is improved to use an itemized list,\nnow ordered alphabetically, with a short description about each item.\n\nAuthor: Laurenz Albe \nReviewed-By: Michael Paquier \nReviewed-By: Fujii Masao \nDiscussion: https://postgr.es/m/5e94c0196084f648ae6a00107125494f5804318a.camel@cybertec.at", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "791843e1591ec2d65a401560bc1fc0a85b32ee32", "date": "2026-07-23T15:18:00+08:00", "subject": "docs: fix examples, links, and inline code formatting", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "08e7756c3900bcfd77a720e7b74e27d6e4ed01a9", "date": "2026-07-23T09:38:02+02:00", "subject": "QUIC server: limit number of pending QUIC channels/connections", "body": "Currently, there is no limit for pending QUIC connections. The port\ndefault packet handler creates channel for every valid initial packet\nwhich does belong to existing channel (a.k.a. connection). The newly\ncreated channel is inserted to list of pending channels where it waits\nto be accepted by local application by call\nto SSL_accept_connection(3ossl).\n\nThis change introduces a limit for pending connection. The pending\nqueue is limited to 256 pending connections. Applications may change\nthe limit by calling SSL_set_feature_request_uint(3ossl)\non SSL server listener object with configurable value\nSSL_VALUE_QUIC_MAX_PENDING_CONNS.\n\nFixes: CVE-2026-14456\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Wed Aug 12 15:00:25 2026\n(Merged from https://github.com/openssl/openssl/pull/32052)\n\n(cherry picked from commit 9416706d408bb84deb7cee4647bff3045d2dc7ba)\n(cherry picked from commit 4084152e040329ca0194c4c1750b9b46d00a5b6b)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4084152e040329ca0194c4c1750b9b46d00a5b6b", "date": "2026-07-23T09:38:02+02:00", "subject": "QUIC server: limit number of pending QUIC channels/connections", "body": "Currently, there is no limit for pending QUIC connections. The port\ndefault packet handler creates channel for every valid initial packet\nwhich does belong to existing channel (a.k.a. connection). The newly\ncreated channel is inserted to list of pending channels where it waits\nto be accepted by local application by call\nto SSL_accept_connection(3ossl).\n\nThis change introduces a limit for pending connection. The pending\nqueue is limited to 256 pending connections. Applications may change\nthe limit by calling SSL_set_feature_request_uint(3ossl)\non SSL server listener object with configurable value\nSSL_VALUE_QUIC_MAX_PENDING_CONNS.\n\nFixes: CVE-2026-14456\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Wed Aug 12 15:00:25 2026\n(Merged from https://github.com/openssl/openssl/pull/32052)\n\n(cherry picked from commit 9416706d408bb84deb7cee4647bff3045d2dc7ba)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9416706d408bb84deb7cee4647bff3045d2dc7ba", "date": "2026-07-23T09:38:02+02:00", "subject": "QUIC server: limit number of pending QUIC channels/connections", "body": "Currently, there is no limit for pending QUIC connections. The port\ndefault packet handler creates channel for every valid initial packet\nwhich does belong to existing channel (a.k.a. connection). The newly\ncreated channel is inserted to list of pending channels where it waits\nto be accepted by local application by call\nto SSL_accept_connection(3ossl).\n\nThis change introduces a limit for pending connection. The pending\nqueue is limited to 256 pending connections. Applications may change\nthe limit by calling SSL_set_feature_request_uint(3ossl)\non SSL server listener object with configurable value\nSSL_VALUE_QUIC_MAX_PENDING_CONNS.\n\nFixes: CVE-2026-14456\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Wed Aug 12 15:00:25 2026\n(Merged from https://github.com/openssl/openssl/pull/32052)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "544d25b7af958ca6c03e98bfbc1538c295b30601", "date": "2026-07-23T16:48:40+09:00", "subject": "Fix socket_putmessage_noblock() to call socket_putmessage()", "body": "socket_putmessage_noblock() used pq_putmessage(), which redirects to\nPqCommMethods->putmessage. In the common cases, this points to\nsocket_putmessage(), but it would become incorrect if PqCommMethods\npoints to a different implementation.\n\nThis change may look like a bug, but as far as I can see this is mostly\ncosmetic. The code is able to work currently, as the repalloc() done in\nthe noblock() call ensures that the blocking path of internal_putbytes()\nis never reached. The issue has gone unnoticed since 2bd9e412f92b.\n\nAuthor: Anthonin Bonnefoy \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAO6_Xqpf5+Rzw_-XOOz-d-R5x6_2JHtpnzXP0nrYWiHyZokA_Q@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b05a190fe598cc1deb6d6e314ab8d1c4dd2d4436", "date": "2026-07-23T09:51:41+02:00", "subject": "test/quicapitest.c: add test for pending connections limit enforcement", "body": "This is a regression test for CVE-2026-14456.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Wed Aug 12 15:00:30 2026\n(Merged from https://github.com/openssl/openssl/pull/32052)\n\n(cherry picked from commit 0461a5636deca9af034db70ff017ad1138836c8c)\n(cherry picked from commit d446963b2622ee371df76971e5e2ddc7289839c4)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d446963b2622ee371df76971e5e2ddc7289839c4", "date": "2026-07-23T09:51:41+02:00", "subject": "test/quicapitest.c: add test for pending connections limit enforcement", "body": "This is a regression test for CVE-2026-14456.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Wed Aug 12 15:00:30 2026\n(Merged from https://github.com/openssl/openssl/pull/32052)\n\n(cherry picked from commit 0461a5636deca9af034db70ff017ad1138836c8c)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0461a5636deca9af034db70ff017ad1138836c8c", "date": "2026-07-23T09:51:41+02:00", "subject": "test/quicapitest.c: add test for pending connections limit enforcement", "body": "This is a regression test for CVE-2026-14456.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Wed Aug 12 15:00:30 2026\n(Merged from https://github.com/openssl/openssl/pull/32052)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "26bdd2b511fbfe01a05759745daf3125fdebd110", "date": "2026-07-23T08:59:04+01:00", "subject": "On macOS, look for utf8proc with no arguments and use it if provided. Only require a flag if it is not available.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "fb88172346b07fe70a385cee593e64091cb749a7", "date": "2026-07-23T08:59:49+01:00", "subject": "Turn ASAN off again on macOS.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "7568c192767a1f34547e0d3ed2db5a9e0ab7f4d1", "date": "2026-07-23T09:12:16+01:00", "subject": "Log jemalloc version if in use.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "3c9499c35f6f178dc9c6e086e6328573d40084a6", "date": "2026-07-23T16:22:21+08:00", "subject": "fix(runtime-vapor): preserve pending async hydration ranges", "body": "- track adopted SSR ranges while async setup is pending\n- move, resolve, and unmount ranges from their live DOM position\n- align Vapor and VDOM interop ownership with regression coverage", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "tmux", "sha": "a10ed3233c82d2b821e34cb87f0278b625dc8ba4", "date": "2026-07-23T09:26:42+01:00", "subject": "Require jemalloc on macOS.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "6049298140210612128219d37c07e64dd230b9e3", "date": "2026-07-23T09:26:57+01:00", "subject": "Show what we have found during configure.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "689e6f8fac4e48c1d25bd61ca3061e9b590f6246", "date": "2026-07-23T09:30:16+01:00", "subject": "Add jemalloc to regress dependencies.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "a99fb915c09b352ac67f24e71af84cda30bcbf3e", "date": "2026-07-23T10:35:26+02:00", "subject": "fix(language-service): account for strictTemplates being enabled by default", "body": "We were raising the suggestion about enabling `strictTemplates` when `strictTemplates` is ommitted, however the option is now enabled by default.\n\nFixes #69905.\n\n(cherry picked from commit e606a020e90cf130f3ab532a7be553bf899313c2)", "style": "conventional", "type": "fix", "scope": "language-service", "breaking": false} {"repo": "angular", "sha": "e606a020e90cf130f3ab532a7be553bf899313c2", "date": "2026-07-23T10:35:26+02:00", "subject": "fix(language-service): account for strictTemplates being enabled by default", "body": "We were raising the suggestion about enabling `strictTemplates` when `strictTemplates` is ommitted, however the option is now enabled by default.\n\nFixes #69905.", "style": "conventional", "type": "fix", "scope": "language-service", "breaking": false} {"repo": "vitest", "sha": "c17677a8302cd08a10339b8d7d7f5c8687ae5b11", "date": "2026-07-23T10:39:17+02:00", "subject": "perf(browser): prewarm the browser while the Vite server starts", "body": "", "style": "conventional", "type": "perf", "scope": "browser", "breaking": false} {"repo": "vitest", "sha": "7e7e35cb4177b83b68ae1b97f42196c639ad8244", "date": "2026-07-23T11:51:58+03:00", "subject": "perf(coverage): reduce RPC data and unnecessary serializations", "body": "", "style": "conventional", "type": "perf", "scope": "coverage", "breaking": false} {"repo": "vitest", "sha": "66d95af84b77b59ce84ef30492768d4c25e886ee", "date": "2026-07-23T04:52:32-04:00", "subject": "fix(browser): mock window.print to avoid hanging (fix #7375)", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "eslint", "sha": "6b8d2f7589b8a7c8b91b8ca2a2ef6d46178760d8", "date": "2026-07-23T10:57:06+02:00", "subject": "fix: escape reserved characters in rule id in `html` formatter", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d307f61f1b341b6eb7d5d07bd6902a939c709bc6", "date": "2026-07-23T11:09:55+02:00", "subject": "Add test for CVE-2026-63072", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMerge-date: Mon Aug 24 14:44:54 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "945a57b06f9158be59b5fec61e238cc67e6f98dd", "date": "2026-07-23T11:09:55+02:00", "subject": "Add test for CVE-2026-63072", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMerge-date: Mon Aug 24 14:44:05 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d384598500595e963383d9a17bf2a2ed0b607434", "date": "2026-07-23T11:09:55+02:00", "subject": "Add test for CVE-2026-63072", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMerge-date: Mon Aug 24 14:42:04 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "35c6d22998b3307429242112ff1ba06aac7ea7e0", "date": "2026-07-23T11:09:55+02:00", "subject": "Add test for CVE-2026-63072", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Mon Aug 24 14:38:35 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dcd63a9a3c6ac61e9ae7e606ab5839f73108f38a", "date": "2026-07-23T11:09:55+02:00", "subject": "Add test for CVE-2026-63072", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Mon Aug 24 14:38:16 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9e55749c83f6b801ef17b069ac6ee74e6487005a", "date": "2026-07-23T11:09:55+02:00", "subject": "Add test for CVE-2026-63072", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Mon Aug 24 14:37:11 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a954d87f0b41e4424f5565f74be5f2bd08d3347b", "date": "2026-07-23T11:20:19+02:00", "subject": "httpsrr: DoH with HTTPS, fix response handling", "body": "Fix handling of DoH response that only asks for HTTPS records.\n\nAdd test 2117 for checking that a HTTPS-RR resolve is processed,\neven though the actual answer is invalid.\n\nCloses #22372", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "de63174f52b2026c53213bf539eb2b43bbc576dd", "date": "2026-07-23T11:30:36+02:00", "subject": "perf(browser): pre-bundle the vitest runtime in optimizeDeps", "body": "", "style": "conventional", "type": "perf", "scope": "browser", "breaking": false} {"repo": "tmux", "sha": "8862390c3426ed31a76c1476089eefcc87932e32", "date": "2026-07-23T09:38:27Z", "subject": "Extend fill-character to allow inside and outside to be separate and use a different default (filled characters rather than dots) for inside.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "a2f3bdb601000c0a6ed436029bbd0664179e5d05", "date": "2026-07-23T10:38:27+01:00", "subject": "fill-character test bits.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "790c57de1ff03a75508bd1aadf76d5dd3f7dc4a8", "date": "2026-07-23T12:00:21+02:00", "subject": "ci: allow ct-validation-daily to be dispatched against a PR head SHA", "body": "Additive: schedule and bare-manual runs unchanged. Adds validated,\nfail-closed pr/head_sha/check_run_id inputs so a caller can dispatch against\nan exact commit; run-name echoes them back for correlation.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) \n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMergeDate: Mon Jul 27 10:16:54 2026\n(Merged from https://github.com/openssl/openssl/pull/32053)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "postgres", "sha": "42b480d286828f94e476ab45a752321602b12c75", "date": "2026-07-23T19:22:36+09:00", "subject": "doc: Improve pg_stat_recovery documentation", "body": "Improve the documentation for pg_stat_recovery in several ways:\n\n- Mention the view in high-availability.sgml as a way to monitor\n recovery state and replay progress, alongside the existing recovery\n information functions.\n- Clarify that the view returns at most one row, not exactly one row,\n and no rows to users who lack the pg_read_all_stats privilege.\n- Correct the description of last_replayed_end_lsn to clarify that it\n is the end LSN of the last replayed record plus one.\n- Document that replay_end_tli equals last_replayed_tli when no WAL\n record is currently being replayed.\n- Clarify that current_chunk_start_time is NULL until streaming WAL\n has been received.\n\nBackpatch to v19, where pg_stat_recovery was introduced.\n\nAuthor: Fujii Masao \nReviewed-by: Michael Paquier \nDiscussion: https://postgr.es/m/CAHGQGwGRavm18HqnQn_f68QB96qk6arhjET1V93OJH09Mgojkg@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "41be67e534ffccd055922482c17b874fb7a74f30", "date": "2026-07-23T19:24:55+09:00", "subject": "Validate subscription conninfo on owner change", "body": "For subscriptions using SERVER, changing the owner can change the\neffective connection string. However, ALTER SUBSCRIPTION ... OWNER TO\ndid not validate the generated conninfo for the new owner.\n\nAs a result, ownership could be transferred to a non-superuser whose\ngenerated connection string did not satisfy password_required=true.\nThe ownership change succeeded, but the subscription would fail later\nwhen the worker or another command tried to connect.\n\nFix this by making ALTER SUBSCRIPTION ... OWNER TO validate the new\nowner's generated conninfo with walrcv_check_conninfo().\n\nBackpatch to v19, where SERVER subscriptions were introduced.\n\nAuthor: Fujii Masao \nReviewed-by: Yuanchao Zhang <145zhangyc@gmail.com>\nReviewed-by: Hayato Kuroda \nDiscussion: https://postgr.es/m/CAHGQGwFGa6+wWVgUmZPFwN=fBY59mYPkMK3=TxT=Pv5C1mNNRQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "91d24bde1feade663441a87f418e03da24a4064d", "date": "2026-07-23T12:25:18+02:00", "subject": "doc: add reference to how to list EC curve names", "body": "EVP_PKEY_keygen() and the related keygen helpers take an EC curve name\nas a string parameter, but the page did not point to any way of\ndiscovering the accepted names. Add a reference to\nEC_get_builtin_curves(3), which lists the built-in curve NIDs, and\nOSSL_EC_curve_nid2name(3), which converts a NID to the name string\naccepted by the keygen functions.\n\nCo-Authored-By: Claude Opus 4.8 \nAssisted-by: Claude:claude-opus-4-8\nReviewed-by: Paul Dale \nReviewed-by: Richard Levitte \nReviewed-by: Tomas Mraz \nMerge-date: Thu Sep 17 13:47:47 2026\nMerged-from: https://github.com/openssl/openssl/pull/29912", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0a3ab822121e3453dc4908dca2cb9e7983cfac1e", "date": "2026-07-23T12:25:18+02:00", "subject": "doc: add reference to how to list EC curve names", "body": "EVP_PKEY_keygen() and the related keygen helpers take an EC curve name\nas a string parameter, but the page did not point to any way of\ndiscovering the accepted names. Add a reference to\nEC_get_builtin_curves(3), which lists the built-in curve NIDs, and\nOSSL_EC_curve_nid2name(3), which converts a NID to the name string\naccepted by the keygen functions.\n\nCo-Authored-By: Claude Opus 4.8 \nAssisted-by: Claude:claude-opus-4-8\nReviewed-by: Paul Dale \nReviewed-by: Richard Levitte \nReviewed-by: Tomas Mraz \nMerge-date: Thu Sep 17 13:47:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/29912", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dd191e6d9deb0b6033617734871155227809beeb", "date": "2026-07-23T12:25:18+02:00", "subject": "doc: add reference to how to list EC curve names", "body": "EVP_PKEY_keygen() and the related keygen helpers take an EC curve name\nas a string parameter, but the page did not point to any way of\ndiscovering the accepted names. Add a reference to\nEC_get_builtin_curves(3), which lists the built-in curve NIDs, and\nOSSL_EC_curve_nid2name(3), which converts a NID to the name string\naccepted by the keygen functions.\n\nCo-Authored-By: Claude Opus 4.8 \nAssisted-by: Claude:claude-opus-4-8\nReviewed-by: Paul Dale \nReviewed-by: Richard Levitte \nReviewed-by: Tomas Mraz \nMerge-date: Thu Sep 17 13:45:52 2026\nMerged-from: https://github.com/openssl/openssl/pull/29912", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3ec437e1b871b651ed5faa754bd09dd0a69ce6c0", "date": "2026-07-23T12:25:18+02:00", "subject": "doc: add reference to how to list EC curve names", "body": "EVP_PKEY_keygen() and the related keygen helpers take an EC curve name\nas a string parameter, but the page did not point to any way of\ndiscovering the accepted names. Add a reference to\nEC_get_builtin_curves(3), which lists the built-in curve NIDs, and\nOSSL_EC_curve_nid2name(3), which converts a NID to the name string\naccepted by the keygen functions.\n\nCo-Authored-By: Claude Opus 4.8 \nAssisted-by: Claude:claude-opus-4-8\nReviewed-by: Paul Dale \nReviewed-by: Richard Levitte \nReviewed-by: Tomas Mraz \nMerge-date: Thu Sep 17 13:45:31 2026\nMerged-from: https://github.com/openssl/openssl/pull/29912", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b2162f96355d26b11359958c24fe5885996fb78d", "date": "2026-07-23T12:25:18+02:00", "subject": "doc: add reference to how to list EC curve names", "body": "EVP_PKEY_keygen() and the related keygen helpers take an EC curve name\nas a string parameter, but the page did not point to any way of\ndiscovering the accepted names. Add a reference to\nEC_get_builtin_curves(3), which lists the built-in curve NIDs, and\nOSSL_EC_curve_nid2name(3), which converts a NID to the name string\naccepted by the keygen functions.\n\nCo-Authored-By: Claude Opus 4.8 \nAssisted-by: Claude:claude-opus-4-8\nReviewed-by: Paul Dale \nReviewed-by: Richard Levitte \nReviewed-by: Tomas Mraz \nMerge-date: Thu Sep 17 13:45:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/29912", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c5f1590574b7f6ba29bf3a6486627d02254fd277", "date": "2026-07-23T13:12:54+02:00", "subject": "ci: allow run-checker-daily and valgrind-daily to be dispatched against a PR head SHA", "body": "Additive: schedule and bare-manual runs unchanged. Adds validated,\nfail-closed pr/head_sha/check_run_id inputs so a caller can dispatch each\nagainst an exact commit; run-name echoes them back for correlation.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) \n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMergeDate: Mon Jul 27 10:16:55 2026\n(Merged from https://github.com/openssl/openssl/pull/32053)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "tmux", "sha": "e476c1230b958df0cb12977517d24b3dc931375b", "date": "2026-07-23T12:16:26+01:00", "subject": "Build with jemalloc on macOS for 3.7c also.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "57dc215230302f0abf0789e0e048f8d29a4e8362", "date": "2026-07-23T16:57:19+05:00", "subject": "refactor(:restart): unnecessary console allocation on Windows #40919", "body": "Problem:\n`:restart` on Windows always forces `AllocConsole`, even though we could\nreuse old server's existing console + unnecessary env.\n\nAnalysis:\nAfter d5516daf121aa718e79bcd423ee24c24492893c0 changed the restart\nlifecycle and 789741bb83d1f31b7162cd8944bc2d846fa2fa51 separated the\nWindows console paths, we no longer need to force `AllocConsole()` for\n`:restart`. The `:detach` path now allocates its own console when it\nactually runs.\n\nSolution:\nTry to attach the replacement server to its parent’s console and fall\nback to `AllocConsole()` if attachment fails. Drop the restart-specific\nenv that forced allocation.\n\nAs documented by Microsoft, [“A console is closed when the last process\nattached to it terminates or calls\n`FreeConsole`.”](https://learn.microsoft.com/en-us/windows/console/attachconsole)\n\nBtw, this way we can preserve `io.stdout:write()`, though if for some\nreason `AttachConsole` would fail then it won't work; but now at least\nit's not always not rendering on `:restart`.", "style": "conventional", "type": "refactor", "scope": ":restart", "breaking": false} {"repo": "curl", "sha": "090056522b494feee8bb2e127a3df087b8753e56", "date": "2026-07-23T13:09:13+01:00", "subject": "configure: only check in the watt library if WATT_ROOT is set", "body": "Only look for gethostbyname in libwatt in $WATT_ROOT/lib if WATT_ROOT\nhas actually been set. This avoids configure trying to search in /lib,\nwhich won't every succeed and can cause problems in cross builds which\ncheck that host paths are not being searched.\n\nCloses #22380", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "e093c67f1cc3fbedefc49fd2e3f9520afcfa769f", "date": "2026-07-23T14:17:13+02:00", "subject": "connect: connection close tweaks", "body": "- connclose/streamclose/connkeep() remove description string that was\n never used anywhere. Add trace statements where reasons for closing\n were not already traced and maybe not obvious.\n- multi_remove_handle: only lookup former connection in pool when\n transfer is set to connect only\n- test1554: adapt expectations now that pool is less often locked\n\nCloses #22379", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "36e93be809b6c46b6596c6a318e086c7a6e59de3", "date": "2026-07-23T15:00:13+02:00", "subject": "ci: centralize dispatch input validation", "body": "Clarify the bot-only workflow inputs and validate any supplied tuple through\na shared reusable workflow without interpolating input values into shell source.\n\nCo-Authored-By: Claude Opus 4.8 (1M context) \n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMergeDate: Mon Jul 27 10:16:56 2026\n(Merged from https://github.com/openssl/openssl/pull/32053)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ef7644b2cd128c76c965e5109ed65eaf086d32c1", "date": "2026-07-23T15:03:53+02:00", "subject": "ci: allow dispatch cancellation", "body": "Reviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMergeDate: Mon Jul 27 10:16:57 2026\n(Merged from https://github.com/openssl/openssl/pull/32053)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "60fbcde982eedbe871a7a2ac9a5854bfbd18e434", "date": "2026-07-23T15:04:14+02:00", "subject": "ci: preserve valgrind fork runs", "body": "Reviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMergeDate: Mon Jul 27 10:16:58 2026\n(Merged from https://github.com/openssl/openssl/pull/32053)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vitest", "sha": "d8b040c63aaffad22aa9921536ae201cff53cc0f", "date": "2026-07-23T15:10:10+02:00", "subject": "perf: reuse compiled code across vm pool contexts and prewarm the module graph", "body": "", "style": "conventional", "type": "perf", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7e4aa397b9ca7893da31fdcd4a881e842d50b14d", "date": "2026-07-23T22:34:17+09:00", "subject": "Fix VS2013 build of QUIC radix tests", "body": "Visual Studio 2013 does not provide the C99 __func__ predefined identifier in C mode. Use OPENSSL_FUNC, which maps to the compiler-supported equivalent.\n\nThis is the only source change required to compile current master with VS2013 Update 5 (cl 18.00.40629), on both x86 and x64.\n\nAssisted-by: OpenCode:GLM-5.2\n\nReviewed-by: Neil Horman \nReviewed-by: Milan Broz \nMergeDate: Fri Jul 31 13:13:04 2026\nReviewed-by: Bob Beck \nReviewed-by: Nikola Pajkovsky \nMerged-from: https://github.com/openssl/openssl/pull/32419", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8027baa90ae77732942b140a0134fe60774fef13", "date": "2026-07-23T22:34:17+09:00", "subject": "Fix VS2013 build of QUIC radix tests", "body": "Visual Studio 2013 does not provide the C99 __func__ predefined identifier in C mode. Use OPENSSL_FUNC, which maps to the compiler-supported equivalent.\n\nThis is the only source change required to compile current master with VS2013 Update 5 (cl 18.00.40629), on both x86 and x64.\n\nAssisted-by: OpenCode:GLM-5.2\n\nReviewed-by: Neil Horman \nReviewed-by: Milan Broz \nMergeDate: Fri Jul 31 13:13:04 2026\n(Merged from https://github.com/openssl/openssl/pull/31988)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "79f0d356fcfcb73e150ff8b96754f25599240d15", "date": "2026-07-23T22:35:19+09:00", "subject": "Fix C99 printf length modifiers with VS2013 CRT", "body": "The Visual Studio 2013 CRT does not support the C99 z, t and j length modifiers. Passing such formats to _vsnprintf_s invokes the invalid-parameter path and normally terminates the process.\n\nFor _MSC_VER < 1900, translate z and t to the pointer-sized I modifier and j to I64 before calling the CRT. Allocate a translated format only when needed, leaving newer MSVC versions and unaffected formats unchanged. Delegate BIO_snprintf to BIO_vsnprintf so all BIO formatting entry points use the same compatibility path.\n\nAdd focused coverage for BIO_snprintf, BIO_vsnprintf and BIO_printf, including the large-output allocation path and truncation behavior.\n\nFixes #31645\n\nAssisted-by: OpenCode:GLM-5.2\n\nReviewed-by: Neil Horman \nReviewed-by: Milan Broz \nMergeDate: Fri Jul 31 13:13:06 2026\nReviewed-by: Bob Beck \nReviewed-by: Nikola Pajkovsky \nMerged-from: https://github.com/openssl/openssl/pull/32419", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2da90c2f053dd97821a518b62a3553888af861e2", "date": "2026-07-23T22:35:19+09:00", "subject": "Fix C99 printf length modifiers with VS2013 CRT", "body": "The Visual Studio 2013 CRT does not support the C99 z, t and j length modifiers. Passing such formats to _vsnprintf_s invokes the invalid-parameter path and normally terminates the process.\n\nFor _MSC_VER < 1900, translate z and t to the pointer-sized I modifier and j to I64 before calling the CRT. Allocate a translated format only when needed, leaving newer MSVC versions and unaffected formats unchanged. Delegate BIO_snprintf to BIO_vsnprintf so all BIO formatting entry points use the same compatibility path.\n\nAdd focused coverage for BIO_snprintf, BIO_vsnprintf and BIO_printf, including the large-output allocation path and truncation behavior.\n\nFixes #31645\n\nAssisted-by: OpenCode:GLM-5.2\n\nReviewed-by: Neil Horman \nReviewed-by: Milan Broz \nMergeDate: Fri Jul 31 13:13:06 2026\n(Merged from https://github.com/openssl/openssl/pull/31988)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3c0f3cf80eba4112bc6abfc24f953d828b5e9a69", "date": "2026-07-23T16:32:23+02:00", "subject": "Add PKCS7_verify fuzzer", "body": "Fuzz PKCS7_verify with PKCS7_NOVERIFY flag\nto exercise signature verification without\nrequiring a trusted certificate store.\n\nMotivated by CVE-2026-45447.\n\nAssisted-by: ChatGPT:gpt-5.6\n\nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMergeDate: Fri Aug 7 13:36:57 2026\n(Merged from https://github.com/openssl/openssl/pull/32093)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0ec289545a054ee21bac7ded9d474288e43554b7", "date": "2026-07-23T09:58:25-07:00", "subject": "configure.ac: don't enable jemalloc if ASAN enabled on macOS, error if both enabled", "body": "ASAN and jemalloc both replace `malloc` with their version.\n\nPreviously, configuring with `--enable-asan` on macOS would enable both\njemalloc and ASAN. I'm not sure which one won, but seemed undesired.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "0a4146d1897f1ead5031ec392403a7e6c52257a7", "date": "2026-07-23T12:05:15-05:00", "subject": "docs: update signal forms skill to match the current implementation", "body": "(cherry picked from commit 7e39b4aedb386e5e69a8c317987d4efb71f9756d)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "7e39b4aedb386e5e69a8c317987d4efb71f9756d", "date": "2026-07-23T12:05:15-05:00", "subject": "docs: update signal forms skill to match the current implementation", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "1caafa6d8abae54dd41e2d48fb9f81c98cd8751a", "date": "2026-07-23T10:21:44-07:00", "subject": "test(compiler-cli): format compliance TEST_CASES.json with prettier", "body": "Reformats the TEST_CASES.json files touched by the following change so they\nsatisfy the repo's prettier check (short inputFiles/files arrays collapsed to a\nsingle line). Pure formatting; the parsed JSON is unchanged. Split into its own\ncommit so the coverage change that follows is easy to review.\n\n(cherry picked from commit 5245ca5ba72fd165b839d0c45b98f7b50d13a4d5)", "style": "conventional", "type": "test", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "5245ca5ba72fd165b839d0c45b98f7b50d13a4d5", "date": "2026-07-23T10:21:44-07:00", "subject": "test(compiler-cli): format compliance TEST_CASES.json with prettier", "body": "Reformats the TEST_CASES.json files touched by the following change so they\nsatisfy the repo's prettier check (short inputFiles/files arrays collapsed to a\nsingle line). Pure formatting; the parsed JSON is unchanged. Split into its own\ncommit so the coverage change that follows is easy to review.", "style": "conventional", "type": "test", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "5d5b2ea72d03ffc2350e66e202cbca0bf491a5f2", "date": "2026-07-24T00:18:28+05:30", "subject": "fix(migrations): correctly detect `then`/`else` keywords in control flow migration", "body": "The control flow migration determines whether an `*ngIf` uses a `then`\nand/or `else` clause by regex matching the raw microsyntax string for\nthe literal keywords `then`/`else`. The regexes only checked that the\nkeyword was preceded by a non-word character, but not that it was\nfollowed by one.\n\nAs a result, a template reference name that merely starts with `then`\n(e.g. `else thenBlock`) or `else` was misidentified as the `then`/`else`\nkeyword itself. This caused the migration to take the wrong code path\n(e.g. then+else instead of else-only), which in turn made\n`getTemplateName()` compute a `slice(start, end)` with `start > end`,\nproducing an empty template name. That empty placeholder was never\nresolved and was silently emitted as an invalid\n``, dropping the\noriginal template content without any warning.\n\nAdd a negative lookahead `(?![\\w\\d])` to both regexes so `then`/`else`\nare only matched as whole keywords, not as a prefix of a longer\ntemplate reference name.\n\nFixes #69914\n\n(cherry picked from commit 5ad823139758b4d3a8a021d378b008c3457f8689)", "style": "conventional", "type": "fix", "scope": "migrations", "breaking": false} {"repo": "angular", "sha": "5ad823139758b4d3a8a021d378b008c3457f8689", "date": "2026-07-24T00:18:28+05:30", "subject": "fix(migrations): correctly detect `then`/`else` keywords in control flow migration", "body": "The control flow migration determines whether an `*ngIf` uses a `then`\nand/or `else` clause by regex matching the raw microsyntax string for\nthe literal keywords `then`/`else`. The regexes only checked that the\nkeyword was preceded by a non-word character, but not that it was\nfollowed by one.\n\nAs a result, a template reference name that merely starts with `then`\n(e.g. `else thenBlock`) or `else` was misidentified as the `then`/`else`\nkeyword itself. This caused the migration to take the wrong code path\n(e.g. then+else instead of else-only), which in turn made\n`getTemplateName()` compute a `slice(start, end)` with `start > end`,\nproducing an empty template name. That empty placeholder was never\nresolved and was silently emitted as an invalid\n``, dropping the\noriginal template content without any warning.\n\nAdd a negative lookahead `(?![\\w\\d])` to both regexes so `then`/`else`\nare only matched as whole keywords, not as a prefix of a longer\ntemplate reference name.\n\nFixes #69914", "style": "conventional", "type": "fix", "scope": "migrations", "breaking": false} {"repo": "openssl", "sha": "ee411aa8d3369c7e7a939e41eb54d7ae0237c660", "date": "2026-07-23T22:02:31+02:00", "subject": "Add a note about internal documentation", "body": "Document internal subdirectory in doc/README.md.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Norbert Pocs \nMergeDate: Fri Aug 7 13:30:30 2026\n(Merged from https://github.com/openssl/openssl/pull/32058)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "3fceb84560f8e7a9b3ea68a20eb3b74d01c24416", "date": "2026-07-23T22:25:57+02:00", "subject": "refactor(options): store dict options in :set-string form", "body": "The parent commit made a valiant effort to store \"dict options\" in their\nreified form, but this is more trouble than it's worth:\n- inconsistent model for developers to understand.\n- string lifetime issues (the \"varp\" convention is to pass around an\n aliased, long-lived option value).\n- lots of extra plumbing to deal with the 2 different option-storage\n paradigms.", "style": "conventional", "type": "refactor", "scope": "options", "breaking": false} {"repo": "vite", "sha": "0f431606b20cc11219c6c6ab9ac02a6d0ec8f0f8", "date": "2026-07-24T05:10:21+08:00", "subject": "fix: don't swallow the positional argument after `--profile`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f95bd7393584e96152e1a2f68736c634ed8de614", "date": "2026-07-24T05:27:14+08:00", "subject": "feat(float): 'previewpopup' option #26799", "body": "Problem:\nThe preview-window (:pedit, etc.) always uses a split, but it would be\nuseful as a floatwin (or \"popup\").\n\nSolution:\nSupport Vim's 'previewpopup' option.", "style": "conventional", "type": "feat", "scope": "float", "breaking": false} {"repo": "openssl", "sha": "8068ac6dd6fa438741b5e28c80ead2f6a26ea533", "date": "2026-07-23T23:44:56+02:00", "subject": "apps: test ec -check key consistency reporting", "body": "Add a test recipe subtest checking that 'openssl ec -check' reports a\nvalid key and an invalid one, using a new data file that combines the\ntestec-p256.pem private scalar with the group generator as the public\nkey, so it loads but fails the pairwise consistency check.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Daniel Kubec \nReviewed-by: Paul Dale \nMergeDate: Thu Jul 30 08:53:23 2026\n(Merged from https://github.com/openssl/openssl/pull/32059)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e8195baaefd8a6cfe403b262af43d00daf709344", "date": "2026-07-23T23:48:57+02:00", "subject": "refactor(options): 'previewpopup' validation", "body": "Use a schema instead of manual parsing.", "style": "conventional", "type": "refactor", "scope": "options", "breaking": false} {"repo": "openssl", "sha": "75b27b995fc4306bbd8b61ec7dd276aba1650b8e", "date": "2026-07-23T17:53:43-04:00", "subject": "Rework how to fetch components in the drbg", "body": "Commit c9a2ce61118c7f73bc4898eedec64c2bde8bb7a0 introduced some code\ninto the drbg in an effort to get it to select the same provider as is\nspecified for the drbg. Unfortunately this creates a problem when a\nuser has altered the identity of a predefined provider (in this case\nfips).\n\nThe DRBG is passed a set of parameters when instantiating, which\nincludes the name of the provider. This provider uses the name that the\ncore knows it as, which may be different than \"fips\", which the fips\nprovider always referrs to itself as.\n\nFix it by skipping the provider name check when operating in the fips\nmodule, as we should only ever get fips approved algs from within that\nspace\n\nFixes a customer issue, who noted that since this was introduced, using\nidentity configurations no longer works.\n\n(cherry picked from commit 71ec72702be26ef9130530624fada6038e59fc01)\nReviewed-by: Mounir Idrassi \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMerge-date: Fri Aug 21 11:51:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32357", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c116a39920be6d40e80f613fa1735ca77786a0a8", "date": "2026-07-23T17:53:43-04:00", "subject": "Rework how to fetch components in the drbg", "body": "Commit c9a2ce61118c7f73bc4898eedec64c2bde8bb7a0 introduced some code\ninto the drbg in an effort to get it to select the same provider as is\nspecified for the drbg. Unfortunately this creates a problem when a\nuser has altered the identity of a predefined provider (in this case\nfips).\n\nThe DRBG is passed a set of parameters when instantiating, which\nincludes the name of the provider. This provider uses the name that the\ncore knows it as, which may be different than \"fips\", which the fips\nprovider always referrs to itself as.\n\nFix it by skipping the provider name check when operating in the fips\nmodule, as we should only ever get fips approved algs from within that\nspace\n\nFixes a customer issue, who noted that since this was introduced, using\nidentity configurations no longer works.\n\nReviewed-by: Bob Beck \nReviewed-by: Simo Sorce \nMergeDate: Wed Aug 12 19:26:01 2026\n\n(cherry picked from commit 71ec72702be26ef9130530624fada6038e59fc01)\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Andrew Dinh \n(Merged from https://github.com/openssl/openssl/pull/32355)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c8e6ab995c72b52f4e0f73e605f84b7eb922fee3", "date": "2026-07-23T17:53:43-04:00", "subject": "Rework how to fetch components in the drbg", "body": "Commit c9a2ce61118c7f73bc4898eedec64c2bde8bb7a0 introduced some code\ninto the drbg in an effort to get it to select the same provider as is\nspecified for the drbg. Unfortunately this creates a problem when a\nuser has altered the identity of a predefined provider (in this case\nfips).\n\nThe DRBG is passed a set of parameters when instantiating, which\nincludes the name of the provider. This provider uses the name that the\ncore knows it as, which may be different than \"fips\", which the fips\nprovider always referrs to itself as.\n\nFix it by skipping the provider name check when operating in the fips\nmodule, as we should only ever get fips approved algs from within that\nspace\n\nFixes a customer issue, who noted that since this was introduced, using\nidentity configurations no longer works.\n\nReviewed-by: Bob Beck \nReviewed-by: Simo Sorce \nMergeDate: Wed Aug 12 19:26:01 2026\n(Merged from https://github.com/openssl/openssl/pull/32060)\n\n(cherry picked from commit 71ec72702be26ef9130530624fada6038e59fc01)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "71ec72702be26ef9130530624fada6038e59fc01", "date": "2026-07-23T17:53:43-04:00", "subject": "Rework how to fetch components in the drbg", "body": "Commit c9a2ce61118c7f73bc4898eedec64c2bde8bb7a0 introduced some code\ninto the drbg in an effort to get it to select the same provider as is\nspecified for the drbg. Unfortunately this creates a problem when a\nuser has altered the identity of a predefined provider (in this case\nfips).\n\nThe DRBG is passed a set of parameters when instantiating, which\nincludes the name of the provider. This provider uses the name that the\ncore knows it as, which may be different than \"fips\", which the fips\nprovider always referrs to itself as.\n\nFix it by skipping the provider name check when operating in the fips\nmodule, as we should only ever get fips approved algs from within that\nspace\n\nFixes a customer issue, who noted that since this was introduced, using\nidentity configurations no longer works.\n\nReviewed-by: Bob Beck \nReviewed-by: Simo Sorce \nMergeDate: Wed Aug 12 19:26:01 2026\n(Merged from https://github.com/openssl/openssl/pull/32060)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a8553739af08e9d26bc573d914785d7118bc6a6e", "date": "2026-07-24T00:02:53+02:00", "subject": "apps: test ecparam -inform, -outform and -conv_form handling", "body": "Add test recipe subtests for the ecparam format options: the PEM/DER\nparameter roundtrip, rejection of DER input without -inform as the input\nformat defaults to PEM, the restriction of -outform to PEM and DER, and\nthe -conv_form selection of the generator point encoding in explicit\nparameters including that it has no effect on named curve parameters.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Daniel Kubec \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 16:55:11 2026\n(Merged from https://github.com/openssl/openssl/pull/32061)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "869b27ac3e421a9c2a4f02bf8abce38670e40b66", "date": "2026-07-23T17:37:35-05:00", "subject": "Add OIDs specified by RFC 9925", "body": "Defines the id-alg-unsigned and id-rdna-unsigned OIDs from RFC 9925\n\nCLA: trivial\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Dmitry Belyavskiy \nMerge-date: Wed Aug 19 10:46:43 2026\nMerged-from: https://github.com/openssl/openssl/pull/32062", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c4013cdb85fcb8fa3f0613b5263a09af21832aef", "date": "2026-07-24T02:44:11+02:00", "subject": "tidy-up: comments, messages, formatting", "body": "- 'null-terminate', sync casing.\n- add an `#endif` comment.\n- avoid a few instances of 'will'.\n- configure: 'aws' -> 'aws-sigv4', where missing.\n- unfold/fold lines.\n- update memzero/strzero comments.\n- uppercase 'CRLF'.\n\nCloses #22443", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "47d64e422366a0272e55fb812183f8faa9f88e89", "date": "2026-07-24T10:36:26+08:00", "subject": "feat: extend global focusOutline support to more components", "body": "* feat(theme): support focusOutline token for borderless controls\n\n* feat(theme): support focusOutline for borderless controls and Rate\n\n* feat(theme): use shared focus outline for Steps and Splitter", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ec444cb914aa3d4b21ce108e7953743998436369", "date": "2026-07-24T10:47:51+08:00", "subject": "dtls: fix DTLSv1_listen record sequence after cookie verify", "body": "DTLSv1_listen() copied the ClientHello record sequence into the\nHelloVerifyRequest, but after a valid cookie it advanced the write\nsequence by one. If the client retransmitted ClientHello, the following\nServerHello could be sent with the wrong record sequence.\n\nSeed the DTLS write record sequence from the valid-cookie ClientHello\nbefore continuing the handshake. Also make DTLS 1.2 write sequence\nincrements fail at the 48-bit wire sequence boundary instead of carrying\ninto the unused high bytes.\n\nAdd tests for the ServerHello record sequence and DTLS 1.2 sequence wrap\nhandling.\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Mounir Idrassi \nMerge-date: Mon Sep 7 08:15:23 2026\nMerged-from: https://github.com/openssl/openssl/pull/32064", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "9af244f6f71991cbad25359ea558d51c2373e3a9", "date": "2026-07-24T11:42:25+08:00", "subject": "docs: bulk fix punctuation, spacing and typos", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "41c465896e8b11b1eb9c5fbdafbdcc528e189a2c", "date": "2026-07-24T11:57:06+08:00", "subject": "fix(bundledDev): print build errors to the terminal when an HMR update fails", "body": "", "style": "conventional", "type": "fix", "scope": "bundledDev", "breaking": false} {"repo": "angular", "sha": "aa6d3189c141a9247615ed14d856813ea1b95c8d", "date": "2026-07-23T23:13:42-05:00", "subject": "fix(core): initialize hydration triggers after late runtime activation", "body": "After incremental hydration became tree-shakable, application bootstrap could finish before a routed component activated the runtime. The one-time trigger scan was then skipped, leaving routed and nested hydration triggers uninitialized.\n\nCoordinate application bootstrap with runtime activation and initialize once both have occurred.\n\nFixes #69908\n\n(cherry picked from commit cfe2cda1100bb84e2ecf954687126c4d30212d07)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "cfe2cda1100bb84e2ecf954687126c4d30212d07", "date": "2026-07-23T23:13:42-05:00", "subject": "fix(core): initialize hydration triggers after late runtime activation", "body": "After incremental hydration became tree-shakable, application bootstrap could finish before a routed component activated the runtime. The one-time trigger scan was then skipped, leaving routed and nested hydration triggers uninitialized.\n\nCoordinate application bootstrap with runtime activation and initialize once both have occurred.\n\nFixes #69908", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "vite", "sha": "95a3cdab83e1125b03d2e8dd942fb6b64209e5fa", "date": "2026-07-24T14:49:07+09:00", "subject": "feat: add `input` to `server.fs.allow`", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vite", "sha": "b1186c36d06bb94941c58e8272fc4acb8512c93b", "date": "2026-07-24T13:49:23+08:00", "subject": "fix(hmr): preserve environment snapshot during server restart", "body": "", "style": "conventional", "type": "fix", "scope": "hmr", "breaking": false} {"repo": "vite", "sha": "23daf0a81ab0e35a8e0168a6c4b2157b4a3b4e95", "date": "2026-07-24T14:03:34+08:00", "subject": "docs(config): rewrite VS Code debugging section", "body": "", "style": "conventional", "type": "docs", "scope": "config", "breaking": false} {"repo": "openssl", "sha": "7d87fe4d745cfe6a0e6d59c20264d96d672c96e7", "date": "2026-07-24T13:37:46+07:00", "subject": "Fix mutcbk_inject_frames after mutation", "body": "mutcbk_inject_frames returned 0 once its one-time mutation was applied, which\ntells the QTX the packet send failed and tears down the connection instead of\ncontinuing the test with unmutated packets. Now it passes subsequent packets\nthrough unmodified by returning the input header, iovec, and count unchanged.\n\nAssisted-by: Claude:claude-sonnet-5\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Neil Horman \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 14:02:07 2026\n(Merged from https://github.com/openssl/openssl/pull/32065)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "836d497803b12f145cc949f94c0b0d625ffa91d6", "date": "2026-07-24T13:37:46+07:00", "subject": "Fix mutcbk_inject_frames after mutation", "body": "mutcbk_inject_frames returned 0 once its one-time mutation was applied, which\ntells the QTX the packet send failed and tears down the connection instead of\ncontinuing the test with unmutated packets. Now it passes subsequent packets\nthrough unmodified by returning the input header, iovec, and count unchanged.\n\nAssisted-by: Claude:claude-sonnet-5\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Neil Horman \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 14:01:46 2026\n(Merged from https://github.com/openssl/openssl/pull/32065)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ebdc3b83bda05156daa05f205da20fc5dace6ee9", "date": "2026-07-24T13:37:46+07:00", "subject": "Fix mutcbk_inject_frames after mutation", "body": "mutcbk_inject_frames returned 0 once its one-time mutation was applied, which\ntells the QTX the packet send failed and tears down the connection instead of\ncontinuing the test with unmutated packets. Now it passes subsequent packets\nthrough unmodified by returning the input header, iovec, and count unchanged.\n\nAssisted-by: Claude:claude-sonnet-5\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Neil Horman \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 14:01:01 2026\n(Merged from https://github.com/openssl/openssl/pull/32065)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "be9fd7be9b132fd5a4304f4604848b096af4fdb1", "date": "2026-07-24T13:37:46+07:00", "subject": "Fix mutcbk_inject_frames after mutation", "body": "mutcbk_inject_frames returned 0 once its one-time mutation was applied, which\ntells the QTX the packet send failed and tears down the connection instead of\ncontinuing the test with unmutated packets. Now it passes subsequent packets\nthrough unmodified by returning the input header, iovec, and count unchanged.\n\nAssisted-by: Claude:claude-sonnet-5\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Neil Horman \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 14:00:23 2026\n(Merged from https://github.com/openssl/openssl/pull/32065)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "2ad9ec262a7bd8e30788df537eb217abe9af82ab", "date": "2026-07-24T14:43:25+08:00", "subject": "test: run playgrounds under bundled dev, skip unsupported cases", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "72207b17daa328d83b1609b4bc5cc4047fe4ba01", "date": "2026-07-24T15:44:56+09:00", "subject": "Fix EXCEPT publication test to check subscriber", "body": "Commit fd366065e06 added tests intended to verify that rows inserted\non the publisher are replicated to the subscriber when using multiple\npublications, with one excluding the target table via EXCEPT and\nanother including it.\n\nHowever, the tests queried the publisher instead of the subscriber.\nSince the rows were inserted directly into the publisher, the checks\nwould always succeed, providing no coverage of replication.\n\nFix this by querying the subscriber so the tests verify the replicated\nstate.\n\nAuthor: Fujii Masao \nReviewed-by: Ayush Tiwari \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/CAHGQGwGfXUO7f4t6KNGurYwg6QsnLtpP0K3EACbAwYWtxGfKfQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b77868f169adcdf31edbc80d8a875204ed7ba191", "date": "2026-07-24T15:46:46+09:00", "subject": "doc: Add missing CREATE/ALTER PUBLICATION parameter descriptions", "body": "Document table_name, column_name, and schema_name in the CREATE\nPUBLICATION and ALTER PUBLICATION reference pages. Also add anchors for\nthe ALTER PUBLICATION parameter list, matching the style already used by\nCREATE PUBLICATION.\n\nAuthor: Peter Smith \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAHut+Ptekz+TO4ui8-fiBm4Y+O2v=HQnkK_cW4G=w9ep8654EA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "c050842b6e9e8adeb73dd7046321eecb79c4ebfc", "date": "2026-07-24T09:45:23+02:00", "subject": "perf: bundle vitest's own dependencies", "body": "", "style": "conventional", "type": "perf", "scope": "", "breaking": false} {"repo": "neovim", "sha": "66441e549d5c560f8b1d6d6b6e71244ce3569604", "date": "2026-07-24T00:52:19-07:00", "subject": "fix(text): validate `opts.expandtab` in `vim.text.indent()` #40936", "body": "", "style": "conventional", "type": "fix", "scope": "text", "breaking": false} {"repo": "angular", "sha": "350ea416b64c98929bdfbcc1dded220bef0df378", "date": "2026-07-24T09:54:04+02:00", "subject": "refactor(core): align navigation types with built in ones", "body": "Aligns our clone of the navigation API types with the built-in TypeScript types. This is related to an internal issue.\n\n(cherry picked from commit d79b3b65e9536accd825620331185706c3339714)", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "d79b3b65e9536accd825620331185706c3339714", "date": "2026-07-24T09:54:04+02:00", "subject": "refactor(core): align navigation types with built in ones", "body": "Aligns our clone of the navigation API types with the built-in TypeScript types. This is related to an internal issue.", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "vite", "sha": "45dc235ecbab3dad3fb66f21ca32471a5aa5cbcf", "date": "2026-07-24T16:28:39+08:00", "subject": "refactor: drop read from the rolldown hook args", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "tmux", "sha": "47f210a9d257b84d624ab6c755c9355a72cf9ac0", "date": "2026-07-24T08:49:23Z", "subject": "Replace outside as well as empty cells with the border.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "e361982acf62f16c4da34ca0afc91724281e4171", "date": "2026-07-24T09:49:32+01:00", "subject": "Fix some tests.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "84d3041da2491a9e0bd4c0c4effd5319a374ada5", "date": "2026-07-24T16:53:10+08:00", "subject": "fix: make table and sider triggers keyboard accessible", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "eslint", "sha": "784dfbe98f0222426fd200153a5ce2e5624151f0", "date": "2026-07-24T18:40:06+09:00", "subject": "docs: Clarify `no-eq-null` description", "body": "* docs: Correct operators in no-eq-null documentation\n\n* docs: Improve clarity in no-eq-null documentation\n\n* docs: correct no-eq-null operators and grammar\n\nCo-authored-by: lumir \n\n---------", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "f6fa9eca7d64dff327635a24ad31ce0c6c0ac49b", "date": "2026-07-24T09:44:28Z", "subject": "docs: update default value for strictTemplates", "body": "Updates the strictTemplates documentation in Angular compiler options to note that the default is true, replacing the reference to the obsolete ng new --strict flag.\n\n(cherry picked from commit 2bcd12a6a59aea228eb112bec138fdbb0659f86f)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "2bcd12a6a59aea228eb112bec138fdbb0659f86f", "date": "2026-07-24T09:44:28Z", "subject": "docs: update default value for strictTemplates", "body": "Updates the strictTemplates documentation in Angular compiler options to note that the default is true, replacing the reference to the obsolete ng new --strict flag.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "eb459f78e8bc26ffe773ecd7e741c2ca67ca7ed2", "date": "2026-07-24T12:51:56+03:00", "subject": "fix(fastify): run middleware on routes excluded from global prefix", "body": "createMiddlewareFactory re-prefixed every non-prefixed middleware path, so a concrete\nroute excluded via setGlobalPrefix(prefix, { exclude }) (e.g. /graphql) was rewritten to\n${prefix}/. Route-scoped middleware was then registered on the wrong path and never\nran on the excluded route. Guard the re-prefix so only a path that collapsed to the root is\nexpanded, leaving concrete excluded routes unprefixed.\n\nFollow-up to #14895 / #14894. Preserves the exclude: ['/'] behavior from #13401.", "style": "conventional", "type": "fix", "scope": "fastify", "breaking": false} {"repo": "eslint", "sha": "6ddf8587e5cfeeb7328332601bb76d98b8a198ea", "date": "2026-07-24T18:55:25+09:00", "subject": "docs: fix broken Specify Parser Options anchor link", "body": "* docs: fix broken Specify Parser Options anchor link\n\n* Update docs/src/rules/rest-spread-spacing.md\r\n\r\nThanks for the clarification. That makes sense.\n\nCo-authored-by: fnx <966276+DMartens@users.noreply.github.com>\n\n* Update docs/src/rules/rest-spread-spacing.md\n\nCo-authored-by: lumir \n\n---------", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "3ac77d9dd742968961af38a5a91ed6b061ceda7d", "date": "2026-07-24T18:18:06+08:00", "subject": "feat(dev): resolve interface name for explicit host in network URLs", "body": "", "style": "conventional", "type": "feat", "scope": "dev", "breaking": false} {"repo": "neovim", "sha": "9604c758b36fd76227997d425a3b8220bdf90679", "date": "2026-07-24T12:32:46+02:00", "subject": "perf(options): no-alloc keysets", "body": "", "style": "conventional", "type": "perf", "scope": "options", "breaking": false} {"repo": "neovim", "sha": "ce9cc43f23d749261d63c06e1e330cf49b01f533", "date": "2026-07-24T03:45:13-07:00", "subject": "fix(dir): open CWD from unnamed buffer with `-` mapping #40926", "body": "", "style": "conventional", "type": "fix", "scope": "dir", "breaking": false} {"repo": "neovim", "sha": "5d76b9836a7c7d49d15bf5530826b7783d098cf4", "date": "2026-07-24T03:48:47-07:00", "subject": "feat(dir): support counts for parent navigation #40927", "body": "", "style": "conventional", "type": "feat", "scope": "dir", "breaking": false} {"repo": "neovim", "sha": "c244a4bace383f0c0e4ba97bda259bb2373426b3", "date": "2026-07-24T13:03:31+02:00", "subject": "refactor(options): rename, simplify", "body": "", "style": "conventional", "type": "refactor", "scope": "options", "breaking": false} {"repo": "openssl", "sha": "b4af0eafe01790c11f4191852f3995625a478d9e", "date": "2026-07-24T13:27:36+02:00", "subject": "ci: accept uppercase hex in dispatch head_sha validation", "body": "Co-Authored-By: Claude Opus 4.8 (1M context) \n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMergeDate: Mon Jul 27 10:17:00 2026\n(Merged from https://github.com/openssl/openssl/pull/32053)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c507351d3d8e94156896ec64505e96d2e53a29a4", "date": "2026-07-24T13:35:00+02:00", "subject": "updating fuzz-corpora submodule", "body": "Reviewed-by: Tomas Mraz \nReviewed-by: Norbert Pocs \nMergeDate: Wed Jul 29 09:24:40 2026\n(Merged from https://github.com/openssl/openssl/pull/32067)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "1adddde12befa8dbf6ff519ecbfae928e18f6cf4", "date": "2026-07-24T13:46:48+02:00", "subject": "feat(cli): support -p shorthand for --project", "body": "", "style": "conventional", "type": "feat", "scope": "cli", "breaking": false} {"repo": "vitest", "sha": "1ffafc0878db61c92ace581b11d926e4699ca079", "date": "2026-07-24T13:47:12+02:00", "subject": "refactor: deduplicate run-level option propagation and document the resolution model", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "curl", "sha": "a479459ea39ac0e7de37237f9f0ba9a0c9d7f470", "date": "2026-07-24T14:20:58+02:00", "subject": "http: fix non-tunneling proxy hostname use", "body": "Make sure hostname used in URL to proxy is IDN decoded form, but keep\nthe original hostname in case it was ipv6.\n\nFixes #22382\nReported-by: RMMoreton on github\nCloses #22385", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "3d5d2ad4a933a9cbd391befe61a4145ae1d1ebb2", "date": "2026-07-24T08:37:47-04:00", "subject": "fix(vim-patch.sh): omit gpg for git-log #37173", "body": "\"vim-patch.sh -L\" includes gpg signature for \"log.showSignature=true\"\nin user's .gitconfig.\nvim-patch.sh expects no signature for 1-line output per patch.", "style": "conventional", "type": "fix", "scope": "vim-patch.sh", "breaking": false} {"repo": "babel", "sha": "e74e391f324d1e63533e75c118a02689cb8aeeb4", "date": "2026-07-24T20:39:19+08:00", "subject": "Simplify AST validators", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "4c853d6462a0a1000f0fd33e57eec8ba12bd898b", "date": "2026-07-24T14:40:47+02:00", "subject": "test: add `@ljharb/tsconfig` as a dependency", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5456f05af55efd9e2e6741e87665d1b5f67968f8", "date": "2026-07-24T14:42:25+02:00", "subject": "ech: check memory failure in vpm allocation in store final check", "body": "This fixes issue found by fuzz test in updated corpora for\nechconfiglist_parser.\n\nReviewed-by: Norbert Pocs \nReviewed-by: Paul Dale \nReviewed-by: Tomas Mraz \nMergeDate: Mon Jul 27 16:36:15 2026\n(Merged from https://github.com/openssl/openssl/pull/32070)\n\n(cherry picked from commit 544e28ee06ac247e8fa02b46e306145d1c52cf17)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "544e28ee06ac247e8fa02b46e306145d1c52cf17", "date": "2026-07-24T14:42:25+02:00", "subject": "ech: check memory failure in vpm allocation in store final check", "body": "This fixes issue found by fuzz test in updated corpora for\nechconfiglist_parser.\n\nReviewed-by: Norbert Pocs \nReviewed-by: Paul Dale \nReviewed-by: Tomas Mraz \nMergeDate: Mon Jul 27 16:36:15 2026\n(Merged from https://github.com/openssl/openssl/pull/32070)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "a31f86af738b2979905f6a61eb5d8848d489eed7", "date": "2026-07-24T16:01:37+02:00", "subject": "fix: coalesce concurrent watch-mode restarts, deflake the pool stderr fixture", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "e25873c82074247233f1782ac5eb7fd4c6457092", "date": "2026-07-24T23:09:12+08:00", "subject": "docs: update Ant Design 5.x site link", "body": "* docs: update Ant Design 5.x site link\n\n* docs: correct Ant Design 5.x site URL\n\n* docs: add mirror links to English FAQ", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "bb104ba4dba42eda89bae817264eee959df5bbf0", "date": "2026-07-24T10:54:12-05:00", "subject": "feat(health): improvements to :checkhealth #40935", "body": "Changes:\n- Removed the blank line before the first healthcheck header.\n- Fixed a bug where the entire \"Terminal\" section of `vim.health` would\n not appear if the `infocmp` executable was not found.\n- Fixed a typo in `vim.lsp` when displaying the LSP log level.", "style": "conventional", "type": "feat", "scope": "health", "breaking": false} {"repo": "openssl", "sha": "f544f4970f2719ddaeb3fa1e78572b95c99dd69f", "date": "2026-07-24T12:14:33-04:00", "subject": "add a test to exercize the identity config directive on fips", "body": "Reviewed-by: Bob Beck \nReviewed-by: Simo Sorce \nMergeDate: Wed Aug 12 19:26:04 2026\n\n(cherry picked from commit 42f8848ac6cbd727264fd3a9bfa97861e4ab71b5)\nReviewed-by: Mounir Idrassi \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMerged-from: https://github.com/openssl/openssl/pull/32357", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "03b0a46a0a49c2ed32877df34d8e58402a037f0d", "date": "2026-07-24T12:14:33-04:00", "subject": "add a test to exercize the identity config directive on fips", "body": "Reviewed-by: Bob Beck \nReviewed-by: Simo Sorce \nMergeDate: Wed Aug 12 19:26:04 2026\n\n(cherry picked from commit 42f8848ac6cbd727264fd3a9bfa97861e4ab71b5)\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Andrew Dinh \n(Merged from https://github.com/openssl/openssl/pull/32355)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f07811dea2769a1e03d066418a038e312977ee43", "date": "2026-07-24T12:14:33-04:00", "subject": "add a test to exercize the identity config directive on fips", "body": "Reviewed-by: Bob Beck \nReviewed-by: Simo Sorce \nMergeDate: Wed Aug 12 19:26:04 2026\n(Merged from https://github.com/openssl/openssl/pull/32060)\n\n(cherry picked from commit 42f8848ac6cbd727264fd3a9bfa97861e4ab71b5)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "42f8848ac6cbd727264fd3a9bfa97861e4ab71b5", "date": "2026-07-24T12:14:33-04:00", "subject": "add a test to exercize the identity config directive on fips", "body": "Reviewed-by: Bob Beck \nReviewed-by: Simo Sorce \nMergeDate: Wed Aug 12 19:26:04 2026\n(Merged from https://github.com/openssl/openssl/pull/32060)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "4fbf46d140120ae838db08cbe49544c985956490", "date": "2026-07-24T18:51:24+02:00", "subject": "test: pin `webpack` version to 5.108.4", "body": "* test: pin `webpack` version to 5.108.4\n\n* remove caret", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "bf3d4210c387a8d5e06db395580d5b3c87015f41", "date": "2026-07-24T13:49:47-04:00", "subject": "refactor: 'previewpopup' #40945", "body": "- drop w_maxwidth/w_maxheight, calculate them on-demand from\n 'previewpopup' (this is cheap bc of `opt_keyset`)", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a99d0c681008a2c1c694780a8685cceda35662eb", "date": "2026-07-25T01:55:36+08:00", "subject": "build: support variant builds via env-var and cpack overrides #40947", "body": "Problem: scripts/genappimage.sh runs `make` with no extra flags and\nemits a fixed nvim-linux-.appimage filename, so packaging a\nvariant build (e.g. with the bundled PUC-Rio Lua interpreter)\nrequires editing the Makefile or post-renaming the AppImage, which\nbreaks the AppImage zsync URL.\n\nSolution: forward DEPS_CMAKE_FLAGS and CMAKE_EXTRA_FLAGS to `make`,\naccept OUTPUT so the AppImage (and its zsync Filename:) follows the\ncaller's chosen name, and mark CPACK_PACKAGE_FILE_NAME as CACHE\nSTRING on Linux so the resulting tarball/deb follow as well. Default\nbehavior is unchanged when none of these env vars are set.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "6f80cca0b8af23b37f24fb8ee0229b6901d01df2", "date": "2026-07-24T12:57:44-05:00", "subject": "fix(platform-server): update domino to latest version", "body": "Updates the domino dependency to the latest version as used in the main branch.\n\nThis update contains fixes for https://github.com/angular/domino/pull/32.", "style": "conventional", "type": "fix", "scope": "platform-server", "breaking": false} {"repo": "curl", "sha": "c4dcdb838883ad060c44a30680328451decdad1b", "date": "2026-07-24T20:26:01+02:00", "subject": "spacecheck: cap number of lines per file", "body": "To prevent merging large text files by accident.\n\nSet the cap at 10k lines. The current line number top list is:\n```\n 5577 configure.ac\n 5561 lib/vtls/openssl.c\n 5077 lib/http.c\n 4517 lib/ftp.c\n 4284 lib/multi.c\n```\n\nCloses #22387", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "a956409d90497abca7a795ef8e61168909043cd4", "date": "2026-07-24T22:32:13+03:00", "subject": "docs: update outdated and broken external links in adev", "body": "Several external links pointed at locations that have moved or gone away:\n\n- ng-packagr, webpack, and node-glob renamed their default branch from\n master to main, so the deep links into those repos redirect.\n- The npm package pages moved from npmjs.org to www.npmjs.com.\n- The Angular Material and CDK docs moved from material.angular.io to\n material.angular.dev.\n- The semver reference pointed at the defunct semver.io service; point it\n at the SemVer spec at semver.org instead.\n- The MDN animationiteration event link had a typo (animationitration).\n- compiler-cli's metadata/schema.ts was removed after v12; pin the link to\n the 12.2.x tag where the file still exists.\n- The angularindepth.com article is gone (its domain is parked); link to its\n live home on the Angular In Depth Medium publication.\n- Google moved the Lighthouse user-timing docs to developer.chrome.com.\n\nPoint the links at their current, working locations.\n\n(cherry picked from commit 058c7f9b0c60c959ad441a43e1db4a782e7d5697)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "058c7f9b0c60c959ad441a43e1db4a782e7d5697", "date": "2026-07-24T22:32:13+03:00", "subject": "docs: update outdated and broken external links in adev", "body": "Several external links pointed at locations that have moved or gone away:\n\n- ng-packagr, webpack, and node-glob renamed their default branch from\n master to main, so the deep links into those repos redirect.\n- The npm package pages moved from npmjs.org to www.npmjs.com.\n- The Angular Material and CDK docs moved from material.angular.io to\n material.angular.dev.\n- The semver reference pointed at the defunct semver.io service; point it\n at the SemVer spec at semver.org instead.\n- The MDN animationiteration event link had a typo (animationitration).\n- compiler-cli's metadata/schema.ts was removed after v12; pin the link to\n the 12.2.x tag where the file still exists.\n- The angularindepth.com article is gone (its domain is parked); link to its\n live home on the Angular In Depth Medium publication.\n- Google moved the Lighthouse user-timing docs to developer.chrome.com.\n\nPoint the links at their current, working locations.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "243cc628d0c2425aca735026afe9c5adcbac2550", "date": "2026-07-24T22:03:02+02:00", "subject": "CMS: Guard against int overflow in cms_kek_cipher()", "body": "Mirror the check from cms_kari.c in cms_kemri.c.\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 16:56:18 2026\n(Merged from https://github.com/openssl/openssl/pull/32072)\n\n(cherry picked from commit 14149949d5c13c12c92a17f1a9eb943a003d0f43)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "14149949d5c13c12c92a17f1a9eb943a003d0f43", "date": "2026-07-24T22:03:02+02:00", "subject": "CMS: Guard against int overflow in cms_kek_cipher()", "body": "Mirror the check from cms_kari.c in cms_kemri.c.\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Paul Dale \nMergeDate: Wed Jul 29 16:56:18 2026\n(Merged from https://github.com/openssl/openssl/pull/32072)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a55731050e8c3dbea0b96205cf916443118f6acb", "date": "2026-07-24T22:49:52+02:00", "subject": "httpsig: add RFC 9421 HTTP Message Signatures support", "body": "Add support for signing outgoing HTTP requests per RFC 9421 using\nEd25519 or HMAC-SHA256 algorithms.\n\nNew libcurl options:\n - CURLOPT_HTTPSIG: signing algorithm (\"ed25519\" or \"hmac-sha256\")\n - CURLOPT_HTTPSIG_KEY: path to hex-encoded key file\n - CURLOPT_HTTPSIG_KEYID: key identifier for Signature-Input\n - CURLOPT_HTTPSIG_HEADERS: space-separated components to sign\n\nNew CLI flags: --httpsig, --httpsig-key, --httpsig-keyid,\n--httpsig-headers\n\nThe crypto layer follows the sha256.c multi-backend pattern with\nimplementations for OpenSSL (EVP_DigestSign) and wolfSSL\n(wc_ed25519_sign_msg). HMAC-SHA256 uses the existing Curl_hmacit()\ninfrastructure which works on all backends.\n\nVerified by test 5000 to 5021\n\nAssisted-by: Daniel Stenberg\nSigned-off-by: Sameeh Jubran \nCloses #22386\nCloses #21239", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "61a7794189f5e94ef4b8a103b14370530886f952", "date": "2026-07-24T23:17:54+02:00", "subject": "apps: test info options of the list command", "body": "Exercise the previously untested -commands, -1, -digest-commands,\n-options, -disabled and -objects options in list app.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Paul Dale \nMerge-date: Fri Sep 4 15:08:52 2026\nMerged-from: https://github.com/openssl/openssl/pull/32073", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "a5bc6abd7dfc34b81f6e5f923f33051955f22de8", "date": "2026-07-24T22:06:41Z", "subject": "fix: package_managers.dependabot.dependabot-missing-cooldown.dependabot-missing-cooldown security vulnerability", "body": "Automated security fix generated by OrbisAI Security", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "b84838073c8d3beb0e870fb4e5e346186a2ecdc0", "date": "2026-07-25T00:11:21+02:00", "subject": "tidy-up: minor code fixes and improvements", "body": "- schannel: drop redundant parentheses.\n- os400sys: drop redundant includes.\n Follow-up to ebc5212dacd3db8f5315b5a2d676415848e936d5 #22374\n- pytest: replace `()` with `[]` to match rest of tests.\n- libtests: constify some local pointers.\n- libtests: drop redundant `(long)` casts.\n- lib650: use `CURL_CSTRLEN()`.\n Follow-up to 59dc2bbe07c3b5889e0b380e5fa384d338d9d24e #22424\n\nCloses #22444", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e63cf571067e8d1777adc354b098dfc91c2ab5de", "date": "2026-07-24T16:28:16-07:00", "subject": "fix(zip): keep a leading dash from becoming a backend option", "body": "", "style": "conventional", "type": "fix", "scope": "zip", "breaking": false} {"repo": "neovim", "sha": "f82cf8fd596fa173a028bd04035d0465d09f8d48", "date": "2026-07-24T16:28:25-07:00", "subject": "fix(zip): list a valid empty archive as empty", "body": "", "style": "conventional", "type": "fix", "scope": "zip", "breaking": false} {"repo": "neovim", "sha": "36204b4323378a32df628a0893b8b1c40bac3b39", "date": "2026-07-24T16:28:36-07:00", "subject": "feat(zip): extract archive members", "body": "", "style": "conventional", "type": "feat", "scope": "zip", "breaking": false} {"repo": "neovim", "sha": "1ffbd3598abafd0cc63fe6e4c5c6346cd5a1b784", "date": "2026-07-24T16:54:49-07:00", "subject": "test(zip): cover difficult member names", "body": "", "style": "conventional", "type": "test", "scope": "zip", "breaking": false} {"repo": "postgres", "sha": "bb6125ca3ac756a44be9ef574e89aa5cff482afb", "date": "2026-07-25T10:30:30+09:00", "subject": "Avoid reporting permission-denied publisher sequences as missing", "body": "Previously, if a sequence synchronization batch contained both a sequence\nthat had been dropped on the publisher and another for which the\nreplication role lacked SELECT privilege, the latter was reported\ntwice: once as a permission failure and again as missing on the\npublisher.\n\nThis happened because the permission-denied sequence was not marked as\nfound on the publisher. As a result, when another sequence in the batch\nwas genuinely missing, the later missing-sequence check incorrectly\nclassified the permission-denied sequence as missing as well.\n\nFix this by marking the permission-denied sequence as found before\nreporting the permission failure, so it is not later reported as\nmissing.\n\nReported-by: Noah Misch \nAuthor: Vignesh C \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CALDaNm3LsUjW7PahuCsbYAxajSF+S328tw5E9rF0erdh7dKOXw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "ea096dc01a5423cf2ef548beda98c9b70e1e90df", "date": "2026-07-25T11:03:18+08:00", "subject": "refactor(demo): replace hardcoded styles with design tokens", "body": "", "style": "conventional", "type": "refactor", "scope": "demo", "breaking": false} {"repo": "angular", "sha": "809253e626c78f2e18f42e1ffb17b4cf9c6e163b", "date": "2026-07-25T00:03:49-05:00", "subject": "docs: correct Angular developer skill guidance", "body": "(cherry picked from commit 3e7094b04991303b9636aa35c8629ee5ea7d5a5f)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "3e7094b04991303b9636aa35c8629ee5ea7d5a5f", "date": "2026-07-25T00:03:49-05:00", "subject": "docs: correct Angular developer skill guidance", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "9658ad0ca875a43870be42305324a44041dc5a2b", "date": "2026-07-25T10:53:25+03:00", "subject": "fix(docs-infra): define missing win95 variables in aria autocomplete examples", "body": "The retro-themed aria autocomplete examples style the clear button with\nvar(--win95-gray), var(--win95-dark-gray), var(--win95-light), and\nvar(--win95-shadow), but none of the three example stylesheets defines\nthose custom properties. The variables resolve to nothing, so the button\nfalls back to inherited theme colors — color: var(--win95-shadow) becomes\nthe light-theme text color in dark mode, leaving the button unreadable.\n\nDefine the four variables in :host of each example, matching the values\nalready used by the sibling aria tree examples, so the clear button keeps\nits Windows 95 look in both light and dark themes.\n\n(cherry picked from commit 8cd4c89c298c63e3b2ad0313f76c3e46ea265fe2)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "8cd4c89c298c63e3b2ad0313f76c3e46ea265fe2", "date": "2026-07-25T10:53:25+03:00", "subject": "fix(docs-infra): define missing win95 variables in aria autocomplete examples", "body": "The retro-themed aria autocomplete examples style the clear button with\nvar(--win95-gray), var(--win95-dark-gray), var(--win95-light), and\nvar(--win95-shadow), but none of the three example stylesheets defines\nthose custom properties. The variables resolve to nothing, so the button\nfalls back to inherited theme colors — color: var(--win95-shadow) becomes\nthe light-theme text color in dark mode, leaving the button unreadable.\n\nDefine the four variables in :host of each example, matching the values\nalready used by the sibling aria tree examples, so the clear button keeps\nits Windows 95 look in both light and dark themes.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "a7d67f9ccc2521794129430bc947ff023e220c75", "date": "2026-07-25T11:46:04+03:00", "subject": "fix(docs-infra): fix malformed --webgl-page-background declaration", "body": "In the light theme block of _colors.scss, --webgl-page-background and\n--webgl-gray-unfilled were written on a single line without a separating\nsemicolon, and --webgl-page-background wrapped --page-background in an\ninvalid nested var(). As a result --webgl-page-background got a garbage\nvalue and --webgl-gray-unfilled was never defined in the light theme.\n\nSplit them into two well-formed declarations and drop the nested var(),\nmatching the structure already used in the dark theme block.\n\n(cherry picked from commit 91822538fe78a02a0fdf59cfb29bf9d49ccd6a96)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "91822538fe78a02a0fdf59cfb29bf9d49ccd6a96", "date": "2026-07-25T11:46:04+03:00", "subject": "fix(docs-infra): fix malformed --webgl-page-background declaration", "body": "In the light theme block of _colors.scss, --webgl-page-background and\n--webgl-gray-unfilled were written on a single line without a separating\nsemicolon, and --webgl-page-background wrapped --page-background in an\ninvalid nested var(). As a result --webgl-page-background got a garbage\nvalue and --webgl-gray-unfilled was never defined in the light theme.\n\nSplit them into two well-formed declarations and drop the nested var(),\nmatching the structure already used in the dark theme block.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "postgres", "sha": "4e19081da88677a6f2e9b6d100768f4f715f95d6", "date": "2026-07-25T19:11:34+09:00", "subject": "psql: Allow pg_read_all_stats to see database size in \\l+", "body": "pg_database_size() allows access to users who have either CONNECT\nprivilege on the target database or privileges of the pg_read_all_stats\nrole. However, previously, psql's \\l+ checked only for CONNECT,\nso users with privileges of pg_read_all_stats still saw \"No Access\" for\ndatabases they could not connect to.\n\nFix this by making \\l+ also check\npg_has_role('pg_read_all_stats', 'USAGE'), matching\npg_database_size()'s permission rules.\n\nFor back branches, emit the pg_read_all_stats check only when\nconnected to PostgreSQL 10 or later, since earlier releases do not have\nthat predefined role.\n\nBackpatch to all supported versions.\n\nAuthor: Christoph Berg \nReviewed-by: Álvaro Herrera \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/amCo6qRmnfPVk4-V@msg.df7cb.de\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "4975a186f26a54aee76b38eedc1d06a87599baa4", "date": "2026-07-25T03:56:52-07:00", "subject": "fix(dir): `:browse edit` opens dir browser #40930", "body": "Problem: `:browse edit` (and variants) without an argument fails instead\nof opening a browser.\n\nSolution: Dispatch to dir.lua using the cwd as the implied directory.\nReplace the inherited GUI-only documentation with the supported\nbehavior. Support vsplit, split, tabedit variants.", "style": "conventional", "type": "fix", "scope": "dir", "breaking": false} {"repo": "curl", "sha": "6f39c854c8e9df067e0acf0cccf631035d40ca7c", "date": "2026-07-25T13:43:33+02:00", "subject": "runtests: fix `mode=\"warn\"` tests passing unconditionally, fix test 1752", "body": "Fix test 1712 to pass curl C by setting `COLUMNS` to the highest\naccepted value, and adjust expected results. To avoid envs with varying\nlengths of `LOGDIR` affect the outcome.\n\nApply the same fix to test 459, though it wasn't affected in curl CI.\n\nAlso sync up test 433 `COLUMNS` value with these two tests for\nconsistency.\n\nRef: #22381\nFollow-up to 8e3a2a64d103a46508e17cde76595993de96ea6c #20666\n\nCloses #22388", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b2a5369e708982d1050ccec76967e8fc42f6e2dd", "date": "2026-07-25T14:18:07+02:00", "subject": "runtests: allow comments in `setenv` section, merge sections in test433", "body": "Closes #22389", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "dc21aa07f7bbe50780e0f18b32f357242bd7c6f2", "date": "2026-07-25T14:48:14+02:00", "subject": "fix(ruler): no window-local highlights for last line 'rulerformat'", "body": "Fixes #38777 in case 'rulerformat' is set.\n\nSee PR 38879. Original message:\nProblem: When the 'ruler' is in the last line of the screen, it takes\n local highlight definitions of the current window, tripping an\n assert (since c1648cf).\nSolution: Don't use window-local highlight definitions when the ruler is\n not part of a statusline.", "style": "conventional", "type": "fix", "scope": "ruler", "breaking": false} {"repo": "neovim", "sha": "78d001cff7f08472e5d59d421a767a6791c4c8fb", "date": "2026-07-25T20:59:12+08:00", "subject": "backport: build: support variant builds via env-var and cpack overrides", "body": "build: support variant builds via env-var and cpack overrides\n\nProblem: scripts/genappimage.sh runs `make` with no extra flags and\nemits a fixed nvim-linux-.appimage filename, so packaging a\nvariant build (e.g. with the bundled PUC-Rio Lua interpreter)\nrequires editing the Makefile or post-renaming the AppImage, which\nbreaks the AppImage zsync URL.\n\nSolution: forward DEPS_CMAKE_FLAGS and CMAKE_EXTRA_FLAGS to `make`,\naccept OUTPUT so the AppImage (and its zsync Filename:) follows the\ncaller's chosen name, and mark CPACK_PACKAGE_FILE_NAME as CACHE\nSTRING on Linux so the resulting tarball/deb follow as well. Default\nbehavior is unchanged when none of these env vars are set.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "5a452a8ba53917d65c9d83c3959ed35c2f890613", "date": "2026-07-25T22:54:17+09:00", "subject": "test: add error locations to `no-const-assign`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "eslint", "sha": "059aa895743639e3ab53d14f9d944ff4be042233", "date": "2026-07-25T23:19:34+09:00", "subject": "test: add error locations to `no-useless-concat`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "curl", "sha": "1d7b8e6c29e22c88750fb53a01e30be03104a3fd", "date": "2026-07-25T16:50:51+02:00", "subject": "gitignore: maintenance updates", "body": "- docs/cmdline-opts/.gitignore: also ignore `manpage.tmp.*`.\n Follow-up to a55731050e8c3dbea0b96205cf916443118f6acb #22386 #21239\n\n- ./.gitignore: drop obsolete entries.\n Follow-up to 4f38db1d28a971f938400f558e968fdffb9233a0 #1923\n\nCloses #22445", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "12532713d654d265bb7cb1c0c62e8a942d3a64d5", "date": "2026-07-25T16:58:54+02:00", "subject": "build: tidy up httpsig options", "body": "- say 'experimental'.\n- cmake: add to documentation.\n- cmake: alpha-sort.\n\nFollow-up to a55731050e8c3dbea0b96205cf916443118f6acb #22386\n\nCloses #22391", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "curl", "sha": "9bcc64c39b9f4b57361d3535829743d6999d4aad", "date": "2026-07-25T17:03:22+02:00", "subject": "curl: make --httpsig-key take a key OR a file name for key", "body": "Verified by test 5022\n\nCloses #22392", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "588a26ddce3c5a20f5b3f3d51ba1353cce7b4b0c", "date": "2026-07-26T00:16:09+09:00", "subject": "test: add error locations to `no-extra-label`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ce3f19e26218283eaff6436e28113b532bfc4a6f", "date": "2026-07-25T12:01:35-04:00", "subject": "Fix another empty nbtree index SSI race.", "body": "Commit f9b7fc65 fixed a race when predicate-locking completely empty\nbtrees: without a buffer lock held, a matching key could be inserted\nbetween _bt_search and the PredicateLockRelation call, so the scan would\nmiss concurrently inserted tuples while the writer wouldn't see the\nreader's predicate lock. That commit only fixed _bt_first's _bt_search\npath, though. Scans without useful insertion scan keys return early\nfrom _bt_first via _bt_endpoint, which still didn't recheck if the\nrelation was empty.\n\nTo fix, add handling to _bt_endpoint that is analogous to the handling\nadded to _bt_search by commit f9b7fc65.\n\nAuthor: Peter Geoghegan \nDiscussion: https://postgr.es/m/CAH2-WzkNoTn3yXY0iGkSuavJ+sL8EROf+kitW+_2v2tJVWuKmA@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "43a0258f55c13b58e95e3a2b75d19a65fef9f87d", "date": "2026-07-25T12:16:03-04:00", "subject": "build(vim-patch): detect N/A tags in vimhelp patches #40942", "body": "Vim patches for Vim help files must use Vim tags as diff hunk header\nto reliably detect if a hunk is N/A or not.\nGit does not know which files are Vim help files\nso that it can use custom \"diff..xfuncname\" regexp\nto create the header.\n\nUpdate Neovim's gitattributes to detect Vim help files\nand then use it on vim-patch.sh to scan Vim patches.\nAdd non-function tags from \":h vim_diff\"\nto detect N/A Vim runtime patches.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "ant-design", "sha": "bfbf564db96098a723b4ccbc7eb7bcb95ff0d382", "date": "2026-07-26T00:35:18+08:00", "subject": "feat(site): add deprecated Badge.Ribbon", "body": "* feat(site): add deprecated Badge.Ribbon\n\n* update", "style": "conventional", "type": "feat", "scope": "site", "breaking": false} {"repo": "neovim", "sha": "715d8887ec7dbcf1ac2d5b9453c726212cafa421", "date": "2026-07-25T12:47:51-04:00", "subject": "docs: misc #40847", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "35e31505d50671b8a783a48e1fc26e5375483e51", "date": "2026-07-25T19:01:52+02:00", "subject": "cms: correctly fail on invalid key length in CMS_decrypt", "body": "Report an invalid content encryption key length instead of decrypting\nwith a random key when this cannot act as an MMA (Bleichenbacher)\noracle, that is with RSA-OAEP or when PKCS#1 v1.5 implicit rejection\nis in effect. Otherwise, such as with third party providers without\nimplicit rejection, keep the random key masking.\n\nThe same applies to a failure of the key decryption itself. With\nimplicit rejection in effect the RSA decryption can only fail for a\npublicly invalid ciphertext, such as a value not smaller than the\nmodulus, and with RSA-OAEP a padding check failure is safe to reveal,\nso the error is reported instead of falling back to the random key\nmasking. This also makes CMS_decrypt with a wrong key deterministic\nin those configurations, which the new test relies on.\n\nFixes #25875\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Alicja Kario \nReviewed-by: Tomas Mraz \nMerge-date: Thu Sep 17 09:29:58 2026\nMerged-from: https://github.com/openssl/openssl/pull/28038", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "603a3335f2b60b2c798da5c627b3bb288b92a7bd", "date": "2026-07-25T19:16:42+02:00", "subject": "Add missing PGDLLIMPORT marker", "body": "Oversight in commit fb23cc7e81db.\n\nReported-by: Anton Voloshin \nDiscussion: https://postgr.es/m/ad5d772e-09d9-4248-97a4-0011afab9e71@postgrespro.ru", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "f5a968b0039ab751690c42a52e1a6a2c7b403760", "date": "2026-07-25T20:23:05+03:00", "subject": "fix(docs-infra): replace undefined --gray-50 in retro aria examples", "body": "The retro variants of the toolbar, select, and multiselect aria examples\ntint the pressed button background with\ncolor-mix(in srgb, var(--retro-button-color) 60%, var(--gray-50)), but\n--gray-50 is defined nowhere. Per the CSS spec, a var() with no fallback\npointing at an undefined property invalidates the whole color-mix(), so the\ndeclaration is dropped and the pressed-state tint never applies.\n\nDefine a local --retro-pressed-tint token in each example's :host (#fbfbfb,\nthe value $gray-50 resolved to) with a :host-context(.docs-dark-mode) override\nto #151417, and use it for the pressed-state mix. This keeps the example\nself-contained rather than depending on adev's global token scope, while\nstaying theme-aware in both light and dark, following the same token-flipping\npattern used in code-editor.component.scss.\n\n(cherry picked from commit 48c1e5bb0f6cd40d543700c19975e644ac885790)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "48c1e5bb0f6cd40d543700c19975e644ac885790", "date": "2026-07-25T20:23:05+03:00", "subject": "fix(docs-infra): replace undefined --gray-50 in retro aria examples", "body": "The retro variants of the toolbar, select, and multiselect aria examples\ntint the pressed button background with\ncolor-mix(in srgb, var(--retro-button-color) 60%, var(--gray-50)), but\n--gray-50 is defined nowhere. Per the CSS spec, a var() with no fallback\npointing at an undefined property invalidates the whole color-mix(), so the\ndeclaration is dropped and the pressed-state tint never applies.\n\nDefine a local --retro-pressed-tint token in each example's :host (#fbfbfb,\nthe value $gray-50 resolved to) with a :host-context(.docs-dark-mode) override\nto #151417, and use it for the pressed-state mix. This keeps the example\nself-contained rather than depending on adev's global token scope, while\nstaying theme-aware in both light and dark, following the same token-flipping\npattern used in code-editor.component.scss.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "neovim", "sha": "1facf1e0818a15588deea4809292cd32cf362bed", "date": "2026-07-25T10:25:56-07:00", "subject": "fix(gen): parameter order follows the prototype, not the docs", "body": "", "style": "conventional", "type": "fix", "scope": "gen", "breaking": false} {"repo": "neovim", "sha": "22d44ef8754448294833fcbc197defaca355f3b5", "date": "2026-07-25T10:26:01-07:00", "subject": "fix(api): add `lhs` option for keymap deletion", "body": "", "style": "conventional", "type": "fix", "scope": "api", "breaking": false} {"repo": "neovim", "sha": "6dc755ee09b8dd1d2f93db8d4cd9396d0e812350", "date": "2026-07-25T13:55:56-04:00", "subject": "build(vim-patch): skip doc for n/a features", "body": "Fix xfuncname to support tags that target Vim mappings, options.\n\n- events\n- options\n - *'luadll'*\n - *'mzschemedll'*\n - *'mzschemegcdll'*\n - *'perldll'*\n - *'pythondll'*\n - *'pythonthreedll'*\n - *'rubydll'*\n - *'tcldll'*\n - *'winptydll'*\n - t_.+, termcap opts\n- internal variables\n- cli flags\n- ':h emacs-tags'", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "postgres", "sha": "62c05d6f2fa64cce44e57871b4cfcd7b34589fcf", "date": "2026-07-25T14:15:33-04:00", "subject": "Add tests for nbtree empty index predicate locking.", "body": "Add coverage for predicate locking of completely empty nbtree indexes,\nwhere we must predicate lock the entire relation (instead of some\nindividual leaf page). Both paths that can find the index empty (and\nmust consider whether it's still empty after PredicateLockRelation\nreturns) are covered by a new isolation test that uses injection points.\n\nCatalog relation scans skip the injection points. The waiting session\nruns catalog queries of its own after arming the (session-local) points,\nand could otherwise suspend itself with nothing lined up to wake it.\n\nFollow-up to bugfix commits ce3f19e2 (the _bt_endpoint fix) and f9b7fc65\n(the _bt_first/_bt_search fix).\n\nAuthor: Peter Geoghegan \nDiscussion: https://postgr.es/m/CAH2-WzkNoTn3yXY0iGkSuavJ+sL8EROf+kitW+_2v2tJVWuKmA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "26fed075ae8a1a6cf70ba13354f7a860f1b84793", "date": "2026-07-26T04:50:31+09:00", "subject": "fix(cmdwin): q: in visual-mode does not insert visual range #40973", "body": "Problem: `q:` in visual mode does not insert `'<,'>` automatically in\nthe current line on cmdwin.\n\nSolution: Insert `'<,'>` in the current line on cmdwin when `q:` is used\nin visual mode.", "style": "conventional", "type": "fix", "scope": "cmdwin", "breaking": false} {"repo": "curl", "sha": "c39193a589164a5ef506385f352b9c904d7107f9", "date": "2026-07-25T13:06:38-07:00", "subject": "tests: improve exception handling in Python test code", "body": "* Explicitly set \"check\" in subprocess.run() to raise an exception\n automatically, where it was done manually before\n* Use contextlib.suppress to ignore exceptions\n* Use custom exceptions for test errors for clarity and flexibility.\n* Replace IOError with OSError\n\nThis fixes ruff rules BLE001, PLW1510, S110, TRY201, TRY203, TRY002,\nUP024.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "163f0cbc8bd1b0ecd2a885594bdb4e41cebb90b3", "date": "2026-07-25T22:22:21+02:00", "subject": "cd2nroff: fix backslashes for 4-space indent lines", "body": "They were previously only properly escaped for ~~~ quotes. Spotted for\nthe CURLOPT_HTTPSIG_KEY man page.\n\nCloses #22393", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "e13362c20accd210de60035825f7195f9679570b", "date": "2026-07-25T13:23:21-07:00", "subject": "tests: address mutable class vars and naive datetime in Python code", "body": "Mark Python mutable class variables with ClassVar, to denote that the\ndanger this can cause has been considered. Since any change made to\nthese in any object affects all other objects, this can cause locality\nerrors. However, as used in the test suite, they are are never modified\nand so they are annotated as being intended.\n\nAlways set a timezone in datetime objects, as mixing naive and\ntimezone-aware object can cause errors.\n\nThese fix ruff rules DTZ005, RUF012.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "3de2777421b25f8bc1dcd115143c3e1b5ae9f15a", "date": "2026-07-25T13:34:38-07:00", "subject": "tests: target Python 3.8 as the minimum Python version", "body": "This version is already two releases out of support, but is \"only\" 7\nyears old so is probably still being used in the real world. Document\nthis version along with some other testing dependencies. Remove code\nsupport for earlier versions. Disable ruff checks that need a newer\nversion.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "4eb691e89c3e324d10d112510bb76f352d655b64", "date": "2026-07-25T13:35:51-07:00", "subject": "tests: change whitespace and comments in Python test code", "body": "* remove an unneeded ruff warning disable\n* remove coding: utf-8 from Python code; PEP 3120 makes UTF-8 the\n default encoding\n* remove unusable shebang lines from Python code\n* remove empty print strings\n* disable warnings when file objects are stored; these instances can't\n be handled with context managers\n* use more consistent whitespace in Python code, fixing flake8 warnings\n* set the executable bit on scorecard.py, making it easier to run\n\nThese fix ruff rules EXE001, FURB105, UP009, SIM115.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "3901c16933c2ab327bf9030e0d49f1161d603fbb", "date": "2026-07-25T13:49:29-07:00", "subject": "tests: simplify by removing unneeded Python code", "body": "* combine separate if statements\n* remove an unneeded encode() call\n* remove unneeded returns\n* simplify code when returning early\n\nThis fixes ruff rules SIM102, SIM114, UP012, PLR1711.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b151a0bb901b8740c0025483c7fe375e65466dae", "date": "2026-07-25T14:05:01-07:00", "subject": "tests: use simpler constructions in Python code", "body": "* call super() without arguments\n* mark an unused variable as such\n* simplify by using dict getter for default values\n* use writelines() when possible\n* use dedent to simplify some text formatting\n* avoid items() on dict in a loop when unnecessary\n* replace most Python format() calls with f-strings\n* use capture_output in subprocess.run\n\nThis fixes ruff rules FLY002, FURB122, PERF102, RUF059, SIM401, UP008,\nUP022, UP030.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c5d683b961d49dbfa297c61c53b58466b960e202", "date": "2026-07-25T23:06:29+02:00", "subject": "tool_getparam: clear the --httpsig-key argument", "body": "To hide it somewhat from process listings.\n\nCloses #22394", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "e395fbd32a07557de4ac98088928c1749d4845d8", "date": "2026-07-25T17:13:00-04:00", "subject": "Add test coverage for nbtree backwards scans.", "body": "Backwards scans have unique concurrency rules: rather than unreservedly\ntrusting a saved left link, the scan optimistically rechecks its\npointed-to leaf page's right link (i.e. whether it still points back to\nthe page that _bt_readpage just read). Usually, the left sibling of the\njust-read page won't have changed, in which case the scan can proceed\nwith reading the left sibling as planned. But it's possible that the\nkey space that the scan needs to read next is no longer covered by the\noriginal left sibling page due to concurrent page splits and/or page\ndeletions. When that happens, the scan must recover by relocating the\nnew/current left sibling of the just-read page.\n\nTest coverage for backwards scans was limited to the happy path. Add an\nisolation test (and associated injection points) that test the recovery\npath. This covers several distinct recovery scenarios (concurrent page\nsplits, concurrent page deletions, and minor variants thereof).\n\nAuthor: Peter Geoghegan \nReviewed-by: Andrey Borodin \nDiscussion: https://postgr.es/m/CAH2-WzmD+jUBOpFS2jrnqqrdPSAjoxqyL9FPKaE1BtnY=8Nntg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "acf4498381cfd4dbaa2a2de6d868eb2f71bfb2b1", "date": "2026-07-25T23:48:07+02:00", "subject": "http: fix httpsig with auth-redir", "body": "Do not let unrelated credentials from a redirected URL bypass the\ncross-host auth boundary\n\nVerified by test 5023 to 5025\n\nFollow-up to a55731050e8c3dbea0\n\nCloses #22395", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "1594ca23682b38bec161685f0c3a6f6548c3c337", "date": "2026-07-25T17:59:35-05:00", "subject": "fix(dir): preserve environment variables in paths", "body": "", "style": "conventional", "type": "fix", "scope": "dir", "breaking": false} {"repo": "neovim", "sha": "ac4e5460b1a4b1bd9bbfe6d948f0c368593b74f9", "date": "2026-07-25T17:59:35-05:00", "subject": "feat(vim.fs): dir(opts.normalize)", "body": "", "style": "conventional", "type": "feat", "scope": "vim.fs", "breaking": false} {"repo": "curl", "sha": "5e75aedab2214c6114f1ad52b1bce6a6cd302480", "date": "2026-07-25T16:41:05-07:00", "subject": "tests: bump ruff version to 0.60.0", "body": "This version enables many more warnings by default.\n\nCloses #22396", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "92469017be4e900e5dafe29f6e25800681142f90", "date": "2026-07-26T08:01:16+08:00", "subject": "fix: align zero-runtime icon styles with runtime styles", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "neovim", "sha": "5ca90d1185592e3b02c0e0d06ea1bbcbc8c9f565", "date": "2026-07-26T03:52:38+03:00", "subject": "build(wasm): use pthreads and correct uv_fs_write offset handling", "body": "", "style": "conventional", "type": "build", "scope": "wasm", "breaking": false} {"repo": "neovim", "sha": "649bf7bcf2d919312a107b92342e27ed68538710", "date": "2026-07-25T22:19:20-04:00", "subject": "build(vim-patch): helptoc is 100% n/a", "body": "Use \"gO\" mapping instead.\nvim9script is N/A.\n\nDetect 'runtime/doc/helphelp.txt' N/A patches for helptoc.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "77f492f29f882e07d57e16920138d991dc9af018", "date": "2026-07-25T22:23:25-04:00", "subject": "oqs noted a double free recently:", "body": "https://github.com/open-quantum-safe/oqs-provider/pull/805#issuecomment-5034437500\n\nIts happening because, recently we made some fixes to account for\nproviders that request no-caching fetches, which all seems to work\nreasonably well.\n\nExcept for the corner case oqs recently exposed, in which the provider\nmay request the caching of some algorithms, but not others. When this\noccurs, the core libcrypto creates a temporary store for those\nrequested, but stored the cachable ones in the libctx store.\n\ninner_evp_generic_fetch treats the presence of an allocated tmp store as\nimplying that all algorithms fetched are non-cacheable, even if some\nare. The result is that, for those cacheable items, they also get\ntreated as non-cacheable, missing the extra ref count taken when they\nare added to the method store cache. As a result, they get freed while\nstill in use, and you get the issue referenced above.\n\nInstead of assuming that a tmp_store means all algorithms are\nnon-cacheable, instead, try to look the method up in the tmp_store\ncache, and cache it if its not found (i.e. its not in the temp store)\n\nFixes https://github.com/open-quantum-safe/oqs-provider/pull/805\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Andrew Dinh \nMergeDate: Sat Aug 1 14:30:49 2026\n(Merged from https://github.com/openssl/openssl/pull/32077)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "79a37f3d0b690a9e9cc893a10cef4ef4cdce4c3b", "date": "2026-07-26T09:38:57+06:00", "subject": "refactor(common): replaces the deprecated positional `subscribe` arguments with the recommended observer object", "body": "(cherry picked from commit ec87f0420000dd6f991b9190ad52754554403c92)", "style": "conventional", "type": "refactor", "scope": "common", "breaking": false} {"repo": "angular", "sha": "ec87f0420000dd6f991b9190ad52754554403c92", "date": "2026-07-26T09:38:57+06:00", "subject": "refactor(common): replaces the deprecated positional `subscribe` arguments with the recommended observer object", "body": "", "style": "conventional", "type": "refactor", "scope": "common", "breaking": false} {"repo": "angular", "sha": "832ff2aff3148b09126dd0c28672652d2b21899a", "date": "2026-07-26T10:40:31+05:30", "subject": "test(platform-server): add SSR integration test for null and normal input values", "body": "Fixes #69785\n\nAdd an Angular SSR integration test in platform-server verifying that null input values do not render string attributes like value=\"null\" during server-side rendering, while normal non-empty string values like value=\"hello\" are properly preserved.\n\n(cherry picked from commit fb698d12cdda1b626d727b35e57411b2a7b75336)", "style": "conventional", "type": "test", "scope": "platform-server", "breaking": false} {"repo": "angular", "sha": "fb698d12cdda1b626d727b35e57411b2a7b75336", "date": "2026-07-26T10:40:31+05:30", "subject": "test(platform-server): add SSR integration test for null and normal input values", "body": "Fixes #69785\n\nAdd an Angular SSR integration test in platform-server verifying that null input values do not render string attributes like value=\"null\" during server-side rendering, while normal non-empty string values like value=\"hello\" are properly preserved.", "style": "conventional", "type": "test", "scope": "platform-server", "breaking": false} {"repo": "tmux", "sha": "afb95f3134567888c03bc45136ac2d6d5dcfd9cc", "date": "2026-07-26T09:02:08Z", "subject": "Do not loop forever if the combining character is too long, GitHub issue 5434 from me at qdrs dot dev.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0768795ce31c064e06cd4dcd65730eaa71bfef25", "date": "2026-07-26T10:02:31+01:00", "subject": "Test for overlong runs of combining characters, GitHub issue 5434.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "55169f12756175d526d7254c53d1701db98e47df", "date": "2026-07-26T09:17:30Z", "subject": "Reset title count when freeing titles, GitHub issue 5429 from Brett Smith.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "3068ae06a88269072affc21e9c6e2cb62246905c", "date": "2026-07-26T09:20:54Z", "subject": "Also set ntitles to 0 when creating screen in the first place.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "10726b1b279da05adf5e5a8e38cc3eaade477850", "date": "2026-07-26T10:21:01+01:00", "subject": "Title stack test.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "3d5de7fd113df5aae395e7fcf23e5105b59e3caf", "date": "2026-07-26T04:36:03-05:00", "subject": "fix(buffer): name becomes empty when it is the cwd #40980", "body": "Problem: Shortening the name of a buffer whose full name is the current\ndirectory + path sep yields empty string -> buf unnamed.\nThis doesn't interact well with things like `:mksession`, which\nfor example could `:badd` with no arg -> E471.\n\nSolution: Keep the full path when shortening yields an empty name, as\n`path_try_shorten_fname()` already does.", "style": "conventional", "type": "fix", "scope": "buffer", "breaking": false} {"repo": "ant-design", "sha": "912a9927357bb9d5b47295fd8f9f408e5f79dd56", "date": "2026-07-26T17:54:53+08:00", "subject": "feat(site): add new component version Badge.Ribbon", "body": "* feat(site): add new component version Badge.Ribbon\n\n* update", "style": "conventional", "type": "feat", "scope": "site", "breaking": false} {"repo": "ant-design", "sha": "4f8c0cf48556fda94c06472c54deba8059835a45", "date": "2026-07-26T17:55:38+08:00", "subject": "type: export missing component props and ref types", "body": "* feat(website): migrate docs from dumi to Next.js\n\n* feat: update\n\n* type: export missing component props and ref types\n\n* chore: revert next.js", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "fcaf15a0a46afa09056b45a6748ba9d8474cbd80", "date": "2026-07-26T13:03:26+03:00", "subject": "fix(docs-infra): fix undefined CSS custom properties in shared-docs styles", "body": "Two custom properties in the shared-docs styles reference tokens that are\ndefined nowhere, so the declarations are invalid at computed-value time.\n\n_kbd.scss sets the text color with var(---tertiary-contrast); the extra\nleading dash points at an undefined property with no fallback, so the color\nresolves to the inherited value instead of the intended --tertiary-contrast.\n\n_colors.scss builds --light-pink-to-light-purple-horizontal-gradient from\nvar(--light-purple), which is not defined anywhere (the sibling token is\n--light-violet, defined in the same file), invalidating the gradient. The\ngradient is not currently referenced, so this corrects a latent malformed\ndeclaration rather than a visible bug.\n\nPoint both at the defined tokens: --tertiary-contrast and --light-violet.\n\n(cherry picked from commit e70994bdc3181e82d8cd5108d6eb7f5942377012)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "e70994bdc3181e82d8cd5108d6eb7f5942377012", "date": "2026-07-26T13:03:26+03:00", "subject": "fix(docs-infra): fix undefined CSS custom properties in shared-docs styles", "body": "Two custom properties in the shared-docs styles reference tokens that are\ndefined nowhere, so the declarations are invalid at computed-value time.\n\n_kbd.scss sets the text color with var(---tertiary-contrast); the extra\nleading dash points at an undefined property with no fallback, so the color\nresolves to the inherited value instead of the intended --tertiary-contrast.\n\n_colors.scss builds --light-pink-to-light-purple-horizontal-gradient from\nvar(--light-purple), which is not defined anywhere (the sibling token is\n--light-violet, defined in the same file), invalidating the gradient. The\ngradient is not currently referenced, so this corrects a latent malformed\ndeclaration rather than a visible bug.\n\nPoint both at the defined tokens: --tertiary-contrast and --light-violet.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "neovim", "sha": "dec3d8215dfa5d1ffd3a7d1d06559290db3cffcf", "date": "2026-07-26T13:24:39+03:00", "subject": "fix(text): indent() ignores `expandtab` when indent unchanged #40932", "body": "Problem:\nAfter expanding tabs with the `expandtab` option, if the old indent\nmatches the requested one, `vim.text.indent()` returns the input\nunchanged.\nThe optimization path assumes that if old_indent == size, the input\nwouldn't be changed, which is not correct when old_indent is formed by\nexpanded tabs.\n\nSolution:\nApply the optimization only when `expandtab` is not set.", "style": "conventional", "type": "fix", "scope": "text", "breaking": false} {"repo": "neovim", "sha": "4a5062cda69fa6cb59b7dc0c488dfb05accf99de", "date": "2026-07-26T07:18:57-04:00", "subject": "fix(path): DOS device path check #40976", "body": "Problem:\nThe current check is intended to match DOS device paths starting with\n`\\\\?` or `\\\\.` but will match any path starting with `?` or `.`,\nincluding relative paths such as `.\\`.\nThis is because the leading slashes are removed before the comparison.\n\nSolution:\nChanges the check to include the path's prefix, assuring the path has\nthe correct amount of separators for a DOS device path.", "style": "conventional", "type": "fix", "scope": "path", "breaking": false} {"repo": "eslint", "sha": "7c1e17543ad396f50f65b003cf9142d43d3c63b0", "date": "2026-07-26T22:40:14+09:00", "subject": "test: add error locations to `require-await`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "eslint", "sha": "688e75ede01b295d51f2f86c9d82da020a4c887e", "date": "2026-07-26T22:50:40+09:00", "subject": "chore: add missing backticks in JSDoc", "body": "* chore: add missing backticks in `lib/languages/js/index.js`\n\n* chore: add missing backticks in `lib/config/config.js`", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "eslint", "sha": "cc01b67061bbf558aa25746883ea1cad847f7cf2", "date": "2026-07-26T23:01:09+09:00", "subject": "test: add error locations to `no-useless-catch`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "eslint", "sha": "d08629382b0a6aaa042823b796e5100e60053b54", "date": "2026-07-26T23:21:31+09:00", "subject": "test: add error locations to `no-undef`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "curl", "sha": "e416e2948b5433251e8939f65f57734bfd4d07bf", "date": "2026-07-26T16:27:34+02:00", "subject": "vquic: initialize new callback slot for nghttp3 v1.18.0+", "body": "Closes #22399", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "e90c4397c8f522e8e395707f3aba0f201e41a1c7", "date": "2026-07-26T16:57:51+02:00", "subject": "vquic: silence `-Wmissing-field-initializers` for nghttp3/ngtcp2 callback tables", "body": "To avoid a breakage in CI and curl-for-win builds on upstream updates\nextending the callback lists. Each such breakage needed patching curl,\nrolling these patches into curl-for-win, and doing it in near real-time,\nto keep CI and builds working (and still causing some red CI jobs).\n\nBring calmness here by suppressing the warnings and allowing time to\nextend the callback tables as/if needed and at a convenient moment.\n\nCloses #22400", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "4cc45aa719a527457ec0d88dcbc7216ca3be2b7e", "date": "2026-07-26T15:08:15Z", "subject": "Do not try to write to clients which have died, GitHub issue 5431 from Stefan Ladstätter.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0255981a546ebb9de8dba2140c4fb0a3773fd811", "date": "2026-07-26T16:08:30+01:00", "subject": "Test run-shell if killed.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "d57d75deee894f2adb9b13fecadb3b2ccadf39bc", "date": "2026-07-26T15:21:53Z", "subject": "Include \\n and \\t in shell special characters, from Nikolas Skarlatos.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "098d28febfc50b529901ac3145dde96ea3b17f8f", "date": "2026-07-26T22:01:50+06:00", "subject": "docs: fix self-referential link in defer block description", "body": "Replace broken link with inline code\nformatting. The link was pointing to the page itself, as no dedicated block concept page exists at that URL.\ncorrects 'everytime' to 'every time'.\n\nUpdate tools/manual_api_docs/blocks/let.md\n\nCo-authored-by: Matthieu Riegler \n(cherry picked from commit 21eed5f8d4855e35f6de8cbfadc865c842d03b58)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "21eed5f8d4855e35f6de8cbfadc865c842d03b58", "date": "2026-07-26T22:01:50+06:00", "subject": "docs: fix self-referential link in defer block description", "body": "Replace broken link with inline code\nformatting. The link was pointing to the page itself, as no dedicated block concept page exists at that URL.\ncorrects 'everytime' to 'every time'.\n\nUpdate tools/manual_api_docs/blocks/let.md", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "1bba96f259b53ff09d40f18949737ba049366eaa", "date": "2026-07-26T11:26:50-05:00", "subject": "fix(test): group zip tests and drop repeated setup", "body": "", "style": "conventional", "type": "fix", "scope": "test", "breaking": false} {"repo": "neovim", "sha": "b6832a53b4555c385e0dd2e0d43083bd21ed918e", "date": "2026-07-26T19:27:30+03:00", "subject": "fix(help): g==: 'No code block found' #40990", "body": "Problem:\nThe code block query relied on a bug fixed in commit\n7ed5609439ba83d75e972beb859245d998df09ea, where a trailing newline was\nnot included in a node's text.\n\nSolution:\nFix the query to expect the newline.\n\n(cherry picked from commit fdc09be03cb10a1e11a31abd2ddd96d3f131b486)", "style": "conventional", "type": "fix", "scope": "help", "breaking": false} {"repo": "neovim", "sha": "fdc09be03cb10a1e11a31abd2ddd96d3f131b486", "date": "2026-07-26T19:27:30+03:00", "subject": "fix(help): g==: 'No code block found' #40990", "body": "Problem:\nThe code block query relied on a bug fixed in commit\n7ed5609439ba83d75e972beb859245d998df09ea, where a trailing newline was\nnot included in a node's text.\n\nSolution:\nFix the query to expect the newline.", "style": "conventional", "type": "fix", "scope": "help", "breaking": false} {"repo": "postgres", "sha": "e01accdb50ce8f879182edb1b50f3bc9bd78bfa7", "date": "2026-07-26T12:49:56-04:00", "subject": "Add _bt_set_startikey row compare test coverage.", "body": "Add pg_regress tests that exercise the row compare logic that commit\n7d9cd2df added to _bt_set_startikey. Also add tests that exercise the\n_bt_set_startikey SAOP array path.\n\nAuthor: Peter Geoghegan \nDiscussion: https://postgr.es/m/CAH2-Wz=KjQsD2W2a=b51uH905=0mF6Le4evhWkN2FL1+uRPhUg@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "deecb301c8aab4c22b80eaa505042311ee938b67", "date": "2026-07-26T21:03:37+03:00", "subject": "fix(core): warn when style property bindings receive invalid values", "body": "Report unsupported style property binding values in development mode while preserving existing binding behavior. Unwrap trusted style values before appending unit suffixes and link NG0318 warnings to the corresponding error guide.\n\n(cherry picked from commit e43eb96341b692525743a1f3aa1380ce21a9dd02)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "e43eb96341b692525743a1f3aa1380ce21a9dd02", "date": "2026-07-26T21:03:37+03:00", "subject": "fix(core): warn when style property bindings receive invalid values", "body": "Report unsupported style property binding values in development mode while preserving existing binding behavior. Unwrap trusted style values before appending unit suffixes and link NG0318 warnings to the corresponding error guide.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "curl", "sha": "6350eb01ed3f4a1ad2e1c6a1d4cdcd855eac3a10", "date": "2026-07-26T20:48:02+02:00", "subject": "cf-ngtcp2-cmn: de-duplicate `ngtcp2_conn_client_new()` call code", "body": "Closes #22401", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "f9dc57a149a8ba38557b5d0e1f7e31ce9b93c810", "date": "2026-07-26T21:15:33+02:00", "subject": "pytest: update two H3 tests for nghttp3 1.18.0+", "body": "Fixing:\n```\nFAILED ../../tests/http/test_02_download.py::TestDownload::test_02_36_looong_urls[65536-h3] -\n AssertionError: expected exit code 0, got 56\nFAILED ../../tests/http/test_14_auth.py::TestAuth::test_14_05_basic_large_pw[h3] -\n AssertionError: expected exit code 0, got 56\n```\nRef: https://github.com/curl/curl/actions/runs/30207198835/job/89807247058?pr=22400\n\nRefs:\nhttps://github.com/ngtcp2/nghttp3/pull/539\nhttps://github.com/ngtcp2/nghttp3/commit/a5872645446cdc07e897fea33b61a3c665563da9\n\nBug: https://github.com/curl/curl/pull/22397#issuecomment-5084927935\n\nCloses #22402", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "c57818bb43903aeb2cbdf2f3604d0c333d4e5641", "date": "2026-07-26T22:40:06+01:00", "subject": "fix(ui2): avoid flickering fileinfo messages #41002", "body": "## Flicker cause\n\nui2's `check_targets()` function creates the hidden ui2 windows with this call chain:\n```\nnvim_open_win()\nwin_set_buf()\ndo_buffer()\nset_curbuf()\nenter_buffer()\n```\n\nand `enter_buffer()` sets `need_fileinfo`:\n\nhttps://github.com/neovim/neovim/blob/fdc09be03cb10a1e11a31abd2ddd96d3f131b486/src/nvim/buffer.c#L1837-L1839\n\nThis is then picked up by `normal_redraw`:\n\nhttps://github.com/neovim/neovim/blob/fdc09be03cb10a1e11a31abd2ddd96d3f131b486/src/nvim/normal.c#L1381-L1385\n\nThe fileinfo message then causes a ui2 window to appear to display it:\n\nhttps://github.com/neovim/neovim/blob/fdc09be03cb10a1e11a31abd2ddd96d3f131b486/runtime/lua/vim/_core/ui2/messages.lua#L325-L330\n\n... which ends up calling `set_pos`, which registers an `on_key` handler:\n\nhttps://github.com/neovim/neovim/blob/fdc09be03cb10a1e11a31abd2ddd96d3f131b486/runtime/lua/vim/_core/ui2/messages.lua#L726-L728\n\nThis handler closes the ui2 window when the user presses, for example, `l` (because the user's not focusing said window).\nThe handler eventually calls `check_targets()`:\n\nhttps://github.com/neovim/neovim/blob/fdc09be03cb10a1e11a31abd2ddd96d3f131b486/runtime/lua/vim/_core/ui2/messages.lua#L594-L595\n\nand `check_targets()` takes us back to the beginning, causing the ui2 window flicker\n\n## Repro\n\n```vim\n=require(\"vim._core.ui2\").enable()\nset ch=0 shm-=F ve=all\ne /tmp/x\n```\n\nThen hit `Ctrl-G` to show fileinfo (if not already shown) and hold `l`\n\n## Fix\n\nEffectively surround the ui2 window creating with `:silent`", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "angular", "sha": "1167626e5b8b295b04197dc760039d1708c162ff", "date": "2026-07-27T00:01:54+02:00", "subject": "refactor(core): Migrate more tests off `fakeAsync`", "body": "This will prevent to polute the agent context with outdated/bad practices.\n\n(cherry picked from commit c1025a0510520fa4398705869bdc73613332b337)", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "c1025a0510520fa4398705869bdc73613332b337", "date": "2026-07-27T00:01:54+02:00", "subject": "refactor(core): Migrate more tests off `fakeAsync`", "body": "This will prevent to polute the agent context with outdated/bad practices.", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "curl", "sha": "16d49ac659659bb4fd0c4d7d0e6add87a5762ec6", "date": "2026-07-27T01:41:07+02:00", "subject": "GHA/windows: bump stunnel to 5.79", "body": "Closes #22403", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e0dd164fb45b9af025fff020650094f608dfa238", "date": "2026-07-27T02:42:41+03:00", "subject": "build(wasm): update wasm_stubs with browser libuv polling support", "body": "Replace the placeholder polling implementation with a browser-backed libuv polling backend and add the required platform compatibility stubs for the WebAssembly build.", "style": "conventional", "type": "build", "scope": "wasm", "breaking": false} {"repo": "curl", "sha": "7484874fdf799398523fc57022c3fd2cc85835f3", "date": "2026-07-27T02:42:09+02:00", "subject": "cmake: verify if options are listed in `INSTALL-CMAKE.md`", "body": "Also:\n- add one debug option to pass the test.\n\nRef: https://github.com/curl/curl/discussions/14885#discussioncomment-10632311\n\nCloses #22404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0962f9e344390c69e44bc55675510b2fa2b3f778", "date": "2026-07-27T09:42:50+09:00", "subject": "Use direct hash lookup in logicalrep_partmap_invalidate_cb()", "body": "This replaces an O(N) hash_seq_search() loop by an O(1) lookup, removing\na TODO item, making the invalidation callback faster when dealing with\nmany relations. This can work because LogicalRepPartMap is keyed by a\npartition OID, and a relmapentry's localreloid matches with it.\n\nAn assertion is added in logicalrep_partition_open() to enforce the fact\nthat localreloid matches with the hash key.\n\nAuthor: DaeMyung Kang \nDiscussion: https://postgr.es/m/20260417174450.4158878-1-charsyam@gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "87f08dbf3499929f4941f224f52ccd6a20081f00", "date": "2026-07-27T09:58:04+09:00", "subject": "Update .gitignore in test/modules/nbtree", "body": "Noticed while doing some routine work.\n\nOversight in e395fbd32a07.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "71a12e8a22032c6ac9d70e45b5e0fa43cef1ee29", "date": "2026-07-27T09:18:14+08:00", "subject": "type: export missing component ref types", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "aa50f4f02920101b84466668bd3b90c156dcda89", "date": "2026-07-27T10:21:18+09:00", "subject": "Fix deparsing of JSON_ARRAY(subquery) with a FORMAT clause", "body": "Commit 8d829f5a0 introduced the JSCTOR_JSON_ARRAY_QUERY constructor\ntype so that ruleutils.c could deparse JSON_ARRAY(subquery) using its\noriginal syntax, storing the transformed subquery in a new orig_query\nfield. However, the input FORMAT clause of JSON_ARRAY(subquery FORMAT\n...) was not preserved for deparsing. The format was recorded only in\nthe executable expression kept in the func field, which ruleutils.c\ndoes not inspect, so it is silently dropped.\n\nThis is more than cosmetic, because FORMAT JSON changes the result:\nwithout it a text value is treated as a string to be quoted, while\nwith it the value is treated as already-formatted JSON.\n\nTo fix, record the input FORMAT in a new deparse-only field of\nJsonConstructorExpr, alongside orig_query, and emit it in ruleutils.c.\n\nBump catalog version.\n\nAuthor: Chao Li \nReviewed-by: Ewan Young \nReviewed-by: Richard Guo \nDiscussion: https://postgr.es/m/4C89B193-7D54-4705-9CF9-F0D484B9E099@gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b8d9cf512c1259f97f9896593cc1c8352c1118ac", "date": "2026-07-27T09:06:05+05:30", "subject": "Fix issues in logical replication sequence synchronization.", "body": "1. Stop a running sequence synchronization worker when\nALTER SUBSCRIPTION ... DISABLE is executed. The worker did not reread its\nsubscription after starting a transaction, so it kept running with a stale\ncopy and missed the disable. It now calls maybe_reread_subscription()\nafter StartTransactionCommand(), matching the apply worker.\n\n2. Restore the invariant that publisher-side synchronization slots are\ndropped last during ALTER SUBSCRIPTION ... REFRESH PUBLICATION. The\nslot-drop loop now runs after the sequence-removal loop, so the\nnon-transactional slot drops happen only after all catalog changes that\ncould still be rolled back on error.\n\n3. Restore psql tab completion for\nALTER SUBSCRIPTION ... REFRESH PUBLICATION WITH (.\n\n4. Make pg_stat_subscription report NULL for the fields that do not apply\nto a sequence synchronization worker, which does not stream from a\nwalsender, and update the documentation accordingly.\n\n5. Update the pg_subscription_rel.srsublsn catalog documentation to\ndescribe its semantics for sequence rows.\n\nReported-by: Noah Misch \nAuthor: vignesh C \nReviewed-by: Hayato Kuroda \nReviewed-by: Amit Kapila \nBackpatch-through: 19, where it was introduced\nDiscussion: https://postgr.es/m/20260710045217.f0.noahmisch@microsoft.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "931a6d2c14d4ed89aeaac569afe148dc93c95b87", "date": "2026-07-27T01:56:38-04:00", "subject": "build(vim-patch): \":h builtin-function-list\" is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "23046608037c73f50a583efdb5f39c60d2c87bd5", "date": "2026-07-27T08:32:48+02:00", "subject": "OSSL_FN: Add OSSL_FN_mod_exp_mont()", "body": "Route odd moduli in OSSL_FN_mod_exp() to a Montgomery sliding-window\npath, a faithful port of BN_mod_exp_mont() in crypto/bn/bn_exp.c. The\nfixed-width deviations, the constant-time caveat, the dispatcher\nrouting, and the OSSL_FN_mod_exp_ctx_size() sizing model are all\ndocumented in the respective function headers.\n\ntest_fn and test_fn_api pass on linux-x86_64 and linux-x86; the odd-\nmodulus cases (the stanza suite and the KAT/alias tests) now exercise\nthe Montgomery path against the BN_mod_exp oracle.\n\nAssisted-by: Pi:z-ai/glm-5.2\nAssisted-by: Pi:moonshotai/kimi-k3\nRelated-to: doc/designs/fixed-size-large-numbers.md\nIssue: https://github.com/openssl/project/issues/2025\nSigned-off-by: Richard Levitte \n\nReviewed-by: Jakub Zelenka \nReviewed-by: Igor Ustinov \nReviewed-by: Matt Caswell \nMergeDate: Fri Jul 31 09:38:11 2026\n(Merged from https://github.com/openssl/openssl/pull/31962)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "d59e443ddd49c3a7c3a551ab072d44b92516cc95", "date": "2026-07-27T15:27:27+08:00", "subject": "fix(DatePicker): prevent unconfirmed range submission", "body": "* fix(DatePicker): prevent unconfirmed range submission\n\n* test(DatePicker): sync picker snapshots", "style": "conventional", "type": "fix", "scope": "DatePicker", "breaking": false} {"repo": "ant-design", "sha": "ee5cd3acd682f3db047d1381773d24502f7e67d4", "date": "2026-07-27T15:30:37+08:00", "subject": "fix(Spin): prevent nested section style leakage", "body": "", "style": "conventional", "type": "fix", "scope": "Spin", "breaking": false} {"repo": "vitest", "sha": "fec001ad3b3a7a9a589b97f46d28c5c201c8f98e", "date": "2026-07-27T09:31:56+02:00", "subject": "feat!: inline projects extend the root config by default", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": true} {"repo": "eslint", "sha": "f9f61dc6112836f46a58a410c38b8c2faf10973f", "date": "2026-07-27T16:42:57+09:00", "subject": "test: add error locations to `no-unreachable`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "curl", "sha": "c9ead9bd1cdc7fa03d5e9eb20ebed37e27110407", "date": "2026-07-27T09:49:33+02:00", "subject": "conncache: conn upkeep/alive: move and enhance", "body": "- move `Curl_conn_seems_dead()` into conncache.c\n- move `Curl_conn_upkeep()` into conncache.c\n- when upkeep gives an error on a connection not in use,\n terminate it\n\nCloses #21806", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "08a70f0a9e3ef10dac3dd3ea3458e2662c81b8f0", "date": "2026-07-27T15:50:48+08:00", "subject": "test: fix flaky playground tests on CI", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "tmux", "sha": "5269e1411363fc91a5efe628e88b41eb11e24872", "date": "2026-07-27T08:03:01Z", "subject": "Add after-swap-window hook from basicalllymiloud at gmail dot com.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "faaa4d57f16d96f529fc08ef5752f4dbcb5665cb", "date": "2026-07-27T03:26:11-05:00", "subject": "test: start plugin specs with `--clean` #40995", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "curl", "sha": "a1ee986c4d775c388edd1464d1d2e84228491d4c", "date": "2026-07-27T10:43:18+02:00", "subject": "more CURLoption", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "7904ec49e5f0cced2c5d90e41720fd08ecc53fa1", "date": "2026-07-27T10:45:48+02:00", "subject": "minor fix", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "633dd4d001c269e7db63c9ec0c1efdfe186b6329", "date": "2026-07-27T10:47:03+02:00", "subject": "fixup return", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "70140f41f379b633e15170b0f37cbf64f36715e5", "date": "2026-07-27T10:47:37+02:00", "subject": "fixup error message", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "05a9370a324efdc9d2fdd21d5d2fa8d5b051ad9c", "date": "2026-07-27T10:49:56+02:00", "subject": "fixup error check", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a1fb6062bda666275ffbc8421c9aa150e8a2195b", "date": "2026-07-27T10:52:57+02:00", "subject": "fixup test mistake", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "359459dec6e1929c0afb56a3111dee7b5a924eea", "date": "2026-07-27T06:25:21-04:00", "subject": "refactor(exmode): Ex-mode as cmdwin + Lua #40991", "body": "Problem:\nPOSIX-compatible Ex-mode requires special-cases all over the codebase to\nmatch various quirks that don't actually matter to users.\n- The main utility of *interactive* Ex-mode is its REPL behavior, and\n that can be achieved with `cmdwin`, which also gains extra UX\n benefits.\n- The main utility of *non-interactive* `nvim -es` is for shell\n scripting, where Ex-mode quirks are mostly unhelpful (e.g. the\n \"Entering Ex mode\" message).\n\nSolution:\n- Reimplement *interactive* Ex-mode as a \"persistent, insert-mode\n cmdwin\" in Lua.\n - \"nvim -e/-E\" is simply an alias to \"gQ\".\n- Reframe *non-interactive* Ex-mode (`nvim -es`) as \"script mode\".\n - Drop POSIX Ex-mode quirks.\n\nImprovements:\n- \"nvim -V1 -es\" output ends with a final newline!\n- \"nvim -V1 -es\" no longer shows the \"Entering Ex mode\" msg. (This was\n pointless noise, unwanted for scripting purposes.)\n- stdin is no longer typeahead. Scripts (\":lua io.read()\") can read\n stdin as data.\n- Empty line is a no-op: a stray blank line no longer moves the cursor\n (deviates from POSIX ex \"+1\"), no longer exits 1 at EOF (E501).\n\nPreserved behavior:\n- cursor starts at \"$\"\n- mode()==\"cv\" (for non-interactive)\n- multiline commands (:append/:function/heredoc pull continuation lines)\n- bare-range print\n- :print=>stdout\n- -V1=>stderr\n- CRLF input\n- continue-after-error and exit codes\n\nDropped (regressed) POSIX behavior (non-interactive):\n- Event loop only ticks while/between commands, not while blocked\n waiting for a stdin line.\n- \":g/pat/visual...Q\"\n- input()/getchar()/\":s/x/y/c\" no longer consume stdin lines as\n answers: Nvim stops at end-of-input, skipping the rest of the script,\n exit 0. Use \":lua io.read()\" instead.\n - If users care about this they should use interactive Ex-mode (`gQ`).\n- \":@r\" stops at end of the register instead of continuing to read\n cmdline input from stdin.", "style": "conventional", "type": "refactor", "scope": "exmode", "breaking": false} {"repo": "vite", "sha": "599c5b02a8b6879b05ede988020f1331e877aaea", "date": "2026-07-27T12:42:13+02:00", "subject": "fix(module-runner): keep stack trace interception working when `Object.prototype` is frozen", "body": "", "style": "conventional", "type": "fix", "scope": "module-runner", "breaking": false} {"repo": "vite", "sha": "6b3210c488fbdee2ed8ee19b7c53c8ca07b4cb68", "date": "2026-07-27T18:44:00+08:00", "subject": "docs(changes): update outdated hotUpdate hook link", "body": "", "style": "conventional", "type": "docs", "scope": "changes", "breaking": false} {"repo": "vite", "sha": "0945bab7307895a1ddca713636585cf73fb26945", "date": "2026-07-27T18:53:50+08:00", "subject": "docs(changes): fix missing commas and language mark in plugin examples", "body": "", "style": "conventional", "type": "docs", "scope": "changes", "breaking": false} {"repo": "vite", "sha": "aafa103af5d71fb59d7c3dd617d0cbef3b222f1f", "date": "2026-07-27T11:53:57+01:00", "subject": "docs(config): correct cacheDir default fallback description", "body": "", "style": "conventional", "type": "docs", "scope": "config", "breaking": false} {"repo": "openssl", "sha": "afa270aa893f2e76d2db97f9b67f068ee28dacf8", "date": "2026-07-27T12:54:03+02:00", "subject": "test: fix flaky -md and -iter mismatch checks in 20-test_enc.t", "body": "The subtests added in f26d632a0e assert that \"openssl enc -d\" exits\nnon-zero when the key derivation parameters do not match. enc has no\nMAC, so a wrong key is not detectable as such; the only signal is\nwhether the garbage happens to form valid PKCS#7 padding, which\nopenssl-enc(1) documents as passing better than one time in 256. With\na random salt the tests thus failed about 0.4% of the time.\n\nAssert instead that mismatching parameters do not recover the\nplaintext: the decryption must either fail or produce differing output.\n\nAssisted-by: Claude:claude-opus-5[1m]\n\nSigned-off-by: Norbert Pocs \n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Saša Nedvědický \nReviewed-by: Neil Horman \nMergeDate: Mon Jul 27 12:17:21 2026\n(Merged from https://github.com/openssl/openssl/pull/32083)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vite", "sha": "a4785c39857640e2035e9c61408108fe5a690a6e", "date": "2026-07-27T16:29:16+05:30", "subject": "docs(create-vite): list react-compiler templates", "body": "", "style": "conventional", "type": "docs", "scope": "create-vite", "breaking": false} {"repo": "neovim", "sha": "69e2fc26f0da27f384c66c738110952324c20e67", "date": "2026-07-27T07:01:54-04:00", "subject": "fix(lua): vim.regex():match_str abort in Luv callback #41008", "body": "(cherry picked from commit 80cd48257721cb4261495131ea806020a309b9c4)", "style": "conventional", "type": "fix", "scope": "lua", "breaking": false} {"repo": "neovim", "sha": "80cd48257721cb4261495131ea806020a309b9c4", "date": "2026-07-27T07:01:54-04:00", "subject": "fix(lua): vim.regex():match_str abort in Luv callback #41008", "body": "", "style": "conventional", "type": "fix", "scope": "lua", "breaking": false} {"repo": "curl", "sha": "c5ad90bf9a8471987ebb47a05881a51108bcf87f", "date": "2026-07-27T13:17:56+02:00", "subject": "typecheck-gcc: allow passing `char[]` as callback data", "body": "E.g. `TEST_DATA_STRING` in test 655.\n\nCherry-picked from #22406\n\nCloses #22409", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "eddbad60158d8cda86830fbe934a9d2282273b32", "date": "2026-07-27T13:49:56+02:00", "subject": "test: fix failing ecosystem test for `eslint-plugin-unicorn`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "angular", "sha": "1a2bcb2295c8b4e3a398c8ecbd92cce0affaff1d", "date": "2026-07-27T15:25:52+03:00", "subject": "fix(devtools): avoid initializing profiler in production mode", "body": "Defer the timing hook subscription until the timing API is explicitly enabled. Keep production module loading and development initialization covered by isolated browser test targets.", "style": "conventional", "type": "fix", "scope": "devtools", "breaking": false} {"repo": "postgres", "sha": "31974f64f029a29a6d7eeeb3f870030beee2066f", "date": "2026-07-27T15:29:51+03:00", "subject": "pg_resetwal: do not allow zero next multixact offset", "body": "Offset 0 is the \"invalid\" marker in pg_multixact/offsets since offsets\nwent 64-bit and the allocator stopped skipping it. pg_resetwal could\nstill produce it via -O 0 or guessed control values, breaking the first\nmultixact created after the reset (\"MultiXact n has invalid offset\",\nand vacuum of the affected table fails from then on). Reject -O 0 like\n-m and -o already do, and guess 1 like initdb does.\n\nAuthor: Zsolt Parragi \nDiscussion: https://www.postgresql.org/message-id/CAN4CZFNoO6MUkg526TmA=mC_RjY2gp4VKCnvK6y12v3ppOkhJA@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "0b3b12da8d076c9f00511e75e5276816d0b68ebc", "date": "2026-07-27T20:49:12+08:00", "subject": "fix(completion): \"preselect\" ignores \"noinsert\", menu order #41007", "body": "Problem: A preselected item is inserted even with \"noinsert\", and the\n first added preselected item wins over the first one shown\n in the menu.\nSolution: Use K_DOWN when \"noinsert\" is set; pick the first preselected\n item that made it into the menu.", "style": "conventional", "type": "fix", "scope": "completion", "breaking": false} {"repo": "neovim", "sha": "97a71ab4846c4b92a1e658ec00ff3eae74bf1256", "date": "2026-07-27T14:20:46+01:00", "subject": "fix(man): use direct lookup without manpath #40421", "body": "Problem:\nOn NetBSD, `man -w open` can return the exact manpage path, but `:Man`\nmay still fail when man directories cannot be discovered from `manpath\n-q`, bare `man -w`, or `$MANPATH`.\n\nSolution:\nFall back to the direct manpage lookup when directory discovery fails.\nAdd a test for resolving `open(2)` through `goto_tag()` without manpath\ndata.", "style": "conventional", "type": "fix", "scope": "man", "breaking": false} {"repo": "postgres", "sha": "2a9541ddfd3a5b069c505e36a91f289126871a21", "date": "2026-07-27T09:37:04-04:00", "subject": "Deparse FOR PORTION OF using the range column's current name.", "body": "Commit 8e72d914c recorded the range column's name in ForPortionOfExpr\nand used that for deparsing FOR PORTION OF. This gives the wrong\nanswer if the ForPortionOfExpr is saved in a rule or SQL function and\nthen the column gets renamed. Drop the ForPortionOfExpr.range_name\nfield; instead fetch the current column name from the catalogs when\nneeded.\n\nAlso drop ForPortionOfState.fp_rangeName, which wasn't being used\nanywhere.\n\nFull disclosure: an earlier draft of this patch was made with\nClaude Opus 4.8.\n\nReported-by: John Naylor \nAuthor: Tom Lane \nReviewed-by: Richard Guo \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/CANWCAZYFEpJ5Oi45gi4q9Y6LYa4_oiAXxuNNWe-1ym-i0fF8Pw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c1b0e770df131a6e294228f88a5bdaa384ea0089", "date": "2026-07-27T22:39:53+09:00", "subject": "quic: fix BIO ownership in test helper", "body": "SSL_set_bio() transfers ownership of cbio to the client SSL object.\nClear the local pointer after the handoff so the error path does not\nrelease an SSL-owned reference before SSL_free().\n\nFixes #32084\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Daniel Kubec \nMergeDate: Wed Jul 29 09:56:04 2026\n(Merged from https://github.com/openssl/openssl/pull/32086)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "460cdd64b344bab0417f9a61beea0f29a029b315", "date": "2026-07-27T22:39:53+09:00", "subject": "quic: fix BIO ownership in test helper", "body": "SSL_set_bio() transfers ownership of cbio to the client SSL object.\nClear the local pointer after the handoff so the error path does not\nrelease an SSL-owned reference before SSL_free().\n\nFixes #32084\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Daniel Kubec \nMergeDate: Wed Jul 29 09:55:49 2026\n(Merged from https://github.com/openssl/openssl/pull/32086)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "20fbf816ab4308fa101e70621bd64c2bcf466f3d", "date": "2026-07-27T22:39:53+09:00", "subject": "quic: fix BIO ownership in test helper", "body": "SSL_set_bio() transfers ownership of cbio to the client SSL object.\nClear the local pointer after the handoff so the error path does not\nrelease an SSL-owned reference before SSL_free().\n\nFixes #32084\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Daniel Kubec \nMergeDate: Wed Jul 29 09:55:36 2026\n(Merged from https://github.com/openssl/openssl/pull/32086)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4ff279eb4992f2971735d437de8826e0d98d6682", "date": "2026-07-27T22:39:53+09:00", "subject": "quic: fix BIO ownership in test helper", "body": "SSL_set_bio() transfers ownership of cbio to the client SSL object.\nClear the local pointer after the handoff so the error path does not\nrelease an SSL-owned reference before SSL_free().\n\nFixes #32084\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Daniel Kubec \nMergeDate: Wed Jul 29 09:55:25 2026\n(Merged from https://github.com/openssl/openssl/pull/32086)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e4dcfebcead9fdab6a220bce12025cc8ca87b265", "date": "2026-07-27T22:39:53+09:00", "subject": "quic: fix BIO ownership in test helper", "body": "SSL_set_bio() transfers ownership of cbio to the client SSL object.\nClear the local pointer after the handoff so the error path does not\nrelease an SSL-owned reference before SSL_free().\n\nFixes #32084\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Daniel Kubec \nMergeDate: Wed Jul 29 09:55:13 2026\n(Merged from https://github.com/openssl/openssl/pull/32086)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "573a6ec16b5822ee209d46c490f7e4200036c7dd", "date": "2026-07-27T16:05:38+02:00", "subject": "urlapi: improved return codes", "body": "- add CURLUE_BACKSLASH that can be returned when a backslash was used\n where a forward one probably was intended.\n\n- make CURLUE_NO_HOST higher priority than port number errors for URLs\n without hostname. Like in \"http://::1\"\n\n- shortened some URL parser error strings\n\nExtend test 1560 to verify.\n\nReported-by: kit-ty-kate on github\nFixes #22337\nCloses #22408", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "f751d3f2c16493cef5e9a6b27fdce7ea0e713f34", "date": "2026-07-27T14:25:46Z", "subject": "When destroying pane, also close pipe and destroy its event. GitHub issue 5424.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "9b97dadc0160313bfc961f17c63a0a19f1d89509", "date": "2026-07-27T15:26:26+01:00", "subject": "Test for pipe-pane.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "fbcb7a056cef7a5efd7aa7f494365a04b1c86895", "date": "2026-07-27T11:12:47-04:00", "subject": "feat(exmode): \"1q:\", :exmode #41010", "body": "Problem:\nWant `gQ` for _le multicursor_.\n\nSolution:\n- Don't use `gQ` for exmode.\n- Introduce `:exmode`.\n- Introduce `[count]q:` as an alias to `:exmode`.", "style": "conventional", "type": "feat", "scope": "exmode", "breaking": false} {"repo": "axios", "sha": "800504c6283115e38f3e2d2a56dc03e71e9bda9a", "date": "2026-07-27T17:29:32+02:00", "subject": "docs: apply v1.19.0 release updates", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "99e47536bbf1a165f5dc8d504f928821ebc8df6a", "date": "2026-07-27T09:10:48-07:00", "subject": "Fix race condition when enabling logical decoding concurrently.", "body": "With wal_level = 'replica', logical decoding is enabled on demand\nwhen the first logical replication slot is created:\n\nWhen enabling logical decoding, EnableLogicalDecoding() flips the\nshared logical_decoding_enabled flag and writes an\nXLOG_LOGICAL_DECODING_STATUS_CHANGE record so that standbys follow the\nstatus change. The initial \"already enabled?\" check and the WAL record\nwrite happen under two separate acquisitions of\nLogicalDecodingControlLock, since the lock must be released while\nwaiting for the ProcSignalBarrier: processes absorbing the barrier\nacquire the same lock in shared mode.\n\nConsequently, if two backends concurrently created the first logical\nslots, both could pass the initial check and both write a\nstatus-change record. The redundant record lands after the decoding\nstart point already reserved by the other backend's slot, so decoding\nthat slot processes the record and fails with \"unexpected logical\ndecoding status change\", as xlog_decode() assumes that no such record\ncan appear within the WAL range any slot decodes.\n\nFix by re-checking the status after re-acquiring the lock, so that\nonly the backend that actually performs the disabled->enabled\ntransition writes the WAL record.\n\nReported-by: Srinath Reddy Sadipiralla \nAuthor: Srinath Reddy Sadipiralla \nReviewed-by: Masahiko Sawada \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/CAFC+b6oYzmAgp7F0ivrhfZT46-CjvCTrU9pWuMNcem-52YjOTw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c36eb5661bc8e5db056cb7def445153a83409631", "date": "2026-07-27T20:22:45+02:00", "subject": "add d2i regression test to asn1 decode test suite", "body": "Reviewed-by: Igor Ustinov \nReviewed-by: Dmitry Belyavskiy \nMergeDate: Fri Sep 4 09:14:13 2026\n(Merged from https://github.com/openssl/openssl/pull/32092)\n\n(cherry picked from commit 573d671485940b68802325f6641005c67aa7abdb)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "573d671485940b68802325f6641005c67aa7abdb", "date": "2026-07-27T20:22:45+02:00", "subject": "add d2i regression test to asn1 decode test suite", "body": "Reviewed-by: Igor Ustinov \nReviewed-by: Dmitry Belyavskiy \nMergeDate: Fri Sep 4 09:14:13 2026\n(Merged from https://github.com/openssl/openssl/pull/32092)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "27ab96b284e71ed82c1a8930dd900260fd1583f0", "date": "2026-07-27T19:15:58Z", "subject": "Store count of how many times a hook is fired and the last time.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5a3b22eb304806d5e492e6b62a34e77d6e060573", "date": "2026-07-27T16:19:24-04:00", "subject": "Further improve the names generated for indexes on expressions.", "body": "Commit 181b6185c failed to do anything useful with a whole-row Var,\ndeeming it \"fishy\". But it is legal to put such a Var into an\nexpression index column, so let's expand it as the name of the table.\n\nAnother problem reachable via that one is that we could generate an\nempty index column name, which isn't really legal although by chance\nnothing complained about it. It's not clear whether any other such\ncases remain, but as cheap insurance let's use \"expr\" if the tree walk\nfails to generate any text.\n\nReported-by: Chauhan Dhruv \nAuthor: Chauhan Dhruv \nCo-authored-by: Tom Lane \nDiscussion: https://postgr.es/m/CANWwWcp_DCJjq8pomeqp6W=fbygvzXXQO028VDJ9_6sLPjQnVA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "677214e7eea83d8bc6e4b79eea871577e1369d5f", "date": "2026-07-28T02:04:38+05:30", "subject": "fix: handle ASI hazards in no-unused-vars removeVar suggestion", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "27a4557c9ea87f1961741fb728387261387b398c", "date": "2026-07-27T23:56:44+02:00", "subject": "EXPERIMENTAL.md: We do not accept vuln reports for experimental features", "body": "Closes #22411", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e3c5974adf25002ea490719a3391d790bd3a3353", "date": "2026-07-27T17:30:54-05:00", "subject": "feat(dir): open cwd with `1-` #40948", "body": "", "style": "conventional", "type": "feat", "scope": "dir", "breaking": false} {"repo": "postgres", "sha": "55adef7abbe8ecfa329335640b598cbb57be0f93", "date": "2026-07-28T08:35:21+09:00", "subject": "Fix propagation of indimmediate flag in index_create_copy()", "body": "index_create_copy is used to create copy definitions of existing indexes.\nCurrently, it passes 0 as constr_flags to index_create(), which results\nin the copied index to always be created as immediate (indimmediate set\nto true). For deferrable unique constraints, it means that the\ntransient index used during the phase 2 of REINDEX CONCURRENTLY forces\nimmediate constraint checks on concurrent inserts, which can cause\nunexpected constraint violations based on the definition of the parent\ntable, inconsistently set in the copied index.\n\nTo fix this without violating the contract of constr_flags (which should\nonly be used when creating constraints) and without relaxing the strict\nassertion in index_create(), this introduces a new index creation flag:\nINDEX_CREATE_DEFERRABLE. If set, a copied index's indimmediate is set\nto false, meaning that unique constraints are not enforced immediately\non insertion, but at transaction commit time.\n\nAn isolation test for REINDEX CONCURRENTLY is added, based on an\ninjection point waiting after phase 1 of the operation, where an index\ncopy has been built and is able to accept DMLs for its validation in\nphase 2. The test is tentatively backpatched down to v17.\nINJECTION_POINT() is outside a transaction context, which should be fine\non HEAD since 8daeaa9b642c but I suspect may cause issues in v19 and\nolder branches due to the wait facility depending on condition variables\nand a DSM setup, but let's see what the buildfarm tells.\n\nAuthor: Nitin Motiani \nDiscussion: https://postgr.es/m/CAH5HC97JmjPpgiQOqW9xm8qXhNiu7zZ1Qh+FfhEESJuDv69kuQ@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "34e4dc166652356772b414c7d263cfca9bef32cf", "date": "2026-07-28T08:59:02+09:00", "subject": "fix(common): exclude pipe options from validator options", "body": "", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "vue-core", "sha": "bf5978815d541e976d0705b7f4864e0ad518fa70", "date": "2026-07-28T09:19:09+08:00", "subject": "fix(runtime-vapor): support vdom slot content in Transition", "body": "Resolve VDOM slot output using the same child selection and transition hook\npreparation as BaseTransition. Prepare both the incoming and previously\nrendered branches so enter and leave transitions work correctly, including\nkeyed fragment identity.", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "123ea94f76b8efc9f45286e2ba55ee02bff5056c", "date": "2026-07-28T10:50:23+09:00", "subject": "Fix portability issue in authentication test 003_peer", "body": "The mapped user name is built upon the OS user name of the environment\nwhere the test is run. Depending on the characters used in the OS user\nname, CREATE ROLE may not get parsed (the author has mentioned hyphens\nas one case), causing a failure of the test.\n\nLet's use double-quotes around the mapped user name, which should be a\nsolution good enough for the environments where this test tends to run.\nThe buildfarm issued no complaint over the years.\n\nOversight in 3c4e26a62c31, so backpatch down to v19. Perhaps\n3c4e26a62c31 and this commit should be backpatched further down, but\nlet's leave that for another day, if it proves necessary.\n\nAuthor: Yugo Nagata \nDiscussion: https://postgr.es/m/20260727133857.fbd23d43d422f10f376a8bee@sraoss.co.jp\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c6f264b8c4b174a5cb90e194dd653941b903b22e", "date": "2026-07-28T11:30:08+09:00", "subject": "test: add precise Valgrind provider init suppression", "body": "The internal provider test can retain provider-init allocations\nwhen the Valgrind suppression workflow intentionally skips OPENSSL_cleanup().\nIts stack calls OSSL_provider_init directly from provider_activate, without\nthe provider_init frame covered by the existing suppression.\n\nAdd a second exact suppression for that stack shape instead of broadening the\nexisting block with an ellipsis.\n\nReviewed-by: Paul Dale \nReviewed-by: Norbert Pocs \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMergeDate: Wed Jul 29 16:58:17 2026\n(Merged from https://github.com/openssl/openssl/pull/32094)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "8a8d972fa734a741df5c42619e4e677ae6a2cacc", "date": "2026-07-28T10:36:23+08:00", "subject": "fix(runtime-vapor): handle custom directives on fragment roots", "body": "Close #15157", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "9887ccf7b3546cab2c4ae825d3f19043e5ae813a", "date": "2026-07-28T11:35:56+08:00", "subject": "fix: hide React demo docs from routing", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "51fb00d8531477bd6a63d9009f1e5309f6fd9d03", "date": "2026-07-28T11:44:05+08:00", "subject": "feat(Input): support global variant config for subcomponents", "body": "", "style": "conventional", "type": "feat", "scope": "Input", "breaking": false} {"repo": "vue-core", "sha": "4616c021f60db9c8ec9f72a59cb9804458cf6ab7", "date": "2026-07-28T11:56:48+08:00", "subject": "fix(runtime-vapor): handle custom directives on async components", "body": "Treat unresolved async wrappers as pending instead of warning about a\nnon-element root. Track fragment updates so directives are applied once\na renderable root becomes available.\n\nFor async setup under Suspense, retry from beforeMount after the resolved\nblock is created. This preserves lifecycle error handling and applies the\ndirective before the resolved child mounts.", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vite", "sha": "51207ba63cc09547aecafaba8d1354cdd8a54871", "date": "2026-07-28T10:49:36+05:30", "subject": "docs: update code references in CONTRIBUTING.md to latest tag", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "0d04351fdc12258c75b9f1cda5780fdb836ed0ef", "date": "2026-07-28T13:32:40+08:00", "subject": "feat(bundled-dev): support worker file update accepted by HMR", "body": "", "style": "conventional", "type": "feat", "scope": "bundled-dev", "breaking": false} {"repo": "ant-design", "sha": "a948e22a784f1fdb117d920276b6be4167456a4f", "date": "2026-07-28T13:41:00+08:00", "subject": "refactor(demo): replace hardcoded styles with design tokens", "body": "* refactor(demo): replace hardcoded styles with design tokens\n\n* test: update demo snapshots\n\n---------", "style": "conventional", "type": "refactor", "scope": "demo", "breaking": false} {"repo": "vite", "sha": "d9f1bf4e5eab4aecdb9a24e2d2a612f4eecdd0ba", "date": "2026-07-28T14:51:24+09:00", "subject": "ci: add semgrep", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vite", "sha": "c08880c6b122e3eb7f92d9316991860e1f1f8850", "date": "2026-07-28T15:17:47+09:00", "subject": "test: use `server.fs.allow: []`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vite", "sha": "8a2c7ea390ec83b16ec5efdc450ef077f20373f9", "date": "2026-07-28T15:35:59+09:00", "subject": "fix: track canonical safe module paths", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vite", "sha": "2e569370f1747d95f016c1cb7059c4a9030dabd0", "date": "2026-07-28T15:36:14+09:00", "subject": "fix(css): register resolved assets as safe modules", "body": "", "style": "conventional", "type": "fix", "scope": "css", "breaking": false} {"repo": "vite", "sha": "91e066322cc34c70d8110ac50aa3a87a5679cf0b", "date": "2026-07-28T15:36:29+09:00", "subject": "fix(worker): register worker URLs as safe modules", "body": "", "style": "conventional", "type": "fix", "scope": "worker", "breaking": false} {"repo": "vite", "sha": "b59106a91e212b122e92285958632481371335e4", "date": "2026-07-28T15:36:43+09:00", "subject": "fix(optimizer): register new URL assets as safe modules", "body": "", "style": "conventional", "type": "fix", "scope": "optimizer", "breaking": false} {"repo": "vite", "sha": "f63b3954f2ec4e555d796e2291a01a6f2fe45364", "date": "2026-07-28T15:36:57+09:00", "subject": "test(ssr): declare server worker input", "body": "", "style": "conventional", "type": "test", "scope": "ssr", "breaking": false} {"repo": "tmux", "sha": "a5ad1914b8b9dae3d1e8f26b2e1440d7355340a3", "date": "2026-07-28T07:04:18Z", "subject": "Add a warning about shell metacharacters, from Nikolas Skarlatos.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "e557360702c5570670c65d09227748dd5578f2fb", "date": "2026-07-28T16:04:39+09:00", "subject": "ci: fix semgrep", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "tmux", "sha": "3c1b2969a8d7e4861985a271513e318bf6700188", "date": "2026-07-28T08:08:26+01:00", "subject": "Tests for hooks time and count.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7bcc460a62bd0ebc11f1614b3b70ba9886747962", "date": "2026-07-28T14:11:58+07:00", "subject": "Remove unneeded lock", "body": "The current lockless design does not need this lock around objs anymore\n\nAssisted-by: Claude:claude-sonnet-5\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Bob Beck \nMergeDate: Thu Jul 30 06:12:46 2026\n(Merged from https://github.com/openssl/openssl/pull/32003)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "1e0d05408bd6c987407b9b54cc8459fdd3218ad3", "date": "2026-07-28T15:42:49+08:00", "subject": "test(bundled-dev): wait for first bundle and client before assertions", "body": "", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "angular", "sha": "a3fb4fe05238b152a4654547a5f2851cb5b7090f", "date": "2026-07-28T09:52:18+02:00", "subject": "docs: remove semicolon text node", "body": "(cherry picked from commit 9373d22a48546bab8de1f4b28c158d6dea617fd8)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "9373d22a48546bab8de1f4b28c158d6dea617fd8", "date": "2026-07-28T09:52:18+02:00", "subject": "docs: remove semicolon text node", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e37405ea52c0bfd7ec91b866520f7abaeeeee398", "date": "2026-07-28T08:25:32Z", "subject": "fix(treesitter): clamp unbounded fold ranges", "body": "Problem:\nTree-sitter uses UINT32_MAX for full-document ranges, which becomes -1 on 32-bit platforms and reaches _foldupdate as an invalid end row.\n\nSolution:\nTreat negative changed-range end rows as unbounded and clamp them to the buffer line count. Add a regression test that simulates the 32-bit sentinel.\n\nAI-assisted: Codex\n(cherry picked from commit 20ff82d9fcdeafe2cdf4824b8af1bdeeffb513b6)", "style": "conventional", "type": "fix", "scope": "treesitter", "breaking": false} {"repo": "neovim", "sha": "20ff82d9fcdeafe2cdf4824b8af1bdeeffb513b6", "date": "2026-07-28T08:25:32Z", "subject": "fix(treesitter): clamp unbounded fold ranges", "body": "Problem:\nTree-sitter uses UINT32_MAX for full-document ranges, which becomes -1 on 32-bit platforms and reaches _foldupdate as an invalid end row.\n\nSolution:\nTreat negative changed-range end rows as unbounded and clamp them to the buffer line count. Add a regression test that simulates the 32-bit sentinel.\n\nAI-assisted: Codex", "style": "conventional", "type": "fix", "scope": "treesitter", "breaking": false} {"repo": "curl", "sha": "53565fbfd60909697ca27a41abc9995c45a9c673", "date": "2026-07-28T10:43:30+02:00", "subject": "conncache: connection healthiness fix", "body": "Update the `lastchecked` timestamp on connection health checks when\nsuccessful to prevent repeated recalcs for a second.\n\nRename `seems_dead` to `seems_healthy` because the world is already\ndepressing enough.\n\nConsider pending input on connection only unhealthy when the connection\nhas no transfers and is not multiplexed.\n\nCloses #22412", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "fd92f74b91d079430ef4544713a90b7b6aa8faee", "date": "2026-07-28T09:44:23+01:00", "subject": "Avoid RETURNING side effects for FOR PORTION OF leftovers.", "body": "UPDATE/DELETE ... FOR PORTION OF inserts leftover rows for the\nuntouched parts of the original row. These hidden inserts should not\naffect the command tag or ROW_COUNT, so they call ExecInsert() with\ncanSetTag set to false.\n\nHowever, ExecInsert() still processed the RETURNING list whenever the\ntarget ResultRelInfo had ri_projectReturning set. That caused\nRETURNING expressions to be evaluated for leftover rows even though\ntheir results were discarded. As a result, expressions with side\neffects and information-leaking functions could be executed on the\nleftover rows, in addition to the visibly updated or deleted row.\n\nFix by having ExecInsert() skip RETURNING processing when it is\nhandling an internal FOR PORTION OF leftover insert. Use both the\npresence of a FOR PORTION OF clause and mtstate->operation ==\nCMD_INSERT for this check, so that the auxiliary INSERT of a\ncross-partition UPDATE with a FOR PORTION OF clause still processes\nRETURNING normally.\n\nBack-patch to v19, where support for FOR PORTION OF was added.\n\nAuthor: Chao Li \nReviewed-by: Dean Rasheed \nReviewed-by: Paul A Jungwirth \nDiscussion: https://postgr.es/m/07C125E5-F6ED-460C-A394-E6503DAE18FB@gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "15fd7a3e2eeb6de915cae6ea30f40a8e9100cbfe", "date": "2026-07-28T10:50:13+02:00", "subject": "Restore vacuum_delay_point() in GIN posting-tree leaf vacuum", "body": "Commit fd83c83d094 turned the recursive posting-tree cleanup in\nginVacuumPostingTreeLeaves() into an iterative sweep that follows the\ntree's leaf pages via their rightlinks. The recursive version called\nvacuum_delay_point() while processing the tree, but that call was removed\nand never re-added to the new loop. As that commit only set out to fix a\ndeadlock, the removal appears to have been unintentional.\n\nConsequently the leaf-page sweep of a single posting tree runs with no\nvacuum_delay_point(), and therefore no CHECK_FOR_INTERRUPTS(). A posting\ntree stores all the TIDs for one indexed key, so for a frequently\noccurring key it can span a large number of leaf pages. While such a\ntree is being vacuumed the operation ignores vacuum_cost_delay and does\nnot respond to query cancellation or statement_timeout; an autovacuum\nworker likewise cannot be interrupted mid-sweep when another backend\nrequests a conflicting lock.\n\nRestore the call, placed after the current page has been unlocked and\nreleased so that no buffer content lock is held across a potential delay\n(cf. 21c27af65fb). The sibling loops in ginbulkdelete() and\nginvacuumcleanup() already call vacuum_delay_point() once per page.\n\nAuthor: Paul Kim \nCo-authored-by: Alexander Korotkov \nReviewed-by: Michael Paquier \nReviewed-by: Andrey Borodin \nReviewed-by: solai v \nDiscussion: https://postgr.es/m/178447127453.110.12276981925360691905%40mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "2aea45d932f20d005f346788d972afa2db6ade9f", "date": "2026-07-28T17:07:19+08:00", "subject": "fix(Upload): isolate default download tabs", "body": "", "style": "conventional", "type": "fix", "scope": "Upload", "breaking": false} {"repo": "angular", "sha": "6dcb534e127d8b9971103596844ca1c0e736596b", "date": "2026-07-28T11:12:27+02:00", "subject": "refactor(common): remove duplicate helper function", "body": "We have `useAutoTick` in our private shared utils.\n\n(cherry picked from commit 36474f7011adf817c3efba841e8a1d630139aa43)", "style": "conventional", "type": "refactor", "scope": "common", "breaking": false} {"repo": "angular", "sha": "36474f7011adf817c3efba841e8a1d630139aa43", "date": "2026-07-28T11:12:27+02:00", "subject": "refactor(common): remove duplicate helper function", "body": "We have `useAutoTick` in our private shared utils.", "style": "conventional", "type": "refactor", "scope": "common", "breaking": false} {"repo": "vite", "sha": "a54f936c1166d6c521a4a37d1170e573426bc153", "date": "2026-07-28T18:15:54+09:00", "subject": "ci: exclude `__tests__` directories from semgrep", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "86a89500d62718fccebbaefa4d47b5bdff72cc96", "date": "2026-07-28T17:42:15+08:00", "subject": "site: refine MUI theme preview styles", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "85503d1372719e0f0adc4a987e14b81ba00694a2", "date": "2026-07-28T15:12:40+05:30", "subject": "fix(Typography): inherit font size in editable textarea", "body": "", "style": "conventional", "type": "fix", "scope": "Typography", "breaking": false} {"repo": "curl", "sha": "c7328740ecad6e5549c210366b8403eca29b4123", "date": "2026-07-28T11:53:42+02:00", "subject": "lib2405: adjust for non-threaded builds", "body": "- Attempt to fix the flakiness set in 9726fc8259ad7ebf1682\n- Reduce macro use\n\nCloses #22414", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "363aacb6ae512af4f393fecc79a42d7d3053a2bd", "date": "2026-07-28T15:29:51+05:30", "subject": "fix(core): resolve barrier immediately when target count is zero", "body": "A Barrier created with targetCount 0 never settled its promise because\nsignal() was never called for classes with zero constructor or property\ndependencies, leaving one forever-pending promise per zero-dependency\nclass and tripping async-leak detection (e.g. Vitest detectAsyncLeaks).\nResolve eagerly in the constructor and make signal() idempotent past\nthe threshold to tolerate over-signaling from error paths.\n\nCloses #17392", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "neovim", "sha": "72e3fc2074cb2bbc68c8475f632f84ef5503142b", "date": "2026-07-28T06:10:36-04:00", "subject": "build(vim-patch): \":version\" doc is n/a #41018", "body": "Nvim diverged from Vim's \":version\".\n\n- no date\n- different compiled features list\n- different config file and directory\n- no compiler flags\n\n* build(vim-patch): n/a doc for balloon feature (8.2.3917)", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "neovim", "sha": "205a1ce4c7c7fd512e3882535171b07e444012eb", "date": "2026-07-28T18:11:42+08:00", "subject": "test(old): map `gQ` to `1q:`", "body": "", "style": "conventional", "type": "test", "scope": "old", "breaking": false} {"repo": "neovim", "sha": "93f427f6031aa6732769ef81e20a3294d4b7994d", "date": "2026-07-28T05:12:01-05:00", "subject": "docs(dir): buffer state discarded by 'bufhidden' #41014", "body": "", "style": "conventional", "type": "docs", "scope": "dir", "breaking": false} {"repo": "neovim", "sha": "9a92f4a001f8262d35dc972735d2f3b5f0bb39cb", "date": "2026-07-28T12:14:32+02:00", "subject": "fix(treesitter): incremental selection: \"vin\" not always starts \"v\" #41016", "body": "(cherry picked from commit 59296a9fef42247f91567a77bd87edd34b14f6cc)", "style": "conventional", "type": "fix", "scope": "treesitter", "breaking": false} {"repo": "neovim", "sha": "59296a9fef42247f91567a77bd87edd34b14f6cc", "date": "2026-07-28T12:14:32+02:00", "subject": "fix(treesitter): incremental selection: \"vin\" not always starts \"v\" #41016", "body": "", "style": "conventional", "type": "fix", "scope": "treesitter", "breaking": false} {"repo": "angular", "sha": "c4b14321292ca59db2b5fe1c178b28d9f058d5f3", "date": "2026-07-28T15:22:03+05:00", "subject": "docs: document overlay stacking change in v21 update guide", "body": "(cherry picked from commit ae303c4a05be72a89b7a8b41736bf9da75c245a7)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "ae303c4a05be72a89b7a8b41736bf9da75c245a7", "date": "2026-07-28T15:22:03+05:00", "subject": "docs: document overlay stacking change in v21 update guide", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "tmux", "sha": "6a13d61b48d01ed288ad033f661df5fdb27cf501", "date": "2026-07-28T10:35:31Z", "subject": "Show hook count and time in customize mode.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "9c402fa7b70c5328ea183fa2f18a7eaf44c0857d", "date": "2026-07-28T11:35:36+01:00", "subject": "Test hook time in customize mode.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "a9aaffe24186f0daf30c0e5d92ac5c6d0b77fba2", "date": "2026-07-28T19:16:19+08:00", "subject": "site: prevent version selector from shifting page scroll", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "5226e26e998f792641590b24b19cdd875b5edd84", "date": "2026-07-28T06:21:45-05:00", "subject": "feat(pos): vim.pos.cursor() without args returns curwin pos #40975", "body": "", "style": "conventional", "type": "feat", "scope": "pos", "breaking": false} {"repo": "neovim", "sha": "916a6e9546444369760e0335b150e74e9511c32d", "date": "2026-07-28T19:24:58+08:00", "subject": "fix(eval): complete_info() equal, preselect, commit_chars #41022", "body": "Problem: complete_info() does not report \"equal\", \"preselect\" or \"commit_chars\".\n\nSolution: Report them for the items that set them.", "style": "conventional", "type": "fix", "scope": "eval", "breaking": false} {"repo": "vite", "sha": "d17d739eedc4e8ede5973f1533826f27554c2078", "date": "2026-07-28T19:28:21+08:00", "subject": "ci: remove warp oz", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f0f6681cea793824a01caaee992272ec10fe1d53", "date": "2026-07-28T19:32:35+08:00", "subject": "fix(floatwin): relative='cursor' may use outdated cursor position #40768", "body": "Problem:\nCursor-relative floats can use stale screen coordinates after a cursor move is restored without a redraw.\n\nSolution:\nValidate the current cursor before converting cursor-relative coordinates.\n\nCo-authored-by: zeertzjq \n(cherry picked from commit 089c415cb236512add6c29372e26640ff401d39b)", "style": "conventional", "type": "fix", "scope": "floatwin", "breaking": false} {"repo": "neovim", "sha": "089c415cb236512add6c29372e26640ff401d39b", "date": "2026-07-28T19:32:35+08:00", "subject": "fix(floatwin): relative='cursor' may use outdated cursor position #40768", "body": "Problem:\nCursor-relative floats can use stale screen coordinates after a cursor move is restored without a redraw.\n\nSolution:\nValidate the current cursor before converting cursor-relative coordinates.", "style": "conventional", "type": "fix", "scope": "floatwin", "breaking": false} {"repo": "vite", "sha": "8c2a87d41fb24536e59643351758084cde4d0dd7", "date": "2026-07-28T07:41:32-04:00", "subject": "fix(importAnalysis): interop imports injected into optimized dep files by plugins", "body": "", "style": "conventional", "type": "fix", "scope": "importAnalysis", "breaking": false} {"repo": "vite", "sha": "8a245726944ed29225920d49be77c33c6e03afc8", "date": "2026-07-28T19:52:18+08:00", "subject": "feat(config): include column in config incompatibility location", "body": "", "style": "conventional", "type": "feat", "scope": "config", "breaking": false} {"repo": "ant-design", "sha": "f561b5ffa62b401fe6e1e5f8d668b8c27149a9a8", "date": "2026-07-28T20:05:18+08:00", "subject": "fix: hide textarea resize handle on touch devices", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "c7b1e10dc1dcacae17fc75fdb6572846605d2ffe", "date": "2026-07-28T14:12:09+02:00", "subject": "h2: bootstrap max streams from multi handle if in use", "body": "Closes #22418", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "079a11bcba5858ffab2cf6e810e9d98024458487", "date": "2026-07-28T14:22:43+02:00", "subject": "vquic: add option to use Apple fast UDP", "body": "Using Apple's secret `SYS_recvmsg_x` and `SYS_sendmsg_x` syscalls for\nreceiving/sending batches of UDP packets.\n\nSince it uses undocumented calls, it's experimental and disabled by\ndefault.\n\nTo enable:\n- autotools: `--enable-apple-fast-udp`\n- cmake: `-DCURL_ENABLE_APPLE_FAST_UDP=ON`\n\nAlso:\n- enable in two H3 CI jobs with both build tools, pytest and clang-tidy.\n\nRefs:\nhttps://max-inden.de/post/fast-udp-io-in-firefox/\nhttps://www.macsyscalls.com/en/syscall/480-recvmsg-x\nhttps://www.macsyscalls.com/en/syscall/481-sendmsg-x\n\nC-code-authored-by: Stefan Eissing\nBuild-code-authored-by: Viktor Szakats\n\nCloses #22341\nCloses #22417", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "7a58550089dc07024e16cc9c404420519fea32d3", "date": "2026-07-28T17:54:48+05:30", "subject": "fix(forms): create controls before listeners", "body": "Create form controls before registering template listeners so custom control values are synchronized before user callbacks run. Cover Signal Forms, reactive forms, and template-driven forms.\n\n(cherry picked from commit 6688a3d5502f159479c4e71cbd44f6713a40fe25)", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "6688a3d5502f159479c4e71cbd44f6713a40fe25", "date": "2026-07-28T17:54:48+05:30", "subject": "fix(forms): create controls before listeners", "body": "Create form controls before registering template listeners so custom control values are synchronized before user callbacks run. Cover Signal Forms, reactive forms, and template-driven forms.", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "curl", "sha": "3974491c97cbff6ca39dd249c02b24e30171f265", "date": "2026-07-28T14:27:13+02:00", "subject": "websocket: pause writing and meta data fix", "body": "When writing a decoded chunk of websocket data, always flush the writer\nchain so that buffered data gets delivered before the ws meta data gets\nupdated.\n\nAdd client writer flags CURL_CW_FLAG_BLOWUP for writer types that may\nsignificantly enlarge write sizes. This flag causes the pause writer to\nbe added and shrinks the write chunk sizes. We do not want that for\ncontent decoders like WS that do not change the size.\n\nAdd test_20_13 to check that large frames are paused/unpaused correctly\nwith the matching meta data.\n\nFixes #22413\nReported-by: Hendrik Hübner\nCloses #22416", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "17d1ed53ffe81ee50835ac2c6de8f5180bdfeb01", "date": "2026-07-28T15:32:17+03:00", "subject": "feat(wasm): implement initial web demo", "body": "Introduce an initial browser-based Neovim demo with automatic startup and UI integration. The demo is functional but still requires persistence, additional testing, and further stabilization.\n\nAcknowledgment: UiState in app.js is based on the implementation from github.com/MuNeNiCK/nvim-wasm, with additional modifications and updates", "style": "conventional", "type": "feat", "scope": "wasm", "breaking": false} {"repo": "neovim", "sha": "3fa5904cf246b783db778a457e5544272b847eff", "date": "2026-07-28T20:39:03+08:00", "subject": "fix(lsp): multibyte fragment treated as commit char #41019", "body": "Problem:\ncommit_chars_str() reads a 0x80-0xBF byte as a 2-byte lead, so\n\"\\xA9x\" makes \"x\" a commit character. nvim_get_hl() returns\na dict, so `#` on it is 0 whether or not the group exists\n\nSolution:\nuse vim.str_utf_end() and drop entries that are not one whole\ncharacter. use next check dict length.", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "vitest", "sha": "0553851f9f299233fbd2e797a76152e98f593ac7", "date": "2026-07-28T14:40:32+02:00", "subject": "chore: add Knip checks", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "cccd76f2aa02b78af0b305a3f548edafb12966e9", "date": "2026-07-28T14:47:44+02:00", "subject": "docs/INTERNALS.md: move a column one byte", "body": "To make the minimum gap two spaces, so that we can separate the columns\nprogrammatically", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "613c7706a5b999b6ba7713feaecc21f052f412ba", "date": "2026-07-28T14:50:47+02:00", "subject": "chore: remove direct `esbuild` devDependency", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "openssl", "sha": "db33cefb81296269c3eb8ecce90ecc9e712076d5", "date": "2026-07-28T15:02:23+02:00", "subject": "pkcs7: use PKCS7_get_octet_string in PKCS7_stream signed arm", "body": "Direct access to `p7->d.sign->contents->d.data` skips the content-type\ncheck: if the inner eContentType is a non-standard OID the ADB sets\nd.other (16 bytes) instead of d.data (24 bytes), and the subsequent\nos->flags write lands out of bounds.\n\nReplace with `PKCS7_get_octet_string()`, which returns NULL for any type\nthat is not `NID_pkcs7_data`, matching the guard already used by\n`PKCS7_ctrl()` and `PKCS7_dataDecode()`.\n\nAdded a unit-test (`pkcs7_stream_non_data_test`) to validate the change.\n\nFixes #31681\n\nReviewed-by: Bob Beck \nReviewed-by: Jakub Zelenka \nReviewed-by: Eugene Syromiatnikov \nMerge-date: Tue Aug 18 07:47:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/31722\n(cherry picked from commit eedaf1c2c46b5e511a603833a65a742085c01bb2)\n(cherry picked from commit 56c1e4fc7625295283c22108746cb974229d9575)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "56c1e4fc7625295283c22108746cb974229d9575", "date": "2026-07-28T15:02:23+02:00", "subject": "pkcs7: use PKCS7_get_octet_string in PKCS7_stream signed arm", "body": "Direct access to `p7->d.sign->contents->d.data` skips the content-type\ncheck: if the inner eContentType is a non-standard OID the ADB sets\nd.other (16 bytes) instead of d.data (24 bytes), and the subsequent\nos->flags write lands out of bounds.\n\nReplace with `PKCS7_get_octet_string()`, which returns NULL for any type\nthat is not `NID_pkcs7_data`, matching the guard already used by\n`PKCS7_ctrl()` and `PKCS7_dataDecode()`.\n\nAdded a unit-test (`pkcs7_stream_non_data_test`) to validate the change.\n\nFixes #31681\n\nReviewed-by: Bob Beck \nReviewed-by: Jakub Zelenka \nReviewed-by: Eugene Syromiatnikov \nMerge-date: Tue Aug 18 07:47:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/31722\n(cherry picked from commit eedaf1c2c46b5e511a603833a65a742085c01bb2)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "eedaf1c2c46b5e511a603833a65a742085c01bb2", "date": "2026-07-28T15:02:23+02:00", "subject": "pkcs7: use PKCS7_get_octet_string in PKCS7_stream signed arm", "body": "Direct access to `p7->d.sign->contents->d.data` skips the content-type\ncheck: if the inner eContentType is a non-standard OID the ADB sets\nd.other (16 bytes) instead of d.data (24 bytes), and the subsequent\nos->flags write lands out of bounds.\n\nReplace with `PKCS7_get_octet_string()`, which returns NULL for any type\nthat is not `NID_pkcs7_data`, matching the guard already used by\n`PKCS7_ctrl()` and `PKCS7_dataDecode()`.\n\nAdded a unit-test (`pkcs7_stream_non_data_test`) to validate the change.\n\nFixes #31681\n\nReviewed-by: Bob Beck \nReviewed-by: Jakub Zelenka \nReviewed-by: Eugene Syromiatnikov \nMerge-date: Tue Aug 18 07:47:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/31722", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "d0d7f57e0810a24ba16dbb1f2ab9f079a096fa3d", "date": "2026-07-28T15:09:03+02:00", "subject": "fix(compiler): scope nested CSS rules", "body": "This is a second attempt at scoping nested CSS rules after not being able to land #50693. I took some extra precautions to try and avoid the crash that was happening last time.", "style": "conventional", "type": "fix", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "a4f1a949486d20ce2bf9add427eea6629228b28b", "date": "2026-07-28T15:10:25+02:00", "subject": "fix(compiler): do not encapsulate nested selectors if parent contains ::ng-deep", "body": "Updates the nested selector encapsulation logic so that it skips encapsulating child selectors if the parent contains `::ng-deep`. The reasoning is that `:host ::ng-deep { .foo {} }` should behave as `:host ::ng-deep .foo {}`.", "style": "conventional", "type": "fix", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "644455b29eade333d38954735fbc5a8d08c14474", "date": "2026-07-28T15:11:16+02:00", "subject": "refactor(core): update acceptance test", "body": "- remove NgModules\n- remove non-necessary TestBed.configureTestingModule\n- remove non-necessary Eager strategy\n- Drop usages of CommonModule\n\n(cherry picked from commit d8a570ec4e56565f4f0185500ebb9a206e894322)", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "d8a570ec4e56565f4f0185500ebb9a206e894322", "date": "2026-07-28T15:11:16+02:00", "subject": "refactor(core): update acceptance test", "body": "- remove NgModules\n- remove non-necessary TestBed.configureTestingModule\n- remove non-necessary Eager strategy\n- Drop usages of CommonModule", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "tmux", "sha": "9f2c535909ae0a369e6006b9cf32d8186c6a3788", "date": "2026-07-28T13:17:45Z", "subject": "Store status position in new mouse event for menus so the position is correct later.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "7cc0c11cde1b9e9ad193cbd2d10e01b0d6dbaa55", "date": "2026-07-28T14:18:02+01:00", "subject": "Add test for mouse menu position.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "0617f7020b286cd8e8211862b5489e585f7fc9a6", "date": "2026-07-28T16:46:13+03:00", "subject": "fix(pack): do not write the lockfile when installing from it #41031", "body": "Problem: Installing plugins during lockfile synchronization always\n writes the lockfile, even though its content is used during install.\n This might be a problem if the lockfile (itself or its parent\n directory) is not writeable (can only be read).\n\nSolution: Do not write the lockfile when installing directly from it.\n This is okay since the `src` and `rev` are used directly from the\n lockfile and don't change at this step. While potential change in\n `version` (that must be written to the lockfile) is handled in other\n code path.\n(cherry picked from commit 4bba83b4b7363bc6b8da04cb1e13817473b028ec)", "style": "conventional", "type": "fix", "scope": "pack", "breaking": false} {"repo": "neovim", "sha": "4bba83b4b7363bc6b8da04cb1e13817473b028ec", "date": "2026-07-28T16:46:13+03:00", "subject": "fix(pack): do not write the lockfile when installing from it #41031", "body": "Problem: Installing plugins during lockfile synchronization always\n writes the lockfile, even though its content is used during install.\n This might be a problem if the lockfile (itself or its parent\n directory) is not writeable (can only be read).\n\nSolution: Do not write the lockfile when installing directly from it.\n This is okay since the `src` and `rev` are used directly from the\n lockfile and don't change at this step. While potential change in\n `version` (that must be written to the lockfile) is handled in other\n code path.", "style": "conventional", "type": "fix", "scope": "pack", "breaking": false} {"repo": "eslint", "sha": "4d7db6628e2badf0857cb88734fe641c3874bce9", "date": "2026-07-28T17:05:42+03:00", "subject": "fix: ignore meta-property names in `id-match`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "eslint", "sha": "6898df9364639ee64b9448a4cb6b08a30c16bd37", "date": "2026-07-28T17:07:16+03:00", "subject": "fix: ignore meta-property names in `id-denylist`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ef382ce53b497fba4fa19306719cf441428e3aab", "date": "2026-07-28T16:23:01+02:00", "subject": "Document the behavior of BN_generate_prime() on failure", "body": "The ret argument is freed on failure which might be\nsurprising to callers.\n\nReported by: Minkyung Park (UNIST)\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Paul Yang \nMergeDate: Mon Aug 10 15:09:34 2026\n(Merged from https://github.com/openssl/openssl/pull/32099)\n\n(cherry picked from commit 6d250710fb8cfd5c9a41af373c0c1ca85ff836c2)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6d250710fb8cfd5c9a41af373c0c1ca85ff836c2", "date": "2026-07-28T16:23:01+02:00", "subject": "Document the behavior of BN_generate_prime() on failure", "body": "The ret argument is freed on failure which might be\nsurprising to callers.\n\nReported by: Minkyung Park (UNIST)\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Paul Yang \nMergeDate: Mon Aug 10 15:09:34 2026\n(Merged from https://github.com/openssl/openssl/pull/32099)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "22e52568536e0009d4493dc888736bc163220c4b", "date": "2026-07-28T23:58:38+09:00", "subject": "ci: add `triage:no` label to Dependabot PRs", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "curl", "sha": "75404bda1a4fe062fdb5c98b31c0be2e1558674e", "date": "2026-07-28T17:08:02+02:00", "subject": "curl_gssapi: document/update feature availability", "body": "- update `GSS_C_DELEG_POLICY_FLAG` comment to include Apple GSS, add\n date, and amend MIT Kerberos version to 1.7+ (was: 1.8+)\n Ref: https://github.com/krb5/krb5/commit/45875a4d7bbd6bb8a943572d84fef5ca2bb18291\n Ref: https://github.com/apple-oss-distributions/Heimdal/commit/1635de38a813f6e1dd3f8fd270683187ad4f02be\n\n- document `HAVE_GSS_SET_NEG_MECHS`/`gss_set_neg_mechs()`.\n Ref: https://github.com/krb5/krb5/commit/079eed2cf749702f75ddc385cf943fbab931f9d8\n It's also committed to Heimdal, but not present in a release\n as of 7.8.0 (current latest).\n Ref: https://github.com/heimdal/heimdal/commit/735039dbdc3aa58d06afdefd214efe3f5e421244\n\nFollow-up to a8881e5e1d2e22d4b085d45ab6c8ce1df251d602 #21315 #22410\nFollow-up to d169ad68faa5ed5ba2375e7502307a6262466d88 #22052\n\nCloses #22419", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "3fd595af47bd7f7d9418eda9b490efa9cd0a4d14", "date": "2026-07-28T18:09:08+03:00", "subject": "fix(microservices): nullify grpcclient on close when tryshutdown fails", "body": "In the close() method, if graceful shutdown is enabled and\ntryShutdown calls back with an error, the promise rejects and\nthis.grpcClient = null was skipped, leaving a dangling reference.\n\nWrapped the shutdown logic in try/finally to guarantee nullification\nregardless of success or failure.", "style": "conventional", "type": "fix", "scope": "microservices", "breaking": false} {"repo": "nestjs", "sha": "8539f8747b39c8d92d19d51d3cce53ada0150f0a", "date": "2026-07-28T18:12:04+03:00", "subject": "refactor(core): remove async promise executor anti-patterns", "body": "Replace new Promise(async (resolve, reject) => { ... }) with\nasync IIFE or standard Promise constructor across 3 files:\n\n- nest-application-context.ts: init() method\n- client-kafka.ts: connect() method\n- module-ref.ts: instantiateClass() method\n\nAsync executors are an anti-pattern: the async function's returned\npromise is ignored, and errors must be manually caught and forwarded.\nThe IIFE form produces the same result with less indirection.", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "eslint", "sha": "65a6519cc8733b4688558a8611397cf9d01dd55d", "date": "2026-07-29T00:13:42+09:00", "subject": "chore: add allowScripts field to package.json", "body": "* chore: add allowScripts field to package.json\n\n* chore: remove version pinning from allowScripts\n\n* chore: set core-js allowScripts to false\n\n* chore: set cypress allowScripts to false\n\n* chore: install cypress binary before running browser tests\n\n* chore: add devEngines to warn on buggy npm versions\n\n* ci: update npm in browser test to avoid allowScripts bin-link bug\n\n* chore: remove devEngines from package.json\n\n* chore: update comment for npm version in CI workflow( test_on_browser)", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "58dc55905cec755aa6f60980e5391b7e39b92f82", "date": "2026-07-28T18:15:07+03:00", "subject": "test(core): add unit tests for module-ref", "body": "Covers:\n- introspect() for DEFAULT, REQUEST, and TRANSIENT scopes\n- registerRequestByContextId() delegation\n- instantiateClass() with dependency resolution and error propagation\n- find() for single, request-scoped, and multi-instance lookups\n- resolve() for static, transient (unique per context), and\n request-scoped providers", "style": "conventional", "type": "test", "scope": "core", "breaking": false} {"repo": "vitest", "sha": "0f86612c6d8b07c3f722f0566057348a75914c0a", "date": "2026-07-28T18:39:10+02:00", "subject": "test(browser): cover instance headless overrides", "body": "", "style": "conventional", "type": "test", "scope": "browser", "breaking": false} {"repo": "vite", "sha": "b2f0f84ef715b1658a66b990fee1cb4f6ec1316e", "date": "2026-07-29T01:05:09+08:00", "subject": "test(bundled-dev): wait for updates after edits in enabled playgrounds", "body": "", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "neovim", "sha": "0d36e61d70a3ab82278e0988ff8985f667aa587c", "date": "2026-07-28T12:37:06-05:00", "subject": "fix(dir): \"-\" on a non-file buffer opens a bogus path #41034", "body": "", "style": "conventional", "type": "fix", "scope": "dir", "breaking": false} {"repo": "postgres", "sha": "6c760f6b6a3a8c1c49cdc54be62363af4308d62c", "date": "2026-07-28T10:39:50-07:00", "subject": "Fix pg_get_publication_tables() failure with concurrent DROP TABLE.", "body": "pg_get_publication_tables() collects the OIDs of the published tables\non its first call, without locking them, and then reopens each table\nlater, once per result row, to compute its column list and fetch its\nrow filter. The reopen used table_open(), which errors out with \"could\nnot open relation with OID\" if the table has been dropped in the\nmeantime. This could happen for any published table without an\nexplicit column list, which is every table in FOR ALL TABLES and FOR\nTABLES IN SCHEMA publications, but also FOR TABLE entries without a\ncolumn list. The failure is common in environments where many tables\nare created and dropped while publication tables are being queried,\ne.g. by table synchronization on a subscriber.\n\nFix by opening every table with try_table_open(), which returns NULL\nif the relation no longer exists, and skipping the table in that\ncase. Concurrently dropped tables are thus simply absent from the\nresult set, which is the expected point-in-time behavior.\n\nAs a side effect, tables with an explicit column list, which were\npreviously returned without being opened, are now also locked with\nAccessShareLock, so the function can block behind concurrent DDL on\nsuch tables where it previously did not.\n\nBackpatch to v16, where we added the table_open() call in\npg_get_publication_tables().\n\nAuthor: Bharath Rupireddy \nReviewed-by: Bertrand Drouvot \nReviewed-by: shveta malik \nReviewed-by: Ajin Cherian \nReviewed-by: Masahiko Sawada \nReviewed-by: Chao Li \nDiscussion: https://www.postgresql.org/message-id/CALj2ACVYYooWH-5tJ6cPKkU%2BmutVxwb_z4S%2BqAi-zdrFqxXE2Q%40mail.gmail.com\nBackpatch-through: 16", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "601d1f6576b296748f56ffb50646bb1808a7c0fc", "date": "2026-07-28T20:42:50+03:00", "subject": "fix(core): throw NG0500 instead of a raw TypeError on element hydration mismatch", "body": "When hydration locates the DOM node for an ɵɵelementStart/ɵɵdomElementStart\ninstruction, locateOrCreateElementNodeImpl assumed the located node was\nalways an Element and called hasSkipHydrationAttrOnRElement(native), which\ndoes native.hasAttribute(...). The check that would normally catch this\nclass of mismatch, validateMatchingNode, is gated behind `ngDevMode &&` and\nis compiled out of production builds. So when a real SSR/hydration\nstructural mismatch located a Text or Comment node instead of the expected\nElement, production builds hit .hasAttribute on a node type that doesn't\nhave it and crashed with a raw, uncoded TypeError instead of a coded\nhydration-mismatch RuntimeError.\n\nAdd a cheap, always-on nodeType check ahead of that call. On mismatch it\nthrows RuntimeError(HYDRATION_NODE_MISMATCH, ngDevMode && '...'), the same\npattern used elsewhere in the codebase, so the descriptive message is only\nbuilt in dev mode and production keeps throwing just the bare NG0500 code\nwithout pulling validateMatchingNode's DOM-printing machinery into the\nproduction bundle (verified via the bundling/hydration golden-symbols test,\nwhich is unchanged).\n\n(cherry picked from commit 4560f4fdcda4d55c9234f71fda2e73938cb55ed6)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "4560f4fdcda4d55c9234f71fda2e73938cb55ed6", "date": "2026-07-28T20:42:50+03:00", "subject": "fix(core): throw NG0500 instead of a raw TypeError on element hydration mismatch", "body": "When hydration locates the DOM node for an ɵɵelementStart/ɵɵdomElementStart\ninstruction, locateOrCreateElementNodeImpl assumed the located node was\nalways an Element and called hasSkipHydrationAttrOnRElement(native), which\ndoes native.hasAttribute(...). The check that would normally catch this\nclass of mismatch, validateMatchingNode, is gated behind `ngDevMode &&` and\nis compiled out of production builds. So when a real SSR/hydration\nstructural mismatch located a Text or Comment node instead of the expected\nElement, production builds hit .hasAttribute on a node type that doesn't\nhave it and crashed with a raw, uncoded TypeError instead of a coded\nhydration-mismatch RuntimeError.\n\nAdd a cheap, always-on nodeType check ahead of that call. On mismatch it\nthrows RuntimeError(HYDRATION_NODE_MISMATCH, ngDevMode && '...'), the same\npattern used elsewhere in the codebase, so the descriptive message is only\nbuilt in dev mode and production keeps throwing just the bare NG0500 code\nwithout pulling validateMatchingNode's DOM-printing machinery into the\nproduction bundle (verified via the bundling/hydration golden-symbols test,\nwhich is unchanged).", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "openssl", "sha": "775693e5c79f85e34aed75f56300470021cedcd9", "date": "2026-07-28T21:50:53+04:00", "subject": "storemgmt: Validate msblob length before buffer allocation", "body": "Reviewed-by: Paul Dale \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 12 09:54:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32102\n(cherry picked from commit 0f5493e60b9e8ec500de5812fe28b9f564408608)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0f5493e60b9e8ec500de5812fe28b9f564408608", "date": "2026-07-28T21:50:53+04:00", "subject": "storemgmt: Validate msblob length before buffer allocation", "body": "Reviewed-by: Paul Dale \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 12 09:54:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32102", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "7be9d5d6b24dd8bec6ee6a3aa5d8b565d6faf91a", "date": "2026-07-28T13:14:43-05:00", "subject": "refactor(forms): modernize signal forms tests to rely on whenStable", "body": "Rely on zoneless test scheduling instead of manually triggering change detection. Keep Signals Forms tests aligned with the async-first testing pattern.\n\n(cherry picked from commit 59f6ef690b82dab96ff8a29acc2a6aa06dcaad02)", "style": "conventional", "type": "refactor", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "59f6ef690b82dab96ff8a29acc2a6aa06dcaad02", "date": "2026-07-28T13:14:43-05:00", "subject": "refactor(forms): modernize signal forms tests to rely on whenStable", "body": "Rely on zoneless test scheduling instead of manually triggering change detection. Keep Signals Forms tests aligned with the async-first testing pattern.", "style": "conventional", "type": "refactor", "scope": "forms", "breaking": false} {"repo": "neovim", "sha": "635acc7dc848853fe4a5db2f0c85dc52e8eabed0", "date": "2026-07-28T13:59:18-05:00", "subject": "fix(terminal): cursor moves on resize when line above is full width #40996", "body": "", "style": "conventional", "type": "fix", "scope": "terminal", "breaking": false} {"repo": "postgres", "sha": "c2d34db0a5eced16cf8d7e1004674b20b8fa383d", "date": "2026-07-28T12:33:50-07:00", "subject": "Fix logical decoding of empty prepared transactions.", "body": "A two-phase transaction that is assigned an XID but produces no change\nto be decoded -- for example, one that only acquires row locks via\nSELECT ... FOR SHARE -- has no base snapshot in the reorder\nbuffer. ReorderBufferReplay() already skips such a transaction at\nPREPARE time and never invokes the begin_prepare/change/prepare\ncallbacks for it, but ReorderBufferFinishPrepared() still called the\ncommit_prepared (or rollback_prepared) callback. As a result a\nspurious COMMIT/ROLLBACK PREPARED was sent to the output plugin with\nno preceding PREPARE. For the built-in subscriber this breaks\nreplication (the apply worker fails to find the prepared transaction),\nand test_decoding could even crash.\n\nFix this by detecting an empty transaction (base_snapshot == NULL) in\nReorderBufferFinishPrepared() and cleaning it up without invoking the\ncommit/rollback prepared callbacks, mirroring the existing empty\ntransaction handling in ReorderBufferReplay().\n\nOn v18 and newer versions, commit 072ee847ad4 changed\nReorderBufferPrepare() to send the prepare whenever it had not already\nbeen sent, which also fires for empty transactions and emits a\nspurious PREPARE. On those branches ReorderBufferPrepare() is\ntherefore additionally guarded with base_snapshot != NULL. This guard\nand the Assert(!rbtxn_sent_prepare()) added in\nReorderBufferFinishPrepared(), are not necessary on v17 and older\nversions: there ReorderBufferPrepare() only sends a prepare for\nconcurrently-aborted transactions (which never applies to an empty\ntransaction) and the RBTXN_SENT_PREPARE flag does not exist.\n\nBack-patch to v14, where decoding of two-phase transactions was\nintroduced.\n\nBug: #19556\nReported-by: Alexander Kozhemyakin \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/19556-daa6d7ea65054d48@postgresql.org\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "4b83d09a94c3bf5102dd9beda70cff8d6e2c440b", "date": "2026-07-28T22:48:33+03:00", "subject": "fix(devtools): invoke selectedNode signal in directive tree keyboard guard", "body": "`isEditingDirectiveState` used `!this.selectedNode` (a bare signal\nreference, always truthy) instead of `!this.selectedNode()`, so the\nguard was dead and ArrowDown with no selection threw a TypeError.", "style": "conventional", "type": "fix", "scope": "devtools", "breaking": false} {"repo": "postgres", "sha": "9eb77f9fc80db071b434aea58681ed2adb9038d3", "date": "2026-07-28T21:52:24+02:00", "subject": "Recheck checksum state before file_copy during CREATE DATABASE", "body": "The file_copy strategy check in createdb() runs during option\nvalidation, before the transaction has an XID and before the\npg_database row exists, so the datachecksumsworker launcher\ncan start in that window and see neither the new database nor\nthe transaction creating it. It then raw-copies a template\nthat was not processed yet, and those files stay unchecksummed,\nfailing verification from then on.\n\nRecheck the state in CreateDatabaseUsingFileCopy(): the XID is\nassigned by then, so a launcher starting after this point waits\nfor the transaction and finds the new database, and the copy\nerrors out instead. Add an injection point before the catalog\ninsert to test the window.\n\nBackpatch to v19 where online checksums were introduced.\n\nAuthor: Zsolt Parragi \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CAN4CZFPEBsz8JeY4ixQ1V4ZL_xOY6pJaZS8ZLGH7R+wF--pEtg@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "e469e4784ea9f632b91e5d988cd336cbf54af58a", "date": "2026-07-28T21:52:27+02:00", "subject": "Handle invalid and dropped databases during checksum enable", "body": "Enable errors out early with a hint when an invalid database exists,\nsince the worker cannot connect to it and its files stay on disk.\n\nA worker that started but failed gets the same dropped-database\nheuristic as one that failed to start, so a concurrent drop during\nprocessing no longer aborts the whole run. The existence check locks\nthe database first, otherwise a DROP DATABASE ... WITH (FORCE) which\nkilled the worker is still only halfway done and the database looks\nlike it is there to stay.\n\nBackpatch to v19 where online checksums were introduced.\n\nAuthor: Zsolt Parragi \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CAN4CZFOGdqxtZ5-6gb4apqmvoH=Z+TNH8RKJ3mVtoR1HirKQWg@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "151ba3f5834a0909e8b9b1736f4889ac694c0104", "date": "2026-07-29T05:01:30+09:00", "subject": "fix: false positives in `getter-return` and `accessor-pairs`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "postgres", "sha": "13b627a3e1811577e830c8b0a4a2cc095b22ac47", "date": "2026-07-28T16:08:46-04:00", "subject": "Fix planner's nullability/strictness logic for ScalarArrayOpExpr.", "body": "find_nonnullable_rels and find_nonnullable_vars mistakenly treated a\nScalarArrayOpExpr that could return FALSE as strict, but that's okay\nonly at top level of a qual expression; further down, we've got to\ninsist on a guaranteed-NULL result. The result was that we could draw\nmistaken conclusions about whether outer joins can be simplified, if\nthe decision hinged on a non-top-level ScalarArrayOpExpr with a\npotentially-empty array argument.\n\nI believe this error dates to commit 72a070a36, which taught\nfind_nonnullable_rels to descend into non-top-level parts of qual\nexpressions. is_strict_saop (added earlier by 72153c058) already had\nenough intelligence to do the case correctly, but it wasn't passed the\nproper flag, ie \"top_level\" needs to be passed for \"falseOK\".\ne006a24ad copied that mistake into find_nonnullable_vars.\n\nLater, over-eager refactoring in commit 2f153ddfd broke\ncontain_nonstrict_functions' handling of ScalarArrayOpExpr by treating\nit as though it were no different from an OpExpr. It is, because\nwe must also prove the array is non-empty before concluding that the\nexpression is strict. This could result in misclassifying an\nexpression as strict when it is not, leading to assorted planning\nmistakes such as inlining a SQL function that shouldn't be inlined.\nWe can almost fix this by just re-adding the previous handling of\nScalarArrayOpExpr in that function, but doing only that would lead to\nalso calling check_functions_in_node() and thus redundantly checking\nthe operator's strictness. Avoid that by turning the if-series into\nan else-if chain, as it arguably should have been all along.\n\nThe reason these errors have escaped detection for decades is that\nthey are exposed only in arcane corner cases. ScalarArrayOpExpr with\nan empty array isn't typical usage, and even when that's possible\nseveral other conditions apply before the planner can reach a mistaken\nconclusion. While it's possible to build test cases demonstrating\nthese mistakes, I (tgl) judged t", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "cb7c0191675454ed81be73aef6c08596b89d9985", "date": "2026-07-28T22:14:21+02:00", "subject": "refactor(options): drop OptVal, use Object", "body": "Problem:\nThe object subsystem has an intermediate representation for no real\nreason. Besides the code cost, this also adds an extra (api <=> OptVal)\nconversion step, which is a (small) perf cost.\n\nSolution:\nWe already have `Object`, so use it instead.\n\n- drop `OptVal`, `OptValData`, `OptValType`, and related boilerplate.\n- add `kObjectTypeUnset`.", "style": "conventional", "type": "refactor", "scope": "options", "breaking": false} {"repo": "curl", "sha": "e2fa29c402b0662ff483a0ccec6b45eab5f7f7b0", "date": "2026-07-28T13:19:40-07:00", "subject": "CI: improve labeler matches", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "curl", "sha": "59dc2bbe07c3b5889e0b380e5fa384d338d9d24e", "date": "2026-07-28T22:36:11+02:00", "subject": "tidy-up: use `CURL_CSTRLEN()` macro on more static strings", "body": "Follow-up to e1450d8fdaf05f27ec75eb15df303fe931755a59 #22406\n\nCloses #22424", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "8186edf85dd35b27fbb2865176b6ac2e8947b49c", "date": "2026-07-28T22:44:50+02:00", "subject": "scripts/badwords.txt: do not recommend using 'will' in rewrites", "body": "Instead suggest \"avoid contraction and rewrite to present tense\"\n\nReported-by: claudex on github\nFixes #22422\nCloses #22423", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "9743720ab93d6f90ef359b3e1c8e55c1cbdbb729", "date": "2026-07-28T16:27:38-07:00", "subject": "tests: fix cert comparison with old cryptography", "body": "The fallback path for cryptography < 42 was broken by commit e13362c2\nthat caused a comparison between offset-naive and offset-aware\ndatetimes. Use the positional form of tz in datetime.now() everywhere.\n\nRef: #22396\n\nPointed out by Codex Security\nCloses #22426", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "dd50eb9145eead17ebd62db2e43a6de7c53102c0", "date": "2026-07-29T08:54:11+09:00", "subject": "Use more strlcpy() in two-phase transaction code", "body": "This commit replaces two calls of strcpy() and one call of strncpy() to\nuse strlcpy(), which are patterns that static analyzers (mostly LLMs, it\nseems) have been complaining regarding buffer overflow risks.\n\nThe existing calls are safe, here are more details for each one of them:\n- MarkAsPreparingGuts()'s strcpy() was guarded by MarkAsPreparing().\n- PrepareRedoAdd()'s strcpy() is safe because the record-level CRC check\nprevents corrupted data from reaching it unless intentionally\ncrafted. The replay code also assumes that the GID is within the allowed\nbounds, as WAL records are trusted.\n- Similarly, ParsePrepareRecord() stores its GID in a buffer bounded by\nGIDSIZE while trusting the length provided by the record.\n\nAs a result, these changes are purely cosmetic. They adopt a more\ndefensive coding style and should also silence some of the static\nanalysis reports received recently.\n\nAuthor: Matt Suiche \nDiscussion: https://postgr.es/m/CAGf6Lfx2kbQfcEnCi99V2i65JSWD6ij_E29F+UkY=TyMUyeG6A@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "b569f6e82b37854564a42966b5fb45f6470b19d2", "date": "2026-07-28T17:11:44-07:00", "subject": "test(compiler-cli): add compliance case for @HostListener on a property", "body": "`@HostListener` is not limited to methods — it is equally valid on a property\nholding a function, which is the idiomatic way to keep `this` bound:\n\n @HostListener('window:beforeunload', ['$event'])\n private onUnload = (event: BeforeUnloadEvent) => {...};\n\nEvery existing host-listener compliance case declares the handler as a method,\nso the property form was uncovered. This adds a case exercising both a public\nand a private function-valued property, one of them with a global (`window:`)\nevent target, and locks in the emitted chained `ɵɵlistener` calls plus\n`ɵɵresolveWindow`.\n\nVerified against all four compliance modes (full, partial/linked,\ndeclaration-only); GOLDEN_PARTIAL.js regenerated via the golden update rule.\n\n(cherry picked from commit d44b3224d960e6a0e3bdd72fa93761e78d47fff0)", "style": "conventional", "type": "test", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "d44b3224d960e6a0e3bdd72fa93761e78d47fff0", "date": "2026-07-28T17:11:44-07:00", "subject": "test(compiler-cli): add compliance case for @HostListener on a property", "body": "`@HostListener` is not limited to methods — it is equally valid on a property\nholding a function, which is the idiomatic way to keep `this` bound:\n\n @HostListener('window:beforeunload', ['$event'])\n private onUnload = (event: BeforeUnloadEvent) => {...};\n\nEvery existing host-listener compliance case declares the handler as a method,\nso the property form was uncovered. This adds a case exercising both a public\nand a private function-valued property, one of them with a global (`window:`)\nevent target, and locks in the emitted chained `ɵɵlistener` calls plus\n`ɵɵresolveWindow`.\n\nVerified against all four compliance modes (full, partial/linked,\ndeclaration-only); GOLDEN_PARTIAL.js regenerated via the golden update rule.", "style": "conventional", "type": "test", "scope": "compiler-cli", "breaking": false} {"repo": "ant-design", "sha": "7690200ffce9e2c7260047efaf3db017618b10c8", "date": "2026-07-29T12:13:50+08:00", "subject": "fix: handle Select clear icon overlap", "body": "* fix: hide Select suffix when clear icon appears\n\n* fix: avoid hiding suffix content when clear icon appears\n\n* fix: align Select clear suffix behavior with DatePicker\n\n* fix: align Select suffix transition easing\n\n* fix: handle Select clear icon edge cases\n\n* test(Select): cover clear suffix hover interactions\n\n* test(Select): group image tests by viewport\n\n---------", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c12c101b0846b1e6488f2dc986a852fbc6bf2e3b", "date": "2026-07-29T09:46:53+05:30", "subject": "Avoid accumulating relation locks during sequence synchronization.", "body": "While collecting the sequences to synchronize, the sequence sync worker\nopened each INIT sequence with RowExclusiveLock and held it until the\ntransaction committed. With many such sequences, this could exhaust the\nshared lock table and fail with \"out of shared memory\".\n\nThe worker only reads each sequence's identity (namespace and name) here\nand needs it to stay stable while read, for which AccessShareLock is\nenough, as it conflicts with the AccessExclusiveLock taken by DROP,\nRENAME, and SET SCHEMA. Take that lock instead and release it as soon as\nthe identity is read. The later synchronization re-opens each sequence, so\nit does not rely on the lock being retained.\n\nReported-by: Noah Misch \nAuthor: vignesh C \nReviewed-by: Hayato Kuroda \nReviewed-by: Amit Kapila \nBackpatch-through: 19, where it was introduced\nDiscussion: https://postgr.es/m/20260710045217.f0.noahmisch@microsoft.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "85b1df6ec17c6164ce604153bc27a6cdd39799c5", "date": "2026-07-29T00:36:40-04:00", "subject": "build(vim-patch): clientserver feature is N/A", "body": "Shared API, documented in remote.txt, is surface-level.\nAssume 0% compatiblity from Vim.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "angular", "sha": "cc4e6119b41a12b381b91ecccc31c5d306745c54", "date": "2026-07-29T06:21:19Z", "subject": "build: update dependency chalk to v6", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "vitest", "sha": "d6ceca9d7613b29ed577a952fcad6b186a2e0269", "date": "2026-07-29T14:25:09+08:00", "subject": "fix(spy): point ESM namespace spy error to module mocking docs", "body": "", "style": "conventional", "type": "fix", "scope": "spy", "breaking": false} {"repo": "vitest", "sha": "a1b0565c9783add9f197be8c77fa82554c4b9cb1", "date": "2026-07-29T11:56:48+05:30", "subject": "fix: prevent `vitest --typecheck` from reporting a false success when the `tsc` process crashes", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "13ec8f9ab2f7f5d6d1711e80c4373cfa742165d2", "date": "2026-07-29T15:01:57+08:00", "subject": "test: update snap", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "curl", "sha": "14cd071484d8abdf41b72d5d862d404381e880b9", "date": "2026-07-29T09:04:46+02:00", "subject": "HISTORY: add when c-ares support was introduced (2004)", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "ac2d46b423bb088dddc915e69588635af5cc770f", "date": "2026-07-29T16:32:29+09:00", "subject": "feat!: enable mocking Temporal without fake timers", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": true} {"repo": "vitest", "sha": "a7bf2b2c876f299b82edbe5118b903d6ee0a3c4b", "date": "2026-07-29T16:34:18+09:00", "subject": "fix: collect in-source tests when the module is cached (fix #10577)", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vitest", "sha": "46a3cf750b51c0f3b0226d0225675ed559882b80", "date": "2026-07-29T00:35:15-07:00", "subject": "fix(browser): ignore channel events in a tester opened as a top-level window (fix #9379)", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "ant-design", "sha": "50b8dcb4c66574ad0fb69f88a49c01af7e98abde", "date": "2026-07-29T15:36:49+08:00", "subject": "fix: snap", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vitest", "sha": "070edc6b32e3b7995e8204ecce1d81559155fe87", "date": "2026-07-29T09:47:36+02:00", "subject": "fix(browser): apply API options without browser config", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "ant-design", "sha": "f8595c7db9367aa0cf2968331a6791dc0ea3ade2", "date": "2026-07-29T16:14:22+08:00", "subject": "fix: update", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "angular", "sha": "420f7c4392a6d58ab579edb4c845139ad9a7eb10", "date": "2026-07-29T11:32:28+03:00", "subject": "docs: invoke signal inputs in adev examples and tutorials", "body": "Several adev example and tutorial files read a signal input as a bare\nreference (this.foo) instead of invoking it (this.foo()). Because an\nInputSignal is a function object, the bare reference is always truthy\nand never yields the underlying value, so the surrounding guard or\nbinding silently did the wrong thing:\n\n- animations open-close(.1/.3): the `!this.logging` guard in\n onAnimationEvent was always false, so the early return never fired.\n- form-validation forbidden-name.directive: the `this.forbiddenName`\n ternary condition was always truthy, so validation ran even when no\n forbidden name was configured.\n- first-app steps 12 and 14 housing-location: `housingLocation.photo`\n read `.photo` off the signal function (undefined), leaving the\n listing image src empty.\n\nInvoke the signals so the examples reflect correct signal-input usage.\n\n(cherry picked from commit 0d26130a4ce34daf9fb62bb19984f263492a219f)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "0d26130a4ce34daf9fb62bb19984f263492a219f", "date": "2026-07-29T11:32:28+03:00", "subject": "docs: invoke signal inputs in adev examples and tutorials", "body": "Several adev example and tutorial files read a signal input as a bare\nreference (this.foo) instead of invoking it (this.foo()). Because an\nInputSignal is a function object, the bare reference is always truthy\nand never yields the underlying value, so the surrounding guard or\nbinding silently did the wrong thing:\n\n- animations open-close(.1/.3): the `!this.logging` guard in\n onAnimationEvent was always false, so the early return never fired.\n- form-validation forbidden-name.directive: the `this.forbiddenName`\n ternary condition was always truthy, so validation ran even when no\n forbidden name was configured.\n- first-app steps 12 and 14 housing-location: `housingLocation.photo`\n read `.photo` off the signal function (undefined), leaving the\n listing image src empty.\n\nInvoke the signals so the examples reflect correct signal-input usage.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "16c3435794befad6499449cbd73696f6d3ca8e24", "date": "2026-07-29T17:39:40+09:00", "subject": "Protect PGPROC lookup when terminating background workers", "body": "TerminateBackgroundWorkersForDatabase() uses BackendPidGetProc() and,\nuntil now, accessed fields of the returned PGPROC after releasing\nProcArrayLock, including its database OID. If the PGPROC slot is\nrecycled during this window, the database OID being checked may belong\nto a different backend, causing an unrelated background worker to be\nterminated.\n\nTriggering this bug requires a very narrow race: the background worker\nidentified by BackendPidGetProc() must exit, its PGPROC slot must be\nreleased and reused, and only then must\nTerminateBackgroundWorkersForDatabase() examine the database OID.\n\nTerminateBackgroundWorkersForDatabase() holds BackgroundWorkerLock,\npreventing parallel workers and dynamically registered workers (such as\nthose created by worker_spi) from reusing the slot. As far as I know,\nthe only plausible scenario is a static background worker that exits and\nis restarted quickly enough to reuse the same PGPROC slot within the\nrace window. In practice, this race is extremely unlikely, still\nreachable in theory.\n\nOversight in f1e251be80a0.\n\nAuthor: Chao Li \nReviewed-by: Aya Iwata \nReviewed-by: Haibo Yan \nDiscussion: https://postgr.es/m/78E81763-EA1D-4788-9741-4092BCB997A5@gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "6f41dac8983cbab517413d4200fde232828eb576", "date": "2026-07-29T08:58:33Z", "subject": "Fix #1 -> #{1} in tmux1, from Rasmus Thystrup Karstensen.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c7fb986b72e14d702172a789111502e34123b305", "date": "2026-07-29T11:25:07+02:00", "subject": "GHA/configure-vs-cmake: dump `./configure --help`", "body": "For visual review when necessary. For the three major platforms for\nsymmetry and good measure.\n\nCloses #22427", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "795246bd9ebb0cf63294c4e8a5606435a1b0edd4", "date": "2026-07-29T17:29:18+08:00", "subject": "test: align snapshots and stabilize Typography tooltip test", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "8f55f60fa168e3154973c7aea141eecffd59cdd4", "date": "2026-07-29T17:42:54+08:00", "subject": "fix(Slider): prevent duplicate focus and blur callbacks", "body": "", "style": "conventional", "type": "fix", "scope": "Slider", "breaking": false} {"repo": "tmux", "sha": "d6c37ce314667c71c6c80cd194706daabf62ae0b", "date": "2026-07-29T11:10:49+01:00", "subject": "Note SIXEL at end of configure as well.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "27b6fe740dc85179552c2d719945b7596b7fb9d1", "date": "2026-07-29T18:32:34+08:00", "subject": "perf(Typography): defer native ellipsis measurement until hover", "body": "", "style": "conventional", "type": "perf", "scope": "Typography", "breaking": false} {"repo": "ant-design", "sha": "4924d63c350e2975c18483c7ed20b8e906fb3ea3", "date": "2026-07-29T13:33:12+03:00", "subject": "fix: mask OTP values when mask is true", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9fa2c1ebd1757db6032cf81b53f973941be77747", "date": "2026-07-29T12:37:02+02:00", "subject": "doc: Add getdatabaseencoding to function docs", "body": "The getdatabaseencoding function was added in bf00bbb0c494 in 1998 but\nwas never documented. While mostly used in tests, there is no reason\nnot to document it as this function isn't going anywhere and is already\nused in extensions.\n\nAuthor: Ian Barwick \nReviewed-by: Thom Brown \nReviewed-by: surya poondla \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CAB8KJ=ij+pznQGub=DkyJuKL=tC=Q=07qSahTyw7TLb0DdNJsg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "182b02afb0402c0473323106ed883cb9a9feb75c", "date": "2026-07-29T18:51:44+08:00", "subject": "fix(insexpand): use String cleanup for compl_leader", "body": "Problem:\ncompl_leader is a String, but it is cleared with XFREE_CLEAR().\n\nSolution:\nReplace XFREE_CLEAR() with API_CLEAR_STRING().", "style": "conventional", "type": "fix", "scope": "insexpand", "breaking": false} {"repo": "curl", "sha": "6e130eb484bf3d974fbcbc704db051c868e35d7f", "date": "2026-07-29T13:11:20+02:00", "subject": "apple-fast-udp: fix sendmsg_x partial results", "body": "When sending with sendmsg_x(), fix handling of last gso chunk being\nsmaller. Handle partial results correctly. Ignore SOCKEMSGSIZE by\nreporting success which drops PMTUD probes into the void.\n\nCloses #22429", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "ef15676abb39704129cb547f07d5f044837868b3", "date": "2026-07-29T13:12:46+02:00", "subject": "EXPERIMENTAL: Apple fast UDP", "body": "Follow-up to 079a11bcba5858ffab2cf6e810e9d980\n\nCloses #22428", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "f76f79696e12f0334275f394f24344f69b6739f7", "date": "2026-07-29T13:21:52+02:00", "subject": "EXPERIMENTAL: cleanups, unify on titles, merge quiche into a single segment", "body": "Closes #22428", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a3b85b9a4d14dad67258313acad1fd1650ff74ad", "date": "2026-07-29T07:52:28-04:00", "subject": "build(vim-patch): ':vim9' is n/a", "body": "1. Vim9 error messages, ex commands, are N/A.\n2. \":scriptversion\" is N/A:\n3. Most `#.*define.*ex_ni$` are N/A\n because `FEAT_` compile guards are N/A.\n\n----\n\nTarget patch: v8.2.2238\n\nActual patch list:\n\n```\nvim-patch:8.2.4240: error for using flatten() in Vim9 script is unclear\nvim-patch:9.0.0170: various minor code formatting issues\nvim-patch:9.1.0356: MS-Windows: --remote may change working directory\nvim-patch:9.2.0251: Link error when building without channel feature\nvim-patch:9.2.0714: Coverity warns for NULL deref\nvim-patch:9.2.0818: tests: client-server test fails without X11 server\nvim-patch:9.2.0868: GTK: Window Manager hint prevents giving focus to dialog\n```\n\n----\n\nIgnore diff blanklines via regex (\"-I\"), not \"--ignore-blank-lines\".\nAlmost useless.\nI tried with/without \"--word-diff=porcelain\", \"--word-diff-regex\",\nand diff algorithms like `--histogram`.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "angular", "sha": "bbbd357bd2b29f0b04578e3fa651418df4aefc67", "date": "2026-07-29T13:56:21+02:00", "subject": "fix(forms): add utility to assert that value is a field tree", "body": "Adds the `isFieldTree` utility that allows users to assert whether a value is a field tree. This is something that has come up on Material recently and will be useful for users as well.\n\nFixes #69984.\n\n(cherry picked from commit 2a141847a5ea2f24a98d5cc672df4e7abe67ec2d)", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "2a141847a5ea2f24a98d5cc672df4e7abe67ec2d", "date": "2026-07-29T13:56:21+02:00", "subject": "fix(forms): add utility to assert that value is a field tree", "body": "Adds the `isFieldTree` utility that allows users to assert whether a value is a field tree. This is something that has come up on Material recently and will be useful for users as well.\n\nFixes #69984.", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "curl", "sha": "172c16b58ccf8bff2122dcde7ea48967198308f9", "date": "2026-07-29T13:59:31+02:00", "subject": "docs/INTERNALS.md -> docs/DEPENDENCIES.md", "body": "With some minor cleanups\n\nCloses #22430", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d6a255a4704ea53c7a3aba8fa3d315a0bbf440bc", "date": "2026-07-29T18:12:22+05:30", "subject": "Add CHANGES.md entry for AES-CBC AVX-512/VAES decryption", "body": "Document the AVX-512 and VAES optimizations for AES-CBC decryption,\nwhich improve decryption performance for large inputs (1024 bytes or\nmore) by 3.5x to 3.8x.\n\nReviewed-by: Simo Sorce \nReviewed-by: Andrew Dinh \nMergeDate: Fri Jul 31 16:33:42 2026\n(Merged from https://github.com/openssl/openssl/pull/30902)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "62fe048fbe324faaf8e11ea9ce3c76be8a085ae7", "date": "2026-07-29T18:12:22+05:30", "subject": "Add AVX-512/VAES optimized AES-CBC decryption", "body": "Implement an AVX-512 VAES-based bulk path for AES-CBC decryption that\nprocesses multiple blocks in parallel, improving decryption performance\nfor large inputs (1024 bytes or more) by 3.5x to 3.8x.\n\n- Add the VAES CBC decrypt intrinsic implementation\n (crypto/aes/aes_cbc_vaes_intrinsic.c) and wire it into the AES-NI\n provider cipher path.\n- Centralize the VAES_CBC_ELIGIBLE eligibility guard in\n include/crypto/aes_platform.h to keep the header and implementation in\n sync, and drop the redundant AES_BLOCK_SIZE define.\n- Guard the intrinsics with refined gcc, clang, and msvc compiler-version\n checks, ensuring clean builds in no-asm, no-deprecated, and Windows\n (shared/minimal) configurations.\n- Add an EVP-level AES-CBC decrypt test exercising the VAES bulk path\n across AES-128/192/256 and a range of input lengths.\n\nReviewed-by: Simo Sorce \nReviewed-by: Andrew Dinh \nMergeDate: Fri Jul 31 16:33:35 2026\n(Merged from https://github.com/openssl/openssl/pull/30902)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "c9389cc65e44b6cd60e56e2a7db1c4d2721501d5", "date": "2026-07-29T20:49:22+08:00", "subject": "fix(compiler-sfc): avoid walking cached type references", "body": "close #15174", "style": "conventional", "type": "fix", "scope": "compiler-sfc", "breaking": false} {"repo": "neovim", "sha": "7b34bca2a6579c4dc8b8767fbfee5f7d98c977d4", "date": "2026-07-29T13:03:53Z", "subject": "docs(health): avoid warnings for dynamic filetypes #41036", "body": "Problem:\nDynamic filetypes defined by `vim.filetype.add` may show health warnings.\n\nSolution:\nDocument a way to suppress these harmless warnings.", "style": "conventional", "type": "docs", "scope": "health", "breaking": false} {"repo": "neovim", "sha": "1a755e4890c08542d9a93301b19d317f971f19e2", "date": "2026-07-29T15:05:19+02:00", "subject": "fix(options): latent codegen bug", "body": "Problem:\nLatent bug from 7f6c14ed54a46e822b80134d39a89f97523a2706 (2015).\nIf an option default is \"false\" (`o.defaults.if_false = false`, e.g.\n'fileignorecase'), codegen does not give it a `.def_val` on the platform\nwhere its `#if` condition is undefined; it is zero-initialized.\n\nThis wasn't noticed until the parent commit, where zero value is\nkObjectTypeNil, which `set_option_varp()` rejects.\n\nSolution:\nCheck `if_false == nil` insteada of \"falsey\", so the `#else` branch gets\ngenerated.", "style": "conventional", "type": "fix", "scope": "options", "breaking": false} {"repo": "vitest", "sha": "41f551bce47a6d9309000daf80b831c495cadf2e", "date": "2026-07-29T23:13:27+10:00", "subject": "feat(types)!: add better promise support in expects and matchers", "body": "", "style": "conventional", "type": "feat", "scope": "types", "breaking": true} {"repo": "openssl", "sha": "9da88ee430a0fd7312fc1390bc4a60809361de76", "date": "2026-07-29T15:51:54+02:00", "subject": "Add X509V3_EXT_add_nconf fuzzer", "body": "Add a fuzz target for parsing X509v3 configuration\nwith both empty and non-empty context.\n\nInspired by : https://github.com/google/boringssl/commit/33b569282ca124c81d9ba74df696a013cb9a80ae#diff-69cdea0423abe6dd8db9db93ce916f9576bf86988b2728ffa1a2178d5d94734b\n\nAssisted-by: ChatGPT:gpt-5.6\nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:50:44 2026\nMerged-from: https://github.com/openssl/openssl/pull/32143", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "415d265728d044a3b7e83cb1fcdae772e08a0fd3", "date": "2026-07-29T14:06:32Z", "subject": "Add a flag to turn off scrollbars for display-panes; GitHub issue 5440 from David Marte.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "b114b7e089d4e85fb50d218cd9f7db83c5a07f6c", "date": "2026-07-29T15:11:38+01:00", "subject": "Update tmux.1.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "57cae40ae5cfbdc928953c074345ef747401be70", "date": "2026-07-29T16:32:40+02:00", "subject": "sws: log the exact closing reason better, to help debugging tests", "body": "Closes #22431", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "202e27f3d66dde86dd5477cf3c1a8b4b2de6448d", "date": "2026-07-29T16:43:51+02:00", "subject": "fix(runner): don't fail collection when accessing `error.stack` throws", "body": "", "style": "conventional", "type": "fix", "scope": "runner", "breaking": false} {"repo": "curl", "sha": "64f746ee1e71f1079424cc173b7ec9d93adbb169", "date": "2026-07-29T17:09:06+02:00", "subject": "urlapi: clear password buffer on error path", "body": "Reported by Copilot\nBug: https://github.com/curl/curl/pull/21637#pullrequestreview-4809702512\nFollow-up to 112a8b5adf36c17e7816a8db701b3e7a22958e52 #21637\nFollow-up to 7c34365ccea19949317878c7fcd5f7376e2e09f1 #21879\n\nCloses #22432", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5fb3c638910cdac2e00861e6f6806d9e03e5831d", "date": "2026-07-29T17:15:45+02:00", "subject": "Fix cascading standby reconnect failure after archive fallback", "body": "A cascading standby could fail to reconnect to its upstream standby with\n\"requested starting point ... is ahead of the WAL flush position\" after\nfalling back to archive recovery. This happened because archive\nrecovery processes whole segment files, so after replaying a segment the\ncascade's next read position lands at the start of the following\nsegment, which is ahead of the upstream's flush position reported by\nGetStandbyFlushRecPtr() (still inside the just-replayed segment).\n\nFix by having the walreceiver check the upstream's current WAL flush\nposition via IDENTIFY_SYSTEM before issuing START_REPLICATION.\nIDENTIFY_SYSTEM already returns this position (as xlogpos), but\nwalrcv_identify_system() previously discarded it; now we have a use for\nit. If the requested start point exceeds the upstream's flush position\non the same timeline, the walreceiver waits for\nwal_retrieve_retry_interval and retries.\n\nThe wait is limited to gaps of at most one WAL segment, which is the\nexpected case from the segment-granularity of archive recovery. Larger\ngaps indicate the upstream is genuinely behind, so START_REPLICATION is\nallowed to proceed (and fail) normally, letting the startup process fall\nback to other WAL sources. The first wait is logged at LOG level;\nsubsequent waits are demoted to DEBUG1 to avoid log noise. The\nwalreceiver honors wal_receiver_timeout during the wait, so it will exit\nif the upstream doesn't catch up in time.\n\nTo preserve ABI compatibility on back branches, the flush position from\nIDENTIFY_SYSTEM is communicated via a new global variable\n(WalRcvIdentifySystemLsn) rather than changing the signature of\nwalrcv_identify_system().\n\nThe bug was introduced in Postgres 9.3 by commit abfd192b1b5b, which\nadded a flush-position check in StartReplication() that rejects requests\nahead of the upstream server's WAL flush position.\n\nAuthor: Marco Nenciarini \nReviewed-by: Xuneng Zhou \nBackpatch-through: 14\nDiscuss", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "23a6e25593729feaafb6f3491a1222247f6e0125", "date": "2026-07-29T11:17:32-04:00", "subject": "Fix autovacuum-induced flakiness in backwards scan test.", "body": "Two of the permutations in backwards-scan-concurrent-splits rely on\ntheir VACUUM step deleting the leaf pages that the waiting backwards\nscan will have to recover from. VACUUM can only do that when it's able\nto remove the index tuples whose heap tuples the concurrent session just\ndeleted. An autovacuum worker holding a snapshot holds back the\nremovable cutoff, which leaves the pages non-empty, and so undeleted,\ncausing the test to fail spuriously.\n\nTo fix, wait for the removable cutoff to advance past the deletions\nbefore the scan acquires its snapshot. This is much like commit\n1c64d2fc, which dealt with the same hazard in nbtree_half_dead_pages by\nadding the wait_prunable() helper that we reuse here.\n\nOversight in commit e395fbd3.\n\nAuthor: Peter Geoghegan \nReported-by: Alexander Lakhin \nDiscussion: https://postgr.es/m/b61d9944-d7a3-45f3-b69a-f18c8bfbbbd0@gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "5496765f971f4e75c6d9589d3e09764a36e6e0ef", "date": "2026-07-29T10:37:14-05:00", "subject": "docs: Adds HTTP communication guidance to Angular Skills", "body": "(cherry picked from commit 840f071566a0c5e44048051135da329e4e3ecaef)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "840f071566a0c5e44048051135da329e4e3ecaef", "date": "2026-07-29T10:37:14-05:00", "subject": "docs: Adds HTTP communication guidance to Angular Skills", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "184c4797b8ce8c0640657981a0f3bea2e89f5828", "date": "2026-07-29T18:38:39+03:00", "subject": "refactor(devtools): rename timing API to performance track", "body": "Rename Timing API to Performance Track to better reflect the purpose of the actual feature;\nMigrate the settings data object to match the new name and drop some redundant keys.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "angular", "sha": "ec16a3d6c6785992197bc0ae81afb1e699232f64", "date": "2026-07-29T10:40:09-05:00", "subject": "fix(http): enable xsrf for root-provided HttpClient", "body": "Include the XSRF interceptor in the root token factory so the automatically provided HttpClient retains the documented default protection without requiring provideHttpClient().\n\n(cherry picked from commit de240a5d0eecdc0c895f581bf3f190f48ad68c18)", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "angular", "sha": "de240a5d0eecdc0c895f581bf3f190f48ad68c18", "date": "2026-07-29T10:40:09-05:00", "subject": "fix(http): enable xsrf for root-provided HttpClient", "body": "Include the XSRF interceptor in the root token factory so the automatically provided HttpClient retains the documented default protection without requiring provideHttpClient().", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "curl", "sha": "9951a2247e5b18a3c961c25ba15cb39f1be11522", "date": "2026-07-29T17:58:08+02:00", "subject": "tool: remove duplicate setopts", "body": "CURLOPT_RESOLVE and CURLOPT_CONNECT_TO were set twice.\nA happy little accident?\n\nCloses #22433", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "1dbc766fa71edee8d040815c5621b94c8ecdf68f", "date": "2026-07-29T09:40:08-07:00", "subject": "fix(ui2): clear empty cmdline after backspace #41043", "body": "Problem:\nCmdline area shows stale \":\" after backspacing out of the command line.\n\nSolution:\nClear the command line for empty commands. Note that `:` will now\nclear the command line too.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "postgres", "sha": "4ee0ccfd55f973160a2ab013bc96488715221270", "date": "2026-07-29T09:52:10-07:00", "subject": "Fix stale comment in parallel_vacuum_main().", "body": "The comment claimed that a parallel vacuum worker has only the\nPROC_IN_VACUUM flag because parallel vacuum is not supported for\nautovacuum, but commit 1ff3180ca01 allowed autovacuum to use parallel\nvacuum workers.\n\nThe assertion itself still holds: the leader, whether a backend\nrunning VACUUM or an autovacuum worker, sets PROC_IN_VACUUM before\ntaking its snapshot, and a parallel worker inherits the flag when\nimporting the leader's snapshot. The leader's other flags don't reach\nthe worker, since the snapshot import copies only the PROC_XMIN_FLAGS\nbits and PROC_IS_AUTOVACUUM is never set on parallel workers, which\nrun as regular background workers. Reword the comment to explain that.\n\nOversight in commit 1ff3180ca01.\n\nAuthor: Bharath Rupireddy \nReviewed-by: Masahiko Sawada \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/CALj2ACVwQ4WABqq8Lnf+VZEJ45jcTFhyFLFr_ctfS4=QLL-r5w@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e29af9c32f8be5767924b79612ce0f73877d2263", "date": "2026-07-29T11:54:43-05:00", "subject": "refactor(zip)!: builtin zip.lua plugin #40846", "body": "Problem:\nThe bundled `zip` plugin is implemented in Vimscript, making it harder to\nmaintain and build on with Nvim's Lua runtime infrastructure.\n\nSolution:\nAdd an opt-out `zip.lua` browser backed by `nvim.dir`, and disable the legacy\nVimscript implementation by moving it to `pack/dist/opt/zip/`. Load it with\n`:packadd zip`.", "style": "conventional", "type": "refactor", "scope": "zip", "breaking": true} {"repo": "neovim", "sha": "726d1a92d243a8b7cb4b221d64e4523b7a89aaf4", "date": "2026-07-29T19:09:48+02:00", "subject": "feat(options): lua closure/function options", "body": "Problem:\nCannot assign Lua functions/closures to \"func\" ('completefunc',\n'tagfun', …) or \"expr\" ('foldexpr', 'indentexpr', …) options.\n\nSolution:\n- Store \"func\"/\"expr\" options as `Callback` instead of string.\n- Delete oceans of copy-pasted code.\n- BREAKING: LuaRef returned via RPC/Vimscript is now represented as\n `\"\"` (like what `:map` shows) instead of `nil`.\n- Note: `man.vim` still uses `v:lua` string, bc it's a vimscript ftplugin.", "style": "conventional", "type": "feat", "scope": "options", "breaking": false} {"repo": "axios", "sha": "712e02373f1c5ac72fdf5da16f04838338ba817e", "date": "2026-07-29T19:39:19+02:00", "subject": "fix(interceptors): trim trailing ejected handlers", "body": "", "style": "conventional", "type": "fix", "scope": "interceptors", "breaking": false} {"repo": "tmux", "sha": "44b8a40b8c8dea2fb938b38263844986b1deb6d1", "date": "2026-07-29T17:42:56Z", "subject": "Correctly skip padding at end of line, GitHub issue 5411.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "axios", "sha": "a75bf44647e5132e39b778730394b0a2b983265c", "date": "2026-07-29T10:58:27-07:00", "subject": "fix(core): tolerate non-string stack when decorating request errors", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "neovim", "sha": "27e196c7d62797b79f5b2cdbcb6cae03f41bb7b3", "date": "2026-07-29T11:19:17-07:00", "subject": "backport: fix(ui2): clear empty cmdline after backspace", "body": "Problem:\nCmdline area shows stale \":\" after backspacing out of the command line.\n\nSolution:\nClear the command line for empty commands. Note that `:` will now\nclear the command line too.\n\n(cherry picked from commit 1dbc766fa71edee8d040815c5621b94c8ecdf68f)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "bc90a53455836a61c4282073866325d3c14954d5", "date": "2026-07-29T14:20:35-04:00", "subject": "doc: remove added space within synopsis replaceable tags", "body": "Restructuring the tags makes the output consistent and doesn't require\nadded spaces.\n\nReported-by: Peter Smith\n\nAuthor: Peter Smith\n\nDiscussion: https://postgr.es/m/CAHut+Pu8JahGm76CMdpzH350pHJedA4R2b8JmOim3+m3yxft3Q@mail.gmail.com\n\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "axios", "sha": "529ce70296334a0397f9e45e7dfa5658f6b6fdba", "date": "2026-07-29T23:51:43+05:30", "subject": "docs: add missing fs import to README stream example", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9e526a14d7cb36c24d6faebf32df31251a97edc9", "date": "2026-07-29T21:13:25+02:00", "subject": "ssl: Remove static var tracking tls_init_hook warnings", "body": "If the TLS init hook is defined in conjunction with ssl_sni we\nissue a warning to help the user re-configure the cluster. To\navoid drowning the log in warnings, we log only once instead of\nonce per host. This removes the static variable tracking the\nwarning to aid future multithreading efforts.\n\nAuthor: Daniel Gustafsson \nReviewed-by: Tristan Partin \nReviewed-by: Andreas Karlsson \nReviewed-by: Yilin Zhang \nDiscussion: https://postgr.es/m/68B9881D-DAA8-467D-A251-C96E98E57BA0@yesql.se", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "05d6b5afdf0fbd0b114534081f05f072584bec8c", "date": "2026-07-29T21:14:08+02:00", "subject": "ssl: Use TLS_method instead of deprecated SSLv23_method", "body": "The SSLv23_method() function has been an alias for TLS_method since\n2015 (OpenSSL commit 32ec41539b5b) so we should use the appropriate\nname to avoid confusion.\n\nAuthor: Daniel Gustafsson \nReviewed-by: Tristan Partin \nReviewed-by: Andreas Karlsson \nReviewed-by: Yilin Zhang \nDiscussion: https://postgr.es/m/68B9881D-DAA8-467D-A251-C96E98E57BA0@yesql.se", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "54dee94e6a71817877dfdeb2c2c811330e58e754", "date": "2026-07-29T21:14:18+02:00", "subject": "ssl: Replace deprecated API to get commonName", "body": "X509_NAME_get_text_by_NID was deprecated in OpenSSL 4.0.0, and could\nbe removed in a future version of OpenSSL. The replacement APIs are\navailable in all versions of OpenSSL and LibreSSL that we support so\nwe can easily change to make the code future proof.\n\nThe reason for the deprecation is that X509_NAME_get_text_by_NID can\nonly grab the first entry in a list, and doesn't handle multibyte\nstrings well. The fix is to get the index of the name entry with\nX509_NAME_get_index_by_NID and use X509_NAME_get_entry to extract\nthe data.\n\nAuthor: Daniel Gustafsson \nReviewed-by: Tristan Partin \nReviewed-by: Andreas Karlsson \nReviewed-by: Yilin Zhang \nDiscussion: https://postgr.es/m/68B9881D-DAA8-467D-A251-C96E98E57BA0@yesql.se", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1ce49fab6026ba53dbe29576a53e67fe9e1557f7", "date": "2026-07-29T21:14:31+02:00", "subject": "ssl: Use the correct feature macros for TLS protocol support", "body": "Our test for if the underlying TLS library supported a specific\nversion tested against the TLSX_Y_VERSION set of macros. These\nare however always defined, regardless of if the library was\nbuilt without support for the specific protocol version. Fix\nby using the feature test macros OPENSSL_NO_TLSX_Y which are\nintended for this usecase.\n\nThe previous coding held no risk of protocol downgrade against\nthe underlying library, a library not supporting the protocol\nversion selected would simply error out as the feature isn't\navailable. This can be easily verified using a modern version\nof LibreSSL, which in version 3.8 disabled TLS1 and 1.1 by\ndefault. Once we bump our minimum supported version of LibreSSL\nto 3.8+ we can add a test for this.\n\nAuthor: Daniel Gustafsson \nReviewed-by: Tristan Partin \nReviewed-by: Andreas Karlsson \nReviewed-by: Yilin Zhang \nDiscussion: https://postgr.es/m/68B9881D-DAA8-467D-A251-C96E98E57BA0@yesql.se", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "dae6efed9e3713e281312697b326868d95fb358c", "date": "2026-07-30T03:19:01+08:00", "subject": "fix(Input): handle OTP mask input types consistently", "body": "* fix(Input): correct OTP mask input type behavior\n\n* fix(Input): preserve OTP mask default type\n\n* test(Input): update OTP mask demo snapshot\n\n* test(Input): update OTP extended demo snapshot", "style": "conventional", "type": "fix", "scope": "Input", "breaking": false} {"repo": "angular", "sha": "949de9c1fab164ebd81a792c31a6b36acf7059a2", "date": "2026-07-29T14:29:53-05:00", "subject": "refactor(forms): modernize reactive forms tests to rely on `whenStable`", "body": "Rely on zoneless scheduling throughout reactive forms tests instead of triggering change detection manually.\n\n(cherry picked from commit 6deb6afcd1af81357a146b51cdea109429af315a)", "style": "conventional", "type": "refactor", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "6deb6afcd1af81357a146b51cdea109429af315a", "date": "2026-07-29T14:29:53-05:00", "subject": "refactor(forms): modernize reactive forms tests to rely on `whenStable`", "body": "Rely on zoneless scheduling throughout reactive forms tests instead of triggering change detection manually.", "style": "conventional", "type": "refactor", "scope": "forms", "breaking": false} {"repo": "postgres", "sha": "01a6119af074a3d346e3f749df7f63cd45acbaaf", "date": "2026-07-29T21:51:46+02:00", "subject": "doc: Add a note that refint will be removed in v20", "body": "refint has been removed from the spi contrib module in v20. Add a note\nto the documentation of the still-supported back branches so that users\nare aware the module is going away.\n\nAuthor: Ayush Tiwari \nReported-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CAJTYsWUHq8Ohc6-N-xamOPYz-q3qUYMtwQX-1=Zi=5N1Q_GSEQ@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "3fdb74748af936c46d883ebc68ac3735e92fa802", "date": "2026-07-29T14:56:26-05:00", "subject": "fix: review", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b43e8f4b3dc5b361e78f237b136c9c1f817e13c4", "date": "2026-07-29T22:02:57+02:00", "subject": "pgindent fix for a84eca6627f", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "265fb7b4ea386be4b2b3cd226113c74a370e8246", "date": "2026-07-29T21:18:38+01:00", "subject": "Test for vi keys and padding.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "8dfb7d029488ec6f6bee7ba47e317b2de78712ab", "date": "2026-07-29T20:43:20Z", "subject": "Fix unzoomed width when scrollbars are hidden.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "458d045bd258cb3ec31cdde8088e758be2bb519e", "date": "2026-07-29T22:10:19+01:00", "subject": "With collapsible output active, line-numbers-on now shows each visible line’s original buffer line number, so hidden lines remain counted.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "ddb22d12c2a3336b33b3e2c9b7fc0da5b893de49", "date": "2026-07-29T23:30:51+02:00", "subject": "lib5004: fix memleak on OOM, check all slist append results (httpsig)", "body": "Detected by torture tests:\n```\ntest 5004...[HTTP RFC 9421 B.2.6: Ed25519 POST with headers (RFC test vector)]\n 105 functions found, but only fail 25 (23.81%)\n** MEMORY FAILURE\nLeak detected: memory still allocated: 99 bytes\nAt 6000022c9408, there is 36 bytes.\n allocated by /Users/runner/work/curl/curl/lib/slist.c:87\nAt 6000039c8e78, there is 31 bytes.\n allocated by /Users/runner/work/curl/curl/lib/slist.c:87\nAt 6000037dd688, there is 16 bytes.\n allocated by /Users/runner/work/curl/curl/lib/slist.c:62\nAt 6000037dd628, there is 16 bytes.\n allocated by /Users/runner/work/curl/curl/lib/slist.c:62\nLIMIT /Users/runner/work/curl/curl/lib/slist.c:62 malloc reached memlimit\n 5004: torture FAILED: function number 10 in test.\n```\nRef: https://github.com/curl/curl/actions/runs/30497660391/job/90730128599?pr=22437#step:16:2331\n\nAlso:\n- enable HTTPSIG in torture tests.\n- NULL check all `curl_slist_append()` results.\n- apply a NULL check to sibling test 5000 also.\n\nCo-authored-by: Daniel Stenberg\nFollow-up to a55731050e8c3dbea0b96205cf916443118f6acb #22386 #21239\n\nCloses #22437", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0a7788b66fc63348215057bcd6633b3b51319027", "date": "2026-07-29T23:25:36+01:00", "subject": "Add Fold View menu item and O key to toggle fold view.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0330d8ac5767d030125312d9e5bcbc968e3a448c", "date": "2026-07-29T23:35:11+01:00", "subject": "Change line-numbers menu to On/Off.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "2c72fe3797ff92ead7a2531d32665fac370b4688", "date": "2026-07-29T22:40:33Z", "subject": "fix(core): allow readonly arrays in RawScopeInfoFromDecorator", "body": "When components pass arrays or readonly tuples to decorator metadata fields (such as ), typechecking generated decorator reflection metadata causes a TS2322 type mismatch error because previously only accepted mutable .\n\nThis change updates to accept , allowing arrays and readonly tuples to be assigned without TypeScript compilation errors.\n\n(cherry picked from commit eee9ef4d09c091e9066ed6074a19c0eee0e84da2)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "eee9ef4d09c091e9066ed6074a19c0eee0e84da2", "date": "2026-07-29T22:40:33Z", "subject": "fix(core): allow readonly arrays in RawScopeInfoFromDecorator", "body": "When components pass arrays or readonly tuples to decorator metadata fields (such as ), typechecking generated decorator reflection metadata causes a TS2322 type mismatch error because previously only accepted mutable .\n\nThis change updates to accept , allowing arrays and readonly tuples to be assigned without TypeScript compilation errors.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "afe529cb2d97e766b4f9014c3b6864420b7c8deb", "date": "2026-07-29T22:40:45Z", "subject": "fix(core): accept readonly arrays for setClassMetadata decorators", "body": "When reflection metadata is emitted via setClassMetadata, passing readonly arrays or const tuples for the decorators parameter causes TypeScript type checking errors because setClassMetadata previously expected decorators to be a mutable any[] or null.\n\nThis change updates the setClassMetadata type signature to accept decorators as readonly any[] or null and casts the parameter internally when mutating the class metadata property.\n\n(cherry picked from commit dc65e3656ff56fc6c5c03cf59991a6e8760b5279)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "dc65e3656ff56fc6c5c03cf59991a6e8760b5279", "date": "2026-07-29T22:40:45Z", "subject": "fix(core): accept readonly arrays for setClassMetadata decorators", "body": "When reflection metadata is emitted via setClassMetadata, passing readonly arrays or const tuples for the decorators parameter causes TypeScript type checking errors because setClassMetadata previously expected decorators to be a mutable any[] or null.\n\nThis change updates the setClassMetadata type signature to accept decorators as readonly any[] or null and casts the parameter internally when mutating the class metadata property.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "curl", "sha": "d3e74f127864777de76016a76d547592e756d479", "date": "2026-07-30T00:52:10+02:00", "subject": "GHA: set `HOMEBREW_NO_INSTALL_CLEANUP=1` where brew is used", "body": "To save work unnecessary in CI context, and to reduce log noise.\n\nCherry-picked from #22437\n\nCloses #22440", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c155cbc6380467487fcbca126366638149f04cf0", "date": "2026-07-30T01:46:15+02:00", "subject": "GHA: work around Homebrew `ca-certificates` install error", "body": "Working around:\n```\n==> Installing libngtcp2 dependency: ca-certificates\n==> Pouring ca-certificates--2026-07-16.all.bottle.1.tar.gz\nWarning: The post-install step did not complete successfully\n[...]\nError: Process completed with exit code 1.\n```\n\nRefs:\nhttps://github.com/Homebrew/homebrew-core/pull/295934\nhttps://github.com/Homebrew/brew/pull/23357\n\nBug: https://github.com/curl/curl/pull/22437#issuecomment-5124151490\n\nCloses #22438", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "7d5398f764ebed25da3bf75afc5384b2a7893399", "date": "2026-07-30T02:24:58+02:00", "subject": "build: fix HTTPSIG option for unsupported TLS backends", "body": "Show a warning and force-disable HTTPSIG when the TLS backend is not\nOpenSSL or wolfSSL. Before this patch this resulted in a mismatched\nfeature list in configure and `curl -V`.\n\nAlso enable HTTPSIG in more CI jobs to cover unsupported ones, Windows\ncompilers, clang-tidy, cmake.\n\nFollow-up to a55731050e8c3dbea0b96205cf916443118f6acb #22386 #21239\nCherry-picked from #22437\n\nCloses #22439", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3e74574a12a1f4672a49ad0e3b9bbde284e1148e", "date": "2026-07-30T09:24:59+09:00", "subject": "Merge event masks for duplicate fds in the RIO poll builder", "body": "ossl_rio_poll_builder_add_fd() documents duplicate registrations as\nequivalent to one registration with the logical OR of the requested\ndirections. The poll(2) backend instead cleared an existing entry,\nmaking duplicate registrations last-writer-wins. The select(2)\nbackend only adds bits to its fd_sets and already merges them.\n\nClear the mask only when populating a new or reusable slot so an\nexisting fd retains its previously requested events.\n\nFixes https://github.com/openssl/openssl/issues/32116\n\nAssisted-by: Codex:gpt-5.6-sol\n\nReviewed-by: Paul Yang \nReviewed-by: Jakub Zelenka \nMergeDate: Tue Aug 11 07:12:57 2026\n(Merged from https://github.com/openssl/openssl/pull/32117)\n\n(cherry picked from commit 363b604bffc131e2ccda48b89d2d3ebf5eb285ad)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "363b604bffc131e2ccda48b89d2d3ebf5eb285ad", "date": "2026-07-30T09:24:59+09:00", "subject": "Merge event masks for duplicate fds in the RIO poll builder", "body": "ossl_rio_poll_builder_add_fd() documents duplicate registrations as\nequivalent to one registration with the logical OR of the requested\ndirections. The poll(2) backend instead cleared an existing entry,\nmaking duplicate registrations last-writer-wins. The select(2)\nbackend only adds bits to its fd_sets and already merges them.\n\nClear the mask only when populating a new or reusable slot so an\nexisting fd retains its previously requested events.\n\nFixes https://github.com/openssl/openssl/issues/32116\n\nAssisted-by: Codex:gpt-5.6-sol\n\nReviewed-by: Paul Yang \nReviewed-by: Jakub Zelenka \nMergeDate: Tue Aug 11 07:12:57 2026\n(Merged from https://github.com/openssl/openssl/pull/32117)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b74ed29f897735d121eafe38ba59be2198a3bc26", "date": "2026-07-30T09:25:18+09:00", "subject": "test: cover duplicate fd event merging in RIO poll builder", "body": "Add a poll-backend unit test which registers the same fd first for\nreading and then for writing. Verify that the builder retains one\npollfd with both POLLIN and POLLOUT set.\n\nWithout the preceding fix, the test observes POLLOUT only. Wire the\ntest into the QUIC-gated test build and skip it on non-poll backends.\n\nAssisted-by: Codex:gpt-5.6-sol\n\nReviewed-by: Paul Yang \nReviewed-by: Jakub Zelenka \nMergeDate: Tue Aug 11 07:12:58 2026\n(Merged from https://github.com/openssl/openssl/pull/32117)\n\n(cherry picked from commit 258f06b1ef57ce147ff2b1da5977d9e9adc55be7)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "258f06b1ef57ce147ff2b1da5977d9e9adc55be7", "date": "2026-07-30T09:25:18+09:00", "subject": "test: cover duplicate fd event merging in RIO poll builder", "body": "Add a poll-backend unit test which registers the same fd first for\nreading and then for writing. Verify that the builder retains one\npollfd with both POLLIN and POLLOUT set.\n\nWithout the preceding fix, the test observes POLLOUT only. Wire the\ntest into the QUIC-gated test build and skip it on non-poll backends.\n\nAssisted-by: Codex:gpt-5.6-sol\n\nReviewed-by: Paul Yang \nReviewed-by: Jakub Zelenka \nMergeDate: Tue Aug 11 07:12:58 2026\n(Merged from https://github.com/openssl/openssl/pull/32117)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "8f191a79617d906e34d041382e6ecfbf78166a65", "date": "2026-07-29T20:25:54-04:00", "subject": "Fix SystemTap dtrace warning about smgr probe argument types.", "body": "Commits ca326e903 and 1f8c504e3 widened the byte-count arguments of\nsmgr__md__read__done and smgr__md__write__done to \"long long int\".\nSystemTap's dtrace(1) fails on that specific spelling and falls back\nwith a warning (misreported near the previous probe). Use ssize_t\nand size_t instead, matching the md.c call sites.\n\nRevise probes.d's note about which types are usable as probe\narguments: recommend using system-supplied type names (macOS dtrace\nrejects names like PostgreSQL's uint64), and call out \"long long int\"\nas a known SystemTap failure case rather than a wider failure mode.\n\nAlso, update the monitoring.sgml entries for these probes,\nwhich were missed by the prior commits.\n\nReported-by: Laurenz Albe \nAuthor: Andrey Rachitskiy \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/697d3c88568442cd637d8453d129f1bb14bbd2a8.camel@cybertec.at", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "1e11696f41a1b4a4a09dc6cfc3d188553c0047e8", "date": "2026-07-30T10:17:37+09:00", "subject": "fix(common): added test for #17398", "body": "", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "vue-core", "sha": "6173a401594a01d382919c17c73882816558b79b", "date": "2026-07-30T11:29:56+08:00", "subject": "fix(runtime-vapor): restore concurrent async hydration context", "body": "- preserve a shared baseline across overlapping async setup continuations\n- restore the baseline on each cleanup instead of restoring sibling state\n- add regression coverage for concurrent hydration and subsequent CSR mounts", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "534579cf56264dc04fc13fea59797e4086bdeaa5", "date": "2026-07-30T12:52:46+08:00", "subject": "fix(form): refresh status context values", "body": "", "style": "conventional", "type": "fix", "scope": "form", "breaking": false} {"repo": "ant-design", "sha": "87805449e90a458618b4954c8321bd81108aa38b", "date": "2026-07-30T12:53:09+08:00", "subject": "fix(table): sync controlled filter selections", "body": "", "style": "conventional", "type": "fix", "scope": "table", "breaking": false} {"repo": "angular", "sha": "84cd4da06c0c32928a7bb2d30dca6bb6129ca1f2", "date": "2026-07-30T04:54:27Z", "subject": "build: update dependency jsdom to v30", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "vite", "sha": "fa005d19af5d847931c6dbefc63841c137383e6c", "date": "2026-07-30T14:15:46+08:00", "subject": "fix(server): strip base in indexHtml module graph lookup", "body": "", "style": "conventional", "type": "fix", "scope": "server", "breaking": false} {"repo": "postgres", "sha": "02decf9a9abc76f04a8117aba524dbe561cb720e", "date": "2026-07-30T11:53:06+05:30", "subject": "Skip SUBSCRIPTION TABLE TOC entries with --no-subscriptions.", "body": "pg_dump in --binary-upgrade mode emits \"SUBSCRIPTION TABLE\" TOC entries to\npreserve pg_subscription_rel state across pg_upgrade. When such a dump\nwas restored with --no-subscriptions, _tocEntryRequired() skipped the\n\"SUBSCRIPTION\" entry but not the associated \"SUBSCRIPTION TABLE\" entries,\nso the restore would try to apply subscription-relation state for a\nsubscription that was never created.\n\nSkip \"SUBSCRIPTION TABLE\" entries as well when no_subscriptions is set.\n\nThis can happen when pg_subscription_rel has entries, the dump is taken\nwith --binary-upgrade, and it is restored with --no-subscriptions.\n\nReported-by: Hayato Kuroda \nAuthor: Hayato Kuroda \nReviewed-by: Shlok Kyal \nReviewed-by: Amit Kapila \nBackpatch-through: 17, where it was introduced\nDiscussion: https://postgr.es/m/OS9PR01MB121493DA4C1A7748B11A646D8F5C02@OS9PR01MB12149.jpnprd01.prod.outlook.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "axios", "sha": "c3f553c740ebf3dff5e22dae24e9caaafafddd2d", "date": "2026-07-30T09:24:04+03:00", "subject": "fix(xhr): reject navigation-canceled requests as ECONNABORTED", "body": "", "style": "conventional", "type": "fix", "scope": "xhr", "breaking": false} {"repo": "openssl", "sha": "933afbd64970ecdd9ccb4936dd3e4fcffc910851", "date": "2026-07-30T08:39:12+02:00", "subject": "include/openssl/macros.h: add whitespace after semicolon in OSSL_DEPRECATED_FOR", "body": "So it concatenates into a properly type set message.\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Bob Beck \nMerge-date: Wed Aug 19 10:35:53 2026\nMerged-from: https://github.com/openssl/openssl/pull/32119", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3f82746fb9e403477298b003adfc1e3f80cebbf7", "date": "2026-07-30T08:43:50+02:00", "subject": "include/openssl: harmonise formatting of OSSL_DEPRECATEDIN_._._FOR messages", "body": "Use lower case (as it is a continuation of a sentence after\na semicolon), no leading space (as it is added\nby OSSL_DEPRECATED_FOR()), and no terminating period (it is a message,\nalso, it mirrors OSSL_DEPRECATED() formatting). Also, shorten some\nmessages for brevity.\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Bob Beck \nMerge-date: Wed Aug 19 10:35:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32119", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "ee0e5ff8dea36e454826d0c9e948bd32dbb18bd9", "date": "2026-07-30T15:20:02+08:00", "subject": "fix(runtime-vapor): schedule async hydration anchors before updates", "body": "- queue structural anchors before updates triggered during async hydration\n- cover both async component and async setup hydration paths\n- add regression tests for beforeMount-triggered branch updates", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "28511528aea3cc319c72a59b559a0d3a06fe061b", "date": "2026-07-30T09:21:31+02:00", "subject": "Add a couple of commits to .git-blame-ignore-revs", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b1aeda3ec939c2867e5c3eb4ee7e1bae4768503e", "date": "2026-07-30T09:26:07+02:00", "subject": "pgindent fix for 4ee0ccfd", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a3eff66e313b89cf721ad70ea34c41b7425200c1", "date": "2026-07-30T02:34:26-05:00", "subject": "fix(tui): xterm default cursor style #41050", "body": "Problem:\nxterm doesn't reset to user configured default with the sequence `\\x1b[0 q`.\n\nSolution:\nRestore old behavior: if the reset cursor style terminfo entry is\nabsent, xterm's sequence should set the cursor to steady block, which is\nthe xterm default. In the near future, they may support another option\n(`\\x1b[7 q`) to reset to the user configured default.", "style": "conventional", "type": "fix", "scope": "tui", "breaking": false} {"repo": "openssl", "sha": "eb2d252133cbe504de1d64a77573974cb3498f60", "date": "2026-07-30T10:58:55+02:00", "subject": "Commentary justifying non-constant-time memcmp", "body": "Reviewed-by: Matt Caswell \nReviewed-by: Shane Lontis \nMergeDate: Fri Aug 7 13:35:40 2026\n(Merged from https://github.com/openssl/openssl/pull/32098)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e6c93b59966141b262432f703d5478ea8ab9a728", "date": "2026-07-30T04:50:39-05:00", "subject": "docs(tui): kitty keyboard protocol flags #41056", "body": "Nvim has sent CSI > 3 u since 4fb3b57a19, but tui.txt still documented\nCSI > 1 u.\n\n(cherry picked from commit c1a65ff353429a987630b926727616590a330539)", "style": "conventional", "type": "docs", "scope": "tui", "breaking": false} {"repo": "neovim", "sha": "c1a65ff353429a987630b926727616590a330539", "date": "2026-07-30T04:50:39-05:00", "subject": "docs(tui): kitty keyboard protocol flags #41056", "body": "Nvim has sent CSI > 3 u since 4fb3b57a19, but tui.txt still documented\nCSI > 1 u.", "style": "conventional", "type": "docs", "scope": "tui", "breaking": false} {"repo": "ant-design", "sha": "393cc77c7b5fc7950f435d0af91ebe44b9926915", "date": "2026-07-30T17:54:47+08:00", "subject": "docs: fix Chinese demo locale redirect", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "41df81a6a4c3eef08f7a9a8ac9530cd136c0eafa", "date": "2026-07-30T19:00:07+09:00", "subject": "fix: support resolving top-level input option with plugins", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "angular", "sha": "6d51b99950edaece0e82cdf8837fab193b21c65f", "date": "2026-07-30T10:10:49Z", "subject": "ci: add branch tracking comments to dev-infra GitHub actions", "body": "Adds `# main` version comments after SHA references for `angular/dev-infra` GitHub Actions across workflows so that Renovate tracks the main branch and proposes updates when digests change.\n\n(cherry picked from commit 32815d5472e10cfb4bab45308536baf092a0cdb7)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "32815d5472e10cfb4bab45308536baf092a0cdb7", "date": "2026-07-30T10:10:49Z", "subject": "ci: add branch tracking comments to dev-infra GitHub actions", "body": "Adds `# main` version comments after SHA references for `angular/dev-infra` GitHub Actions across workflows so that Renovate tracks the main branch and proposes updates when digests change.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b835cdba9aa96cc338ba8d9dc979a29b48fca88f", "date": "2026-07-30T12:41:40+02:00", "subject": "Make sure to detach injection points for re-attaching", "body": "The new test for enabling data checksums with concurrent CREATE\nDATABASE calls use the same injection points as a previous test\nbut accidentally missed detaching the injection point first.\n\nFix by detaching the injection point in the PG_TEST_EXTRA SKIP\nblock to make it can be reused. Pointed out by buildfarm member\nporpoise which failed with:\n\n die: error running SQL: 'psql::1:\n\t ERROR: injection point \"datachecksumsworker-fake-temptable-wait\"\n\t\t already defined'\n\nBackpatch to v19 where online checksums were introduced.\n\nAuthor: Daniel Gustafsson \nReported-by: Buildfarm member porpoise\nReviewed-by: Jonathan Gonzalez V. \nDiscussion: https://postgr.es/m/28CF6FD9-E1C4-4C04-8270-E3305AC46171@yesql.se\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "ec367cf2a6c955da8304e8cea935d1f3dc034a98", "date": "2026-07-30T12:44:54+02:00", "subject": "feat!: support nested projects", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": true} {"repo": "angular", "sha": "51244173700ed1468e8f24c533c5891c876b6a0c", "date": "2026-07-30T14:01:20+03:00", "subject": "refactor(devtools): hide search button when there is not a signal graph", "body": "The button is not functional when there isn't a graph. No need to keep it.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "vite", "sha": "b24381d741941b9ce2b1c07db62cc5f4d7bad981", "date": "2026-07-30T19:05:33+08:00", "subject": "feat(bundled-dev): reload once after rebuild instead of via the fallback page", "body": "", "style": "conventional", "type": "feat", "scope": "bundled-dev", "breaking": false} {"repo": "vite", "sha": "94fc91be25005db8bb38c43d5c8e86cd381f61a7", "date": "2026-07-30T20:05:51+09:00", "subject": "test: use top-level input option", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vite", "sha": "8b17656bb1b8db6f6003a0145c4aea1e0f512d6f", "date": "2026-07-30T19:29:43+08:00", "subject": "test(bundled-dev): add deterministic settle harness for page reloads", "body": "Squash of the settle-harness branch rebased onto main after #23106:\na test-only plugin (watch-change acknowledgment, server-side reload\nledger split by decision kind, client instrumentation with loud-fail\ninstall checks), waitForBundledDevSettled, withPageReload, and a global\nafterEach settle guard. Zero production-source changes.", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "vite", "sha": "d5bb819f1760864f5435d4c92dbbd7124476d59a", "date": "2026-07-30T19:31:56+08:00", "subject": "test(bundled-dev): reduce withPageReload to a load-event anchor", "body": "Since #23106 a reload-shaped edit produces exactly one navigation, sent\nonly after the rebuild completes. The helper now just registers the load\nlistener before the edit and waits for that navigation; the settle call\nand the watcher-acknowledgment pre-wait are gone with it.\n\nPlaywright findings behind this (verified against playwright-core 1.62.0\nand vitest 4.1.10, plus an empirical reload-hammer script):\n- selector reads (page.textContent, locator) retry internally through a\n navigation and never see \"Execution context was destroyed\"\n- expect.poll retries probes that throw, which covers evaluate-based\n helpers like getColor\n- the one unsafe pattern is polling an ElementHandle captured before the\n edit — the helper's doc comment now warns about it", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "curl", "sha": "5eb2f0757ebe9fa0328acc2bdb4df155a40942cf", "date": "2026-07-30T13:47:03+02:00", "subject": "curl: help category cleanups", "body": "- add 'mqtt' as a category\n- add more protocol categories to several options\n- make --data worded better to also cover MQTT\n\nCloses #22447", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "73a05428d4c659d950bcf26aad9d7bd3164b8f7e", "date": "2026-07-30T13:49:58+02:00", "subject": "dnscache: use Curl_peer in resolve and dnscache operations", "body": "Removes unused/duplicate members in async/ares/doh structs.\n\nCloses #22446", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d4420a97206cad5bcf63405062db5b4ebaa7f2b1", "date": "2026-07-30T14:06:20+02:00", "subject": "Initialize bs_reltuples in parallel GIN builds", "body": "Index builds update pg_class.reltuples for the table. In parallel GIN\nbuilds, workers track the number of processed rows, and report it to\nthe leader, who then updates the pg_class with a total. However,\ngin_parallel_build_main failed to initialize the bs_reltuples field,\nleaving it set to whatever happens to be on the stack (which may be\nbogus values like Infinity or NaN, or just impossibly high values).\n\nIf such values get reported to the leader and stored in pg_class, that\ncan have serious consequences. The pg_class.reltuples field is used to\ndecide when a table is due for autovacuum or autoanalyze, and if it\nhappens to be set to a bogus value, that may never happen. The field is\nalso used by the optimizer when calculating costs.\n\nFixed by initializing bs_reltuples together with the rest of the build\nstate. The bs_numtuples was initialized later, but it seems cleaner to\njust initialize all the fields at once.\n\nAfter a bogus value gets persisted in pg_class, affected systems are\nunlikely to self-heal. That would require an ANALYZE, but preventing\nthat is one of the consequences. We have considered forcing autoanalyze\nin these cases, but there's not a good way to reliably identify bogus\nvalues (except for a small minority like Infitiny/NaN).\n\nA manual ANALYZE on (possibly) affected tables is the only solution.\n\nBackpatch to 18, where parallel GIN builds were introduced.\n\nReported-by: Jan Nidzwetzki \nDiscussion: https://postgr.es/m/518BA772-8026-412A-AA8F-A7FE4C6B3717@planetscale.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e3387d3a5af5feee25c33ea33d0c7916c1b9120d", "date": "2026-07-30T15:14:22+03:00", "subject": "Fix encrypt decrypt fetch for cms from privoder side", "body": "For algorithms without NID (without registration via the engine), cms_enc attempts to fetch via OBJ_obj2txt.\n\npassing cipher instead of NULL in EVP_PKEY_CTX_ctrl in apps/cms.c\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Paul Dale \nReviewed-by: Andrew Dinh \nMergeDate: Sat Aug 1 03:02:06 2026\n(Merged from https://github.com/openssl/openssl/pull/31512)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "cb21a37a68d347e79e06bec648cfa1be44e3944d", "date": "2026-07-30T14:20:59+02:00", "subject": "build: assume POSIX `select()` is available", "body": "This change effectively replaces an explicit compile-time #error with\na missing prototype error in environments not offering `select()`, and\nsaves curl-compatible systems from performing an explicit feature check.\n\nRefs:\nhttps://pubs.opengroup.org/onlinepubs/009695399/functions/pselect.html\nhttps://linux.die.net/man/2/select\n\nCloses #22448", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a5355cef42944093801aef35bd50e73b8dd8608a", "date": "2026-07-30T23:12:26+10:00", "subject": "Reject clearly degenerate RSASVE parameters.", "body": "- A public exponent <= 1\n- Ciphertext <= 1 or equal to n-1, where n is the public modulus.\n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Bob Beck \nMergeDate: Mon Aug 10 06:27:45 2026\n(Merged from https://github.com/openssl/openssl/pull/32124)\n\n(cherry picked from commit e5d2eaee133290cce1b01bc7cead839bd7e18b8e)\n(cherry picked from commit 69de45cd12e51d69a098b7c81c2bd7ee80f060b5)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a201731333daee5a13d9ca7f010a230b0f467a63", "date": "2026-07-30T23:12:26+10:00", "subject": "Reject clearly degenerate RSASVE parameters.", "body": "- A public exponent <= 1\n- Ciphertext <= 1 or equal to n-1, where n is the public modulus.\n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Bob Beck \nMergeDate: Mon Aug 10 06:27:45 2026\n(Merged from https://github.com/openssl/openssl/pull/32124)\n\n(cherry picked from commit 69de45cd12e51d69a098b7c81c2bd7ee80f060b5)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "69de45cd12e51d69a098b7c81c2bd7ee80f060b5", "date": "2026-07-30T23:12:26+10:00", "subject": "Reject clearly degenerate RSASVE parameters.", "body": "- A public exponent <= 1\n- Ciphertext <= 1 or equal to n-1, where n is the public modulus.\n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Bob Beck \nMergeDate: Mon Aug 10 06:27:45 2026\n(Merged from https://github.com/openssl/openssl/pull/32124)\n\n(cherry picked from commit e5d2eaee133290cce1b01bc7cead839bd7e18b8e)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e5d2eaee133290cce1b01bc7cead839bd7e18b8e", "date": "2026-07-30T23:12:26+10:00", "subject": "Reject clearly degenerate RSASVE parameters.", "body": "- A public exponent <= 1\n- Ciphertext <= 1 or equal to n-1, where n is the public modulus.\n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Bob Beck \nMergeDate: Mon Aug 10 06:27:45 2026\n(Merged from https://github.com/openssl/openssl/pull/32124)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d711cba7fad8271979ceb0790e95b42b2befd3d8", "date": "2026-07-30T09:27:01-04:00", "subject": "Fix FIPS to require a derivation function", "body": "FIPS-140-3 requires that if a CTR-DRBG is allocated from the fips\nprovider that either:\n\na) The entropy source must be NIST validated and exist within the FIPS\nboundary\n\nor\n\nb) The CTR-DRBG must use a derivation function with an entropy source\noutside the FIPS boundary\n\nGiven that we have no approved noise source inside the FIPS boundary, we\nneed to enforce the fact that FIPS allocated CTR-DRBGS only allocate\ninstances in which a derivation function is requested (i.e. the USE_DF\nparameter is asserted and set to one)\n\nFollow path b, and ensure that FIPS CTR-DRBG allocations assert the use\nof USE_DF or fail if an allocation does not\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nMergeDate: Mon Aug 10 14:00:06 2026\n(Merged from https://github.com/openssl/openssl/pull/32249)\n\n(cherry picked from commit 28b27fe53b975e7a86310c9becffd738482b8ddb)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "28b27fe53b975e7a86310c9becffd738482b8ddb", "date": "2026-07-30T09:27:01-04:00", "subject": "Fix FIPS to require a derivation function", "body": "FIPS-140-3 requires that if a CTR-DRBG is allocated from the fips\nprovider that either:\n\na) The entropy source must be NIST validated and exist within the FIPS\nboundary\n\nor\n\nb) The CTR-DRBG must use a derivation function with an entropy source\noutside the FIPS boundary\n\nGiven that we have no approved noise source inside the FIPS boundary, we\nneed to enforce the fact that FIPS allocated CTR-DRBGS only allocate\ninstances in which a derivation function is requested (i.e. the USE_DF\nparameter is asserted and set to one)\n\nFollow path b, and ensure that FIPS CTR-DRBG allocations assert the use\nof USE_DF or fail if an allocation does not\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nMergeDate: Mon Aug 10 14:00:06 2026\n(Merged from https://github.com/openssl/openssl/pull/32249)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f8917c2263db4abd0202b62f69847876947bee70", "date": "2026-07-30T09:27:01-04:00", "subject": "Fix FIPS to require a derivation function", "body": "FIPS-140-3 requires that if a CTR-DRBG is allocated from the fips\nprovider that either:\n\na) The entropy source must be NIST validated and exist within the FIPS\nboundary\n\nor\n\nb) The CTR-DRBG must use a derivation function with an entropy source\noutside the FIPS boundary\n\nGiven that we have no approved noise source inside the FIPS boundary, we\nneed to enforce the fact that FIPS allocated CTR-DRBGS only allocate\ninstances in which a derivation function is requested (i.e. the USE_DF\nparameter is asserted and set to one)\n\nFollow path b, and ensure that FIPS CTR-DRBG allocations assert the use\nof USE_DF or fail if an allocation does not\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMergeDate: Mon Aug 10 12:01:38 2026\n(Merged from https://github.com/openssl/openssl/pull/32235)\n\n(cherry picked from commit 3d8b51d27714921c74bdebc0b39e4b387c442a63)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3d8b51d27714921c74bdebc0b39e4b387c442a63", "date": "2026-07-30T09:27:01-04:00", "subject": "Fix FIPS to require a derivation function", "body": "FIPS-140-3 requires that if a CTR-DRBG is allocated from the fips\nprovider that either:\n\na) The entropy source must be NIST validated and exist within the FIPS\nboundary\n\nor\n\nb) The CTR-DRBG must use a derivation function with an entropy source\noutside the FIPS boundary\n\nGiven that we have no approved noise source inside the FIPS boundary, we\nneed to enforce the fact that FIPS allocated CTR-DRBGS only allocate\ninstances in which a derivation function is requested (i.e. the USE_DF\nparameter is asserted and set to one)\n\nFollow path b, and ensure that FIPS CTR-DRBG allocations assert the use\nof USE_DF or fail if an allocation does not\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMergeDate: Mon Aug 10 12:01:38 2026\n(Merged from https://github.com/openssl/openssl/pull/32235)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "188d35133f14798a4173422edd00b37586a428eb", "date": "2026-07-30T15:29:28+02:00", "subject": "Update corpora submodule", "body": "Reviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMergeDate: Fri Aug 7 13:36:58 2026\n(Merged from https://github.com/openssl/openssl/pull/32093)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a0369dd8448fba99d22068f533764bf2fb001278", "date": "2026-07-30T15:30:19+02:00", "subject": "Reject non-finite reltuples when restoring stats", "body": "When restoring relation stats, pg_restore_relation_stats() rejected\ncalls with (reltuples < -1.0). But that is insufficient - Infinity and\nNaN values both pass that check, and get stored in pg_class verbatim.\nThis can have various undesirable consequences.\n\nFixed by rejecting non-finite reltuple values, in the same non-fatal way\nas for the existing checks (emit WARNING and skip the update). Adds a\nregression test to stats_import for these non-finite values, and to\ncheck the -1.0 special value is still accepted.\n\nBackpatch to 18, where pg_restore_relation_stats() was introduced.\n\nPatch by Jan Nidzwetzki, minor commit message tweaks by me.\n\nAuthor: Jan Nidzwetzki \nDiscussion: https://postgr.es/m/518BA772-8026-412A-AA8F-A7FE4C6B3717@planetscale.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a7f422fa484b41f4999e340db1c97bfc78cbee24", "date": "2026-07-30T09:50:26-04:00", "subject": "refactor(options): naming #41046", "body": "- \"obj\" refers to structured (non-scalar `Object`) option values.\n- \"optval\" generally refers to the internal scalar (legacy \":set\"-style)\n option value.", "style": "conventional", "type": "refactor", "scope": "options", "breaking": false} {"repo": "openssl", "sha": "c9159efd6ec853c8e86677a089e339254cf11c7c", "date": "2026-07-30T16:02:57+02:00", "subject": "SSL_dup() should copy additional members", "body": "currently those members are missing to be copied in SSL_dup(3ossl)\n - server_cert_type\n - client_cert_type\n - ct_validation_callback\n - retsc->ext.status_type (typically needed for OCSP)\n\nAll those missing members are usually set before handshake starts,\nso SSL_dup() needs to copy them so behavior of duplicated SSL\nobject is consistent with its original.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMerge-date: Wed Aug 19 14:26:25 2026\nMerged-from: https://github.com/openssl/openssl/pull/32122\n(cherry picked from commit 52ee7fdf278c2316f1ddd94526f8730e014dd58f)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "52ee7fdf278c2316f1ddd94526f8730e014dd58f", "date": "2026-07-30T16:02:57+02:00", "subject": "SSL_dup() should copy additional members", "body": "currently those members are missing to be copied in SSL_dup(3ossl)\n - server_cert_type\n - client_cert_type\n - ct_validation_callback\n - retsc->ext.status_type (typically needed for OCSP)\n\nAll those missing members are usually set before handshake starts,\nso SSL_dup() needs to copy them so behavior of duplicated SSL\nobject is consistent with its original.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMerge-date: Wed Aug 19 14:26:25 2026\nMerged-from: https://github.com/openssl/openssl/pull/32122", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "68a644d648403ebb16e903d6a233c0006742d047", "date": "2026-07-30T17:07:00+02:00", "subject": "ci: add local skills to the fw-general scope", "body": "They were caught by dev-infra which isn't the best scope for the imho\n\n(cherry picked from commit 613c51e324fd0b6a2c19aac940fb7608009f2b5b)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "613c51e324fd0b6a2c19aac940fb7608009f2b5b", "date": "2026-07-30T17:07:00+02:00", "subject": "ci: add local skills to the fw-general scope", "body": "They were caught by dev-infra which isn't the best scope for the imho", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "2f96c8020f85ccb715a76de4b79a0c680c2c7264", "date": "2026-07-30T10:42:54-05:00", "subject": "fix(core): sanitize host bindings on concrete hosts", "body": "Compute host binding security contexts against concrete hosts, including host directives, inheritance, dynamic directives, and createComponent hostElement usage.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "1113faf2df8d0230162a438636995a335fa830c6", "date": "2026-07-30T18:44:59+03:00", "subject": "refactor(devtools): create Angie component", "body": "Introduce a component that provides shared styling for Angie illustrations and predefined sizes for ease of use.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "angular", "sha": "eb10e937fbcf1f34e399bf1f064010849923a25f", "date": "2026-07-30T10:56:25-05:00", "subject": "refactor(common): modernize pipes & non bindable tests to rely on `whenStable`", "body": "Rely on zoneless scheduling throughout reactive forms tests instead of triggering change detection manually.\n\n(cherry picked from commit d068fc1ea0314eaf81b5f02f170484018492e4e9)", "style": "conventional", "type": "refactor", "scope": "common", "breaking": false} {"repo": "angular", "sha": "d068fc1ea0314eaf81b5f02f170484018492e4e9", "date": "2026-07-30T10:56:25-05:00", "subject": "refactor(common): modernize pipes & non bindable tests to rely on `whenStable`", "body": "Rely on zoneless scheduling throughout reactive forms tests instead of triggering change detection manually.", "style": "conventional", "type": "refactor", "scope": "common", "breaking": false} {"repo": "curl", "sha": "decc6090851236ec6b09ce79433283867d4d9746", "date": "2026-07-30T21:59:16+05:30", "subject": "h3-proxy: fix NULL deref when non-:status header arrives before :status", "body": "Closes #22449", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "da04aa96abc0b9b5700e1146edb613d8ceb1b6e0", "date": "2026-07-30T18:34:18+02:00", "subject": "tidy-up: comments and formatting", "body": "Closes #22558", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "01eeee4bd1b178f1e6909592c6a9986f1ab5a8a0", "date": "2026-07-30T19:58:07+03:00", "subject": "docs: replace Firebase Studio references with Antigravity", "body": "(cherry picked from commit a0f943d53da69c0ed756b12fb2e50fa833b6c796)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "a0f943d53da69c0ed756b12fb2e50fa833b6c796", "date": "2026-07-30T19:58:07+03:00", "subject": "docs: replace Firebase Studio references with Antigravity", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "489fdb0da31364f969415c8619b353d64d8e3cab", "date": "2026-07-30T21:58:24+03:00", "subject": "docs: fix self-referential link and merge intro tips in inputs guide", "body": "The \"Customizing model inputs\" section linked \"standard input\" back to\nthe inputs guide itself, a dead-end self-link. Point the sentence at the\nrelevant in-page sections instead (#required-inputs and #input-aliases).\n\nAlso merge the two stacked intro TIP callouts into a single tip.\n\n(cherry picked from commit 5dd634d18d40f6085bbf93c2ad8ab19420adfd74)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "5dd634d18d40f6085bbf93c2ad8ab19420adfd74", "date": "2026-07-30T21:58:24+03:00", "subject": "docs: fix self-referential link and merge intro tips in inputs guide", "body": "The \"Customizing model inputs\" section linked \"standard input\" back to\nthe inputs guide itself, a dead-end self-link. Point the sentence at the\nrelevant in-page sections instead (#required-inputs and #input-aliases).\n\nAlso merge the two stacked intro TIP callouts into a single tip.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "8c6f5e7c30bbed4e3e826db891d85adadbe09bd0", "date": "2026-07-30T20:58:27+02:00", "subject": "docs: remove duplicate aria-hidden attributes in aria examples", "body": "(cherry picked from commit a37c79d7d4ce7fc5ad81396d2c968b72e06850e1)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "a37c79d7d4ce7fc5ad81396d2c968b72e06850e1", "date": "2026-07-30T20:58:27+02:00", "subject": "docs: remove duplicate aria-hidden attributes in aria examples", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "6cdab150bc161a35ca6356a4d1b4add17c52eb4b", "date": "2026-07-30T15:09:19-04:00", "subject": "docs: misc #41063", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "6b6160d3d4ba3e642feb705defb2aae1b5bab5bb", "date": "2026-07-30T22:17:00+03:00", "subject": "fix(Tabs): center bare `` icons at any size", "body": "A tab `icon` from a third-party library renders as a bare `` inside\n`.ant-tabs-tab-icon`, which the `.anticon` reset never reaches. An `` has\nno baseline, so it is aligned by its bottom margin edge (CSS 2.1 §10.8.1) and\nrides above the label — by an amount that grows with icon size.\n\nCentre it with `vertical-align: middle` + `margin-block-end: 0.2em` (x-height\ncentre lifted onto the cap-height centre), which is size-independent. The\n`> svg` selector only matches a bare ``, never an `.anticon`.", "style": "conventional", "type": "fix", "scope": "Tabs", "breaking": false} {"repo": "postgres", "sha": "6a80179f6b09a6c71a48e864cc4e85f3ea2f4e3c", "date": "2026-07-30T12:47:08-07:00", "subject": "Fix races between deactivation of logical decoding and slot creation.", "body": "On standbys, logical decoding can be deactivated while a logical slot\nis being created: either by replaying an\nXLOG_LOGICAL_DECODING_STATUS_CHANGE record, or by the end-of-recovery\ntransition upon promotion, which deactivates logical decoding if no\nvalid logical slot exists. Both could interleave with a check of the\nlogical decoding status performed before creating a new slot because\nthe slot invalidation executed as part of the deactivation cannot find\na slot being created.\n\nFor regular slot creation on standbys, EnsureLogicalDecodingEnabled()\nassumed that logical decoding must still be enabled during recovery\nsince the caller had already checked it, tripping an assertion failure\nif a concurrent deactivation interleaved.\n\nFor slot synchronization, the local slot could be created and\npersisted based on the remote slot information fetched before the\ndeactivation was replayed, leaving a valid slot whose restart_lsn\nprecedes the deactivation.\n\nFix both paths by re-checking the logical decoding status after the\nnew slot has been created: regular slot creation raises an error, and\nslot synchronization skips persisting the slot. If the deactivation\nhappens after the re-check instead, it is guaranteed to invalidate the\nnewly created slot.\n\nReviewed-by: Srinath Reddy Sadipiralla \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/CAD21AoDEB99VtNbQdDrNd=1gQupJNGMfW_5kdnxq03Q82EK3ag@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "3e01dcddbe8b8a286c78af0a7a63d7412be8443f", "date": "2026-07-30T21:49:32+02:00", "subject": "refactor(options): memory management", "body": "Problem:\nNumerous callers have to manually check optval ownership (i.e. whether,\nand how, to release) via `is_callback_option`, `option_is_global_local`,\netc. This is fragile, hard to use correctly; and if we introduce another\noptval variant in the future, we'll have to redo all of these careful\nchecks and boilerplate again.\n\nSolution:\nProvide a unified system and use it everywhere:\n\n optval_free_owned\n optval_is_owned\n optval_own", "style": "conventional", "type": "refactor", "scope": "options", "breaking": false} {"repo": "tmux", "sha": "f72e789240f1f1383a06bc680f7217879aad20ef", "date": "2026-07-30T22:15:53+01:00", "subject": "Initial commit. Implement generalised image support.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "d601d0d8c567a30274e3efdcfcbd8d405ee2137f", "date": "2026-07-30T22:47:31+01:00", "subject": "Fix image flickering.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "ca48bd076dfe108b22137340a9d050bb1c09ec24", "date": "2026-07-31T00:26:37+02:00", "subject": "curl_ed25519: tidy-up backend fallback", "body": "Sync fallback logic with other crypto algos to:\n\n- allow falling back to the next backend candidate when wolfSSL does not\n have ed25519 built in.\n\n- de-duplicate fallback code.\n\nFollow-up to a55731050e8c3dbea0b96205cf916443118f6acb #22386 #21239\n\nCloses #22450", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5c808b8c5c395a7d7afea7efaaa3328b4bc4e004", "date": "2026-07-30T16:17:13-07:00", "subject": "Fix background psql session cleanup in 051_effective_wal_level.pl.", "body": "Commit 6aba42c660c added quit() calls for two background psql sessions\nwhose slot creation is canceled by pg_cancel_backend(). Both sessions\nran with the default ON_ERROR_STOP=1 and ended their script with \\q,\nso psql exited as soon as the cancellation error arrived. quit() then\nwrote another \\q to the already-closed pipe, making the test die with\n\"ack Broken pipe\".\n\nRun both sessions with on_error_stop => 0 and drop the trailing \\q, so\nthat psql stays at the prompt after reporting the error and quit() can\nshut it down cleanly.\n\nDiscussion: https://postgr.es/m/CAD21AoCZY1fKYgfkvHGWGiXpatUKd23FSLnDCL4m9bWFjdXNZw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "e9f66196948e9e29e7cfd1d66357072c8ab1da11", "date": "2026-07-31T01:35:56+02:00", "subject": "curl_ed25519: drop unused wolfSSL random generator", "body": "Follow-up to a55731050e8c3dbea0b96205cf916443118f6acb #22386 #21239\n\nCloses #22451", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "4a791db9bdbd48280e752578b93909995f93fe49", "date": "2026-07-31T13:16:29+12:00", "subject": "Fix incorrect Result node flattening logic", "body": "This fixes some incorrect flattening of nested Result nodes during\ncreate_plan that was introduced by f2bae51df. That commit failed to\nmaintain the logic that checks for subplans and gating quals from the\nnested Result node before flattening, and that could result in the nested\ngating qual and subplan being lost, which could produce incorrect results.\n\nBug: #19579\nReported-by: Viktor Leis \nAuthor: Ayush Tiwari \nReviewed-by: David Rowley \nDiscussion: https://postgr.es/m/19579-e6296b6c9fc0591c@postgresql.org\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "209142c30048c322f8d5a8b71b2115d79a797f2b", "date": "2026-07-31T09:34:49+08:00", "subject": "docs(Tabs): clarify global configuration for more icon", "body": "", "style": "conventional", "type": "docs", "scope": "Tabs", "breaking": false} {"repo": "vue-core", "sha": "d179213a37e6a7c4a426eb28a0dbd3ff4af50320", "date": "2026-07-31T10:14:30+08:00", "subject": "fix(runtime-vapor): defer KeepAlive updates while async setup is pending", "body": "- defer Vapor KeepAlive root-chain updates while async setup is pending, avoiding transient branch renders and matching VDOM behavior\n- preserve deferred updates across component-root wrappers, nested KeepAlive roots, and nested async roots in the same Suspense pending cycle\n- restore normal Vapor scheduler ordering and component update error handling when flushing deferred effects, with direct-owner hydration support", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "d64c805dea27d189904394ece7576762bece278c", "date": "2026-07-31T11:48:18+09:00", "subject": "Use [re]palloc_array() in buffile.c and fd.c", "body": "The code paths patched in this commit fix some of remnants not addressed\nby 1b105f9472bd. We should have more holes in the tree that could\nbenefit from stronger type safety guarantees. These hypothetical holes\ncould be addressed later; this finishes the job in storage/file/ for the\nbackend code.\n\nAuthor: Tristan Partin \nReviewed-by: Daniel Gustafsson \nReviewed-by: Sami Imseih \nDiscussion: https://postgr.es/m/DKBAWPZ2QDOS.10Y3JDUNKFMXX@partin.io", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "6098f35f44ef7df3ec829c4c1e69d015a0136ad3", "date": "2026-07-31T15:39:21+12:00", "subject": "Fix issue with RANGE's DEFAULT partition pruning", "body": "Partition pruning for RANGE-partitioned tables could mistakenly prune\nthe DEFAULT partition in some cases when it was not valid to do so,\nwhich could lead to rows missing from query results.\n\nThe only known cases where this could happen is when combining pruning\nsteps from an IS NOT NULL clause with other steps that matched to the\nDEFAULT partition. This could occur due to RANGE partitioned tables\nhaving two distinct internal representations for marking if the DEFAULT\npartition should be scanned. The IS NOT NULL steps would mark the\n\"scan_default\" boolean, but other steps created for different purposes\ncould mark a bound_offset Bitmapset, which would ultimately translate into\nalso scanning the default partition. This could all fail after multiple\nsteps were combined with a combine intersect operator, as that will\nintersect the bound_offset bits and only set scan_default if all pruning\nsteps have that flag set. When both input steps to the intersect operator\nhad different representations of whether to scan the DEFAULT partition,\nthe resulting intersect step result would contain neither representation.\n\nHere, we fix this by having the IS NOT NULL pruning result mark the\nbound_offsets so that it uses both representations to mark that the\nDEFAULT partition must be scanned.\n\nReported-by: Jacob Brazeal \nDiagnosed-by: Jacob Brazeal \nAuthor: David Rowley \nDiscussion: https://postgr.es/m/CA+COZaDXrfTaBjLE=Z79MTaH6Xun1V4PeKxLvCNv8mXS8wn0rw@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "745becce3321267a4530af286433d1dc6f043360", "date": "2026-07-31T12:44:51+09:00", "subject": "Fix error handling in port's getopt_long() for missing argument", "body": "In the long option error path, the code previously returned BADARG\nimmediately when optstring[0] == ':' for a missing required argument,\nwithout advancing \"optind\" or resetting \"place\". This error handling\nwas inconsistent with the short option path, where both updates are\nperformed before returning BADARG (and inconsistent with libc,\nadditionally..).\n\nThis affects platforms where our port version of getopt_long() is used,\na concept that should be limited to WIN32 these days. An argument could\nbe made in favor of a backpatch, but this could lead to a slight\ndifferent error handling, for a report that would only show up when\nusing incorrect option combinations.\n\nAuthor: Japin Li \nDiscussion: https://postgr.es/m/SY7PR01MB10921AF81F18A8BCFA06388BEB6C22@SY7PR01MB10921.ausprd01.prod.outlook.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "00b3e50054dad648e74d19cc199555381ab366d2", "date": "2026-07-31T09:58:46+05:30", "subject": "Improve wording of sequence origin warning in logical replication.", "body": "check_publications_origin_sequences() warns when a subscription with\norigin = NONE synchronizes sequence values that may have originated from\nanother subscription. The existing warning is phrased in terms of\ncopy_data and copying data, which is appropriate for table synchronization\nbut misleading for sequence synchronization.\n\nReword the warning, detail, and hint to describe sequence synchronization\nand the associated origin = NONE semantics more accurately.\n\nAlso fix a typo (\"rathen\" -> \"rather\") in a comment in sequencesync.c.\n\nReported-by: Noah Misch \nReported-by: Peter Smith \nAuthor: vignesh C \nReviewed-by: Amit Kapila \nBackpatch-through: 19, where it was introduced\nDiscussion: https://postgr.es/m/20260710045217.f0.noahmisch@microsoft.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "d2c21187ca8a58f5509ab3269fbad687f9c9b04a", "date": "2026-07-31T06:30:36Z", "subject": "build: update dependency node to v22.23.2", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "eslint", "sha": "d3266fb26c719c6b90a6cbe54bb120c9335ad19c", "date": "2026-07-31T09:34:59+02:00", "subject": "chore: unpin `webpack` dependency", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vite", "sha": "9cbeecbd359179bd7ae2b658687e093e1f2f2dd4", "date": "2026-07-31T16:46:53+09:00", "subject": "test: add relative path in assetsInclude case", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "ffdba77fe745da38fd8d44f29dbf3d9bb8e650cb", "date": "2026-07-31T15:53:58+08:00", "subject": "feat(Tree): support auto expand in scrollTo", "body": "* feat(Tree): support auto expand in scrollTo\n\n* refactor(Tree): simplify scrollTo demo\n\n* refactor(Tree): simplify scrollTo handling\n\n* refactor(Tree): simplify controlled scrollTo demo\n\n* test(Tree): skip scrollTo image demo\n\n* docs(Tree): show scrollTo target key\n\n* chore(deps): upgrade @rc-component/virtual-list to 1.5.1\n\n* fix(TreeSelect): use valid value in internal panel demo\n\n* Revert \"fix(TreeSelect): use valid value in internal panel demo\"\n\nThis reverts commit e9e989fb27229f327aa3e19e009e89c23b65519b.\n\n* test(TreeSelect): expect scrollTo retry warning", "style": "conventional", "type": "feat", "scope": "Tree", "breaking": false} {"repo": "vite", "sha": "17f7b2f193a110d0b47742ad296d182cb4666ce7", "date": "2026-07-31T16:56:19+09:00", "subject": "fix: handle shebang ending with uncommon line terminators", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vite", "sha": "e6b6b167afa0a80548829d1f24a0712f9194389a", "date": "2026-07-31T16:56:34+09:00", "subject": "docs: extract VersionsList component", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "c0ff020fbb201f777681858a0302cd2f63f36b28", "date": "2026-07-31T16:25:18+08:00", "subject": "docs(Tree): add version to scrollTo demo", "body": "* docs(Tree): add version to scrollTo demo\n\n* docs(Tree): add version to English scrollTo demo", "style": "conventional", "type": "docs", "scope": "Tree", "breaking": false} {"repo": "neovim", "sha": "4cdd6d76c5ad2714aa5c056c8b16fca408120e61", "date": "2026-07-31T11:28:11+03:00", "subject": "feat(pack): introduce plugin manifest support", "body": "Problem: There is no agreed way for plugins to provide extra information\n about themselves. Like required Neovim version, install/update/delete\n hooks, and dependencies.\n\nSolution: Introduce basic concept of plugin manifest file `pkg.json`\n based on https://packspec.org/.", "style": "conventional", "type": "feat", "scope": "pack", "breaking": false} {"repo": "neovim", "sha": "02a25caa1b5211e349c41ac2e0642de905ae7be9", "date": "2026-07-31T05:09:43-04:00", "subject": "build(vim-patch): N/A diffs from vim9 err msgs #41070", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vue-core", "sha": "135c406e5dcd8b6bf7d59cd4daeb41ff5d100890", "date": "2026-07-31T17:32:37+08:00", "subject": "fix(runtime-vapor): properly prune interop keep-alive entries", "body": "- sync KeepAlive state before interop deactivation\n- fully unmount max-pruned component types and VNodes", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "angular", "sha": "3a293291911ada3cfac1ba898e1257a9b8da3b7b", "date": "2026-07-31T11:34:40+02:00", "subject": "fix(devtools): Extension name for non googlers.", "body": "For corps urls the promise is actually resolved but returns a `status:0`.\n\nfixes #70023", "style": "conventional", "type": "fix", "scope": "devtools", "breaking": false} {"repo": "tmux", "sha": "7c9fdcb770eb4e8b6c93a4e37d90ba8e57e9287b", "date": "2026-07-31T10:39:09+01:00", "subject": "Add k=c and a helper function to check for secondary prompt.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "e144e697ff5f6cdb901dd73fdef7096c3cd16ba3", "date": "2026-07-31T13:24:03+03:00", "subject": "refactor(devtools): substitute remaining Angie images with ng-angie", "body": "Substitute the remaining ``s with the Angie component.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "openssl", "sha": "dfee8502a49fc1f6ab5326ee1924109c3591f188", "date": "2026-07-31T12:29:17+02:00", "subject": "Error reason fixes - mostly a small reordering", "body": "'make update' doesn't quite agree with the ordering of error reasons\nfrom recent development / merge.\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Igor Ustinov \nMergeDate: Fri Jul 31 10:50:04 2026\n(Merged from https://github.com/openssl/openssl/pull/32131)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "527573490eb2564b3d7c9dd51d8bff963b5d6303", "date": "2026-07-31T12:51:25+02:00", "subject": "GHA/http3-linux: enable HTTPSIG in jobs running tests", "body": "To test HTTPSIG with all supported OpenSSL forks.\n\nFollow-up to a55731050e8c3dbea0b96205cf916443118f6acb #22386 #21239\n\nCloses #22453", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f70acc8a2b96e0b565836799b6ccc3bacba8068a", "date": "2026-07-31T23:24:46+12:00", "subject": "Fix Hash Join performance issue when hashing NULL values", "body": "adf97c156 allowed expression evaluation to perform hashing, and\nsubsequently 9ca67658d fixed a memory stomping bug in that commit\nthat caused unrelated-to-hashing expression op steps to stomp on the\nintermediate hash value. The intermediate hash value needs to be\nmaintained when hashing multiple hash keys. 9ca67658d didn't quite get\nthings right when in \"strict\" mode when it aborted hashing early after\nencountering a NULL hash key. What was meant to happen was that the\nexpression returns NULL directly to indicate to the caller the value\nhashed to NULL. The problem was that any EEOP_HASHDATUM_FIRST_STRICT or\nEEOP_HASHDATUM_NEXT32_STRICT op step that didn't belong to the final\nkey to be hashed would have its op->resnull and op->resvalue pointing to\nthe location to store the intermediate hash value. That's correct for\nnon-NULLs since we bit-rotate the intermediate value and continue hashing,\nbut with the strict case, when we get a NULL key, we immediately jump to\nthe \"jumpdone\" step. The problem is the jumpdone step expects the\nExprState resnull and resvalue fields to be set (as they would be if we\ndidn't abort hashing early due to the NULL), but when we aborted early,\nthe ExprState fields never got set. This would result in inserting\nrecords into the hash table that would never match to any join partner,\nwhich is a waste of CPU and memory.\n\nHere we fix this by having EEOP_HASHDATUM_FIRST_STRICT and\nEEOP_HASHDATUM_NEXT32_STRICT populate the ExprState resnull and resvalue\nfields directly when the value to hash is NULL.\n\nAlthough Hash Agg and Hashed Subplans do use hashing from ExprStates,\nthose were unaffected by this bug, as neither of those uses the STRICT op\nsteps.\n\nThanks to Tomas Vondra for finding the offending commit.\n\nReported-by: Dan Stefura \nAuthor: David Rowley \nDiscussion: https://postgr.es/m/YQBPR0101MB89738FB972FBD02A3640C6D3D6C92@YQBPR0101MB8973.CANPRD01.PROD.OUTLOOK.COM\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9d12df0398499d22bb509ad5c209dd608095596d", "date": "2026-07-31T15:17:00+03:00", "subject": "feat(pack): add support for sourcing manifest scripts", "body": "Problem: No way for plugins to define hooks that would be executed\n during plugin lifecycle. Like after install/update or before\n update/delete.\n\nSolution: Automatically source scripts defined in plugin manifest file\n (if any) after triggering corresponding `PackChanged{,Pre}` events.\n During sourcing make some termporary adjustments:\n - Current directory is set to plugin's root to simplify execution of\n CLI calls like `make build`.\n - Plugin's path is ensured to be inside 'runtimepath' to allow using\n `require('plugin-module')` inside manifest scripts.\n\n The reason to execute after triggering event is so that there is a\n possibility for users to execute code both before the script (exactly\n on event) and after the script (more-or-less via `vim.schedule` called\n on the event).", "style": "conventional", "type": "feat", "scope": "pack", "breaking": false} {"repo": "angular", "sha": "3d0067ad982995881aa9b6e492a413b6ba965b38", "date": "2026-07-31T14:37:28+02:00", "subject": "refactor(animations): remove outdated comment", "body": "This isn't necessary anymore.\n\n(cherry picked from commit 1b13db7a1b6eb2451d28a66dde23a52fadf45ad6)", "style": "conventional", "type": "refactor", "scope": "animations", "breaking": false} {"repo": "angular", "sha": "1b13db7a1b6eb2451d28a66dde23a52fadf45ad6", "date": "2026-07-31T14:37:28+02:00", "subject": "refactor(animations): remove outdated comment", "body": "This isn't necessary anymore.", "style": "conventional", "type": "refactor", "scope": "animations", "breaking": false} {"repo": "angular", "sha": "d6ec7c7c30eccf05ffcb979fc108e23c9190e12f", "date": "2026-07-31T16:25:13+03:00", "subject": "docs: document deprecated allowSignalWrites option in v19 update guide", "body": "The `allowSignalWrites` option of `effect()` became a no-op and was\ndeprecated in v19, but the update guide never mentioned it. Applications\nupgrading from v18 or earlier still carry the option at their `effect()`\ncall sites, where it now only produces a dev-mode warning.\n\nAdd an entry to the v19 recommendations instructing those applications to\nremove it, mirroring the wording of the deprecation notice on\n`CreateEffectOptions`.\n\n(cherry picked from commit 690eecda6b3ec2a3f03410b5d0b244e9723d3fd5)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "690eecda6b3ec2a3f03410b5d0b244e9723d3fd5", "date": "2026-07-31T16:25:13+03:00", "subject": "docs: document deprecated allowSignalWrites option in v19 update guide", "body": "The `allowSignalWrites` option of `effect()` became a no-op and was\ndeprecated in v19, but the update guide never mentioned it. Applications\nupgrading from v18 or earlier still carry the option at their `effect()`\ncall sites, where it now only produces a dev-mode warning.\n\nAdd an entry to the v19 recommendations instructing those applications to\nremove it, mirroring the wording of the deprecation notice on\n`CreateEffectOptions`.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f869ff08b20d5a824e3535fbd1a7df3561e67536", "date": "2026-07-31T23:30:59+09:00", "subject": "Fix Poly1305 SVE2 symbol visibility", "body": "poly1305-armv8 takes the address of poly1305_blocks_sve2 with a direct\nADRP/ADD sequence. Mark the SVE2 entry point hidden so ELF linkers bind\nit locally when building shared libraries.\n\nFixes #31670\n\nCLA: trivial\n\nReviewed-by: Milan Broz \nReviewed-by: Kurt Roeckx \nReviewed-by: Simo Sorce \nReviewed-by: Tom Cosgrove \nReviewed-by: Andrew Dinh \nMergeDate: Wed Aug 26 15:15:36 2026\n(Merged from https://github.com/openssl/openssl/pull/32132)\n\n(cherry picked from commit 6bfa751d0e382a293a8572b7b05e4a65fb521ca6)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6bfa751d0e382a293a8572b7b05e4a65fb521ca6", "date": "2026-07-31T23:30:59+09:00", "subject": "Fix Poly1305 SVE2 symbol visibility", "body": "poly1305-armv8 takes the address of poly1305_blocks_sve2 with a direct\nADRP/ADD sequence. Mark the SVE2 entry point hidden so ELF linkers bind\nit locally when building shared libraries.\n\nFixes #31670\n\nCLA: trivial\n\nReviewed-by: Milan Broz \nReviewed-by: Kurt Roeckx \nReviewed-by: Simo Sorce \nReviewed-by: Tom Cosgrove \nReviewed-by: Andrew Dinh \nMergeDate: Wed Aug 26 15:15:36 2026\n(Merged from https://github.com/openssl/openssl/pull/32132)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a71825973acddde12c24501d5fe203d8214e82fa", "date": "2026-07-31T09:57:24-05:00", "subject": "Remove code for pre-v10 servers from AdjustUpgrade.pm.", "body": "We recently removed support for upgrading from pre-v10 servers, so\nwe no longer need to handle older versions in this helper module.\n\nOversight in commit 14d8418083.\n\nDiscussion: https://postgr.es/m/ak7Ekv2-L-G55-YD%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "102e806320a4238a9d1d744bd8ea312960b7b9c2", "date": "2026-07-31T17:12:56+02:00", "subject": "refactor(options): simplify \"unset\" logic", "body": "", "style": "conventional", "type": "refactor", "scope": "options", "breaking": false} {"repo": "angular", "sha": "caf616670fd20d528aa69e0131cc17d60f0cc27d", "date": "2026-07-31T10:32:14-05:00", "subject": "fix(http): run root interceptors in the terminal request chain", "body": "Represent withRequestsMadeViaParent() with an internal delegating backend so the interceptor handler can distinguish delegated clients from independent child configurations.", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "angular", "sha": "a13b9684510dfd04e18f3e1d6686d556c051f34e", "date": "2026-07-31T10:32:14-05:00", "subject": "fix(http): run root interceptors in the terminal request chain", "body": "Represent withRequestsMadeViaParent() with an internal delegating backend so the interceptor handler can distinguish delegated clients from independent child configurations.\n\n(cherry picked from commit bb78286e5e729d1e1fcfd2733d88351a38083ef5)", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "postgres", "sha": "288d4e83f1dab9fc9744eed7129e4f77d82be7ad", "date": "2026-07-31T10:34:40-05:00", "subject": "Fix autovacuum's database sorting.", "body": "When db_comparator() was updated to use pg_cmp_s32(), the arguments\nwere listed in the wrong order. This caused autovacuum to sort the\ndatabases by their scores in ascending order instead of descending\norder. To fix, swap the arguments to pg_cmp_s32().\n\nOversight in commit 3b42bdb471.\n\nReported-by: Хамидуллин Рустам \nAuthor: Хамидуллин Рустам \nDiscussion: https://postgr.es/m/5c5a7984-b149-b505-7ad9-2a7766c65b55%40postgrespro.ru\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "499d9eabb5698ff6091b5a186acd7d05cf9d4c2b", "date": "2026-07-31T11:57:44-04:00", "subject": "Prevent walsummarizer from getting stuck at a timeline switch.", "body": "As previously coded, walsummarizer only wants to read WAL from a file\nwhere the TimeLineID in the filename exactly matches the TimeLineID being\nsummarized. But in some cases, when a timeline switch occurs, the WAL file\nfrom the old timeline is not archived, because it's never completely\nfilled, so the only way to obtain the contents of that last partial\nsegment is to read from the first segment on the new timeline. Teach\nWAL summarizer to do that, and add a test case to make sure that it\nworks.\n\nReported-by: Nick Ivanov \nReviewed-by: Andrey Borodin \nTested-by: Amit Kapila \nReviewed-by: Srinath Reddy Sadipiralla \nReviewed-by: Zhijie Hou \nReviewed-by: Thom Brown \nDiscussion: http://postgr.es/m/CA+Tgmobr27GpKDZx3_ezW2+C5_g18i+jSK3sGF_cR-_ESv5N5A@mail.gmail.com\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "fff8c86759659a070e014ffd7295e93a11a1479c", "date": "2026-07-31T11:19:37-07:00", "subject": "oauth: Add unit tests for multiplexer handling", "body": "(This is a late cherry-pick of 4e1e41733, now that the new suite has\nproven its stability, to ensure coverage for this code in PG18 as the\nlater branches diverge.)\n\nTo better record the internal behaviors of oauth-curl.c, add a unit test\nsuite for the socket and timer handling code. This is all based on TAP\nand driven by our existing Test::More infrastructure.\n\nThis commit is a replay of 1443b6c0e, which was reverted due to\nbuildfarm failures. Compared with that, this version protects the build\ntargets in the Makefile with a with_libcurl conditional, and it tweaks\nthe code style in 001_oauth.pl.\n\nReviewed-by: Dagfinn Ilmari Mannsåker \nReviewed-by: Andrew Dunstan \nDiscussion: https://postgr.es/m/CAOYmi+nDZxJHaWj9_jRSyf8uMToCADAmOfJEggsKW-kY7aUwHA@mail.gmail.com\nDiscussion: https://postgr.es/m/CAOYmi+m=xY0P_uAzAP_884uF-GhQ3wrineGwc9AEnb6fYxVqVQ@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "74169d3a1d695556ad81ef7a9c256daf0d554da1", "date": "2026-07-31T11:19:44-07:00", "subject": "libpq-oauth: Avoid overflow for very large intervals", "body": "The slow_down interval parsing code checks explicitly for overflow, but\nsince it does that after the signed overflow has already occurred, we\nend up inviting undefined behavior from the compiler anyway.\n\nUse checked arithmetic instead. set_timer() takes a long int in order to\ninterface nicely with libcurl, so use an int32 as the interval counter\nand clamp to LONG_MAX during conversion to milliseconds.\n\nBackpatch to 18, where libpq-oauth was introduced.\n\nReported-by: Andres Freund \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/qtclihmrkq67ach3xjxyi4qcksstin5qxwsnkqefkmotxwh4g6%40ae2bj6jvcmry\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "481052c7754013d44a89f50912ba6845398f88df", "date": "2026-07-31T14:39:29-04:00", "subject": "On Windows, make link(2) report ENOTSUP when appropriate.", "body": "CreateHardLinkA reports ERROR_INVALID_FUNCTION if the target file\nis on a filesystem that doesn't support hard links. _dosmaperr\nmaps that to EINVAL, which confuses zic.c into failure. zic.c is\nexpecting ENOTSUP if the filesystem lacks link support, and will\nproperly fall back to making a physical copy if it gets that.\nHence, add code to map ERROR_INVALID_FUNCTION to ENOTSUP.\n\n(We could instead teach _dosmaperr to do that, but it's far from clear\nthat this would be appropriate as a global behavior: intuitively\nit seems like EINVAL should be appropriate most of the time.)\n\nWe didn't need this before commit aeb07c55f, because the tzcode\nversion we were using before that didn't have this particular\nerror-handling logic. Hence, no back-patch for now; but if we\ndecide to back-patch tzcode 2026b or later, we'll need this too.\n\nAuthor: Vladlen Popolitov \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/e6122f9b2eef9096f1f11ecc058bcd91@postgrespro.ru", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "3b5c798072a3cf43b5e0ec0b3f4abc18bc6196a5", "date": "2026-07-31T21:59:35+03:00", "subject": "fix(forms): keep radio inputs in sync when values change", "body": "Recompute the checked state when a reused radio input receives a new value while the form model remains unchanged.\n\nHandle this case for both reactive forms and signal forms, and add regression tests covering reused radio elements.\n\n(cherry picked from commit 9e37a58e14bd7955d40255e42e20aa84fcde02f4)", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "9e37a58e14bd7955d40255e42e20aa84fcde02f4", "date": "2026-07-31T21:59:35+03:00", "subject": "fix(forms): keep radio inputs in sync when values change", "body": "Recompute the checked state when a reused radio input receives a new value while the form model remains unchanged.\n\nHandle this case for both reactive forms and signal forms, and add regression tests covering reused radio elements.", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "postgres", "sha": "0c57a40694d88cdda588b5f8a6e5568861d1759d", "date": "2026-07-31T14:00:37-05:00", "subject": "Add list of major features to the v19 release notes.", "body": "Reviewed-by: Michael Banck \nDiscussion: https://postgr.es/m/akWIxtcathhoUuCQ%40nathan\nBackpatch-through: 19 only", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "6b643ab037db087a1e6a005db1332ee9d3181879", "date": "2026-07-31T16:40:00-05:00", "subject": "refactor(common): modernize directive tests to rely on whenStable", "body": "Replace synchronous detectChanges calls with zoneless-compatible scheduling and stability waits. Preserve NgComponentOutlet coverage for components declared by NgModules.\n\n(cherry picked from commit c9c4f2afc60f9d7d5c9f61c8b978691d9793d229)", "style": "conventional", "type": "refactor", "scope": "common", "breaking": false} {"repo": "angular", "sha": "c9c4f2afc60f9d7d5c9f61c8b978691d9793d229", "date": "2026-07-31T16:40:00-05:00", "subject": "refactor(common): modernize directive tests to rely on whenStable", "body": "Replace synchronous detectChanges calls with zoneless-compatible scheduling and stability waits. Preserve NgComponentOutlet coverage for components declared by NgModules.", "style": "conventional", "type": "refactor", "scope": "common", "breaking": false} {"repo": "postgres", "sha": "d6792cc58e9c0aead136f381052151e1c246b534", "date": "2026-07-31T17:45:45-04:00", "subject": "Allow IO time to be counted without a matching IO operation in pg_stat_io", "body": "Since 999dec9ec6a816680, pg_stat_io can show read time with zero reads\nfor an IO Context: a foreign IO is counted as a read only in the\ninitiating backend, while other waiters record only the wait time. That\nviolates pgstat_bktype_io_stats_valid(). Relax the check to allow time\nwithout a matching operation count, since we want to count read wait\ntime even in backends that did not initiate the read. This also enables\nfuture accounting of waits on IO resources (e.g., AIO handles) in\nbackends that didn't start the IO.\n\nAuthor: Andrey Rachitskiy \nReported-by: Justin Pryzby \nReviewed-by: Melanie Plageman \nReviewed-by: Andrey Borodin \nDiscussion: https://postgr.es/m/ak5lccE4qiQpOBHn@pryzbyj2023\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a368fbe968f6aa4ef88c7d3469eb4f90124e5dbc", "date": "2026-08-01T00:45:57+02:00", "subject": "curl_ed25519: add GnuTLS support (via nettle, hogweed)", "body": "The necessary cryptography API is provided by nettle 3.1+, via its\n'hogweed' library. The minimum GnuTLS version required by curl is 3.6.5,\nwhich requires nettle 3.4.1+, so the API is always available.\n\nAlso:\n- autotools: detect and use nettle's hogweed library.\n- cmake/FindNettle: add support for the hogweed library.\n- GHA/http3-linux: enable in the autotools/cmake GnuTLS jobs.\n\nRef: https://github.com/gnutls/gnutls/commit/4353ea025ae032887f3e8cf5aadace25662c6b35\n\nCloses #22456", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "5cc795db5965eff25b014f5854bef6eea9921a14", "date": "2026-08-01T08:22:57+08:00", "subject": "fix(Table): skip disabled rows across pages when selecting all data", "body": "* fix(Table): skip disabled rows across pages when selecting all data\n\n* test(Table): verify disabled row state across pages", "style": "conventional", "type": "fix", "scope": "Table", "breaking": false} {"repo": "ant-design", "sha": "67d059466dc5c1ee68d0dff22f712a885efb4fe3", "date": "2026-08-01T09:40:27+08:00", "subject": "chore(ci): fix master branch ci/cd", "body": "* chore(ci): fix master ci/cd\n\n* chore: update", "style": "conventional", "type": "chore", "scope": "ci", "breaking": false} {"repo": "ant-design", "sha": "6ee813c0bca7a6f77997b78255bd37542ecc1d4c", "date": "2026-08-01T09:51:49+08:00", "subject": "fix(Transfer): correct deselect all after filtering", "body": "", "style": "conventional", "type": "fix", "scope": "Transfer", "breaking": false} {"repo": "ant-design", "sha": "9e6c0181073ae0ee8c313069167f4e0347df0a2d", "date": "2026-08-01T04:52:40+03:00", "subject": "fix(Segmented): center bare `` icons at any size", "body": "A Segmented item `icon` from a third-party library renders as a bare ``\ninside `.ant-segmented-item-icon`, which the `.anticon` reset never reaches.\nThe item label is a block with a tall line-height (not flex), so its inline\ncontent is baseline-aligned; an `` has no baseline, so it is aligned by\nits bottom margin edge (CSS 2.1 §10.8.1) and rides above the label — by an\namount that grows with icon size.\n\nCentre it with `vertical-align: middle` + `margin-block-end: 0.2em` (x-height\ncentre lifted onto the cap-height centre), which is size-independent. The\n`&-icon > svg` selector only matches a bare ``, never an `.anticon`, and\nleaves item height unchanged for 1em icons.", "style": "conventional", "type": "fix", "scope": "Segmented", "breaking": false} {"repo": "ant-design", "sha": "40b25ab28fba27042fdc5ec763b573ceeb3fb89c", "date": "2026-08-01T07:38:57+05:00", "subject": "fix: apply disabled text color to AutoComplete input", "body": "* fix: apply disabled text color to AutoComplete input\n\n* test: assert AutoComplete disabled color on inner input node\n\n* test: add debug demo for AutoComplete disabled color in Form\n\n* test: add extend snapshot for AutoComplete disabled-in-form-debug demo\n\n---------", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "a1b698dfe5a0aae40adabaf03995a96a2dc416c6", "date": "2026-08-01T11:13:34+08:00", "subject": "update snap", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "3478df716ec42fdb80bbcc209c9a5c3401ac6ead", "date": "2026-08-01T10:44:03+06:00", "subject": "refactor(compiler): remove unused code and exports", "body": "Removes unused code and exports that are no longer referenced anywhere\nin the compiler codebase.\n\n(cherry picked from commit 337053ef1ab6400c3641984511b45c341d306e87)", "style": "conventional", "type": "refactor", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "337053ef1ab6400c3641984511b45c341d306e87", "date": "2026-08-01T10:44:03+06:00", "subject": "refactor(compiler): remove unused code and exports", "body": "Removes unused code and exports that are no longer referenced anywhere\nin the compiler codebase.", "style": "conventional", "type": "refactor", "scope": "compiler", "breaking": false} {"repo": "neovim", "sha": "8787b8b9b6a3c745905cd11ba93205e8da9697ed", "date": "2026-07-31T23:52:00-05:00", "subject": "fix(zip): do not run unzip from the current directory", "body": "Problem:\nWindows searches the current directory before $PATH, so opening an archive\nin a directory that also contains an `unzip` executable runs that one. The\nlegacy plugin refused this; the port dropped the check.\n\nSolution:\nRefuse to run `unzip` when it resolves to the current directory.", "style": "conventional", "type": "fix", "scope": "zip", "breaking": false} {"repo": "neovim", "sha": "886a270e7f4162933e180d3d96c3ef7ea162d4a0", "date": "2026-07-31T23:52:07-05:00", "subject": "refactor(zip)!: address entries with zip:// paths", "body": "Problem:\n`zipfile://{archive}::{entry}` is ambiguous: `::` is legal in both an\narchive path and an entry path, so the separator cannot be identified.\nSplitting at the last `::` reads archives correctly but breaks entries\nthat contain it, and the plugin then emits buffer names it cannot read\nback.\n\nSolution:\nAddress entries as `zip://{archive}/{entry}`, joining the two paths.\nResolve the split by walking components: the first one that is a regular\nfile is the archive, because a regular file cannot have children on disk.\nEntry paths may then contain any character, and no escaping is needed.\n\nzipPlugin.vim keeps its own scheme, so the Java ftplugin emits whichever\nform matches the active plugin until the legacy package is removed.", "style": "conventional", "type": "refactor", "scope": "zip", "breaking": true} {"repo": "neovim", "sha": "af1b82fc7bc747ad916408fb01b64c1500f68193", "date": "2026-08-01T01:07:59-04:00", "subject": "build(vim-patch): vim \"mch_job_\" api is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "d0aa08287cae993bc60b9638892a93c77ce86865", "date": "2026-08-01T02:35:37-03:00", "subject": "Document SM4 table-lookup side-channel risk", "body": "Note in EVP_CIPHER-SM4(7) that SM4 implementations may use\nsecret-dependent table lookups and are not guaranteed to be\nconstant-time. Mention the portable C path as a concrete example\nand point to CPU capability man pages for code-path selection.\n\nFixes: https://github.com/openssl/openssl/issues/31911\n\nAssisted-by: Cursor:composer\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Paul Yang \nMergeDate: Tue Aug 11 07:03:10 2026\n(Merged from https://github.com/openssl/openssl/pull/32136)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d5e78dfc824eb8d99de1d861992eaad83de53fda", "date": "2026-08-01T02:35:37-03:00", "subject": "Document SM4 table-lookup side-channel risk", "body": "Note in EVP_CIPHER-SM4(7) that SM4 implementations may use\nsecret-dependent table lookups and are not guaranteed to be\nconstant-time. Mention the portable C path as a concrete example\nand point to CPU capability man pages for code-path selection.\n\nFixes: https://github.com/openssl/openssl/issues/31911\n\nAssisted-by: Cursor:composer\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Paul Yang \nMergeDate: Tue Aug 11 07:03:00 2026\n(Merged from https://github.com/openssl/openssl/pull/32136)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7f9d9d35b945ce0d128b995368e55e4070be1362", "date": "2026-08-01T02:35:37-03:00", "subject": "Document SM4 table-lookup side-channel risk", "body": "Note in EVP_CIPHER-SM4(7) that SM4 implementations may use\nsecret-dependent table lookups and are not guaranteed to be\nconstant-time. Mention the portable C path as a concrete example\nand point to CPU capability man pages for code-path selection.\n\nFixes: https://github.com/openssl/openssl/issues/31911\n\nAssisted-by: Cursor:composer\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Paul Yang \nMergeDate: Tue Aug 11 07:02:49 2026\n(Merged from https://github.com/openssl/openssl/pull/32136)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a223c031147ae054ca51ce9fc8f37d8af52dc6cf", "date": "2026-08-01T02:35:37-03:00", "subject": "Document SM4 table-lookup side-channel risk", "body": "Note in EVP_CIPHER-SM4(7) that SM4 implementations may use\nsecret-dependent table lookups and are not guaranteed to be\nconstant-time. Mention the portable C path as a concrete example\nand point to CPU capability man pages for code-path selection.\n\nFixes: https://github.com/openssl/openssl/issues/31911\n\nAssisted-by: Cursor:composer\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Paul Yang \nMergeDate: Tue Aug 11 07:02:37 2026\n(Merged from https://github.com/openssl/openssl/pull/32136)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3d9325c028624cef8979164fa4f1d899696f3439", "date": "2026-08-01T02:35:37-03:00", "subject": "Document SM4 table-lookup side-channel risk", "body": "Note in EVP_CIPHER-SM4(7) that SM4 implementations may use\nsecret-dependent table lookups and are not guaranteed to be\nconstant-time. Mention the portable C path as a concrete example\nand point to CPU capability man pages for code-path selection.\n\nFixes: https://github.com/openssl/openssl/issues/31911\n\nAssisted-by: Cursor:composer\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Paul Yang \nMergeDate: Tue Aug 11 07:02:12 2026\n(Merged from https://github.com/openssl/openssl/pull/32136)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0f29c47278a0d8c30d019628c084118d9419ce71", "date": "2026-08-01T08:26:05+01:00", "subject": "Cleanup. Rename things for consistency. Fix indenting.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "3e6e0a286085e0474a507e706d0a114fd513bab4", "date": "2026-08-01T08:27:42+01:00", "subject": "New image related files.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "355814931141537e51418a276c74e91283da9631", "date": "2026-08-01T19:21:46+09:00", "subject": "pg_verifybackup: Improve some error handling around strtoul() calls", "body": "Three code paths checking the size, timeline ID and system identifier\nstored in a manifest now check for an empty value. Values are always\nexpected in these parts of a backup banifest. A couple of tests are\nadded to validate this behavior\n\nAdditionally, precheck_tar_backup_file() checked that \"endptr\" is NULL.\nBased on the C standard, strtoul() never sets an \"endptr\" to NULL when\ngiven a value (that is the case here), returning a pointer to the\noriginal value if there is nothing to convert. The pre-tar validation\ncode is adjusted to do so.\n\nAuthor: Tristan Partin \nDiscussion: https://postgr.es/m/DKBS2Z9CGARC.2T07O6TJYSE8B@partin.io", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9fbcb28ea697a116280fbfd4c7f3fb5eaf4c4b46", "date": "2026-08-01T06:23:08-04:00", "subject": "build(vim-patch): N/A prop func defs, n/a vim9 type vars", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "neovim", "sha": "81261804a8e335f9a97f0c4b2e8e70c59593ccca", "date": "2026-08-01T12:29:54+02:00", "subject": "fix(ruler): always clear ruler when disabled", "body": "Problem: the ruler is not cleared in the following circumstances:\n- ui1 is running\n- 'rulerformat' is configured\n- the default ruler was not previously visible in the last line, for\n example because 'rulerformat' is configured in init.lua\n- 'ruler' is disabled without using the command-line, for example via\n key-binding (entering the command-line would clear the ruler)\n\nThe corresponding test case did not fail because 'rulerformat' was set\nwhile the default ruler was shown.\n\nSolution: use a dedicated variable for tracking whether the ui1 ruler\nwas previously shown in the last line. `did_ruler_col` is now only used\nfor setting `msg_col`, which is not implemented in the case where\n'rulerformat' is configured.\n\nReorder the test code to make the individual checks more independent\nfrom each other, and to reflect the future where 'rulerformat' will\nnever be empty. Note that the check where 'rulerformat' was configured\nrelied on the default ruler not being cleared and a stale \"0,\" still\nbeing shown in front of the new ruler - this is also fixed with ui2.", "style": "conventional", "type": "fix", "scope": "ruler", "breaking": false} {"repo": "angular", "sha": "79a819379f172d681227325ec5364a5b8b132963", "date": "2026-08-01T13:38:01+03:00", "subject": "docs: fix roadmap link text and preview-site link", "body": "Two link issues on the roadmap page:\n\n- \"Introduce deferred loading\" linked next.angular.dev, the preview build of\n this same site, for a page that exists here at guide/templates/defer. Use the\n relative path so readers stay on the stable docs.\n- \"Local template variables\" linked \"`@let` docs\", putting a code chip and a\n plain word in one anchor. The anchor's underline crosses the chip and the two\n halves render in different colors. Link only the symbol, as the rest of adev\n does.\n\nAlso on the same page:\n\n- Drop a stray space before a period in the Components card.\n- Give the older completion labels the same \"Completed in \" wording the\n newer cards already use.\n\n(cherry picked from commit f2f12925bdea3a83fb49a4b6398b54ff63456202)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "f2f12925bdea3a83fb49a4b6398b54ff63456202", "date": "2026-08-01T13:38:01+03:00", "subject": "docs: fix roadmap link text and preview-site link", "body": "Two link issues on the roadmap page:\n\n- \"Introduce deferred loading\" linked next.angular.dev, the preview build of\n this same site, for a page that exists here at guide/templates/defer. Use the\n relative path so readers stay on the stable docs.\n- \"Local template variables\" linked \"`@let` docs\", putting a code chip and a\n plain word in one anchor. The anchor's underline crosses the chip and the two\n halves render in different colors. Link only the symbol, as the rest of adev\n does.\n\nAlso on the same page:\n\n- Drop a stray space before a period in the Components card.\n- Give the older completion labels the same \"Completed in \" wording the\n newer cards already use.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "tmux", "sha": "e23219ec047e8f7bb8a8bba318e8c06b53bb54be", "date": "2026-08-01T11:47:25+01:00", "subject": "Add comments during my review including one fix, separator text after D also expandable.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "c5a5b99788894e471d2cc989e3faf0592fe269c5", "date": "2026-08-01T11:59:18+01:00", "subject": "Rename output to fold in places for consistency.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "0043b3fb8cf930825c3043bd72af2f57ab5d6719", "date": "2026-08-01T13:08:30+02:00", "subject": "DEPENDENCIES.md: document minimum nettle version: 3.4.1 (2018-12-04)", "body": "It comes as a transitive requirement by the minimum GnuTLS version.\nBecause libcurl uses nettle directly (in GnuTLS builds), I figure it is\nuseful to document explicitly.\n\nRefs:\nhttps://github.com/gnutls/gnutls/commit/4353ea025ae032887f3e8cf5aadace25662c6b35\nhttps://github.com/curl/curl/pull/22456#discussion_r3695417678\nhttps://github.com/gnutls/nettle/releases/tag/nettle_3.4.1_release_20181204\n\nCloses #22457", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "36c2197539f7275a8688a073161435f56d880b9a", "date": "2026-08-01T16:28:25+05:00", "subject": "fix(zone.js): guard against null handle when detecting refreshable timers", "body": "A timer wrapper such as a browser extension, policy, or automation harness can block a call and return null instead of a handle. patchTimer then read `.refresh` on that null value and threw synchronously.\n\nTreat a nullish handle as non-refreshable so the patched timer keeps working.\n\nFixes #70044", "style": "conventional", "type": "fix", "scope": "zone.js", "breaking": false} {"repo": "neovim", "sha": "9be130304d39e2223ec976c26e69b87b9ef7a0e2", "date": "2026-08-01T13:33:52+02:00", "subject": "refactor(statusline): simplify and remove dead code", "body": "- Remove code that is never executed because the ruler is never rendered\n separately for a window with visible statusline or when `ls=3`, in\n other words: the variables `part_of_status` and `in_status_line` were\n always false for the ruler and badly named for the statusline itself.\n- Leave `maxwidth` unchanged after calling `stl_alloc_click_defs`.\n- Don't reuse the window's click definitions for the ruler. Even though\n it doesn't seem to be a bug because statusline and ruler are never\n shown at the same time for the same window, the ruler currently\n doesn't support clicks anyway.\n- Disentangle the different cases, in particular statusline- from ruler-\n specific code. Another example: `wp->w_wincol`, `wp->w_winrow`, and\n `wp->w_width` are always used together, but this was hard to see.", "style": "conventional", "type": "refactor", "scope": "statusline", "breaking": false} {"repo": "nestjs", "sha": "4b51491f0f810b07795b9dd6c1caa7e613c46961", "date": "2026-08-01T20:50:36+08:00", "subject": "fix(deps): remove duplicate version @fastify/static", "body": "#17382", "style": "conventional", "type": "fix", "scope": "deps", "breaking": false} {"repo": "neovim", "sha": "d95b0fc12c6fc17a06b122c53d4d7a4f2ad85a00", "date": "2026-08-01T08:51:39-04:00", "subject": "build(vim-patch): n/a vim9 struct from *.h", "body": "Target v8.2.0543\nThis can catch stray '#define' but they're most likely 'TTFLAG' for\nVim9 types. N/A.\nTTFLAG is used in 'static_types[]'.\nSee v9.0.0623.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "neovim", "sha": "b11f1b1f76296e0296bb21a54315e2a18c8d6a6e", "date": "2026-08-01T14:09:21+01:00", "subject": "feat(vim.fs): abspath({cwd, plain}) #40597", "body": "Adds optional parameters to `vim.fs.abspath`:\n1. `cwd` to specify directory akin to `--relative-from` in `realpath(1)`.\n2. `plain` disables expansion of tilde (~) in paths.", "style": "conventional", "type": "feat", "scope": "vim.fs", "breaking": false} {"repo": "neovim", "sha": "e02755cd9ff29277d14421c9df627e5dc48e4f67", "date": "2026-08-01T10:07:26-04:00", "subject": "fix(options): \":set foo?\" for Lua callbacks #41100", "body": "Problem:\n`:set foo=` and `:set foo?` for func/expr options set to a Lua\nfunction, always displays \"v:lua\".\n\nThe existing \"\" form was avoided because the ref id\nN changes on every read of the same option.\n\nSolution:\n- Don't attempt to tab-complete Lua functions.\n- Show \"\" (or \"\" if source file is unknown).\n\nExample:\n\n :set operatorfunc?\n operatorfunc=", "style": "conventional", "type": "fix", "scope": "options", "breaking": false} {"repo": "neovim", "sha": "259e6fa9ccb7354cb541cf77c64c0f8196cbe2ec", "date": "2026-08-01T18:01:00+02:00", "subject": "fix(ruler)!: consistent width in last line and statusline", "body": "Problem: Traditionally, the ruler in the last line is one cell shorter\nthan in the statusline, leaving the last cell of the screen blank.\nAccording to code comments, this is in order to prevent unwanted\nscrolling on \"some\" (unspecified, but presumably ancient) terminals.\nBerkeley vi is more specific in its `vs_modeline` function: dumb\nterminals with hardware scroll, SunOS 4.1.1 and Ultrix 4.2 curses.\n(n)curses still has a similar limitation in `(w)addstr`, but apparently\nonly for historical reasons.\n\nMaintaining the different widths leads to awkward inconsistencies when\nthe ruler is configured with 'rulerformat', except for the special case\nwhere it contains a top-level `%=`. Shifting the ruler in the last line\nto the left would be a solution, but the empty cell at the end doesn't\nseem to be relevant anymore.\n\nSolution: extend the ruler in the last line all the way to the right\nedge of the screen, just like in the statusline. The exact same amount\nof place will be available to the rest of the UI as before.\n\nBREAKING CHANGE:\n- the default ruler width is now 18 cells\n- the last cell of the screen is no longer empty\n\nCloses #41076", "style": "conventional", "type": "fix", "scope": "ruler", "breaking": true} {"repo": "neovim", "sha": "fe8e6a7efa33dc9ea1c57aa3f65be9469f05f287", "date": "2026-08-01T18:03:15+02:00", "subject": "test(lsp-selection): of by one", "body": "The test uses the position col=5,row=5; but such a position\nis outside of the buffer.\n\nPresume that it was meant to be col=5,row=4...", "style": "conventional", "type": "test", "scope": "lsp-selection", "breaking": false} {"repo": "neovim", "sha": "77dfa02f3ec74555039a20b9c68ea9e27eebf03c", "date": "2026-08-01T18:04:09+02:00", "subject": "refactor: combine all visual_selection funcions", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "neovim", "sha": "8695579636b4f76561f54b92aac1a05ca00d09a1", "date": "2026-08-01T18:13:21+02:00", "subject": "fix(treesitter): handle `&selection=exclusive`", "body": "", "style": "conventional", "type": "fix", "scope": "treesitter", "breaking": false} {"repo": "neovim", "sha": "d060d91ef8aa2febbc8f1186b12eadab58ae4ec0", "date": "2026-08-01T12:15:02-04:00", "subject": "fix(bufwrite): coverity \"uninitialized member\" #41104", "body": "CID 652079: (UNINIT)\n /src/nvim/bufwrite.c: 1413 in buf_write()\n 1407 // only makes sense at the start of the file.\n 1408 if (buf->b_p_bomb && !write_bin && (!append || perm < 0)) {\n 1409 write_info.bw_len = make_bom(buffer, fenc);\n 1410 if (write_info.bw_len > 0) {\n 1411 // don't convert\n 1412 write_info.bw_flags = FIO_NOCONVERT | wb_flags;\n >>> CID 652079: (UNINIT)\n >>> Using uninitialized value \"write_info.bw_first\" when calling \"buf_write_bytes\".\n 1413 if (buf_write_bytes(&write_info) == FAIL) {\n 1414 end = 0;\n 1415 } else {\n 1416 nchars += write_info.bw_len;\n 1417 }\n 1418 }\n /src/nvim/bufwrite.c: 1453 in buf_write()\n 1447 *s = c;\n 1448 }\n 1449 s++;\n 1450 if (++write_info.bw_len != bufsize) {\n 1451 continue;\n 1452 }\n >>> CID 652079: (UNINIT)\n >>> Using uninitialized value \"write_info.bw_first\" when calling \"buf_write_bytes\".\n 1453 if (buf_write_bytes(&write_info) == FAIL) {\n 1454 end = 0; // write error: break loop\n 1455 break;\n 1456 }\n 1457 nchars += bufsize - write_info.bw_len;\n 1458 s = buffer + write_info.bw_len;", "style": "conventional", "type": "fix", "scope": "bufwrite", "breaking": false} {"repo": "neovim", "sha": "92544471e04fd439e8a04380bd205e80c459d9c8", "date": "2026-08-01T13:03:20-04:00", "subject": "build(vim-patch): \"build_argv_from_\" api is n/a", "body": "Meant for Vim job, terminal, cscope, etc.\nIncompatible with Nvim implementation.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "tmux", "sha": "cce56fa19b29ae2368390bdf2d464e97d61b59f3", "date": "2026-08-01T17:04:12Z", "subject": "Pessimize compiler flags for screen-redraw.c to prevent tmux from dumping core upon startup on landisk.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "51da48c8f89751ebb7e36aee42c55ae480872ada", "date": "2026-08-01T13:22:49-04:00", "subject": "build(vim-patch): libvterm color_ramp.vim is n/a for nvim", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "postgres", "sha": "602f19c84ca1db471b6264d4965e2becfc00cbbc", "date": "2026-08-01T21:35:19+02:00", "subject": "doc: Fix glossary entry for data checksums workers", "body": "The glossary entry for data checksums workers incorrectly stated that\nthey were auxiliary processes, but they are implemented as background\nworkers. Fix, and while there, simplify the entry by combining the\nworker and launcher into a single glossary term. Backpatch down to\nv19 where online checksums were introduced.\n\nAuthor: Daniel Gustafsson \nReported-by: Fujii Masao \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAHGQGwEv-C9ia+rBYyePzO8F=5FVvS412ZqcOupazuOb5RafNg@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "4afa9788b51625d19eee08a1420b48b04c0699fa", "date": "2026-08-01T21:35:51+02:00", "subject": "Add a comment to distinguish backend types", "body": "The data checksums entries were seemingly auxiliary processes from\nreading the code, but they are in fact background workers. Add a\ncomment to clarify. Backpatch down to v19 where online checksums\nwere introduced.\n\nAuthor: Daniel Gustafsson \nReported-by: Fujii Masao \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAHGQGwFsBjQs2fv7b72hxzGV_fJMh6LAg4E83pNfDOu1jVgWCA@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d03b4cfb7ab03bd00c04749d93ba630c267a3e82", "date": "2026-08-01T14:55:02-05:00", "subject": "feat(zip): checkhealth and encrypted entries", "body": "Problem:\nThere is no health check. Encrypted entries cannot be read at all, since\nInfo-ZIP takes a password only from a terminal and `-P` would expose it\nin the process arguments.\n\nSolution:\nAdd `:checkhealth nvim.zip`, reporting the backend and which\nimplementation is handling archives. Prompt for the password on a pty,\nextracting to a file so the entry's bytes stay off the terminal. Report\nInfo-ZIP's exit code rather than inferring a cause, so an archive using\nAES is not reported as a failed decryption.", "style": "conventional", "type": "feat", "scope": "zip", "breaking": false} {"repo": "neovim", "sha": "98d767cd5303f116c3f97eb44e05bb4aaad5b965", "date": "2026-08-01T17:48:32-04:00", "subject": "docs: func/expr options, misc #41102", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "baff4d07aa49881de2f3b43560cc2df26a1ec396", "date": "2026-08-02T00:50:21+03:00", "subject": "fix(docs-infra): keep inline code in links as an atomic inline box", "body": "Inline code inside a link renders as a chip with its own background. Text\ndecoration propagates from the anchor and cannot be cancelled by a descendant,\nso with the chip laid out as a regular inline box the link underline is drawn\nstraight across it.\n\nLay the chip out as an atomic inline box, which the propagated decoration is not\ndrawn across. This applies to the link rule only; code outside of links keeps\n`display: inline` for wrapping in multiline paragraphs, and the link rule\nalready sets `white-space: nowrap`.\n\n(cherry picked from commit efe1aae1cd31a6c107ec95d7d79114137d93e54d)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "efe1aae1cd31a6c107ec95d7d79114137d93e54d", "date": "2026-08-02T00:50:21+03:00", "subject": "fix(docs-infra): keep inline code in links as an atomic inline box", "body": "Inline code inside a link renders as a chip with its own background. Text\ndecoration propagates from the anchor and cannot be cancelled by a descendant,\nso with the chip laid out as a regular inline box the link underline is drawn\nstraight across it.\n\nLay the chip out as an atomic inline box, which the propagated decoration is not\ndrawn across. This applies to the link rule only; code outside of links keeps\n`display: inline` for wrapping in multiline paragraphs, and the link rule\nalready sets `white-space: nowrap`.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "neovim", "sha": "ad2ae824ecd4a8a8f9814de8b9d50944950442b2", "date": "2026-08-02T07:18:29+08:00", "subject": "fix(move): using wrong window width", "body": "", "style": "conventional", "type": "fix", "scope": "move", "breaking": false} {"repo": "openssl", "sha": "a0c8ec557d9cac078f032d76cdf684fe743eb382", "date": "2026-08-02T00:23:39Z", "subject": "Fix heap buffer overflow (8-byte OOB write) in AES-WRAP-PAD unwrap", "body": "On its integrity-failure paths that primitive writes and cleanses up to inlen\nbytes of the output buffer. Size the buffer for that worst case so a failed\nunwrap cannot write past the allocation.\n\nFixes CVE-2026-63072\n\nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMerge-date: Mon Aug 24 14:44:55 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9ec2f6d2ae2bcad907cf7ee38584855bafe4979a", "date": "2026-08-02T00:23:39Z", "subject": "Fix heap buffer overflow (8-byte OOB write) in AES-WRAP-PAD unwrap", "body": "On its integrity-failure paths that primitive writes and cleanses up to inlen\nbytes of the output buffer. Size the buffer for that worst case so a failed\nunwrap cannot write past the allocation.\n\nFixes CVE-2026-63072\n\nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMerge-date: Mon Aug 24 14:44:05 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "87784ad619af36b8807c2044b3940006fccc1e42", "date": "2026-08-02T00:23:39Z", "subject": "Fix heap buffer overflow (8-byte OOB write) in AES-WRAP-PAD unwrap", "body": "On its integrity-failure paths that primitive writes and cleanses up to inlen\nbytes of the output buffer. Size the buffer for that worst case so a failed\nunwrap cannot write past the allocation.\n\nFixes CVE-2026-63072\n\nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMerge-date: Mon Aug 24 14:42:05 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "45e594bd87056a2d765d3d69f0362e342078edc4", "date": "2026-08-02T08:35:15+08:00", "subject": "fix(runtime-vapor): constrain expose to Exposed type", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "88fcf1ec723e4e52c1f2cb874e9332a9397287bb", "date": "2026-08-01T17:54:43-07:00", "subject": "fix(runtime-vapor): remove special boolean attributes for falsy values", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "2d0e207175ea02a7eb938b38452e4215d59cb9c7", "date": "2026-08-02T12:58:32+08:00", "subject": "fix(runtime-vapor): defer KeepAlive branch removal until cache pruning", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "60e3b57d1c953fca597f29fa4148d57d196eabe0", "date": "2026-08-02T13:26:51+08:00", "subject": "fix(runtime-vapor): unmount leaving KeepAlive cache entries", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "openssl", "sha": "9b53f8873a96574553f169c58e2f4b6c66070be7", "date": "2026-08-02T14:41:30+09:00", "subject": "doc: document QUIC thread cancellation consequences", "body": "Describe the domain-wide consequences of interrupting QUIC API\nprocessing while synchronisation is held or notification state is\nregistered.\n\nClarify that SSL_poll() can leave stale registrations in multiple\nnotifier-enabled domains, and provide cooperative cancellation and\napplication-driven event-loop guidance.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Wed Aug 19 10:32:14 2026\nMerged-from: https://github.com/openssl/openssl/pull/32139", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dba14047ed3b1b96a139d535e7d4d8b5e4ade24f", "date": "2026-08-02T14:41:30+09:00", "subject": "doc: document thread cancellation and non-local exits", "body": "Document that OpenSSL provides no general guarantee that an\nin-progress call can be abandoned safely through thread cancellation,\nforced termination, or another non-local exit.\n\nExplain indirect cancellation points, the limits of thread-local\ncleanup, POSIX cancellation masking, and cooperative cancellation.\nClarify in OPENSSL_init_crypto(3) that automatic thread cleanup does\nnot unwind an interrupted call.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Wed Aug 19 10:32:13 2026\nMerged-from: https://github.com/openssl/openssl/pull/32139", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e375de560d4ef053002db1d1cc71c89e898a87eb", "date": "2026-08-02T14:41:30+09:00", "subject": "doc: document QUIC thread cancellation consequences", "body": "Describe the domain-wide consequences of interrupting QUIC API\nprocessing while synchronisation is held or notification state is\nregistered.\n\nClarify that SSL_poll() can leave stale registrations in multiple\nnotifier-enabled domains, and provide cooperative cancellation and\napplication-driven event-loop guidance.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Wed Aug 19 10:31:46 2026\nMerged-from: https://github.com/openssl/openssl/pull/32139", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "84e613df3a0b660280fb037b9070e866fbe013b8", "date": "2026-08-02T14:41:30+09:00", "subject": "doc: document thread cancellation and non-local exits", "body": "Document that OpenSSL provides no general guarantee that an\nin-progress call can be abandoned safely through thread cancellation,\nforced termination, or another non-local exit.\n\nExplain indirect cancellation points, the limits of thread-local\ncleanup, POSIX cancellation masking, and cooperative cancellation.\nClarify in OPENSSL_init_crypto(3) that automatic thread cleanup does\nnot unwind an interrupted call.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Wed Aug 19 10:31:45 2026\nMerged-from: https://github.com/openssl/openssl/pull/32139", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "475cd6c34bc7111b6beea35dc74742fce86145dd", "date": "2026-08-02T14:41:30+09:00", "subject": "doc: document QUIC thread cancellation consequences", "body": "Describe the domain-wide consequences of interrupting QUIC API\nprocessing while synchronisation is held or notification state is\nregistered.\n\nClarify that SSL_poll() can leave stale registrations in multiple\nnotifier-enabled domains, and provide cooperative cancellation and\napplication-driven event-loop guidance.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Wed Aug 19 10:31:04 2026\nMerged-from: https://github.com/openssl/openssl/pull/32139", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e99e4e89c1c30b40502d61efbf4b16a98c22f406", "date": "2026-08-02T14:41:30+09:00", "subject": "doc: document thread cancellation and non-local exits", "body": "Document that OpenSSL provides no general guarantee that an\nin-progress call can be abandoned safely through thread cancellation,\nforced termination, or another non-local exit.\n\nExplain indirect cancellation points, the limits of thread-local\ncleanup, POSIX cancellation masking, and cooperative cancellation.\nClarify in OPENSSL_init_crypto(3) that automatic thread cleanup does\nnot unwind an interrupted call.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Wed Aug 19 10:31:03 2026\nMerged-from: https://github.com/openssl/openssl/pull/32139", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a0605a85b24e135f2da4d870c62b6a7b163e18f2", "date": "2026-08-02T14:41:30+09:00", "subject": "doc: document QUIC thread cancellation consequences", "body": "Describe the domain-wide consequences of interrupting QUIC API\nprocessing while synchronisation is held or notification state is\nregistered.\n\nClarify that SSL_poll() can leave stale registrations in multiple\nnotifier-enabled domains, and provide cooperative cancellation and\napplication-driven event-loop guidance.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Wed Aug 19 10:28:50 2026\nMerged-from: https://github.com/openssl/openssl/pull/32139", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "76ef06e5266f79d46b273ad8146cafeee22bd86b", "date": "2026-08-02T14:41:30+09:00", "subject": "doc: document thread cancellation and non-local exits", "body": "Document that OpenSSL provides no general guarantee that an\nin-progress call can be abandoned safely through thread cancellation,\nforced termination, or another non-local exit.\n\nExplain indirect cancellation points, the limits of thread-local\ncleanup, POSIX cancellation masking, and cooperative cancellation.\nClarify in OPENSSL_init_crypto(3) that automatic thread cleanup does\nnot unwind an interrupted call.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Wed Aug 19 10:28:49 2026\nMerged-from: https://github.com/openssl/openssl/pull/32139", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "03c0a53bcab16a1a074f89aa411bec77dbd0a4c4", "date": "2026-08-02T14:27:27+08:00", "subject": "fix(runtime-vapor): preserve move semantics for VDOM roots", "body": "- forward fragment move types to VDOM interop roots\n- apply current transition hooks during enter and leave moves", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "c59b06c99ce1663560caf0147a11eb05c4b30689", "date": "2026-08-02T11:09:40+02:00", "subject": "sshserver.pl: bump an sshd config to use its modern name", "body": "Ref: https://github.com/openssh/openssh-portable/commit/ee9c0da8035b3168e8e57c1dedc2d1b0daf00eec\n\nCloses #22460", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "fe10c5bc361c12b455db2617cf094eb2b3655631", "date": "2026-08-02T05:09:58-04:00", "subject": "test(fs): \"fs.find() follows symlinks\" #41105", "body": "Problem:\nTest fails if other \"build*/\" dirs exist with a \"nvim\" file, e.g.\n`build-asan/bin/nvim`.\n\nSolution:\nDo the test in an isolated dir.\nThis also fixes the test for the zig build.", "style": "conventional", "type": "test", "scope": "fs", "breaking": false} {"repo": "tmux", "sha": "80f2ab72377aeaa4ccf9e75e6562f9d1c45e8fd6", "date": "2026-08-02T10:36:49+01:00", "subject": "Add missing .h file.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "825b2c39b8f8fb7b34a85baaf72dd4d6edbe9906", "date": "2026-08-02T10:38:26+01:00", "subject": "Fix copyrights.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0852a39e31f18a5b44224567809c6312e2420a72", "date": "2026-08-02T10:41:59+01:00", "subject": "Move image-diacritics.h into image-kitty.c.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "8fc9c3a5f946f2d354e25e015eb358794af2ff73", "date": "2026-08-02T12:37:09+01:00", "subject": "Add missing regression test files.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "7d2249c579b0a9a19982170823675e25f3fba070", "date": "2026-08-02T14:00:48+02:00", "subject": "docs: misc, :bcd, slug()", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "3b85a689ab55312e42646a569252545bfdd161dc", "date": "2026-08-02T08:19:29-04:00", "subject": "docs(api): rpc examples #41077", "body": "", "style": "conventional", "type": "docs", "scope": "api", "breaking": false} {"repo": "tmux", "sha": "5432f2ef0a42945ef66ad7209d36f3edb20e5452", "date": "2026-08-02T16:14:12+01:00", "subject": "Add tool to generate the kitty diacritics directly from the public UnicodeData.txt file.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "812cc1a734bb86ab6cedfd168e3e415eed463e3b", "date": "2026-08-02T11:26:30-04:00", "subject": "Update time zone data files to tzdata release 2026c.", "body": "Alberta (America/Edmonton) moved to permanent UTC-06 on\n2026-06-18, which will affect their clocks beginning on 2026-11-01.\nFor lack of any clarity on the point, assume their TZ abbreviation\nwill be CST from that time forward.\n\nMorocco (Africa/Casablanca) will move to permanent UTC+00,\nwithout daylight saving time transitions, on 2026-09-20.\n\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "cfc720ef4b1d1643f73ffd60621a7a332a7faf7b", "date": "2026-08-02T13:22:39-04:00", "subject": "Fix memory-safety bugs in the ispell/hunspell dictionary loader.", "body": "Allocate CompoundAffix with room for its terminator, initialize the\nold-format flag buffer before NIAddAffix(), and reject incomplete or\nmissing Hunspell AF aliases. None of these errors would be likely to\ntrigger on real dictionary files, accounting for the lack of previous\nreports; but they're certainly bugs.\n\nBug: #19595\nReported-by: Michael Malis \nAuthor: Andrey Rachitskiy \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/19595-7dc18b4e212c4757@postgresql.org\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9975b73dfb7490fa9237f3213e998a8c59eebf35", "date": "2026-08-02T20:19:50+02:00", "subject": "fix(optwin): handle tabpage local options", "body": "", "style": "conventional", "type": "fix", "scope": "optwin", "breaking": false} {"repo": "neovim", "sha": "ec957ffd8c378918b8f544fc9fa588b9d56b1c6c", "date": "2026-08-02T20:22:40+02:00", "subject": "refactor(optwin): use current window as fallback", "body": "", "style": "conventional", "type": "refactor", "scope": "optwin", "breaking": false} {"repo": "neovim", "sha": "ecae8c994e136733239c542720535d3921cfc780", "date": "2026-08-03T02:30:23+08:00", "subject": "ci: check FreeBSD, OpenBSD builds #41091", "body": "FreeBSD and OpenBSD are Tier 2 supported platforms (:help\nsupported-platforms) but have had no CI coverage since Cirrus CI was\ndropped in #39321. Restore build coverage for both using vmactions,\nwhich boots the VM on a regular ubuntu-latest runner.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "neovim", "sha": "2f09090134455a75d09621821ab31f87f2c9c908", "date": "2026-08-02T19:37:26+01:00", "subject": "fix(ui2): `:$q` should not target hidden windows #40992", "body": "Similar to #36123, the ui2 windows confuse `:$q`, which doesn't close\nthe last window on the tab, but targets a hidden ui2 window.\n\nThe problem seems to be a combination of using\n`FOR_ALL_WINDOWS_IN_TAB()`, which includes hidden windows, and then\ncounting the windows without factoring in `win_has_winnr()`.\n\nThere are examples of using `win_has_winnr()`, such as:\n- [`eval/buffer.c`'s `buf_win_common()`](https://github.com/neovim/neovim/blob/4a5062cda69fa6cb59b7dc0c488dfb05accf99de/src/nvim/eval/buffer.c#L474-L475)\n- [`eval/window.c`'s `f_getwininfo()`](https://github.com/neovim/neovim/blob/4a5062cda69fa6cb59b7dc0c488dfb05accf99de/src/nvim/eval/window.c#L130-L135)\n- [`window.c`'s `win_get_tabwin()`](https://github.com/neovim/neovim/blob/4a5062cda69fa6cb59b7dc0c488dfb05accf99de/src/nvim/window.c#L7860-L7869)\n\nBut possibly problematic ones:\n- `ex_docmd.c`'s:\n\t- [`invalid_range()`](https://github.com/neovim/neovim/blob/4a5062cda69fa6cb59b7dc0c488dfb05accf99de/src/nvim/ex_docmd.c#L3891-L3893)\n\t\t- ^ should this include hidden windows? Can a user address them by number? `win_has_winnr()` suggests not\n\t- [`ex_close()`](https://github.com/neovim/neovim/blob/4a5062cda69fa6cb59b7dc0c488dfb05accf99de/src/nvim/ex_docmd.c#L5233-L5244)\n\t- [`ex_hide()`](https://github.com/neovim/neovim/blob/4a5062cda69fa6cb59b7dc0c488dfb05accf99de/src/nvim/ex_docmd.c#L5478-L5484)\n- `eval/window.c`'s:\n\t- [`f_winrestcmd()`](https://github.com/neovim/neovim/blob/4a5062cda69fa6cb59b7dc0c488dfb05accf99de/src/nvim/eval/window.c#L797-L808)\n\t- [`find_win_by_nr()`](https://github.com/neovim/neovim/blob/4a5062cda69fa6cb59b7dc0c488dfb05accf99de/src/nvim/eval/window.c#L172-L180)", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "neovim", "sha": "ebcb61b0f9d4d52e1d1b5efb6a5765f4f0be2542", "date": "2026-08-02T11:42:47-07:00", "subject": "fix(treesitter): make TSHighlighter.new() idempotent for an active buffer #41090", "body": "Problem:\nCalling vim.treesitter.start() a second time on a buffer that already\nhas an active TSHighlighter creates a brand new instance instead of\nreusing it, whether the parser tree is unchanged (e.g. calling start()\ntwice) or different (e.g. switching languages). Either way the old\ninstance is silently discarded without calling :destroy() on it, so\nits on_bytes/on_changedtree/on_detach callbacks stay registered and\nits buffer-local state (spelloptions, decoration namespace) is never\nrestored, both leaking indefinitely for an orphaned instance that\nnothing references anymore.\n\nSolution:\nReturn the existing instance when TSHighlighter.active[source] is\nalready set for the same parser tree, instead of unconditionally\nconstructing a new one. When the tree differs instead (e.g. a language\nswitch), destroy() the old instance first, matching stop() semantics,\nbefore constructing the new one.\n\n(cherry picked from commit 67839a72f7eba1454a1a3755a55bca5e6562e461)", "style": "conventional", "type": "fix", "scope": "treesitter", "breaking": false} {"repo": "neovim", "sha": "67839a72f7eba1454a1a3755a55bca5e6562e461", "date": "2026-08-02T11:42:47-07:00", "subject": "fix(treesitter): make TSHighlighter.new() idempotent for an active buffer #41090", "body": "Problem:\nCalling vim.treesitter.start() a second time on a buffer that already\nhas an active TSHighlighter creates a brand new instance instead of\nreusing it, whether the parser tree is unchanged (e.g. calling start()\ntwice) or different (e.g. switching languages). Either way the old\ninstance is silently discarded without calling :destroy() on it, so\nits on_bytes/on_changedtree/on_detach callbacks stay registered and\nits buffer-local state (spelloptions, decoration namespace) is never\nrestored, both leaking indefinitely for an orphaned instance that\nnothing references anymore.\n\nSolution:\nReturn the existing instance when TSHighlighter.active[source] is\nalready set for the same parser tree, instead of unconditionally\nconstructing a new one. When the tree differs instead (e.g. a language\nswitch), destroy() the old instance first, matching stop() semantics,\nbefore constructing the new one.", "style": "conventional", "type": "fix", "scope": "treesitter", "breaking": false} {"repo": "ant-design", "sha": "beb04531191dfdf0ce316d7ccd06049ddfd9e3ac", "date": "2026-08-02T22:16:00+03:00", "subject": "fix(Breadcrumb): align and space bare `` icons", "body": "A breadcrumb item `title` from a third-party icon library (lucide, react-icons,\nphosphor) renders as a bare `` inside `.ant-breadcrumb-link`, which the\n`.anticon` reset never reaches. An `` has no baseline of its own, so it is\naligned by its bottom margin edge (CSS 2.1 §10.8.1) and rides above the link\ntext — by an amount that grows with the icon's size, so a constant nudge like\n`.anticon`'s `-0.125em` cannot fix it. The existing icon-to-text spacing rule\nwas also `.anticon`-only, so a bare `` sat flush against the label.\n\nCenter and space it like an `.anticon`: `.ant-breadcrumb-link > svg { display:\ninline-block; vertical-align: middle; margin-block-end: 0.2em }`, and extend the\nicon+text spacing selector to `> svg + span` / `> svg + a`. `display: inline-block`\nkeeps the svg an atomic inline box so `vertical-align` still applies even under a\nCSS reset that forces `svg { display: block }` (e.g. Tailwind Preflight), which\nwould otherwise drop the icon onto its own line. `middle` centres the margin box\non the x-height line; `margin-block-end` lifts it half its own value onto the\ncap-height centre (capHeight − xHeight ≈ 0.2em), keeping it centred at any size.\nThe `> svg` child selector never matches an `.anticon` (its `` is one level\ndeeper), so existing icons stay byte-identical. Mirrors the Tag fix (#58723).", "style": "conventional", "type": "fix", "scope": "Breadcrumb", "breaking": false} {"repo": "ant-design", "sha": "38ca44523d722d7e0b4fa6dc8ff201c5af3125a4", "date": "2026-08-02T22:17:38+03:00", "subject": "fix(Collapse): center bare `` icons at any size", "body": "A panel title `label` from a third-party icon library (lucide, react-icons,\nphosphor) renders as a bare `` inside `.ant-collapse-title`, which the\n`.anticon` reset never reaches. An `` has no baseline of its own, so it is\naligned by its bottom margin edge (CSS 2.1 §10.8.1) and rides above the title\ntext — by an amount that grows with the icon's size, so a constant nudge like\n`.anticon`'s `-0.125em` cannot fix it.\n\nCenter it instead: `.ant-collapse-title > svg { display: inline-block; vertical-\nalign: middle; margin-block-end: 0.2em }`. `display: inline-block` keeps the svg\nan atomic inline box so `vertical-align` still applies even under a CSS reset\nthat forces `svg { display: block }` (e.g. Tailwind Preflight), which would\notherwise drop the icon onto its own line. `middle` centres the margin box on the\nx-height line; `margin-block-end` then lifts it half its own value onto the\ncap-height centre (capHeight − xHeight ≈ 0.2em), keeping it centred at any size.\nThe `> svg` child selector never matches an `.anticon` (its `` is one level\ndeeper), and the expand arrow lives in a separate `.ant-collapse-expand-icon`,\nso both stay byte-identical.", "style": "conventional", "type": "fix", "scope": "Collapse", "breaking": false} {"repo": "ant-design", "sha": "49c4a03cc93035087534d2b34f6ba257ac28e2a0", "date": "2026-08-02T22:24:28+03:00", "subject": "fix: keep bare `` icons aligned under a `display: block` reset", "body": "The bare-`` icon-alignment fixes shipped for Tag (#58723), Tabs (#58847)\nand Segmented (#58862) rely on the icon being an inline-level box so that\n`vertical-align: middle` applies. Under a CSS reset that forces\n`svg { display: block }` — most notably Tailwind Preflight, and the ant.design\ndocs site's own reset — the bare `` becomes block-level, which *ignores*\n`vertical-align` entirely and breaks the icon onto its own line. The icon then\nrides a full line-height above the label (measured -18 to -24px), which is why\nthe released Tag/Tabs icon demos look broken on ant.design even though the fix\nworks in a reset-free app.\n\nAdd `display: inline-block` to each `> svg` rule so the icon stays an atomic\ninline box: `vertical-align` applies again and the icon no longer wraps. The\nrule's specificity (`.ant-x > svg`) beats a bare `svg { display: block }`, and\n`@ant-design/icons` icons are unaffected (an `.anticon` keeps its `` one\nlevel deeper, so `> svg` never matches it). No behavior change without a reset.", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "postgres", "sha": "dda622edcdba673e90d7b2cafe747ab78249368b", "date": "2026-08-02T16:49:18-04:00", "subject": "Tighten up TS dictionary cache entry creation.", "body": "In the not-too-likely scenario where we successfully created a hash\ntable entry for a TS dictionary, but then failed to make a small\nmemory context for it, we left the hash entry in existence but with\na garbage value for dictCtx. This confused the code the next time\nthrough, leading to a crash. Rearrange things so that we leave\nthe hash entry in a well-defined state with dictCtx == NULL, and\nthen the next try knows it still needs to make a memory context.\n\nReported-by: Alexander Lakhin \nAuthor: Tom Lane \nDiscussion: https://postgr.es/m/0f3ddeb5-0dbd-479c-9d0e-ae254758e624@gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "54371bca75051f6a0af4b5edcb2f1fe59af1198e", "date": "2026-08-03T00:53:00+02:00", "subject": "lib: update mentions of the legacy \"sessionhandle\"", "body": "It is now \"Curl_easy\"\n\nFollow-up to 434f8d0389 (June 2016)\n\nCloses #22463", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "8a8ff47b63aff1d72c1796dbdb4955f243beac23", "date": "2026-08-03T02:28:33+02:00", "subject": "tidy-up: C and CI", "body": "- drop redundant casts for `CURL_UNCONST()` pointers (Windows).\n- GHA/linux: delete redundant/dupe `-g` C flag in memory sanitizer job.\n Spotted-by: Daniel Stenberg\n- examples/http2-upload: drop local macro.\n- examples/http2-upload: drop leading underscore from union name.\n- examples/http2-upload: drop interim macro.\n- tests/server/util: reapply patch lost in a rebase.\n- sync `main()` declarations across the codebase.\n\nCloses #22559", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "c49b688b2e51d66995beb69f57adf0563158e741", "date": "2026-08-03T09:25:10+08:00", "subject": "fix(types): allow Vapor JSX Element in defineComponent render functions", "body": "", "style": "conventional", "type": "fix", "scope": "types", "breaking": false} {"repo": "ant-design", "sha": "31b11bc9702deb99141378a7a4efa2801bc69301", "date": "2026-08-03T09:28:39+08:00", "subject": "fix(site): avoid Markdown route conflicts in Surge previews", "body": "* fix(site): avoid Markdown route conflicts in Surge previews\n\n* update", "style": "conventional", "type": "fix", "scope": "site", "breaking": false} {"repo": "ant-design", "sha": "8c16d19c3ad1260350d58ef9313ec7089dbcca42", "date": "2026-08-03T09:28:49+08:00", "subject": "type: typeScript definition improvement", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "40a85fe9340150efe467006f7f044ee82cb45ebd", "date": "2026-08-03T09:43:39+08:00", "subject": "fix(site): avoid Markdown route conflicts in Surge previews", "body": "", "style": "conventional", "type": "fix", "scope": "site", "breaking": false} {"repo": "vue-core", "sha": "dff8dd5badaddea39f91da674be84d6af6f0f857", "date": "2026-08-03T10:36:52+08:00", "subject": "fix(vapor): align component v-text and v-html with vdom", "body": "Compile component v-text and v-html directives as textContent and innerHTML\nprops so reactive updates and attribute fallthrough follow the normal\ncomponent prop flow.\n\nPreserve values that VDOM considers constant, escape literal v-text content\nembedded in native element templates, and remove the obsolete setBlockText\nand setBlockHtml runtime helpers.", "style": "conventional", "type": "fix", "scope": "vapor", "breaking": false} {"repo": "angular", "sha": "2a2585708a68ffa565a839ac17003d5b69b6d885", "date": "2026-08-02T22:38:19-04:00", "subject": "fix(devtools): clarify disabled View Source tooltip and update demo app", "body": "Before: The View Source button tooltip in signal details static ('View source') even when disabled.\n\nAfter: The tooltip explains 'Source location is not available for this node' when disabled.", "style": "conventional", "type": "fix", "scope": "devtools", "breaking": false} {"repo": "vite", "sha": "7673c02e53343ae9356c1f496c1c1da2eb732ac1", "date": "2026-08-03T11:48:27+09:00", "subject": "test: reduce logs", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "729d4367d750b665c993c734c965472f608f9ab5", "date": "2026-08-03T10:59:07+08:00", "subject": "docs: update dingding qr-code", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "1331b0b438b1e7193effb7d2341660bccb9c3155", "date": "2026-08-03T12:03:29+09:00", "subject": "perf(css): look up pure CSS chunks through a Set", "body": "", "style": "conventional", "type": "perf", "scope": "css", "breaking": false} {"repo": "vite", "sha": "b4bf59686a7ac238929e91a6e1708c739b843a2f", "date": "2026-08-03T11:10:20+08:00", "subject": "fix: don't mutate the user config when resolving the lib entry from the top-level `input`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "24e29bcec975e3783c141abb38770d73847c6670", "date": "2026-08-03T11:13:17+08:00", "subject": "fix(App): warn about ignored root props", "body": "* fix(App): remove css variable warning for component={false}\n\n* update\n\n* update", "style": "conventional", "type": "fix", "scope": "App", "breaking": false} {"repo": "vite", "sha": "40a3999e66b5b51f45ec0ead60265f3b2184e780", "date": "2026-08-03T17:01:15+12:00", "subject": "docs: remove nonexistent `RemoteEnvironmentTransport` import in environment API example", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "721ce4f6bd3c1f9400d45231b16a059531a16f6c", "date": "2026-08-03T14:12:31+08:00", "subject": "fix(compiler-vapor): generate v-model after prop effects", "body": "Generate built-in v-model operations after initial DOM prop operations and\nreactive effects. This ensures dynamic input types are applied before\napplyDynamicModel selects the text, checkbox, or radio implementation.\n\nAdd codegen coverage for constant and reactive input types, along with runtime\nregressions for initially dynamic checkbox and radio inputs.", "style": "conventional", "type": "fix", "scope": "compiler-vapor", "breaking": false} {"repo": "vite", "sha": "89620f09afcfef6b35e7bb8660132ab5b4d0cd3b", "date": "2026-08-03T14:20:35+08:00", "subject": "docs(static-deploy): remove kinsta", "body": "", "style": "conventional", "type": "docs", "scope": "static-deploy", "breaking": false} {"repo": "ant-design", "sha": "ba159013cb12237c9994dabde0d54ad502f9c738", "date": "2026-08-03T14:34:41+08:00", "subject": "fix(BackTop): disable scroll animation for reduced motion", "body": "", "style": "conventional", "type": "fix", "scope": "BackTop", "breaking": false} {"repo": "postgres", "sha": "3fb599b26411a1322dec89759349ce4920e1055a", "date": "2026-08-03T15:51:59+09:00", "subject": "Fix nullability check for a sub-select's upper-level Vars", "body": "When checking whether a sub-select's output columns can produce NULL,\nso as to decide whether a NOT IN can be converted to an anti-join,\nquery_outputs_are_not_nullable() falls back on find_nonnullable_vars()\nfor targetlist entries that are plain Vars: if the sub-select's own\nquals prove the Var non-null, the output is non-nullable. But that\ntest compared only varno and varattno, without checking varlevelsup.\nAn outer reference in the targetlist could thus be matched against a\nVar of the sub-select's own range table that happens to share the same\nvarno and varattno, wrongly proving the output non-nullable and\nallowing an invalid conversion to an anti-join, which yields wrong\nanswers when the outer reference is NULL.\n\nTo fix, restrict the fallback to Vars of the current query level.\n\nAuthor: Rui Zhao \nReviewed-by: Tender Wang \nReviewed-by: Richard Guo \nDiscussion: https://postgr.es/m/CAHWVJhGuaFFRpmq4j+mcMcm_HC5QOT7LZsC9bf9b7BCBmvbfMA@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "329938c1c4e9d1559dcd8123ac093f76b64f8d0c", "date": "2026-08-03T17:27:09+10:00", "subject": "Additional ML-KEM cleansing", "body": "Also, mark X448MLKEM1024 as FIPS non-approved (backport of #26220)\n\nReviewed-by: Tim Hudson \nReviewed-by: Milan Broz \nMergeDate: Tue Aug 11 15:06:57 2026\n(Merged from https://github.com/openssl/openssl/pull/32167)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "65c4a6765e49f8b7c58790f6fda1addd567fc7fa", "date": "2026-08-03T17:27:09+10:00", "subject": "Additional ML-KEM cleansing", "body": "Also, mark X448MLKEM1024 as FIPS non-approved (backport of #26220)\n\nReviewed-by: Milan Broz \nReviewed-by: Neil Horman \nReviewed-by: Tim Hudson \nMergeDate: Tue Aug 11 15:05:35 2026\n(Merged from https://github.com/openssl/openssl/pull/32152)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "99969c550d5ea189144ed5fcf59daf7a5632123d", "date": "2026-08-03T17:27:09+10:00", "subject": "Additional ML-KEM cleansing", "body": "Also, mark X448MLKEM1024 as FIPS non-approved (backport of #26220)\n\nReviewed-by: Neil Horman \nReviewed-by: Milan Broz \nMergeDate: Tue Aug 11 13:56:01 2026\n(Merged from https://github.com/openssl/openssl/pull/32159)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e0b716092d83adb21524c10af4254c043c2a47c7", "date": "2026-08-03T17:27:09+10:00", "subject": "Additional ML-KEM cleansing", "body": "Also, mark X448MLKEM1024 key management as FIPS non-approved (completion of #26220)\n\nReviewed-by: Tim Hudson \nReviewed-by: Bob Beck \nReviewed-by: Milan Broz \nMergeDate: Tue Aug 11 06:57:33 2026\n(Merged from https://github.com/openssl/openssl/pull/32148)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "d8da5278822783ed05635685e75b636f3877c07d", "date": "2026-08-03T09:13:45+01:00", "subject": "Refactor to centralize image-specific rendering and lifecycle code in image.c and the image-* backends.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d2ac26eb9b1e8ad19fd526efb0ffff4fae5ed56f", "date": "2026-08-03T10:14:30+02:00", "subject": "Fix missing space before WHERE in GRAPH_TABLE deparse", "body": "get_graph_pattern_def() emitted the pattern-level WHERE keyword as\n\"WHERE \" with no leading space, so reverse-parsing produced output\nlike \"(o IS orders)WHERE (...)\". The element-level WHERE deparse in\nget_path_pattern_expr_def() already prepends a separating space; the\npattern-level branch was inconsistent with it. Emit \" WHERE \" to\nmatch.\n\nThe output still re-parses to the same tree, so this is cosmetic.\n\nFor test coverage, add a whole-pattern WHERE clause to the existing\ncustomers_us view, which is already reverse-parsed with\npg_get_viewdef().\n\nAuthor: Dhruv Chauhan \nReviewed-by: Ashutosh Bapat \nDiscussion: https://www.postgresql.org/message-id/flat/CANWwWcpHb0h7tg6otRnL-FV83jwQpAiyw1bhvv8T78kpwZ-0ow%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "4e379f38d683eeab237bd81343b03ae445976ed4", "date": "2026-08-03T09:24:16+01:00", "subject": "Remove local image ASCII test tool references", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "08679d89f53f76dd8e182286c7bac9b8710684b4", "date": "2026-08-03T10:37:28+02:00", "subject": "multi: remove #if 0'ed code that uses old struct", "body": "Closes #22467", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "fa8c5789a6efb552ef6dd1690ed9534b43e8ccbb", "date": "2026-08-03T09:53:43+01:00", "subject": "Sleep after kill.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "87400d67b0fc6a28eedad2ef28e237aa19c9a746", "date": "2026-08-03T09:55:25+01:00", "subject": "Improve image fallback rendering", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "b1e88d1ed4a382f7f3e9756600150954a8f55f45", "date": "2026-08-03T02:20:04-07:00", "subject": "fix(ui): highlight bleeds into 'linebreak' filler #41072", "body": "Problem:\nWhen 'linebreak' pushes a word entirely to the next screen row, the\nfiller cells left on the current row keep whatever highlight was set\nby the last real character before the break, even when that highlight\nshould not extend past it (e.g. an underline, which looks broken drawn\nover blank cells).\n\nSolution:\nReset decor_attr and area_attr at the filler when their attribute has\nan underline, undercurl, strikethrough, or overline; otherwise leave\nthem, since a plain background or reverse-video highlight looks\ncorrect extending through blank filler cells, regardless of where the\npushed-down word happens to end. search_attr keeps the same check,\nplus its pre-existing on_last_col case (its own match ending exactly\nhere).\n\n(cherry picked from commit 9ff302d0ca260ffb06b1f1815ac4db1c63d1a9b3)", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "neovim", "sha": "9ff302d0ca260ffb06b1f1815ac4db1c63d1a9b3", "date": "2026-08-03T02:20:04-07:00", "subject": "fix(ui): highlight bleeds into 'linebreak' filler #41072", "body": "Problem:\nWhen 'linebreak' pushes a word entirely to the next screen row, the\nfiller cells left on the current row keep whatever highlight was set\nby the last real character before the break, even when that highlight\nshould not extend past it (e.g. an underline, which looks broken drawn\nover blank cells).\n\nSolution:\nReset decor_attr and area_attr at the filler when their attribute has\nan underline, undercurl, strikethrough, or overline; otherwise leave\nthem, since a plain background or reverse-video highlight looks\ncorrect extending through blank filler cells, regardless of where the\npushed-down word happens to end. search_attr keeps the same check,\nplus its pre-existing on_last_col case (its own match ending exactly\nhere).", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "vue-core", "sha": "0aca1e0d84efaffacac03cabc00af92730985d2f", "date": "2026-08-03T17:27:10+08:00", "subject": "fix(runtime-vapor): avoid tracking v-show transition hooks", "body": "close #15202", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "3f08c4e3e800987c474b1ed64cc1ade31aaefd82", "date": "2026-08-03T17:27:55+08:00", "subject": "fix(runtime-vapor): avoid tracking v-show source in fragment effects", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "openssl", "sha": "54c77a1a85e0958db4ca4da26e39cab97d55160a", "date": "2026-08-03T19:28:48+10:00", "subject": "Additional ML-DSA cleansing", "body": "Reviewed-by: Tim Hudson \nReviewed-by: Milan Broz \nMergeDate: Tue Aug 11 15:06:58 2026\n(Merged from https://github.com/openssl/openssl/pull/32167)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8fc318aa18f761c24c8a23ef0f3db15c582764b7", "date": "2026-08-03T19:28:48+10:00", "subject": "Additional ML-DSA cleansing", "body": "Reviewed-by: Milan Broz \nReviewed-by: Neil Horman \nReviewed-by: Tim Hudson \nMergeDate: Tue Aug 11 15:05:36 2026\n(Merged from https://github.com/openssl/openssl/pull/32152)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "873179bdc668dd32fa022bb367c0e3fcbb97dbfc", "date": "2026-08-03T19:28:48+10:00", "subject": "Additional ML-DSA cleansing", "body": "Reviewed-by: Neil Horman \nReviewed-by: Milan Broz \nMergeDate: Tue Aug 11 13:56:04 2026\n(Merged from https://github.com/openssl/openssl/pull/32159)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "63c17a531c7cd7f6aae0ebef4eb318505a4af87e", "date": "2026-08-03T19:28:48+10:00", "subject": "Additional ML-DSA cleansing", "body": "Reviewed-by: Tim Hudson \nReviewed-by: Bob Beck \nReviewed-by: Milan Broz \nMergeDate: Tue Aug 11 06:57:34 2026\n(Merged from https://github.com/openssl/openssl/pull/32148)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "e8f781a447640c10f1fcb1220915fb59e0bc3a87", "date": "2026-08-03T17:30:34+08:00", "subject": "feat(Listy): add virtual list component", "body": "* feat: add Listy virtual list component\n\n* update\n\n* Update components/listy/index.en-US.md\n\nCo-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>\nSigned-off-by: Koyomi <1844749591@qq.com>\n\n* add demo\n\n* test(Listy): skip big-data demos in snapshot tests\n\n* test: update antd exports snapshot with Listy\n\n* fix(Listy): include item padding and border in itemHeight estimation\n\n* chore(Listy): remove code comments\n\n* test: update exports, site and vitest snapshots for Listy\n\n* test(Listy): skip demo snapshot tests for virtual rendering\n\n* test: support boolean skip option in node ssr demo test\n\n* docs(Listy): add global config column to API table\n\n* remove deprecated api\n\n* update api docs\n\n* update\n\n* fix(Listy): improve virtual config, itemHeight token and type support\n\n* test(Listy): exclude jest.mock based token test from vitest\n\n* docs: add 6.6.0 version\n\n* fix: apply code review\n\n* fix: fix lint\n\n* test: update snap\n\n* update demo\n\n* test: update Listy Vitest snapshot\n\n* docs: update docs\n\n* docs(Listy): adjust demos for non-virtual default and enable demo snapshots\n\n* docs(Listy): mark scroll-to demo as debug\n\n* test(Listy): add image snapshot test\n\n* docs(Listy): complete API metadata\n\n* fix(Listy): adjust itemHeight calculation\n\n---------", "style": "conventional", "type": "feat", "scope": "Listy", "breaking": false} {"repo": "neovim", "sha": "0aed30b4de042dce9941f789b6df1035b0fa73c8", "date": "2026-08-03T05:33:32-04:00", "subject": "build(vim-patch): misc n/a funcs, headers, err msgs #41116", "body": "* build(vim-patch): n/a func in version.c\n\ninit_longVersion() sets datetime for \":version\".\nNvim removed it for more deterministic builds.\n\n* build(vim-patch): n/a header from amiga\n\n* build(vim-patch): n/a vim9 err msg for declaring vars\n\n* build(vim-patch): nvim support.txt replaces *os-support*\n\n* build(vim-patch): vim xdg is n/a\n\n* build(vim-patch): n/a 1-line err msg\n\nVim error message can be 1-line (globals.h), not multi-line (errors.h).\nTarget v8.2.1032", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vue-core", "sha": "1f7d4cfe3a83858640ac50e23c73eb4dbae6922b", "date": "2026-08-03T17:34:55+08:00", "subject": "chore(runtime-vapor): remove duplicated import of setActiveSub", "body": "", "style": "conventional", "type": "chore", "scope": "runtime-vapor", "breaking": false} {"repo": "tmux", "sha": "4e7f3d5520fbb472f47a17fd58c5188e927f6739", "date": "2026-08-03T10:39:11+01:00", "subject": "Use sixel images for WezTerm", "body": "Enable the sixel feature when WezTerm is detected and clarify that the kitty terminal feature requires Unicode placeholder support rather than only the basic graphics protocol.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "52863f0fe963e23ef293cf39210d9bb5902cb991", "date": "2026-08-03T10:49:41+01:00", "subject": "Redraw images after pixel geometry changes", "body": "A resize may temporarily leave the terminal cell pixel dimensions unknown while tmux queries the terminal. Schedule another client redraw when the pixel geometry response arrives so the graphical backend replaces the temporary text fallback.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f4898759c5c311b7a3d2801e6363a62f51cd8671", "date": "2026-08-03T19:03:40+09:00", "subject": "doc: Clarify VERBOSE output for ANALYZE and VACUUM", "body": "ANALYZE VERBOSE and VACUUM VERBOSE both print per-table information\nat INFO level, including the table currently being processed and\nvarious statistics. However, the documentation described ANALYZE\nVERBOSE only as displaying generic \"progress messages\", while VACUUM\nVERBOSE used different wording about a vacuum activity report.\n\nReword the VERBOSE option descriptions and Outputs sections to describe\nthe actual output more consistently.\n\nAuthor: Shinya Kato \nReviewed-by: Surya Poondla \nReviewed-by: solai v \nReviewed-by: Maciek Sakrejda \nReviewed-by: David G. Johnston \nReviewed-by: Yushu Chen \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAOzEurTpMTUEW8kHu-zKB0EBtuPfpvyoJ--8pxKe87p24BGrpg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "4918360ea12cdc15f2a43ce021f7e0631eb46856", "date": "2026-08-03T10:07:57Z", "subject": "Exit failure on socket failure, from Rayan Salhab.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "fc9270c9430c18c4e114580c1005e1ae878cc9ec", "date": "2026-08-03T11:09:04+01:00", "subject": "Test for socket error from Rayan Salhab.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b3cd31965520616c77822347a1c7da8dfb46df84", "date": "2026-08-03T12:14:17+02:00", "subject": "curl_trc: remove unused expire timers", "body": "The expire timers\n- DNS_PER_NAME\n- DNS_PER_NAME2\n- HAPPY_EYEBALLS_DNS\n\nare unused since we changed our happy eyeballing and handling of partial\nresolve results.\n\nCloses #22468", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "0e44ef231a7b5d541fc0cc7ca89a3db1fa1b838a", "date": "2026-08-03T18:31:15+08:00", "subject": "fix(dir): do not swallow errors raised after the list callback", "body": "Problem: An error raised after the list handler ran (e.g. while\n rendering) unwinds through the pcall around provider.list()\n and is routed back into the now no-op handler: silently\n discarded.\nSolution: Re-raise the error when the handler has already run.", "style": "conventional", "type": "fix", "scope": "dir", "breaking": false} {"repo": "neovim", "sha": "d80121327fe15663d7f5367f03bb150c748243db", "date": "2026-08-03T18:31:15+08:00", "subject": "fix(dir): keep a listing buffer when the initial listing fails", "body": "Problem: A failed initial directory listing discards the listing state:\n the user is left in an empty modifiable non-listing buffer, and\n the BufEnter autocmd retries the failing open, repeating the\n error on every re-entry.\nSolution: Render the failure as an empty listing with state and handlers\n intact: the error is reported once and \"R\" or :edit retries the\n listing. Keep the error on the listing state so a failed\n listing is distinguishable from an empty one; the next list\n clears it.", "style": "conventional", "type": "fix", "scope": "dir", "breaking": false} {"repo": "curl", "sha": "1f860394e611e937b02d5ac72d4622583ec50e84", "date": "2026-08-03T12:51:02+02:00", "subject": "lib1560: add CURLU_NO_GUESS_SCHEME tests", "body": "Closes #22469", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "c4bdd8eca16cb2325a9a1b64f0bf6f76f7e99e38", "date": "2026-08-03T11:56:25+01:00", "subject": "Make sixel images grid resident", "body": "Replace screen-coordinate image placements with immutable pixel images referenced by ordinary grid cells. Draw visible marker runs through the redraw scene using per-client sixel or text backends so images follow history, copy mode, clipping and overwrites.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "762e329e83f2194969a6ea859ddd08361a18ac98", "date": "2026-08-03T20:05:23+09:00", "subject": "Add dshash_get_dsa_area(), able to retrieve the DSA area of a dshash table", "body": "Code using GetNamedDSHash() to retrieve a dshash table may also want to\nlimit its size, but there is actually no easy way to do so because the\ndsa_area of a hash table is hidden within dshash.c.\n\nThis commit adds dshash_get_dsa_area() to close the gap, which returns\nthe dsa_area for a given dshash_table. The DSA area retrieved can then\nbe passed to dsa_set_size_limit(), to limit its size. This accessor\nroutine also opens the door for allocating additional objects in the\nsame DSA area as a hash table, which may be useful for some.\n\nAuthor: Sami Imseih \nDiscussion: https://postgr.es/m/CAA5RZ0tKfCVqFnMZtavM42H63ha2Haf_C4mbJNWqkaW30cPW1w@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "9731adb30062fd156c1b666d166506e4f32ed5ca", "date": "2026-08-03T12:08:28+01:00", "subject": "Correct SIXEL HLS colour conversion", "body": "Use the DEC hue origin and channel order, and round percentage and HLS conversions to the nearest byte value.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "643bb662a00d1a851e438776d6491e5e7abb18fc", "date": "2026-08-03T12:14:49+01:00", "subject": "Preserve SIXEL pixel dimensions", "body": "Retain the padded canonical pixel canvas separately from the image content. Scale only the populated portion of partial edge cells so an image keeps its exact raster size on the originating terminal while retaining its cell footprint.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "2c7a5f73c4aa3afe5ae1f0b7c42818159c554065", "date": "2026-08-03T12:21:54+01:00", "subject": "Improve SIXEL colour quantization", "body": "Build an adaptive 256-colour palette with median cut instead of using a fixed colour cube. Apply Floyd-Steinberg error diffusion while mapping pixels to reduce banding and preserve image detail.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "c8fe307682a107ba96fd9f7e5781155d856cae2e", "date": "2026-08-03T12:31:11+01:00", "subject": "Support Kitty source rectangle placements", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a2b178d378c0e307bda0bd82028afbe546057f1a", "date": "2026-08-03T13:31:33+02:00", "subject": "tests: convert unit test 1396 and 1398 into libtests", "body": "They were previously unit tests but used only public library functions.\n\nCloses #22471", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dbb7cb4a983e5329b6a8a74898815533bcedb5f8", "date": "2026-08-03T21:45:36+10:00", "subject": "slh-dsa: cleanse temporary copies of sensitive data", "body": "Erase the temporaries that hold secrets or data derived from them:\nWOTS+ chain secrets, Merkle tree nodes, the message digest and\ncandidate FORS keys in sign and verify, the MGF1 seed and PRF-msg\noutput buffers, the encoded message (which may contain confidential\ncontent), and the caller's signature buffer when signing fails part\nway through. Also cleanse a private key of unexpected length that\nossl_slh_dsa_key_fromdata() copied into the key before rejecting it.\n\nThe SHA2 hash functions kept the digest and the concatenated child\nnodes in local stack buffers, abandoned on every call. Erasing them\nper call measurably slows the innermost functions, so move them into\na scratch buffer in SLH_DSA_HASH_CTX that is erased when the context\nis freed. The SHAKE hash functions write straight to the caller's\nbuffer and need no scratch.\n\nThe key material copies built by slh_dsa_export() were freed with\nOSSL_PARAM_free(), which only wipes the secure-heap data block. This\nbranch has no OSSL_PARAM_clear_free(), so wipe the copies explicitly\nbefore freeing.\n\nThe FORS roots buffer in ossl_slh_fors_pk_from_sig() was sized as if\nit held k * a nodes, but only one root per tree is ever written.\nShrink it to k * n bytes maximum and erase only the used length.\n\nReviewed-by: Tim Hudson \nReviewed-by: Milan Broz \nMergeDate: Tue Aug 11 15:07:00 2026\n(Merged from https://github.com/openssl/openssl/pull/32167)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7e5852f1a162706d6f5b40442e07fc6070ba6aea", "date": "2026-08-03T21:45:36+10:00", "subject": "slh-dsa: cleanse temporary copies of sensitive data", "body": "The hash functions worked on stack copies of the prehashed PK.seed context.\nDuring signing these absorb SK.seed and WOTS+ chain secrets, and every call\nleft the final hash state on the stack. Erasing it per call would burden the\ninnermost functions, which dominate signing time.\n\nGive SLH_DSA_HASH_CTX a heap-allocated scratch context and use it in place of\nthe stack copies. The working state now sits in one reusable buffer, erased\nwhen the hash context is freed; the per-call copy is unchanged and stack use\ngoes down. For the SHA2 parameter sets the buffer is sized to double as the\nSHA-512 context of security categories 3 and 5. The prehashed contexts are now\nfreed with OPENSSL_clear_free().\n\nErase the remaining temporaries holding secrets or data derived from them:\nWOTS+ and FORS secret values, Merkle tree nodes, the message digest and\ncandidate FORS keys in sign and verify, the encoded message, and the caller's\nsignature buffer when signing fails part way. Cleanse a wrong-length private\nkey that ossl_slh_dsa_key_fromdata() copied in before rejecting it. The FORS\nroots buffer held one root per tree but was sized for k * a nodes; shrink it to\nk * n bytes and erase only the used length.\n\nPerformance is unchanged within noise; the SHAKE parameter sets measure a few\npercent faster, likely because the working Keccak state now stays at one fixed\naddress rather than a fresh stack copy at each recursion depth.\n\nReviewed-by: Milan Broz \nReviewed-by: Neil Horman \nReviewed-by: Tim Hudson \nMergeDate: Tue Aug 11 15:05:38 2026\n(Merged from https://github.com/openssl/openssl/pull/32152)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3a95a436ee527d6f172c64d8cf3add9ed17bac06", "date": "2026-08-03T21:45:36+10:00", "subject": "slh-dsa: cleanse temporary copies of sensitive data", "body": "Erase the temporaries that hold secrets or data derived from them:\nWOTS+ chain secrets, Merkle tree nodes, the message digest and\ncandidate FORS keys in sign and verify, the MGF1 seed and PRF-msg\noutput buffers, the encoded message (which may contain confidential\ncontent), and the caller's signature buffer when signing fails part\nway through. Also cleanse a private key of unexpected length that\nossl_slh_dsa_key_fromdata() copied into the key before rejecting it.\n\nThe SHA2 hash functions kept the digest and the concatenated child\nnodes in local stack buffers, abandoned on every call. Erasing them\nper call measurably slows the innermost functions, so move them into\na scratch buffer in SLH_DSA_HASH_CTX that is erased when the context\nis freed. The SHAKE hash functions write straight to the caller's\nbuffer and need no scratch.\n\nThe key material copies built by slh_dsa_export() were freed with\nOSSL_PARAM_free(), which only wipes the secure-heap data block. This\nbranch has no OSSL_PARAM_clear_free(), so wipe the copies explicitly\nbefore freeing.\n\nThe FORS roots buffer in ossl_slh_fors_pk_from_sig() was sized as if\nit held k * a nodes, but only one root per tree is ever written.\nShrink it to k * n bytes maximum and erase only the used length.\n\nReviewed-by: Neil Horman \nReviewed-by: Milan Broz \nMergeDate: Tue Aug 11 13:56:06 2026\n(Merged from https://github.com/openssl/openssl/pull/32159)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7970e288ee43697956ae0a48ccb045155357a541", "date": "2026-08-03T21:45:36+10:00", "subject": "slh-dsa: cleanse temporary copies of sensitive data", "body": "The hash functions worked on stack copies of the prehashed PK.seed context.\nDuring signing these absorb SK.seed and WOTS+ chain secrets, and every call\nleft the final hash state on the stack. Erasing it per call would burden the\ninnermost functions, which dominate signing time.\n\nGive SLH_DSA_HASH_CTX a heap-allocated scratch context and use it in place of\nthe stack copies. The working state now sits in one reusable buffer, erased\nwhen the hash context is freed; the per-call copy is unchanged and stack use\ngoes down. For the SHA2 parameter sets the buffer is sized to double as the\nSHA-512 context of security categories 3 and 5. The prehashed contexts are now\nfreed with OPENSSL_clear_free().\n\nErase the remaining temporaries holding secrets or data derived from them:\nWOTS+ and FORS secret values, Merkle tree nodes, the message digest and\ncandidate FORS keys in sign and verify, the encoded message, and the caller's\nsignature buffer when signing fails part way. Cleanse a wrong-length private\nkey that ossl_slh_dsa_key_fromdata() copied in before rejecting it. The FORS\nroots buffer held one root per tree but was sized for k * a nodes; shrink it to\nk * n bytes and erase only the used length.\n\nPerformance is unchanged within noise; the SHAKE parameter sets measure a few\npercent faster, likely because the working Keccak state now stays at one fixed\naddress rather than a fresh stack copy at each recursion depth.\n\nReviewed-by: Tim Hudson \nReviewed-by: Bob Beck \nReviewed-by: Milan Broz \nMergeDate: Tue Aug 11 06:57:36 2026\n(Merged from https://github.com/openssl/openssl/pull/32148)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "507978799367a621631e3be63a05d54ce682d5bd", "date": "2026-08-03T14:45:49+03:00", "subject": "docs: mention wasm32-emscripten build target in news.txt", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d04a24d9ead0d47f772d5a416813259af5a36e98", "date": "2026-08-03T13:52:41+02:00", "subject": "Remove unused arg and dead code in set_attnotnull()", "body": "The is_valid parameter was never referenced in the function body, and\nthe 'thisatt' local variable is set but never used. Remove both.\n\nOversight in a379061a22a8.\n\nAuthor: Sami Imseih \nBackpatch-through: 18\nDiscussion: https://postgr.es/m/CAA5RZ0tHnvSrfUy4jWJchjvkL_aJe0hCnZpMsFRdLrSxCne5qQ@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "260cde9f48c3521a04a5c5deddaa9cea79956f30", "date": "2026-08-03T12:55:58+01:00", "subject": "Remove obsolete Kitty diacritics comment", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "a4d6004d40196aceb8137ead7fb3c6e6f44b7a9d", "date": "2026-08-03T12:56:12+01:00", "subject": "Correct image copyrights", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "5fc24e373a91f2b5acfabaf74b0ae4cca5e52ae2", "date": "2026-08-03T13:14:04+01:00", "subject": "Rename image damage functions to redraw", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "56457f838c0b8e4e83d73fa350bddb28cf81061e", "date": "2026-08-03T14:38:38+02:00", "subject": "test557: test curl_mv*printf() functions", "body": "These functions were previously untested in the test suite. This is just\na set of basic invokes to make sure they work. The core of these\nfunctions is identical and is tested already.\n\n - curl_mvfprintf\n - curl_mvprintf\n - curl_mvsnprintf\n - curl_mvsprintf\n - curl_mvaprintf\n\nCloses #22472", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e80aae70ac46990141cc1b82acb3d4a4f24bde9d", "date": "2026-08-03T08:54:45-04:00", "subject": "fix: guard MD5-dependent X509_ALGOR test with OPENSSL_NO_MD5", "body": "Reviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Neil Horman \nMergeDate: Thu Sep 10 21:14:58 2026\n(Merged from https://github.com/openssl/openssl/pull/30431)", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "tmux", "sha": "01a775a46c250761cf650058010fc0c6fe128732", "date": "2026-08-03T12:58:53Z", "subject": "Store background colour in padding cells and correctly clear adjacent cells when tabs are overwritten, GitHu issue 5441 from Ayman Bagabas.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "433a5cf580f44c0d88dbc1752a8bfb459162e14d", "date": "2026-08-03T13:59:27+01:00", "subject": "Add regress test from Ayman Bagabas.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "0a05812adb12598b32e85297b98df5ad14501d60", "date": "2026-08-03T22:07:31+09:00", "subject": "docs: add missing backticks to `no-duplicate-imports.js`", "body": "chore: add missing backticks to `no-duplicate-imports.js`", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "tmux", "sha": "6db5175e4edaaf3d57822282a980813e388667a8", "date": "2026-08-03T13:38:42Z", "subject": "Queue control mode notifications rather than emitting them inside %begin/%end, GitHub issue 5458 from George Nachman.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "d568f8ce3739b532d5bf2c1ee1e45e8a8a473d09", "date": "2026-08-03T16:40:43+03:00", "subject": "fix(cli): exit on `CTRL+c` even when `globalSetup` runs blocked code", "body": "", "style": "conventional", "type": "fix", "scope": "cli", "breaking": false} {"repo": "openssl", "sha": "a429bea81818d8db4b44a96818c357cde1595f33", "date": "2026-08-03T22:47:08+09:00", "subject": "ssl: Remove shadowed variable in write retry path", "body": "Commit d137c9e4f722 \"Fix false success on zero BIO write\" introduced\na function-scope left variable while the app-buffer recovery block\nalready declared one. This breaks builds using -Werror=shadow.\n\nAssign the current buffer length to the existing function-scope variable\nin the recovery block. This preserves the previous re-read at that point\nwhile eliminating the shadowed declaration.\n\nFixes: d137c9e4f722 \"Fix false success on zero BIO write\"\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nReviewed-by: Jakub Zelenka \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Mon Aug 3 15:46:22 2026\n(Merged from https://github.com/openssl/openssl/pull/32153)\n\n(cherry picked from commit a2eea872d0af948e5b889083f248e270d7326be7)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "88cdff90e4af0f0b6732ba4bb395b0e8a831edb7", "date": "2026-08-03T22:47:08+09:00", "subject": "ssl: Remove shadowed variable in write retry path", "body": "Commit 8d46334ee4ef \"Fix false success on zero BIO write\" introduced\na function-scope left variable while the app-buffer recovery block\nalready declared one. This breaks builds using -Werror=shadow.\n\nAssign the current buffer length to the existing function-scope variable\nin the recovery block. This preserves the previous re-read at that point\nwhile eliminating the shadowed declaration.\n\nFixes: 8d46334ee4ef \"Fix false success on zero BIO write\"\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nReviewed-by: Jakub Zelenka \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Mon Aug 3 15:46:22 2026\n(Merged from https://github.com/openssl/openssl/pull/32153)\n\n(cherry picked from commit a2eea872d0af948e5b889083f248e270d7326be7)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "813a8cfc31ffd6bdc2624c7c87a5b024391459f8", "date": "2026-08-03T22:47:08+09:00", "subject": "ssl: Remove shadowed variable in write retry path", "body": "Commit fae9fd3b6109 \"Fix false success on zero BIO write\" introduced\na function-scope left variable while the app-buffer recovery block\nalready declared one. This breaks builds using -Werror=shadow.\n\nAssign the current buffer length to the existing function-scope variable\nin the recovery block. This preserves the previous re-read at that point\nwhile eliminating the shadowed declaration.\n\nFixes: fae9fd3b6109 \"Fix false success on zero BIO write\"\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nReviewed-by: Jakub Zelenka \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Mon Aug 3 15:46:22 2026\n(Merged from https://github.com/openssl/openssl/pull/32153)\n\n(cherry picked from commit a2eea872d0af948e5b889083f248e270d7326be7)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c40d0e386e49d874141e76990e74ab8e40404b77", "date": "2026-08-03T22:47:08+09:00", "subject": "ssl: Remove shadowed variable in write retry path", "body": "Commit ebfcd4cfd4b7 \"Fix false success on zero BIO write\" introduced\na function-scope left variable while the app-buffer recovery block\nalready declared one. This breaks builds using -Werror=shadow.\n\nAssign the current buffer length to the existing function-scope variable\nin the recovery block. This preserves the previous re-read at that point\nwhile eliminating the shadowed declaration.\n\nFixes: ebfcd4cfd4b7 \"Fix false success on zero BIO write\"\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nReviewed-by: Jakub Zelenka \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Mon Aug 3 15:46:22 2026\n(Merged from https://github.com/openssl/openssl/pull/32153)\n\n(cherry picked from commit a2eea872d0af948e5b889083f248e270d7326be7)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a2eea872d0af948e5b889083f248e270d7326be7", "date": "2026-08-03T22:47:08+09:00", "subject": "ssl: Remove shadowed variable in write retry path", "body": "Commit 5f3db0d81132 \"Fix false success on zero BIO write\" introduced\na function-scope left variable while the app-buffer recovery block\nalready declared one. This breaks builds using -Werror=shadow.\n\nAssign the current buffer length to the existing function-scope variable\nin the recovery block. This preserves the previous re-read at that point\nwhile eliminating the shadowed declaration.\n\nFixes: 5f3db0d81132 \"Fix false success on zero BIO write\"\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nReviewed-by: Jakub Zelenka \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Mon Aug 3 15:46:22 2026\n(Merged from https://github.com/openssl/openssl/pull/32153)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "6538d0aadb32625e2d6969c835ebd061cad8a780", "date": "2026-08-03T19:18:33+05:30", "subject": "fix(ui2): retain substitute confirmation highlight with nohlsearch #41067", "body": "Problem:\nui2 clears the substitute confirmation match when it updates its prompt buffer with hlsearch disabled.\n\nSolution:\nOnly invalidate the match highlight when the current buffer changes.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "openssl", "sha": "e7c84542e25bca93ec2d3e4819433b8b5e0c483f", "date": "2026-08-03T09:54:19-04:00", "subject": "test: cover quoted property case variants", "body": "Reviewed-by: Paul Dale \nReviewed-by: Andrew Dinh \nMergeDate: Mon Aug 10 12:23:43 2026\n(Merged from https://github.com/openssl/openssl/pull/31597)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "700019fb6689cafb3c9ec035af5adb4ba667e80d", "date": "2026-08-03T16:05:25+02:00", "subject": "doc/internal/man7/deprecation.pod: document OSSL_DEPRECATEDIN_._._FOR macros", "body": "Signed-off-by: Eugene Syromiatnikov \n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Tue Aug 11 07:00:52 2026\n(Merged from https://github.com/openssl/openssl/pull/32155)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "874a54de5dcb5b68e22e896787d8b9bd9cda8aaa", "date": "2026-08-03T16:12:03+02:00", "subject": "multi: use index list for expire timeouts", "body": "Replace the `timeoutlist` and time nodes for expire timers with `struct\nexpire_timers` that keeps a time and next array.\n\nThere is only ever one timeout per timer id and a fully dynamic list is\nnot needed. Do the sorting by keeping an array of 'next' indiced and a\n'first' starting point.\n\nAdded tracking of splay tree status.\n\nFixed missing splay removal when last timeout was removed via\nCurl_expire_done().\n\nCloses #22473", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "7acf124614179f18a66a6d54eef911e75a12db53", "date": "2026-08-03T16:49:40+02:00", "subject": "url: rename Curl_init_do => Curl_init_transfer", "body": "And correct some comments\n\nCloses #22474", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "f131c034ad91bbf06bcbb6b5e931447a8e419a46", "date": "2026-08-03T17:52:45+03:00", "subject": "chore: improve ecosystem test failure reporting", "body": "* chore: improve ecosystem test failure reporting\n\n* move ecosystem test inline scripts to tools\n\n* ignore ecosystem test scripts in knip\n\n* use process.cwd() and explicit issue state\n\n* update ecosystem-tests.yml\n\n* remove blank lines\n\n* refactor\n\n* move ecosystem test tools into `tools/test-ecosystem`\n\n* remove redundant async\n\n* Update .github/workflows/ecosystem-tests.yml\n\nCo-authored-by: Francesco Trotta \n\n---------", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "angular", "sha": "11b4089acf6c1ef19f733c88e32f5b2e09362a14", "date": "2026-08-03T16:56:51+02:00", "subject": "test(compiler): add test for imports with unquoted URL", "body": "Adds a test to cover the case where an `@import` might have a URL with special characters that isn't quoted.", "style": "conventional", "type": "test", "scope": "compiler", "breaking": false} {"repo": "curl", "sha": "26fdb92c0a48337a9f696b79a8922e504243a823", "date": "2026-08-03T20:51:52+05:30", "subject": "mbedtls: enforce verifyhost when verifypeer is disabled", "body": "Verify in test 2118\n\nCloses #22475", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "c34d02ac79c7d8f01ccf544f2fe611b497df8269", "date": "2026-08-03T08:32:55-07:00", "subject": "fix(decor): 'breakindent'/'showbreak' gap ignores active highlight #41073", "body": "Problem:\n'breakindent' and 'showbreak' draw their own padding with no\nreference to whatever decoration or syntax highlight is currently\nactive, so it goes unhighlighted even mid-highlight, not just past a\nreal EOL. Gating this on the decoration's `hl_eol` flag (as an\nearlier version of this fix did) missed plain highlights with no\n`hl_eol` at all, which have the exact same problem.\n\nSolution:\nSnapshot decor_attr into decor_attr_save right before it can be\nreset by 'linebreak' filler handling, and pass it into\nhandle_breakindent()/handle_showbreak_and_filler() to extend into\ntheir padding: it is not a real end of the highlight, just screen\ncells with no buffer text. Like the 'linebreak' filler, an\nunderline/strikethrough/overline is excluded, since it looks like a\nbroken line drawn over the gap. This also fixes 'breakindent' losing\nthe highlight right after a real 'linebreak' word-push, since that\nreset otherwise leaked into the next row.\n\n(cherry picked from commit 866d61e91abaa73561f91ddd8f6e27bd50f9bc72)", "style": "conventional", "type": "fix", "scope": "decor", "breaking": false} {"repo": "neovim", "sha": "866d61e91abaa73561f91ddd8f6e27bd50f9bc72", "date": "2026-08-03T08:32:55-07:00", "subject": "fix(decor): 'breakindent'/'showbreak' gap ignores active highlight #41073", "body": "Problem:\n'breakindent' and 'showbreak' draw their own padding with no\nreference to whatever decoration or syntax highlight is currently\nactive, so it goes unhighlighted even mid-highlight, not just past a\nreal EOL. Gating this on the decoration's `hl_eol` flag (as an\nearlier version of this fix did) missed plain highlights with no\n`hl_eol` at all, which have the exact same problem.\n\nSolution:\nSnapshot decor_attr into decor_attr_save right before it can be\nreset by 'linebreak' filler handling, and pass it into\nhandle_breakindent()/handle_showbreak_and_filler() to extend into\ntheir padding: it is not a real end of the highlight, just screen\ncells with no buffer text. Like the 'linebreak' filler, an\nunderline/strikethrough/overline is excluded, since it looks like a\nbroken line drawn over the gap. This also fixes 'breakindent' losing\nthe highlight right after a real 'linebreak' word-push, since that\nreset otherwise leaked into the next row.", "style": "conventional", "type": "fix", "scope": "decor", "breaking": false} {"repo": "postgres", "sha": "072b962d99c6317e5cd5a5ac11b21fae2cab6f8a", "date": "2026-08-03T12:25:01-04:00", "subject": "Undo inadvertent loosening of archive filename checking.", "body": "Commit c8a350a439826267186c187dbfbf1f839f7521aa attempted to consolidate\ncode for identify possibly-compressed tar archives by suffix into a new\nfunction parse_tar_compress_algorithm(). Unfortunately, the refactoring\nwasn't perfect, and slightly changed the behavior at both existing call\nsites.\n\nIn CreateBackupStreamer(), the previous code required the filename to\nconsist of more than just a suffix, so the aforementioned commit had the\neffect of allowing pg_basebackup to accept a file from the server whose\nentire name was something like .tar.gz -- which should never happen, but\nlet's reject it as previous releases did.\n\nIn precheck_tar_backup_file(), the previous code required the suffix to\nbe immediately adjacent to the prefix already checked, so the commit\nin question allowed pg_verifybackup to accept not only filenames like\nbase.tar.gz but also filenames like baseFOOBARBAZ.tar.gz. While such\nfilenames are perhaps unlikely, rejecting them is correct, so let's go\nback to that behavior.\n\nDiscussion: http://postgr.es/m/CA+TgmoYJY8FkoeYKGF_YF1S6uOK7fd0Bd3zrw0XY_oZXbmVFpQ@mail.gmail.com\nReported-by: Sarath Kumar \nReviewed-by: Andrew Dunstan \nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "1e757a2feb57819c9c8d2e84dd0ef4787a4118bb", "date": "2026-08-03T11:53:12-05:00", "subject": "fix(zip): do not warn when zip.lua is disabled", "body": "", "style": "conventional", "type": "fix", "scope": "zip", "breaking": false} {"repo": "openssl", "sha": "0e3a4ea8aae66d61551c328b80e61977ec4fd3b7", "date": "2026-08-03T17:00:11Z", "subject": "ECH: Fix keylog callback to use inner client hello random from ECH", "body": "If the client attempted ECH, secrets are logged against the inner ClientHello\nrandom. CLIENT_EARLY_LABEL and EARLY_EXPORTER_SECRET_LABEL are special-cased\nbecause they are logged before the server's acceptance of ECH is known, but are\nalways derived from the inner ClientHello (see tls13_change_cipher_state()), so\nthey use the inner random regardless of ech.success.\n\nOn rejection the early secrets therefore carry the inner random while later ones\ncarry the outer: each line is tagged with the random of the ClientHello its\nsecret was derived from, and since a rejecting server never derives the early\nsecrets, both peers still log identical material for every label they both log.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Mounir Idrassi \nReviewed-by: Tomas Mraz \nMerge-date: Wed Sep 16 09:36:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32149", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5ad0d4b84b6402ab552dadb4d4c86cd25c3c9301", "date": "2026-08-03T17:00:11Z", "subject": "ECH: Fix keylog callback to use inner client hello random from ECH", "body": "If the client attempted ECH, secrets are logged against the inner ClientHello\nrandom. CLIENT_EARLY_LABEL and EARLY_EXPORTER_SECRET_LABEL are special-cased\nbecause they are logged before the server's acceptance of ECH is known, but are\nalways derived from the inner ClientHello (see tls13_change_cipher_state()), so\nthey use the inner random regardless of ech.success.\n\nOn rejection the early secrets therefore carry the inner random while later ones\ncarry the outer: each line is tagged with the random of the ClientHello its\nsecret was derived from, and since a rejecting server never derives the early\nsecrets, both peers still log identical material for every label they both log.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Mounir Idrassi \nReviewed-by: Tomas Mraz \nMerge-date: Wed Sep 16 09:35:56 2026\nMerged-from: https://github.com/openssl/openssl/pull/32149", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d39b221fd73fd1cce500105c140c33b3dba11d5f", "date": "2026-08-03T17:00:11Z", "subject": "ECH: Fix keylog callback to use inner client hello random from ECH", "body": "If the client attempted ECH, secrets are logged against the inner ClientHello\nrandom. CLIENT_EARLY_LABEL and EARLY_EXPORTER_SECRET_LABEL are special-cased\nbecause they are logged before the server's acceptance of ECH is known, but are\nalways derived from the inner ClientHello (see tls13_change_cipher_state()), so\nthey use the inner random regardless of ech.success.\n\nOn rejection the early secrets therefore carry the inner random while later ones\ncarry the outer: each line is tagged with the random of the ClientHello its\nsecret was derived from, and since a rejecting server never derives the early\nsecrets, both peers still log identical material for every label they both log.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Mounir Idrassi \nReviewed-by: Tomas Mraz \nMerge-date: Wed Sep 16 09:35:44 2026\nMerged-from: https://github.com/openssl/openssl/pull/32149", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "3910e7041068100ee5c1a9380367c89d72230fba", "date": "2026-08-03T19:42:35+02:00", "subject": "fix(Checkbox): filter out Checkbox.Group options with nullish values", "body": "* fix(CheckboxGroup): add support for undefined value\n\n* test: null use case\n\n* fix: change key definition strategy to avoid key conflic with 'null' string\n\n* fix: add different prefixes\n\n* update\n\n---------", "style": "conventional", "type": "fix", "scope": "Checkbox", "breaking": false} {"repo": "postgres", "sha": "529fc19ccc9c6044e6123274ffeef606bbf6a6b5", "date": "2026-08-03T11:34:58-07:00", "subject": "Do not log subscription conninfo.", "body": "Logging connection information, even at DEBUG1, creates unnecessary\nrisks. Remove the entire log message because it had no other useful\ncontent.\n\nAddresses finding 14 in report from linked discussion.\n\nReported-by: Noah Misch \nDiscussion: https://postgr.es/m/20260710195902.4f.noahmisch@microsoft.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "1e6d4479d23721e9a3f4751a0ccd941a854721a5", "date": "2026-08-03T11:43:33-07:00", "subject": "fix(lsp): only send `didSave` to clients attached to the buffer", "body": "", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "postgres", "sha": "343d98c3601abf3060ff82b7b2a8bc7903105bf3", "date": "2026-08-03T20:44:57+02:00", "subject": "Don't skip invalid databases when enabling data checksums", "body": "When enabling checksums cannot process a database, the launcher uses\nDatabaseExists to tell a concurrent drop (benign) from a real failure.\nSince 1df361e3d82 that check also treats a present, but-invalid, data-\nbase as non-existent. An interrupted DROP DATABASE flush the invalid\nmarker before the row and files are removed, so a crash or ERROR can\nleave an invalid row whose files remain on disk.\n\nReport a database as existing whenever its catalog row is found to\nensure that checksums cannot be enabled if there are invalid databases.\nThe AccessShareLock in DatabaseExists already waits out an in-flight\ndrop, so an invalid-but-present row can only be an interrupted drop\nleftover whose files still need checksums; enabling then aborts until\nit is dropped.\n\nBackpatch to v19 where online checksums were introduced.\n\nAuthor: Ayush Tiwari \nReviewed-by: Zsolt Parragi \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CAN4CZFOGdqxtZ5-6gb4apqmvoH=Z+TNH8RKJ3mVtoR1HirKQWg@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "e61d44fdee7e299dcbd4e168bda1e6670680aebc", "date": "2026-08-03T12:21:05-07:00", "subject": "Fix lock release for role membership grants in DROP OWNED BY.", "body": "Commit 6566133c5f5 added a case for AuthMemRelationId in\nAcquireDeletionLock(), but not ReleaseDeletionLock(). The fall-through\ncase would go to UnlockDatabaseObject(), which would raise a WARNING;\nand the lock would be retained until the end of the transaction.\n\nAdd the missing branch.\n\nDiscussion: https://postgr.es/m/2487ddcd737d4fc8e408e87aa9ad4365eed3bbb3.camel@j-davis.com\nBackpatch-through: 16", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "cd3f577c8520035ab4b51dd62d3b033a253a3911", "date": "2026-08-03T20:27:10+01:00", "subject": "Add test for control mode events and guards.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "f99e916ba7fc0f6f80fcf579f0a9a47b91241325", "date": "2026-08-03T14:42:45-05:00", "subject": "refactor(common): modernize directive tests to rely on whenStable", "body": "Replace synchronous detectChanges calls with zoneless-compatible scheduling and stability waits in NgClass, NgStyle, and NgOptimizedImage tests.\n\n(cherry picked from commit 2f0be5bef83ce15668fd401a470b4a5abe9c7f88)", "style": "conventional", "type": "refactor", "scope": "common", "breaking": false} {"repo": "angular", "sha": "2f0be5bef83ce15668fd401a470b4a5abe9c7f88", "date": "2026-08-03T14:42:45-05:00", "subject": "refactor(common): modernize directive tests to rely on whenStable", "body": "Replace synchronous detectChanges calls with zoneless-compatible scheduling and stability waits in NgClass, NgStyle, and NgOptimizedImage tests.", "style": "conventional", "type": "refactor", "scope": "common", "breaking": false} {"repo": "tmux", "sha": "114aa6808218c68e69dcd0730d67567fe44681ef", "date": "2026-08-03T20:18:20Z", "subject": "Do not let a stuck client prevent the server from exiting - give up after 10 seconds. GitHub issue 5444 from Ben Maurer.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "ca0d77855fe6aa2429054c693f25349d6face95b", "date": "2026-08-03T21:18:40+01:00", "subject": "Client exit test from Ben Maurer.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "608704adee75fd4316a10f4aa9a93aa8c6dc7eeb", "date": "2026-08-03T13:21:46-07:00", "subject": "Improve DROP SERVER handling of dependent subscriptions.", "body": "We do not allow a DROP SERVER ... CASCADE to implicitly drop a\nsubscription, because it's in a shared catalog and dropping a\nsubscription has side effects. Instead we throw an error and the user\nmust drop the subscription explicitly. Document this behavior and add\na HINT to the error message.\n\nGeneralize AcquireDeletionLock()/ReleaseDeletionLock() to use shared\nobject locks for all shared catalogs, which includes AuthMemRelationId\nand now SubscriptionRelationId.\n\nMove error message after AcquireDeletionLock() to avoid an unnecessary\nerror if there's a concurrent DROP SUBSCRIPTION.\n\nAddresses finding 10 & 15 in report from linked discussion.\n\nReported-by: Noah Misch \nDiscussion: https://postgr.es/m/20260710195902.4f.noahmisch@microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "d2322d0114c1d3313aa84151804a399e1c0b6ddf", "date": "2026-08-03T20:29:52Z", "subject": "Check for floating only for flags that require a floating pane, reported by Ilya Grigoriev.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "f07a76d959c3bd49410ea9d69c29e45338554941", "date": "2026-08-03T21:30:44+01:00", "subject": "Use move-pane for a test.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "755abbc8d5101b1ea79291635a91896fd752e8ef", "date": "2026-08-03T15:35:30-05:00", "subject": "fix(zip): only refuse a cwd unzip on Windows", "body": "", "style": "conventional", "type": "fix", "scope": "zip", "breaking": false} {"repo": "postgres", "sha": "8c1723f550a7adf6373c8c507e819fc3249d793a", "date": "2026-08-03T13:41:17-07:00", "subject": "postgres_fdw: reject use_scram_passthrough for subscriptions.", "body": "The subscription is initiated from a logical replication worker, so\nSCRAM pass-through won't work.\n\nPartially addresses finding 3 in report from linked discussion.\n\nReported-by: Noah Misch \nDiscussion: https://postgr.es/m/20260710195902.4f.noahmisch@microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "824d5f6241ea7a0a85c9d2b3d27beb78e42a36ab", "date": "2026-08-03T16:02:07-05:00", "subject": "Handle concurrently-dropped relations in database-wide VACUUM.", "body": "When VACUUM or ANALYZE is run without a table list, we scan\npg_class to build the list of relations to process, and we\ncheck the privileges on each relation we find. Since we don't\ntake any locks on the relations at this point, it's possible\nfor one to be concurrently dropped, in which case the privilege\ncheck fails with an ERROR such as the following:\n\n\tERROR: relation with OID 16388 does not exist\n\nThis unnecessarily aborts the entire command. To fix, use\npg_class_aclcheck_ext() for the privilege check so that we can\ndetect concurrently-dropped relations and silently skip them.\nThere's no need to emit a WARNING for such relations because\nthey weren't explicitly named, and a drop at this point is no\ndifferent than one that happened before our pg_class scan\nbegan. Note that concurrent drops that occur later on are\nalready handled gracefully by vacuum_open_relation().\n\nThe new missing_ok behavior is limited to get_all_vacuum_rels().\nAll other callers of vacuum_is_permitted_for_relation() should\nhold a lock on the relation that prevents it from being\nconcurrently dropped, so this commit also adds an assertion to\nthat effect.\n\nOversight in commit a556549d7e.\n\nThis is a bug fix and could be back-patched, but given the\npresumed rarity of the race condition and the lack of field\nreports, I'm not going to bother.\n\nReported-by: ChangAo Chen \nAuthor: ChangAo Chen \nCo-authored-by: Nathan Bossart \nReviewed-by: Kyotaro Horiguchi \nReviewed-by: Surya Poondla \nReviewed-by: Bharath Rupireddy \nReviewed-by: Michael Paquier \nDiscussion: https://postgr.es/m/tencent_F9D483523BB0D082C2EFDA80142F192DBC07%40qq.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "688a0a7118215bcab10004a3bf2667773fefdf74", "date": "2026-08-04T01:37:22+02:00", "subject": "fix(http): respect content-type charset in fetch backend text decoder", "body": "Extract the charset parameter from the Content-Type response header in FetchBackend and pass it to TextDecoder when decoding text and json responses. When no valid charset is provided or supported, gracefully fall back to default utf-8 decoding.\n\nFixes #70061\n\n(cherry picked from commit 6a0789dc7f0ac9d1249c4d9f315e83424867169b)", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "angular", "sha": "6a0789dc7f0ac9d1249c4d9f315e83424867169b", "date": "2026-08-04T01:37:22+02:00", "subject": "fix(http): respect content-type charset in fetch backend text decoder", "body": "Extract the charset parameter from the Content-Type response header in FetchBackend and pass it to TextDecoder when decoding text and json responses. When no valid charset is provided or supported, gracefully fall back to default utf-8 decoding.\n\nFixes #70061", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "openssl", "sha": "5a5542983d334f9be002aa97b2c723cbee399e3b", "date": "2026-08-04T10:49:57+10:00", "subject": "Fix FIPS to require a derivation function with CTR-DRBG", "body": "FIPS-140-3 requires that if a CTR-DRBG is allocated from the fips\nprovider that either:\n\na) The entropy source must be NIST validated and exist within the FIPS\nboundary\n\nor\n\nb) The CTR-DRBG must use a derivation function with an entropy source\noutside the FIPS boundary\n\nGiven that we have no approved noise source inside the FIPS boundary, we\nneed to enforce the fact that FIPS allocated CTR-DRBGS only allocate\ninstances in which a derivation function is requested (i.e. the USE_DF\nparameter is asserted and set to one)\n\nFollow path b, and ensure that FIPS CTR-DRBG allocations assert the use\nof USE_DF or fail if an allocation does not\n\nAlso fix up the evp tests to skip the 144 tests which allocate a DRBG\nwithout a derivation function when testing the fips provider\n\nReviewed-by: Neil Horman \nReviewed-by: Viktor Dukhovni \nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMergeDate: Mon Aug 10 12:01:41 2026\n(Merged from https://github.com/openssl/openssl/pull/32235)\n\n(cherry picked from commit 9ef31bc73697701219ff439b47f57de5cf61c1bf)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9ef31bc73697701219ff439b47f57de5cf61c1bf", "date": "2026-08-04T10:49:57+10:00", "subject": "Fix FIPS to require a derivation function with CTR-DRBG", "body": "FIPS-140-3 requires that if a CTR-DRBG is allocated from the fips\nprovider that either:\n\na) The entropy source must be NIST validated and exist within the FIPS\nboundary\n\nor\n\nb) The CTR-DRBG must use a derivation function with an entropy source\noutside the FIPS boundary\n\nGiven that we have no approved noise source inside the FIPS boundary, we\nneed to enforce the fact that FIPS allocated CTR-DRBGS only allocate\ninstances in which a derivation function is requested (i.e. the USE_DF\nparameter is asserted and set to one)\n\nFollow path b, and ensure that FIPS CTR-DRBG allocations assert the use\nof USE_DF or fail if an allocation does not\n\nAlso fix up the evp tests to skip the 144 tests which allocate a DRBG\nwithout a derivation function when testing the fips provider\n\nReviewed-by: Neil Horman \nReviewed-by: Viktor Dukhovni \nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMergeDate: Mon Aug 10 12:01:41 2026\n(Merged from https://github.com/openssl/openssl/pull/32235)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5ea1022e4a987b6cffe4ca6014fd1389b3311f88", "date": "2026-08-04T13:29:42+12:00", "subject": "Remove unused field from PartitionDescData struct", "body": "3592e0ff9 added caching logic to ExecFindPartition() to allow faster\npartition lookups for LIST and RANGE partitions when we hit the same\npartition multiple times in a row. The last_found_part_index field was\nadded, but it was never used, so remove it. Caching works by caching the\nlast matching index into the PartitionBoundInfo.datums, not by caching the\nlast found partition index.\n\nUpdate comments to reflect this, and perform other general improvements to\nthe comments in this area.\n\nAuthor: Aleksander Alekseev \nAuthor: David Rowley \nDiscussion: https://postgr.es/m/CAJ7c6TMRxO6vo1tNgLZs4nTNizpjCSWdUviB8Mf3nkYiNHNhzg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "431d6698882b71f73585890387c274e7b8695be0", "date": "2026-08-04T10:05:50+08:00", "subject": "refactor(Table): simplify filter trigger semantics", "body": "", "style": "conventional", "type": "refactor", "scope": "Table", "breaking": false} {"repo": "ant-design", "sha": "73c1b39800bbd8984a0afd3aeff49dd7edfafde3", "date": "2026-08-04T11:08:49+08:00", "subject": "fix(Carousel): keep current slide when children are added", "body": "* fix(Carousel): keep current slide when children are added\n\nfix issue #24879\n\nCo-Authored-By: Claude \n\n* test(Carousel): add test case for children removal not overflow\n\n* test(Carousel): verify slide after children removal\n\n---------", "style": "conventional", "type": "fix", "scope": "Carousel", "breaking": false} {"repo": "openssl", "sha": "6c1f1c8fdb4131117ffaf97deae5bb6c90574d68", "date": "2026-08-04T12:11:39+09:00", "subject": "apps/x509.c: failing exit status on every failure path reaching err:", "body": "With -multi, a later certificate failure could return success after an\nearlier certificate succeeded because ret retained the earlier result\non reaching err:.\n\nSet ret = 1 at err: so every path reaching it reports failure. Paths\nbypassing err: remain unchanged, including the pre-existing\nX509_VERIFY_PARAM_new() failure under -checkend, which will be addressed\nseparately.\n\nAdd a regression test using -multi -checkhost with a two-certificate\nchain whose second certificate fails.\n\nFixes #32189\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Todd Short \nMerge-date: Mon Sep 7 14:01:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32190\n(cherry picked from commit 8da3880874acaac2e338665f216e430f78f976f1)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8da3880874acaac2e338665f216e430f78f976f1", "date": "2026-08-04T12:11:39+09:00", "subject": "apps/x509.c: failing exit status on every failure path reaching err:", "body": "With -multi, a later certificate failure could return success after an\nearlier certificate succeeded because ret retained the earlier result\non reaching err:.\n\nSet ret = 1 at err: so every path reaching it reports failure. Paths\nbypassing err: remain unchanged, including the pre-existing\nX509_VERIFY_PARAM_new() failure under -checkend, which will be addressed\nseparately.\n\nAdd a regression test using -multi -checkhost with a two-certificate\nchain whose second certificate fails.\n\nFixes #32189\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Todd Short \nMerge-date: Mon Sep 7 14:01:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32190", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a3fed7608f75336ba748ab2784b2c5fa8cbf284d", "date": "2026-08-04T08:58:32+05:30", "subject": "Validate publisher for retain_dead_tuples in the apply worker.", "body": "Enabling retain_dead_tuples requires the publisher to run PostgreSQL 19 or\nlater and to not be in recovery. Previously this was checked only at DDL\ntime. That forced ALTER SUBSCRIPTION ... ENABLE to connect to the\npublisher, so pg_upgrade (which re-enables subscriptions during restore)\nfailed if the publisher was unreachable. It was also not authoritative,\nsince the publisher's version or recovery status can change afterwards,\nfor example after a failover.\n\nPerform the check authoritatively in the apply worker when it connects,\nand stop doing it when enabling a subscription. ENABLE is the only command\nissued during restore that triggered it, so this also fixes the pg_upgrade\nfailure. The DDL-time check is kept as a convenience for the other paths,\nnone of which are issued during restore.\n\nReported-by: Noah Misch \nAnalyzed-by: Jeff Davis \nAuthor: Amit Kapila \nReviewed-by: Jeff Davis \nReviewed-by: Hayato Kuroda \nBackpatch-through: 19, where it was introduced\nDiscussion: https://postgr.es/m/20260710195902.4f.noahmisch@microsoft.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "dd19ff1a965cd78b871027acc0746de0aa2a309c", "date": "2026-08-04T09:22:50+05:30", "subject": "fix(core): convert adjacent wildcard segments in legacy route converter", "body": "The middle-wildcard branch used replaceAll('/*/', ...), which consumes the\nslash shared by two adjacent '/*/*/' segments, so only the first segment was\nconverted and the second was left as an unnamed '*' that path-to-regexp still\nrejects. Match '/*' with a lookahead for the following slash instead, so every\nwildcard segment is converted. Existing single-wildcard output is unchanged.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "b48a8cc0e50ad2cfa604ca89765b9fbec58b0717", "date": "2026-08-04T16:10:39+12:00", "subject": "Fix missing MCXT_ALLOC_NO_OOM handling in MemoryContextAllocAligned", "body": "Fix missing NULL check in MemoryContextAllocAligned(). The underlying\ncall to MemoryContextAllocExtended() could return NULL when\nflags contains MCXT_ALLOC_NO_OOM and the underlying malloc fails.\n\nThere are no current callers using MemoryContextAllocAligned() that pass\nthe MCXT_ALLOC_NO_OOM in core, so no live bug fix in core here. However,\nan extension might use this pattern, so we'd better fix.\n\nFix this so we correctly pass the NULL to the caller rather than trying\nto write to a NULL memory address.\n\nThis also fixes the same bug in AlignedAllocRealloc(), which is also\nunused in core.\n\nBackpatch to v16, where these functions first appeared.\n\nAuthor: Chao Li \nDiscussion: https://postgr.es/m/07DAC4C3-120D-4F3C-8FEE-BA236F7E9C1D@gmail.com\nBackpatch-through: 16", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "278dc80436919b09473581ab45913c006779c54c", "date": "2026-08-04T12:24:43+08:00", "subject": "test(bundled-dev): un-skip long-tail playground cases", "body": "", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "ant-design", "sha": "5040df92921d404b5b494eea911d24516062e813", "date": "2026-08-04T13:46:10+08:00", "subject": "fix(Alert): show close button when closable contains onClose", "body": "", "style": "conventional", "type": "fix", "scope": "Alert", "breaking": false} {"repo": "postgres", "sha": "fec40878c8702ff4e58241d00a1914d8568bec03", "date": "2026-08-04T18:01:55+12:00", "subject": "Fix missing money overflow checks for INT64_MIN / -1", "body": "Similar to what 1f7cb5c30 did for the INT types, protect against\noverflow when dividing the lowest possible money value by -1. This\ncannot be represented on a two's complement machine.\n\nWithout this check, the result depends on the machine, and in the worst\ncase, could result in a crash. With the fix installed, this will now\nresult in:\n\nERROR: money out of range\n\nBug: #19585\nAuthor: Andrey Rachitskiy \nReported-by: Michael Malis \nReviewed-by: Tristan Partin \nReviewed-by: Rafia Sabih \nDiscussion: https://postgr.es/m/19586-bb603bf5ad9934dd%40postgresql.org\nDiscussion: https://postgr.es/m/CAB8bMisnXJVXte6s3kUOpuuAY9%3D9kehG6MMX-%2BTQoFsSGan22Q%40mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "06cd77f0c6e92586bb727100785a71e65a9a30c1", "date": "2026-08-04T14:30:32+08:00", "subject": "docs: fix broken links in Environment API pages", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "e72036eed2e28936ed824971b18aeaa3900857f6", "date": "2026-08-04T14:45:45+08:00", "subject": "refactor(bundled-dev): remove rolldown lazy stub module workaround", "body": "", "style": "conventional", "type": "refactor", "scope": "bundled-dev", "breaking": false} {"repo": "postgres", "sha": "cad17745edbf1234cd53293260b6e681eb8f6dd3", "date": "2026-08-04T09:06:46+02:00", "subject": "Fix ALTER COLUMN ... DROP EXPRESSION with subpartitions", "body": "Per commit 8bf6ec3ba3a4, a column can be GENERATED only if it is such in\nthe whole inheritance tree.\n\nFor this reason, ATPrepDropExpression refuses to be called with ONLY on\na partitioned table. To detect this, the current implementation checks\nwhether recurse is set to false and the rel has direct children.\n\nRecursion is implemented with ATSimpleRecursion, which calls ATPrepCmd\nwith recurse = false for every node in the tree. Inner nodes (for\nexample a partition which itself has subpartitions) then fail the check,\naccidentally preventing the command from working on inheritance trees of\ndepth > 2.\n\nThis commit fixes it by also checking that we're at the top level of the\nrecursive calls using the recursing parameter, which is always true when\ncalled through ATSimpleRecursion, always false when invoked on the root\nrel.\n\nAlso, remove a comment claiming that DROP EXPRESSION could be\nimplemented with some effort. It cannot, as the commit message for\n8bf6ec3ba3a4 explains.\n\nAuthor: Alberto Piai \nBackpatch-through: 14\nDiscussion: https://postgr.es/m/DHMT78XOD8BK.341V3H87KZ7NO@gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "e602dfa771d36388af8ba80c408a645b8048b296", "date": "2026-08-04T08:25:34+01:00", "subject": "Keep SIXEL palettes across redraw spans", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "0516c43284084f0526032a1cc312fb6916bf04fe", "date": "2026-08-04T15:31:12+08:00", "subject": "fix(types): preserve defineModel inference with factory defaults", "body": "close #15096", "style": "conventional", "type": "fix", "scope": "types", "breaking": false} {"repo": "vite", "sha": "fe37d670ed1f335ec6df7b92304aefd4fd43fa23", "date": "2026-08-04T16:32:13+09:00", "subject": "test: fix flaky legacy watch failure by workaround", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "b6191cbf95c680e076164a319dd987a14de658c1", "date": "2026-08-04T09:35:25+02:00", "subject": "fix(runtime-core): restore SSR setup state when handling async setup result", "body": "fix #15113", "style": "conventional", "type": "fix", "scope": "runtime-core", "breaking": false} {"repo": "vue-core", "sha": "a7513a1a3ed411ccd90bf135c6cba9c9cb2ef10b", "date": "2026-08-04T08:40:21+01:00", "subject": "fix(runtime-core): avoid re-fetching resource props with unchanged values", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-core", "breaking": false} {"repo": "vue-core", "sha": "71f9ff58a2a8748b2c1ec26c55c89f7343b86110", "date": "2026-08-04T15:40:44+08:00", "subject": "fix(custom-element): preserve nested async mount order", "body": "Closes #15153", "style": "conventional", "type": "fix", "scope": "custom-element", "breaking": false} {"repo": "vue-core", "sha": "c04a45d15498ebba0de3ceae674b88ac43cb9e76", "date": "2026-08-04T04:41:05-03:00", "subject": "fix(scheduler): avoid stack overflow when flushing very large post cb arrays", "body": "close #15142", "style": "conventional", "type": "fix", "scope": "scheduler", "breaking": false} {"repo": "vue-core", "sha": "a4440c01b96798a23bef12cb6f9ada5ffaba0d3a", "date": "2026-08-04T15:41:33+08:00", "subject": "fix(compiler-sfc): handle transformed template AST after cache invalidation", "body": "Closes #15126\nCloses #15128", "style": "conventional", "type": "fix", "scope": "compiler-sfc", "breaking": false} {"repo": "vue-core", "sha": "77061fe88608a5ab77fec705ff65cb81779eb01c", "date": "2026-08-04T15:42:23+08:00", "subject": "fix(transition): support transition to teleport component child", "body": "fix #11910", "style": "conventional", "type": "fix", "scope": "transition", "breaking": false} {"repo": "vue-core", "sha": "2506700b289c5277eb465bf959ce0ea23329094d", "date": "2026-08-04T10:44:17+03:00", "subject": "fix(slots): handle nullish v-bind slot props", "body": "", "style": "conventional", "type": "fix", "scope": "slots", "breaking": false} {"repo": "vue-core", "sha": "684a8defd2fdcba7eafa0c817b97ae1fab2c8b75", "date": "2026-08-04T16:53:42+09:00", "subject": "dx(runtime-core): warn when innerHTML/textContent overrides children", "body": "close #5081", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "fa62d4f6ea7529aa75105aa1ed3b43477ab365a0", "date": "2026-08-04T09:55:03+02:00", "subject": "Prohibit GRANT ... ON TABLE on a property graph", "body": "We allowed GRANT ... ON TABLE on sequences for backward compatibility.\nWe don't need to consider backward compatibility in case of property\ngraphs since we will be prohibiting its usage on property graph from\nthe very release which introduced property graphs.\n\nChange regression tests that used GRANT ... ON [TABLE] on property\ngraphs to use GRANT ... ON PROPERTY GRAPH instead.\n\nWhile here, add the missing RELKIND_PROPGRAPH cases in\npg_class_aclmask_ext() and in the object-type switch in\nExecGrant_Relation() so that the default ACL and the objtype passed to\nrestrict_and_check_grant() are correct.\n\nAuthor: Ashutosh Bapat \nReviewed-by: Noah Misch \nReviewed-by: Tom Lane \nDiscussion: https://www.postgresql.org/message-id/20260630023308.c7.noahmisch@microsoft.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "80e45d86e3c2d4d1e3bafdfb4311c1fc11b0ff52", "date": "2026-08-04T10:03:17+02:00", "subject": "feat(Table): support force rendering expanded rows", "body": "* feat(table): support force rendering expanded rows\n\n* docs(table): add force render expanded rows demo\n\n* test(table): add force render demo extend snapshots\n\n* docs(Table): remove forceRender demo\n\n* build: increase bundle size limit\n\n---------", "style": "conventional", "type": "feat", "scope": "Table", "breaking": false} {"repo": "postgres", "sha": "6b46a5d1b6164f2fdb15229dea9f3ed94ef20760", "date": "2026-08-04T17:03:27+09:00", "subject": "Fix error handling in getCopyDataMessage() and pqFunctionCall3()", "body": "Commit f6f0542266f0 changed getNotify(),\ngetParameterStatus(), and related libpq message-processing paths to\nabandon the connection on out-of-memory errors.\n\nHowever, getCopyDataMessage() and pqFunctionCall3() did not handle\nthis new fatal-error state. Both can process asynchronous\nNotificationResponse and ParameterStatus messages while waiting for\nother responses. If one of those messages triggered a fatal error, these\nloops continued processing instead of reporting it immediately.\n\nFix this by checking for a saved fatal error after processing an\nasynchronous message. If the connection has been abandoned, return the\nappropriate error immediately instead of continuing to parse input.\n\nBackpatch to v18, where commit f6f0542266f0 introduced this issue.\n\nAuthor: Anthonin Bonnefoy \nReviewed-by: Ewan Young \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAO6_XqpGfm+XHE1OzS=_+jroeDOxhhGa11P3cbm9q2gT05yorA@mail.gmail.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "69ed7fd7e9da1cff2f04af04f630287971fe99fe", "date": "2026-08-04T17:06:38+09:00", "subject": "Checkpoint replication slots later in the checkpoint cycle", "body": "Previously, CheckPointReplicationSlots() ran at the start of\nCheckPointGuts(), while WAL cleanup occurred much later in\nCreateCheckPoint() and CreateRestartPoint(), after the buffer write\nand ProcessSyncRequests() phases. During a spread checkpoint, this gap\ncould be several minutes.\n\nDuring that time, active replication slots could advance their\nrestart_lsn. However, replicationSlotMinLSN had already been\ncomputed from the older saved values. As a result, KeepLogSeg() could\nretain WAL segments that were no longer needed, causing unnecessary\npg_wal growth until the next checkpoint or restartpoint.\n\nFix this by moving CheckPointReplicationSlots(),\nCheckPointSnapBuild(), and CheckPointLogicalRewriteHeap() to just\nbefore CheckPointTwoPhase(), after the buffer write and\nProcessSyncRequests() phases. This makes WAL retention decisions use\nthe latest replication slot state. The logical snapshot and rewrite heap\ncleanup decisions also benefit from the updated saved restart_lsn.\n\nAuthor: Ants Aasma \nAuthor: Hüseyin Demir \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CANwKhkPCBcTQ_pk06MD5W5YYNnuYHp8dLNuOUz8-5pMBMPY1Bw%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "2112f185c0d906e02939048e24a85f60a8671327", "date": "2026-08-04T10:17:03+02:00", "subject": "TODO: do not consider APPDATA for netrc", "body": "See #22462\nCloses #22480", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "63c3bbb39f9cb2d2bd830c8e286e03356e6e44a7", "date": "2026-08-04T10:31:55+02:00", "subject": "doc: Add PROPERTY GRAPH to the access privilege tables", "body": "The SELECT privilege can be granted on a property graph, and its\nprivileges can be examined with psql's \\dp command, but property\ngraphs were missing from both summary tables in the \"Privileges\"\nsection: the applicable object types for SELECT in the privilege\nabbreviations table, and the per-object-type row in the summary of\naccess privileges table. Add the missing entries so the tables match\nthe actual behavior described for the SELECT privilege and in GRANT.\n\nWhile at it, also add a test for \\dp on a property graph.\n\nAuthor: Shinya Kato \nReviewed-by: Ashutosh Bapat \nDiscussion: https://www.postgresql.org/message-id/flat/CAOzEurScgwLDXQmNFnDZANGAMMiva9GnLH_kO8qGtGVHUvVk2A%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "30b4c1b40d3c5a69cc2d53727e8e3f53bdb6be70", "date": "2026-08-04T16:50:36+08:00", "subject": "fix(ctx): missing options copy when loading hidden buffer #41149", "body": "It's necessary to copy the global 'fileencoding' to the buffer-local\nvalue before entering the buffer, otherwise 'fileencoding' is changed\nwhen reading the file, which will mark the file as modified.", "style": "conventional", "type": "fix", "scope": "ctx", "breaking": false} {"repo": "eslint", "sha": "56110356652dd614b3cf0933538abb551fc6bd6f", "date": "2026-08-04T17:59:24+09:00", "subject": "test: add error locations info to `no-void`", "body": "test: add location info to no-void tests", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d0644230f12b1ade01f5cad35531019ac974e225", "date": "2026-08-04T05:14:15-04:00", "subject": "fix(lsp): ignore (and log once) invalid documentSelector globs #41055", "body": "Problem:\nClient:_get_registrations() errors if a documentSelector pattern is not\na valid glob (e.g. \"**/**.dart\" from dartls). The error discards\nalready-matched registrations and propagates to supports_method() and\nthe client/registerCapability handler, disabling capabilities unrelated\nto the offending pattern.\n\nSolution:\nSkip invalid globs. Treat them as non-matching, and log once per client.\n\n(Similar approach was taken for filewatcher globs: 4d9e5acfb5ee081b5b0761205c8a8c36ed7e49e5)", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "vite", "sha": "ff4ab6c14dc403e255abdaff079ed063d445c4dd", "date": "2026-08-04T17:14:26+08:00", "subject": "test(bundled-dev): enable assets playgrounds", "body": "", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "vite", "sha": "5290aea2ea9bb2be8fe52c1541dc50619d19253c", "date": "2026-08-04T17:15:35+08:00", "subject": "test(bundled-dev): enable css playgrounds", "body": "", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "neovim", "sha": "a29297ac538f9a86bbcd5e59cc8a9e3494b263c6", "date": "2026-08-04T04:19:54-05:00", "subject": "docs(api): buffer columns are byte indices #41137", "body": "", "style": "conventional", "type": "docs", "scope": "api", "breaking": false} {"repo": "vite", "sha": "23b8a088dec9dcc3f1c1353f2074f8644b3cc21f", "date": "2026-08-04T18:22:04+09:00", "subject": "refactor(bundled-dev): avoid injecting server values in the bundle", "body": "", "style": "conventional", "type": "refactor", "scope": "bundled-dev", "breaking": false} {"repo": "postgres", "sha": "e9d53cf459419cc7fd32dc85c59ee08ee5bbf1f4", "date": "2026-08-04T11:44:11+02:00", "subject": "pg_surgery: Fix infinite loop on large TID arrays", "body": "heap_force_common() tracked the current position in the caller-supplied\ntid[] using OffsetNumber, which is only 16 bits wide, so when the array\nheld more than 65535 entries, the updated index wrapped around and the\nouter loop never reached the exit condition. A SQL call with a\nsufficiently large TID array would then run until interrupted.\n\nFix by tracking the tid[] position using int instead of OffsetNumber.\nA regress case based on the report is included.\n\nAuthor: Andrey Rachitskiy \nReviewed-by: Andrey Borodin \nReported-by: Yuelin Wang <1217816127@qq.com>\nBackpatch-through: 14\nBug: #19607\nDiscussion: https://postgr.es/m/19607-2f256a66481c514b@postgresql.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "137bcf0a170e16a6c934f83b7e67082ab04bd400", "date": "2026-08-04T11:47:03+02:00", "subject": "ci: name the second reader of the dispatch run-name", "body": "The run-name of a dispatched workflow is parsed by two services, not one: the CI\nbot correlates a run with its check-run, and the Actions statistics collector\nattributes CI load to the pull request that caused it. Neither validates the\nstring, so a mis-copied run-name is silent in both. The comment named only the\nfirst reader, which understates what editing that line costs.\n\nAssisted-by: Claude Opus 5 (1M context) \n\nReviewed-by: Dmitry Misharov \nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nReviewed-by: Andrew Dinh \nMergeDate: Sat Aug 8 15:34:01 2026\n(Merged from https://github.com/openssl/openssl/pull/32174)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "eb5626c572691d50ad7a580e1a66291400114af6", "date": "2026-08-04T11:47:30+02:00", "subject": "ci: let the CI bot dispatch the compiler, merge and AVX512 zoos", "body": "These three only run after the fact: compiler-zoo and run-checker-merge on a\npush, avx512-sde on a nightly cron. Their coverage therefore lands once a pull\nrequest is already merged, which is when it is most expensive to act on.\n\nGive all three the dispatch entry point the daily workflows already have:\nvalidated pr/head_sha/check_run_id inputs, the run-name both services correlate\non, concurrency keyed on the pull request, and every checkout pinned to the\nrequested commit.\n\nExisting behaviour is untouched. A push or a cron carries no inputs, so the\nvalidator is skipped and each checkout falls back to github.sha -- the build jobs\nstill run on a push to a fork branch, the only pre-merge signal a contributor\ngets here. The jitter library stays pinned to its release tag: a dispatched run\nmay choose the OpenSSL tree, never its dependencies.\n\nRefs: openssl/project#2027\n\nAssisted-by: Claude Opus 5 (1M context) \n\nReviewed-by: Dmitry Misharov \nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nReviewed-by: Andrew Dinh \nMergeDate: Sat Aug 8 15:34:06 2026\n(Merged from https://github.com/openssl/openssl/pull/32174)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "curl", "sha": "69a224d6b48edb43df29cb69881ca8edc90f1527", "date": "2026-08-04T11:52:00+02:00", "subject": "vdns: directory for all DNS related sources", "body": "Move all DNS related source files from lib to lib/vdns. Fix\ninclude paths, no furher changes.\n\nCloses #22482", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e6cc3e5f63eb5035cefef5d715085f29bfc893d1", "date": "2026-08-04T11:52:15+02:00", "subject": "ci: let the CI bot dispatch the platform, cross-compile, fuzzing and interop suites", "body": "These five suites only run nightly, so a break in a platform, a cross-compiled\ntarget, a fuzzer or an interop peer is found the morning after it merged. Give\neach the dispatch entry point the daily workflows already have, and adapt the\nones that assumed they only ever see their own default branch.\n\nos-zoo's self-hosted legs (ppc64le, s390x, riscv64) are restricted to the\nnightly cron and plain manual runs, keeping pull-request code off them; the\nGitHub-hosted legs gain the repository guard the rest of the file already had.\nThe cross-compile suites' test steps, gated by event name, now also run on a\ndispatch instead of silently skipping. oss-fuzz checks out the requested tree\nso CIFuzz fuzzes it instead of the default branch. interop-tests skips\nopenssh_interop on a dispatch, since it never checks this repo out.\n\nRefs: openssl/project#2027\n\nAssisted-by: Claude Opus 5 (1M context) \n\nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMergeDate: Fri Aug 7 13:31:12 2026\n(Merged from https://github.com/openssl/openssl/pull/32175)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "postgres", "sha": "01805b7d16b9a06d5223071215db0337bc8e6d92", "date": "2026-08-04T12:16:55+02:00", "subject": "Do not reuse rd_smgr in fork loop when enabling data checksums", "body": "ProcessSingleRelationByOid called RelationGetSmgr(rel), discarded the\nresult, and then read rel->rd_smgr directly when looping over forks.\nOnly RelationGetSmgr is authorized to read that field since a relcache\ninvalidation resets rd_smgr to NULL.\n\nBackpatch to v19 where online checksums were introduced.\n\nAuthor: Mihail Nikalayeu \nReviewed-by: ChangAo Chen \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CADzfLwXGvb4Y-mqy8T+O0f_tkXR1sTDBGzP5Z=V_qcGnZ46rWg@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "85894471594936afe90b08c3a382c36dd166fa47", "date": "2026-08-04T06:20:16-04:00", "subject": "feat(detach): opt-in to \"server keeps running\" #41133", "body": "Problem:\nBy default, `nvim` does not survive if its host terminal dies. This is\ninconvenient if you want to use Nvim as a \"session manager\" (like tmux).\n\nSolution:\nLet users opt-in to the \"survive\" behavior via `:detach!` (bang \"!\").\nThis marks the current UI as \"detachable\", so the server will not\nself-exit if the UI channel closes.", "style": "conventional", "type": "feat", "scope": "detach", "breaking": false} {"repo": "openssl", "sha": "0321286f0cc7f2c5eef419cbf8b130a95af072e7", "date": "2026-08-04T11:58:39+01:00", "subject": "ssl/record: lower the DTLS unprocessed_rcds queue limit", "body": "100 buffered next-epoch records is far more than a normal handshake\never needs. A peer that has already completed its side of the epoch\ntransition may send more than one record under the new epoch before\nwe catch up and bump our own receive epoch - for example, application\ndata sent immediately once the peer considers the handshake done -\nbut real-world bursts like that are still small. Now that each entry\nonly costs as much memory as the record actually received, the limit\nmainly serves as a ceiling on worst-case per-connection memory use, so\nlower it to 16 to keep that ceiling smaller while still leaving ample\nheadroom over legitimate usage.\n\ndtls1_buffer_record() serves three queues here, so take the limit as an\nargument rather than tightening all of them. Only unprocessed_rcds gets\nthe lower limit: processed_rcds is filled from unprocessed_rcds and so\ncannot exceed it in any case, and buffered_app_data holds application\ndata that arrived between ChangeCipherSpec and Finished, which is not\nthe queue this ceiling is aimed at.\n\nAssisted-by: Claude:claude-opus-5\nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMerge-date: Mon Aug 24 15:40:45 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "597555ebc09dd5bf7b0349b27cdcb308638d79ce", "date": "2026-08-04T07:06:47-04:00", "subject": "fix(options): crash on \":let &t_Co = v:true\" #41152", "body": "Problem:\nAssigning a Boolean or special value (v:true/v:false/v:null/v:none) to a TTY\noption aborts:\n\n Assertion failed: (curval.type == newval.type), function ex_let_option, file vars.c, line 1408.\n 3 libsystem_c.dylib __assert_rtn + 284\n 4 nvim ex_let_one + 3308\n 5 nvim ex_let_vars + 112\n 6 nvim ex_let + 2356\n 7 nvim execute_cmd0 + 252\n 8 nvim do_cmdline + 9076\n 9 nvim call_user_func + 3320\n 10 nvim call_func + 2076\n 11 nvim get_func_tv + 696\n 12 nvim eval_func + 380\n 20 nvim eval_to_string_eap + 276\n 21 nvim eval_map_expr + 444\n 22 nvim vgetorpeek + 3172\n 23 nvim vgetc + 764\n\nSolution:\nApply the string-type check to TTY options too, so a Boolean/special value\ngives \"E928: String required\" instead of aborting. Valid string/number\nassignments to `t_*` pseudo-options still silently no-op.", "style": "conventional", "type": "fix", "scope": "options", "breaking": false} {"repo": "nestjs", "sha": "4a4097836fd867e1339c5ba9d820d8d6c16caefe", "date": "2026-08-04T14:22:00+03:00", "subject": "fix(express): map missing multer field nesting error", "body": "", "style": "conventional", "type": "fix", "scope": "express", "breaking": false} {"repo": "nestjs", "sha": "2a532e7786a4cee8e50d1f47f05cbdf1243b3c8b", "date": "2026-08-04T14:34:55+03:00", "subject": "chore: fix misspelled spec filenames", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "openssl", "sha": "119ab9555dc62275bbd71f6f49529b1a44feba42", "date": "2026-08-04T07:35:48-04:00", "subject": "Check the tag on EVP_Cipher() finalize: Poly1305 and OCB AEADs", "body": "For the affected OpenSSL built-in provider AEAD implementations,\nEVP_Cipher(ctx, out, NULL, 0) reaches the ccipher callback as a\nNULL-input terminal call. OCB and ChaCha20-Poly1305 took an early exit\non an empty message, with or without AAD, and returned success without\ncomparing an explicitly supplied tag. Consequently a corrupted tag was\naccepted before this change.\n\nMake these built-in callbacks perform their terminal tag operation,\naligning their explicit-tag handling with the streaming Final path\nwithout defining NULL input as part of the generic EVP_Cipher()\ncontract.\n\nAES-GCM-SIV also failed to generate a tag when Final was its first\nempty-message operation. Generate the tag in that case and propagate\nfailures from the matching empty-message decrypt operation.\n\nThe stable ChaCha20-Poly1305 implementation aliases Update to the\none-shot cipher callback, so this backport introduces a dedicated Update\ncallback to preserve zero-length Update as a no-op.\n\nFollow-up to #31555\nFixes #32258\nFixes CVE-2026-75803\n\nAssisted-by: Claude:claude-opus-4-8\nAssisted-by: Codex:gpt-5.6-sol\n\n(cherry picked from commit 5741d29a5f356e05262cd0936a472a9961398d53)\n\nCo-authored-by: Mounir IDRASSI \nReviewed-by: Bob Beck \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 19 17:43:24 2026\nMerged-from: https://github.com/openssl/openssl/pull/32417", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "bf95f5f772e9362f87b25cfa2f8cb15d984865b9", "date": "2026-08-04T07:35:48-04:00", "subject": "Check the tag on EVP_Cipher() finalize: Poly1305 and OCB AEADs", "body": "For the affected OpenSSL built-in provider AEAD implementations,\nEVP_Cipher(ctx, out, NULL, 0) reaches the ccipher callback as a\nNULL-input terminal call. OCB and ChaCha20-Poly1305 took an early exit\non an empty message, with or without AAD, and returned success without\ncomparing an explicitly supplied tag. Consequently a corrupted tag was\naccepted before this change.\n\nMake these built-in callbacks perform their terminal tag operation,\naligning their explicit-tag handling with the streaming Final path\nwithout defining NULL input as part of the generic EVP_Cipher()\ncontract.\n\nAES-GCM-SIV also failed to generate a tag when Final was its first\nempty-message operation. Generate the tag in that case and propagate\nfailures from the matching empty-message decrypt operation.\n\nThe stable ChaCha20-Poly1305 implementation aliases Update to the\none-shot cipher callback, so this backport introduces a dedicated Update\ncallback to preserve zero-length Update as a no-op.\n\nFollow-up to #31555\nFixes #32258\nFixes CVE-2026-75803\n\nAssisted-by: Claude:claude-opus-4-8\nAssisted-by: Codex:gpt-5.6-sol\n\n(cherry picked from commit 5741d29a5f356e05262cd0936a472a9961398d53)\n\nCo-authored-by: Mounir IDRASSI \nReviewed-by: Bob Beck \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 19 17:41:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/32416", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3621257986e27e540bf96a11570929a6e5a9e05b", "date": "2026-08-04T07:35:48-04:00", "subject": "Check the tag on EVP_Cipher() finalize: Poly1305 and OCB AEADs", "body": "For the affected OpenSSL built-in provider AEAD implementations,\nEVP_Cipher(ctx, out, NULL, 0) reaches the ccipher callback as a\nNULL-input terminal call. OCB and ChaCha20-Poly1305 took an early exit\non an empty message, with or without AAD, and returned success without\ncomparing an explicitly supplied tag. Consequently a corrupted tag was\naccepted before this change.\n\nMake these built-in callbacks perform their terminal tag operation,\naligning their explicit-tag handling with the streaming Final path\nwithout defining NULL input as part of the generic EVP_Cipher()\ncontract.\n\nAES-GCM-SIV also failed to generate a tag when Final was its first\nempty-message operation. Generate the tag in that case and propagate\nfailures from the matching empty-message decrypt operation.\n\nThe stable ChaCha20-Poly1305 implementation aliases Update to the\none-shot cipher callback, so this backport introduces a dedicated Update\ncallback to preserve zero-length Update as a no-op.\n\nFollow-up to #31555\nFixes #32258\nFixes CVE-2026-75803\n\nAssisted-by: Claude:claude-opus-4-8\nAssisted-by: Codex:gpt-5.6-sol\n\n(cherry picked from commit 5741d29a5f356e05262cd0936a472a9961398d53)\n\nCo-authored-by: Mounir IDRASSI \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 19 15:58:27 2026\nMerged-from: https://github.com/openssl/openssl/pull/32259", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6c7aa6f8f6449b7fe0137ee8be65fcd239bd7d6a", "date": "2026-08-04T07:35:48-04:00", "subject": "Check the tag on EVP_Cipher() finalize: Poly1305 and OCB AEADs", "body": "For the affected OpenSSL built-in provider AEAD implementations,\nEVP_Cipher(ctx, out, NULL, 0) reaches the ccipher callback as a\nNULL-input terminal call. OCB and ChaCha20-Poly1305 took an early exit\non an empty message, with or without AAD, and returned success without\ncomparing an explicitly supplied tag. Consequently a corrupted tag was\naccepted before this change.\n\nMake these built-in callbacks perform their terminal tag operation,\naligning their explicit-tag handling with the streaming Final path\nwithout defining NULL input as part of the generic EVP_Cipher()\ncontract.\n\nAES-GCM-SIV also failed to generate a tag when Final was its first\nempty-message operation. Generate the tag in that case and propagate\nfailures from the matching empty-message decrypt operation.\n\nOpenSSL 4.0 already has a dedicated ChaCha20-Poly1305 Update callback,\nso keep zero-length Update as a no-op there while allowing the\nNULL-input ccipher call to perform the terminal tag operation.\n\nFollow-up to #31555\nFixes #32258\nFixes CVE-2026-75803\n\nAssisted-by: Claude:claude-opus-4-8\nAssisted-by: Codex:gpt-5.6-sol\n\n(cherry picked from commit 5741d29a5f356e05262cd0936a472a9961398d53)\n\nCo-authored-by: Mounir IDRASSI \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 19 15:56:02 2026\nMerged-from: https://github.com/openssl/openssl/pull/32300", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5741d29a5f356e05262cd0936a472a9961398d53", "date": "2026-08-04T07:35:48-04:00", "subject": "Check the tag on EVP_Cipher() finalize: Poly1305 and OCB AEADs", "body": "At the EVP level, for AEADs EVP_Cipher(ctx, out, NULL, 0) performs\nfinalization. For consistency across AEADs (GCM, etc.), on decrypt it\nmust check the tag. OCB and ChaCha20-Poly1305 took an early exit on the\nempty message (with or without AAD) and returned success without\nchecking, so a forged tag (with or without AAD) was accepted before this\nchange.\n\nFollow-up to #31555\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Dmitry Belyavskiy \nMergeDate: Fri Aug 7 13:34:35 2026\n(Merged from https://github.com/openssl/openssl/pull/32173)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "387bd0fbe78756b030884805e61754cee1be4bb4", "date": "2026-08-04T19:39:59+08:00", "subject": "refactor(window): use window handle in frame layout log", "body": "Problem: The frame layout debug log uses the obsolete window ID field.\n\nSolution: Use the window handle field when printing window information.", "style": "conventional", "type": "refactor", "scope": "window", "breaking": false} {"repo": "vite", "sha": "4352047a47a87a04de0b9e4c68da0fbfd742e125", "date": "2026-08-04T21:11:51+09:00", "subject": "feat: CSS sourcemaps for build", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "postgres", "sha": "16aba1f9b3addecc6f8463ac2ab783551627d148", "date": "2026-08-04T14:36:47+02:00", "subject": "doc: Make synopsis placeholders consistent", "body": "Existing text (e.g., create_foreign_table.sgml) uses server_name, not\nservername.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "8126fe4739b52f930a1bf2b9f205536e3b26f81b", "date": "2026-08-04T15:10:22+02:00", "subject": "psql: Message style fixes", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "3894397b07fd8473a7bb315b1aaa7605bd6d4447", "date": "2026-08-04T13:16:03Z", "subject": "Switch from imsg_get to imsgbuf_get and switch type of n to int.", "body": "OK tb@", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "db2e86fba4f732dab43f70dd7f6d0a089ea9e98c", "date": "2026-08-04T15:16:17+02:00", "subject": "refactor(editor): cleanup change-directory (:bcd) logic", "body": "", "style": "conventional", "type": "refactor", "scope": "editor", "breaking": false} {"repo": "openssl", "sha": "31f7e4eb7a907fb4e6b8ec577400801b93d46e7b", "date": "2026-08-04T15:26:32+02:00", "subject": "Fix NULL dereference in v2i_AUTHORITY_KEYID()", "body": "X509V3_EXT_d2i() may return NULL on malformed input\nor on allocation failure. v2i_AUTHORITY_KEYID() passed\nthat NULL to ASN1_STRING_length_ex() causing NULL dereference.\n\nFound by x509v3 fuzzer. Example fuzzer input:\n```\n[default]\nsubjectKeyIdentifier = DER:05:00\nauthorityKeyIdentifier = keyid\n```\n\nReviewed-by: Milan Broz \nReviewed-by: Viktor Dukhovni \nMergeDate: Fri Aug 7 14:03:37 2026\n(Merged from https://github.com/openssl/openssl/pull/32177)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f10e9176e6ac7944e55ea33ef0ac4cfb01036f4e", "date": "2026-08-04T15:26:32+02:00", "subject": "Fix NULL dereference in v2i_AUTHORITY_KEYID()", "body": "X509V3_EXT_d2i() may return NULL on malformed input\nor on allocation failure. v2i_AUTHORITY_KEYID() passed\nthat NULL to ASN1_STRING_length_ex() causing NULL dereference.\n\nFound by x509v3 fuzzer. Example fuzzer input:\n```\n[default]\nsubjectKeyIdentifier = DER:05:00\nauthorityKeyIdentifier = keyid\n```\n\nReviewed-by: Milan Broz \nReviewed-by: Viktor Dukhovni \nMergeDate: Fri Aug 7 14:03:05 2026\n(Merged from https://github.com/openssl/openssl/pull/32177)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "08b18612b65fc1b8e57fd59cf4334b56aafe92c1", "date": "2026-08-04T15:26:32+02:00", "subject": "Fix NULL dereference in v2i_AUTHORITY_KEYID()", "body": "X509V3_EXT_d2i() may return NULL on malformed input\nor on allocation failure. v2i_AUTHORITY_KEYID() passed\nthat NULL to ASN1_STRING_length_ex() causing NULL dereference.\n\nFound by x509v3 fuzzer. Example fuzzer input:\n```\n[default]\nsubjectKeyIdentifier = DER:05:00\nauthorityKeyIdentifier = keyid\n```\n\nReviewed-by: Milan Broz \nReviewed-by: Viktor Dukhovni \nMergeDate: Fri Aug 7 13:59:14 2026\n(Merged from https://github.com/openssl/openssl/pull/32177)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "378dd47e95cadfdb5fb0e783aed7d8bb55786381", "date": "2026-08-04T15:26:32+02:00", "subject": "Fix NULL dereference in v2i_AUTHORITY_KEYID()", "body": "X509V3_EXT_d2i() may return NULL on malformed input\nor on allocation failure. v2i_AUTHORITY_KEYID() passed\nthat NULL to ASN1_STRING_length_ex() causing NULL dereference.\n\nFound by x509v3 fuzzer. Example fuzzer input:\n```\n[default]\nsubjectKeyIdentifier = DER:05:00\nauthorityKeyIdentifier = keyid\n```\n\nReviewed-by: Milan Broz \nReviewed-by: Viktor Dukhovni \nMergeDate: Fri Aug 7 13:43:20 2026\n(Merged from https://github.com/openssl/openssl/pull/32177)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "cb1e161353deaea65fb0f5484aab06d1d76dc2d0", "date": "2026-08-04T15:26:32+02:00", "subject": "Fix NULL dereference in v2i_AUTHORITY_KEYID()", "body": "X509V3_EXT_d2i() may return NULL on malformed input\nor on allocation failure. v2i_AUTHORITY_KEYID() passed\nthat NULL to ASN1_STRING_length_ex() causing NULL dereference.\n\nFound by x509v3 fuzzer. Example fuzzer input:\n```\n[default]\nsubjectKeyIdentifier = DER:05:00\nauthorityKeyIdentifier = keyid\n```\n\nReviewed-by: Milan Broz \nReviewed-by: Viktor Dukhovni \nMergeDate: Fri Aug 7 13:42:59 2026\n(Merged from https://github.com/openssl/openssl/pull/32177)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "ac3728e79fd8fae7a4b38e20d152c5a734a6e3fc", "date": "2026-08-04T15:29:31+02:00", "subject": "fix(http): avoid aborting completed requests in FetchBackend", "body": "Prevent AbortController.abort() from executing during Observable teardown when a FetchBackend HTTP request has already completed successfully or errored.\nPreviously, FetchBackend unconditionally called abort() upon stream termination. When requests completed normally, calling abort() after delivery caused Chromium-based browsers to mark the resolved request as net::ERR_ABORTED in DevTools due to a race condition, leading to missing response body payloads (\"Failed to load response data\"). By tracking whether the request has already settled—similar to XhrBackend checking for xhr.readyState !== xhr.DONE—we ensure abort() is only called for unsettled, in-flight requests upon unsubscription.\n\nFixes #70071\n\n(cherry picked from commit ef4dfead83ea8b3843b98a17ac572f2a380b1c46)", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "angular", "sha": "ef4dfead83ea8b3843b98a17ac572f2a380b1c46", "date": "2026-08-04T15:29:31+02:00", "subject": "fix(http): avoid aborting completed requests in FetchBackend", "body": "Prevent AbortController.abort() from executing during Observable teardown when a FetchBackend HTTP request has already completed successfully or errored.\nPreviously, FetchBackend unconditionally called abort() upon stream termination. When requests completed normally, calling abort() after delivery caused Chromium-based browsers to mark the resolved request as net::ERR_ABORTED in DevTools due to a race condition, leading to missing response body payloads (\"Failed to load response data\"). By tracking whether the request has already settled—similar to XhrBackend checking for xhr.readyState !== xhr.DONE—we ensure abort() is only called for unsettled, in-flight requests upon unsubscription.\n\nFixes #70071", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "vitest", "sha": "a0b20bc8688fd0b056540add0b354dd35d5a054b", "date": "2026-08-04T15:37:23+02:00", "subject": "fix!: use `>` as separator in `-t`, calculate `only` once", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": true} {"repo": "postgres", "sha": "589eb4c3b309f5eaa7c16592ff4edbbf780671fe", "date": "2026-08-04T15:42:56+02:00", "subject": "Reindent conditional to improve readability", "body": "Found via readability-misleading-indentation warning in clangd.\n\nAuthor: Tristan Partin \nDiscussion: https://postgr.es/m/DKFO99FP6SLY.2O6A6BUK6P3OQ@partin.io", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "46ca2365256e9d9fa7d845d6757b15ea96e47fbc", "date": "2026-08-04T15:45:45+02:00", "subject": "fix(cwd): validate getcwd(…, -1)", "body": "", "style": "conventional", "type": "fix", "scope": "cwd", "breaking": false} {"repo": "vitest", "sha": "96e40feeae35a35e185ba2dd718253459bda2b30", "date": "2026-08-04T15:57:58+02:00", "subject": "docs: explain vm pool worker recycling cost", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "101d6148945450e59b283c8b11962747e4a882d2", "date": "2026-08-04T16:45:37+02:00", "subject": "CI: specify RISC-V vector extension version in QEMU", "body": "Made according to the recomimendation of @ZenithalHourlyRate\n\nFixes #32229\n\nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Milan Broz \nMerge-date: Thu Aug 13 07:02:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32270\n(cherry picked from commit 5597d48bee6d7feb3b8c87fef7bcc52f3a3da91c)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5597d48bee6d7feb3b8c87fef7bcc52f3a3da91c", "date": "2026-08-04T16:45:37+02:00", "subject": "CI: specify RISC-V vector extension version in QEMU", "body": "Made according to the recomimendation of @ZenithalHourlyRate\n\nFixes #32229\n\nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Milan Broz \nMerge-date: Thu Aug 13 07:02:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32270", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "curl", "sha": "5050196dcc74b8369bcfad35d3b0926d0ff96180", "date": "2026-08-04T16:52:46+02:00", "subject": "build: require `!NDEBUG` for debug-enabled (aka development) builds", "body": "To prevent creating a curl-development (aka debug-enabled, aka\n`-DENABLE_DEBUG=ON`, `--enable-debug` or `DEBUGBUILD`) build without\n`assert()`s. Since it may break expectations by missing to catch error\ncases. Also to make it less likely to build an inherently insecure\ndevelopment build by accident.\n\n- in CMake, for non-Multi-Config builds, show this error unless Debug\n configuration is set at configure-time:\n ```\n CMake Error at CMakeLists.txt:263 (message):\n Debug-enabled (aka development mode) curl requires the Debug configuration\n ```\n Ref: https://github.com/curl/curl/actions/runs/31159065911/job/92805128331?pr=22513#step:6:56\n The other 3 predefined configs (`Release`, `RelWithDebInfo`,\n `MinSizeRel`) set `NDEBUG` automatically.\n\n- or this message in other cases, e.g. when using CMake Multi-Config\n with non-Debug, e.g. `--config Release`, or passed `NDEBUG` manually\n with either build tools:\n ```\n curl_setup.h(1062): fatal error C1189: #error:\n \"Debug-enabled builds cannot be combined with NDEBUG\"\n ```\n Ref: https://github.com/curl/curl/actions/runs/31159628749/job/92807878479#step:9:28\n\nAlso:\n- INSTALL-CMAKE.md: document.\n- GHA/windows: adjust CI jobs to meet the above requirement, where\n missing.\n\nReported-by Stefan Eissing\nBug: https://github.com/curl/curl/pull/22484#discussion_r3730045572\nRef: ea984456804ab3ff1656a668fabdf734a32e379e #13592\nFixes #22481\nFixes #22484\n\nCloses #22513", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ef9a76655168506596240a5a747f65ac08b9df53", "date": "2026-08-04T09:03:33-06:00", "subject": "Don't treat a CRL distribution point URI as a NUL terminated string", "body": "get_dp_url() returned the internal ASN1_STRING data of a DIST_POINT URI\ndirectly. That data has an explicit length and is not necessarily NUL\nterminated, yet load_crl() and the HTTP client it hands the URI to consume\nit as a C string, reading past the buffer for a certificate carrying a\nprimitive-form IA5String URI in its CRL distribution points.\n\nReturn an allocated NUL terminated copy of the URI and free it in\nload_crl_crldp().\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:39 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9cd4dd1c1934b21cf4acc7950da6b2b811c41b71", "date": "2026-08-04T17:13:32+02:00", "subject": "fix(:bcd): do not \"inherit\" buffer-local dir", "body": "Problem:\nBuffer-local CWD (:bcd) is \"sticky\", similar to window-local CWD (:lcd).\nBut this contradicts one of its main benefits: per-buffer \"project root\"\nfor LSP, OSC7.\n\nOther problems:\n- A buffer created with :edit/:enew/:new silently inherits b_localdir\n (and b_prevdir) from the previous buffer.\n- curbuf_reusable() refuses to recycle a scratch buffer that has\n `b_localdir`.\n- After :new/:vnew/:tabnew the CWD sticks to previous buffer's\n `b_localdir` even though the new curbuf has none, so :new is not\n equivalent to \":split | enew\", and getcwd() disagrees with\n haslocaldir().\n- Requires \"which buffer spawned this buffer\" semantics that no other\n buffer-local state has.\n\nSolution:\nDrop sticky/inherit behavior of buffer-local CWD (:bcd).\n\n- do_ecmd: always apply the new curbuf's dir (`fix_current_dir`), like\n `do_autochdir` already does. :tabnew from a :bcd buffer now reverts to\n global CWD (and fires DirChanged), same as :tabnew from a :lcd window.\n- curbuf_reusable(): recycling a scratch buffer frees its b_localdir.\n\nTo get sticky/inherit behavior of CWD, use `:lcd`.", "style": "conventional", "type": "fix", "scope": ":bcd", "breaking": false} {"repo": "axios", "sha": "f02b722c082f55486d708fe6aea290a8d1ac3441", "date": "2026-08-04T17:16:53+02:00", "subject": "docs: update sponsorship link and add new sponsor information", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "32af9b525efe005998a9cd02aebfcc08e6793cf2", "date": "2026-08-04T17:29:46+02:00", "subject": "fix(http): strip RFC 6265 DQUOTE characters and handle URIError in parseCookieValue", "body": "Previously, `parseCookieValue` did not strip enclosing double quotes (`DQUOTE`) from quoted cookie values as specified in RFC 6265 Section 4.1.1. In addition, malformed percent-encoding in cookie values caused an unhandled `URIError` when calling `decodeURIComponent`.\n\n(cherry picked from commit 280d09b16011e4f53f9e696fd95964943c6dcbfa)", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "angular", "sha": "280d09b16011e4f53f9e696fd95964943c6dcbfa", "date": "2026-08-04T17:29:46+02:00", "subject": "fix(http): strip RFC 6265 DQUOTE characters and handle URIError in parseCookieValue", "body": "Previously, `parseCookieValue` did not strip enclosing double quotes (`DQUOTE`) from quoted cookie values as specified in RFC 6265 Section 4.1.1. In addition, malformed percent-encoding in cookie values caused an unhandled `URIError` when calling `decodeURIComponent`.", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "openssl", "sha": "def638aa2d6895d36648cbccfb16444e24311683", "date": "2026-08-04T22:55:07+07:00", "subject": "fix module extension on non-.so platforms", "body": "pathed.cnf and nocache-and-default.cnf hardcoded \"p_test.so\", but\nprovider modules use the platform's DSO extension (e.g. .dylib on\nmacOS), so the file never existed there.\n\nConvert both configs to build-time templates to\nfill in the correct extension.\nUpdate prov_config_test.c's presence check to use DSO_EXTENSION, and\nhave 30-test_prov_config.t read the generated build-tree file.\n\nAssisted-by: Claude:claude-sonnet-5\n\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Bob Beck \nMergeDate: Fri Aug 7 01:45:49 2026\n(Merged from https://github.com/openssl/openssl/pull/32192)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1cb8e8dffaaac0a22060b098ae2baa9b7b65b83d", "date": "2026-08-04T18:00:37+02:00", "subject": "statem: fix missing SSLfatal in TLSv1.3 ticket construction", "body": "tls_construct_new_session_ticket() assumed ssl_session_dup() calls\nSSLfatal() on failure but it never does, unlike all its other call\nsites which call SSLfatal() themselves. An allocation failure there\nmade the construct function return CON_FUNC_ERROR without entering\nthe fatal state, tripping the check_fatal assertion in\nwrite_state_machine() on debug builds.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Tomas Mraz \nMergeDate: Mon Aug 10 13:55:51 2026\n(Merged from https://github.com/openssl/openssl/pull/32183)\n\n(cherry picked from commit d271efc353856df91d28c8f54ead56b298c852f0)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d271efc353856df91d28c8f54ead56b298c852f0", "date": "2026-08-04T18:00:37+02:00", "subject": "statem: fix missing SSLfatal in TLSv1.3 ticket construction", "body": "tls_construct_new_session_ticket() assumed ssl_session_dup() calls\nSSLfatal() on failure but it never does, unlike all its other call\nsites which call SSLfatal() themselves. An allocation failure there\nmade the construct function return CON_FUNC_ERROR without entering\nthe fatal state, tripping the check_fatal assertion in\nwrite_state_machine() on debug builds.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Tomas Mraz \nMergeDate: Mon Aug 10 13:55:51 2026\n(Merged from https://github.com/openssl/openssl/pull/32183)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "308a361f71bd7c6966cc9502a0521da76b22eedf", "date": "2026-08-04T16:09:58Z", "subject": "build: update domino digest to e0779df", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c6674c2f9a0b5a006dd4aef9902ac29ce1a0692e", "date": "2026-08-04T12:35:19-04:00", "subject": "Update pkcs11-provider and drop test disablement", "body": "This version of pkcs11-provider has been fixed to work with\nOpenSSL ASN1 changes and the test passes again.\n\nSigned-off-by: Simo Sorce \n\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Milan Broz \nMergeDate: Thu Aug 6 12:16:37 2026\n(Merged from https://github.com/openssl/openssl/pull/32180)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "6f86d666cabd98e8286eeeafe98e7acae911113f", "date": "2026-08-04T21:44:36+05:00", "subject": "docs: fix supported Node.js version in installation prerequisites", "body": "(cherry picked from commit 96b80424c787f7a278b543c19f2c2364041761f4)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "96b80424c787f7a278b543c19f2c2364041761f4", "date": "2026-08-04T21:44:36+05:00", "subject": "docs: fix supported Node.js version in installation prerequisites", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "axios", "sha": "ac2813149c7ffaf7bcc6dfde3731eb61cdaa5a5d", "date": "2026-08-04T19:18:17+02:00", "subject": "ci: use locked ESM smoke dependencies", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "neovim", "sha": "ceccca7780c54d932dd183517c4c04a340486dd7", "date": "2026-08-04T13:28:26-04:00", "subject": "fix(options): crash on \":let &l:autoread = v:true\" #41158", "body": "Problem:\nAssigning to the local value of a global-local boolean option\n('autoread', 'autocomplete', 'fsync') aborts:\n\n Assertion failed: (curval.type == newval.type), function\n ex_let_option, file vars.c, line 1408.\n\n ex_let_one\n ex_let_vars\n ex_let\n execute_cmd0\n do_cmdline\n call_user_func\n ...\n eval_map_expr\n vgetorpeek\n vgetc\n state_enter\n main\n\nA global-local option may have local value `kObjectTypeUnset`, but\n`ex_let_option()` guards only `kObjectTypeNil`.\n\nSolution:\nWhen curval is Unset, resolve it to the inherited global value.", "style": "conventional", "type": "fix", "scope": "options", "breaking": false} {"repo": "axios", "sha": "f64e4586e9a6a6458af9364bc0b5839ce8cac571", "date": "2026-08-04T18:30:14+01:00", "subject": "docs: add localized global search", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "axios", "sha": "ceb4176f5440faebc76c9560798cda010b3f118b", "date": "2026-08-05T02:38:23+09:00", "subject": "docs(readme): fix broken interceptor test link", "body": "", "style": "conventional", "type": "docs", "scope": "readme", "breaking": false} {"repo": "neovim", "sha": "f6de99d028c27337cdbac2e1d1c1b954e9badc48", "date": "2026-08-04T23:13:22+05:30", "subject": "fix(highlight): make Ignore hidden by default #41115", "body": "Problem:\nIgnore is linked to Normal by default, making the text visible instead\nof hidden.\n\nSolution:\nReplace the default link with an explicit highlight definition using\nctermfg=0 guifg=bg.", "style": "conventional", "type": "fix", "scope": "highlight", "breaking": false} {"repo": "openssl", "sha": "6fbc443ad53299972c5eb71ea78f6422de23d3d1", "date": "2026-08-04T19:52:50+02:00", "subject": "Fix memory leaks in x509v3 config handling", "body": "Duplicate integer value fields in the policyConstraints,\nbasicConstraints and basicAttConstraints extensions overwrite\npreviously allocated ASN1_INTEGER values without freeing them.\n\nFound by : x509v3 fuzzer\n\nExample 1: (crypto/x509/v3_pcons.c)\n```\n[default]\npolicyConstraints = requireExplicitPolicy:1,requireExplicitPolicy:2\n```\n\nExample 2: (crypto/x509/v3_bcons.c)\n```\n[default]\nbasicConstraints = pathlen:1,pathlen:2\n```\n\nExample 3: (crypto/x509/v3_battcons.c)\n```\n[default]\nbasicAttConstraints = pathlen:1,pathlen:2\n```\nAssisted-by: ChatGPT:gpt-5.6\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nMergeDate: Sun Aug 9 04:06:32 2026\n(Merged from https://github.com/openssl/openssl/pull/32181)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "451871b740d7ad252b8aeaf16a20bcf831a39254", "date": "2026-08-04T10:57:43-07:00", "subject": "Support UUIDv6 in uuid_extract_timestamp().", "body": "Previously, uuid_extract_timestamp() returned NULL for UUIDv6 values,\neven though it already extracted timestamps from UUIDv1 and\nUUIDv7.\n\nThis commit adds a UUIDv6 branch that reassembles the timestamp from\nthat layout and then converts it exactly as the UUIDv1 branch does. We\nprovide no uuidv6() generator, but such values can be produced by\nclient applications and stored in the database, just as UUIDv1 values\nalready are.\n\nAuthor: Tristan Partin \nReviewed-by: Miłosz Bieniek \nReviewed-by: Masahiko Sawada \nDiscussion: https://postgr.es/m/DJHJW50YWKIT.XIOT4NOZRQ03@partin.io", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9c5afe660667537f876d11c0486bedd7eb363e61", "date": "2026-08-04T14:00:13-04:00", "subject": "test: unreliable \"nvim.zip … incorrect password\" #41159", "body": "Problem:\n\n FAILED …/plugin/zip_spec.lua @ 442: nvim.zip reports an incorrect archive password\n Expected values to be equal.\n Expected:\n true\n Actual:\n false\n stack traceback:\n …/plugin/zip_spec.lua:453: in function <…/plugin/zip_spec.lua:442>\n\nSolution:\nThe message is scheduled, so poll for it.", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "838f130e6a0b018853c63e7908f98e53a2b7fa7e", "date": "2026-08-04T20:00:51+02:00", "subject": "fix(types): emmylua warning about `params` field", "body": "", "style": "conventional", "type": "fix", "scope": "types", "breaking": false} {"repo": "axios", "sha": "39955a6ebed53217a524a3c8c7e5568fc0faa4bd", "date": "2026-08-04T23:47:16+05:00", "subject": "feat(HttpStatusCode): add RFC 9110 aliases for 413 and 422", "body": "", "style": "conventional", "type": "feat", "scope": "HttpStatusCode", "breaking": false} {"repo": "neovim", "sha": "7b03df5d5402bc0564b77b5e660a3224bcbb484f", "date": "2026-08-05T03:24:35+08:00", "subject": "feat(lua): vim.fs.slug() #41005", "body": "Problem:\nSeveral subsystems need to derive a short, filesystem-safe identifier from an\narbitrary path, and each reinvents it ad-hoc:\n- `'undodir'` and `swapfiles` encode the full path into a single filename, which\n may exceed filesystem length-limits.\n- `:connect ssh://` needs the SSH ControlPath socket name to stay under the\n 104-byte `sun_path` limit on macOS; today the path overflows it.\n- the upcoming :terminal state dir.\n- arbitrary plugin purposes.\n\nSolution:\nProvide `vim.fs.slug()`, which generates a bounded, one-way filename from an\narbitrary string. The input is normalized so equivalent paths produce the\nsame result. An 8-char hash is appended for uniqueness", "style": "conventional", "type": "feat", "scope": "lua", "breaking": false} {"repo": "axios", "sha": "e8147e6ae25a74793e48b4e00e2f98bc303f5b29", "date": "2026-08-04T15:31:05-04:00", "subject": "fix(http): stop per-socket error listener from retaining the installing request's context (memory leak)", "body": "", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "neovim", "sha": "2f9ef98a331959ea6fdb12337be5337df561ac06", "date": "2026-08-04T15:38:22-04:00", "subject": "test(chdir): cleanup #41161", "body": "Problem:\nSome assertions are erroneously skipped for `not is_os('win')`.\n\nSolution:\nUpdate tests. Deduplicate logic.", "style": "conventional", "type": "test", "scope": "chdir", "breaking": false} {"repo": "postgres", "sha": "3f2fdd150eb04fe1da60e2a99f4ba2dd2be6443b", "date": "2026-08-04T13:41:02-07:00", "subject": "Add comments to uuid_extract_timestamp().", "body": "The timestamp-reassembly expressions in uuid_extract_timestamp() are\nchains of shifted octets whose shift counts encode the field layout of\neach UUID version, which is not otherwise recoverable from the code.\n\nThis commit documents, for UUIDv1, UUIDv6, and UUIDv7 alike, which\noctets carry which part of the timestamp and where each part lands in\nthe reassembled value, citing the RFC section that defines it.\n\nAuthor: Tristan Partin \nReviewed-by: Miłosz Bieniek \nReviewed-by: Masahiko Sawada \nDiscussion: https://postgr.es/m/DJHJW50YWKIT.XIOT4NOZRQ03@partin.io", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "df183c63f671c2fdf8152446f2867bedcb17fac4", "date": "2026-08-04T16:55:00-04:00", "subject": "DTLS 1.3 Update CHANGES.md", "body": "Updating the CHANGES.md for all the changes added via the DTLS1.3\nfeature branch.\n\nFixes: openssl/project#2039\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Tue Aug 11 14:52:30 2026\n(Merged from https://github.com/openssl/openssl/pull/32182)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "18992dc9d98022a43831da8e1868abf3ce5cd66a", "date": "2026-08-05T06:53:10+09:00", "subject": "Improving handling of oversized records in xlogreader.c", "body": "XLogRecordAssemble() refuses records larger than XLogRecordMaxSize when\ngenerating records, but the reader-side only checked a minimal number\nfor xl_tot_len.\n\nA crafted multi-page record with xl_tot_len near UINT32_MAX could bypass\nthe contrecord length checks, overflow allocate_recordbuf()'s size math,\nand corrupt memory during reassembly (or hit related asserts under\ncassert).\n\nxlogreader.c is updated to check that the received record is never\nlarger than XLogRecordMaxSize, when first receiving the first bytes of\nxl_tot_len, then again when validating the record header.\n\nWAL is a source of data trusted by the server, with CRC32 checksums\nproviding a level of protection before replaying the records if data is\ncorrupted. These limits could be internally reached only with crafted\nWAL records, which is something that analyzers (named as LLMs) like\ncomplaining about these days.\n\n039_end_of_wal.pl is adjusted according to the new maximum limit, a test\nfor XLogRecordMaxSize is added.\n\nAuthor: David Karapetyan \nAuthor: Matthias van de Meent \nDiscussion: https://postgr.es/m/CALmTjZBddzeGVU2o1JNZb3mPvV68PqDskWPAUQBVb-35yVkiwA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2f768dda531c94c071972a68d77163fba35684a6", "date": "2026-08-04T18:04:28-04:00", "subject": "Silence Coverity warning about unused visibilitymap_clear() result", "body": "Commit c0d9864f5ce made all but one caller of visibilitymap_clear() check\nits return value, causing Coverity to flag the remaining unchecked call\nin heap_page_fix_vm_corruption().\n\nThis VM clear is not WAL-logged, so the caller doesn't need the return\nvalue of visibilitymap_clear(). Add an explicit void cast and comment to\ndocument that the return value is intentionally ignored.\n\nBackpatch to 19 when the number of callers discarding the return value\ndropped low enough to trigger Coverity's warning.\n\nDiscussion: https://postgr.es/m/1065814.1784514869%40sss.pgh.pa.us\nReported-by: Tom Lane \nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "6d8401f8d51ff020855932d9ebd78c79ed6ab4a4", "date": "2026-08-05T06:42:39+08:00", "subject": "fix(ui): missing vertical separator for a winbar only window", "body": "Problem: A winbar-only window with zero text height still occupies one row,\n but win_update() returns early on w_view_height == 0 and skips the\n vertical separator.\n\nSolution: Also draw the vertical separator in the early return path.", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "openssl", "sha": "70cebd74d3592f5272945501b58a60374c4e13af", "date": "2026-08-05T00:56:25+02:00", "subject": "Avoid double free of qrx in port_default_packet_handler()", "body": "port_default_packet_handler() may perform double free of qrx\nwhen channel creation fails. The port_default_packet_handler()\ntransfers ownership of qrx to channel/connection via call to\nport_bind_channel(). The port_bind_channel() however may\nrelease the qrx when channel can not be bound. The error is\nthen detected in port_default_packet_handler() which then agains\nreleases qrx for the second time.\n\nThe fix is to add a reference counter to QRX object so transfer\nof ownership between port_default_packet_handler() and QUIC_CHANNEL\ncan be handled safely.\n\nFixes CVE-2026-18798\n\nReviewed-by: Igor Ustinov \nReviewed-by: Andrew Dinh \nMerge-date: Mon Aug 24 15:47:02 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "967582d5037f01a26b6d19beae19af62a1b15c3c", "date": "2026-08-05T00:56:25+02:00", "subject": "Avoid double free of qrx in port_default_packet_handler()", "body": "port_default_packet_handler() may perform double free of qrx\nwhen channel creation fails. The port_default_packet_handler()\ntransfers ownership of qrx to channel/connection via call to\nport_bind_channel(). The port_bind_channel() however may\nrelease the qrx when channel can not be bound. The error is\nthen detected in port_default_packet_handler() which then agains\nreleases qrx for the second time.\n\nThe fix is to add a reference counter to QRX object so transfer\nof ownership between port_default_packet_handler() and QUIC_CHANNEL\ncan be handled safely.\n\nFixes CVE-2026-18798\n\nReviewed-by: Igor Ustinov \nReviewed-by: Andrew Dinh \nMerge-date: Mon Aug 24 15:48:03 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a14a1deac403522fbeafabcb198503cf6caa7dc4", "date": "2026-08-05T00:56:25+02:00", "subject": "Avoid double free of qrx in port_default_packet_handler()", "body": "port_default_packet_handler() may perform double free of qrx\nwhen channel creation fails. The port_default_packet_handler()\ntransfers ownership of qrx to channel/connection via call to\nport_bind_channel(). The port_bind_channel() however may\nrelease the qrx when channel can not be bound. The error is\nthen detected in port_default_packet_handler() which then agains\nreleases qrx for the second time.\n\nThe fix is to add a reference counter to QRX object so transfer\nof ownership between port_default_packet_handler() and QUIC_CHANNEL\ncan be handled safely.\n\nFixes CVE-2026-18798\n\nReviewed-by: Igor Ustinov \nReviewed-by: Andrew Dinh \nMerge-date: Mon Aug 24 15:46:33 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b6011a9ee87ab25f47d999a716a1567da00e1878", "date": "2026-08-05T00:56:25+02:00", "subject": "Avoid double free of qrx in port_default_packet_handler()", "body": "port_default_packet_handler() may perform double free of qrx\nwhen channel creation fails. The port_default_packet_handler()\ntransfers ownership of qrx to channel/connection via call to\nport_bind_channel(). The port_bind_channel() however may\nrelease the qrx when channel can not be bound. The error is\nthen detected in port_default_packet_handler() which then agains\nreleases qrx for the second time.\n\nThe fix is to add a reference counter to QRX object so transfer\nof ownership between port_default_packet_handler() and QUIC_CHANNEL\ncan be handled safely.\n\nFixes CVE-2026-18798\n\nReviewed-by: Igor Ustinov \nReviewed-by: Andrew Dinh \nMerge-date: Mon Aug 24 15:46:13 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "2468464798a0fc050f86f8fa0bb96a34851bd824", "date": "2026-08-05T07:57:47+08:00", "subject": "fix(v-model): preserve text input before hydration", "body": "close #14403", "style": "conventional", "type": "fix", "scope": "v-model", "breaking": false} {"repo": "angular", "sha": "4f7e9987fa35888de43895bfd859382dfa436e51", "date": "2026-08-04T19:58:51-05:00", "subject": "fix(http): always decode JSON responses as UTF-8", "body": "Keep the charset handling added in #70062 limited to text responses. JSON\nbytes must stay UTF-8 regardless of the Content-Type charset.\n\nFetch defines Body.json() using \"parse JSON from bytes\". Infra specifies that\nstep as: \"Let string be the result of running UTF-8 decode on bytes.\"\n\nhttps://fetch.spec.whatwg.org/#dom-body-json\nhttps://infra.spec.whatwg.org/#parse-json-bytes-to-a-javascript-value\n(cherry picked from commit 09bc90003ed6e9f6c166e885503728af49d1af67)", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "angular", "sha": "09bc90003ed6e9f6c166e885503728af49d1af67", "date": "2026-08-04T19:58:51-05:00", "subject": "fix(http): always decode JSON responses as UTF-8", "body": "Keep the charset handling added in #70062 limited to text responses. JSON\nbytes must stay UTF-8 regardless of the Content-Type charset.\n\nFetch defines Body.json() using \"parse JSON from bytes\". Infra specifies that\nstep as: \"Let string be the result of running UTF-8 decode on bytes.\"\n\nhttps://fetch.spec.whatwg.org/#dom-body-json\nhttps://infra.spec.whatwg.org/#parse-json-bytes-to-a-javascript-value", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "vue-core", "sha": "02421cdbc4da5dd2eaf39e6c51aa790f9310db62", "date": "2026-08-05T09:43:39+08:00", "subject": "fix(compiler-core): preserve vnode lifecycle in stable v-for", "body": "close #9239\nclose #12569", "style": "conventional", "type": "fix", "scope": "compiler-core", "breaking": false} {"repo": "openssl", "sha": "1c4aa2ec079771f87fde0a7cba9c5093bc666ecf", "date": "2026-08-05T12:06:40+10:00", "subject": "FIPS: key gen PCT tests no longer cause FIPS provider to enter a unrecoverable error state.", "body": "Reviewed-by: Viktor Dukhovni \nReviewed-by: Paul Dale \nMergeDate: Fri Aug 7 09:58:28 2026\n(Merged from https://github.com/openssl/openssl/pull/32191)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5be1a2a1432947ca0a381848d71f5e852d85929d", "date": "2026-08-05T12:06:40+10:00", "subject": "FIPS: key gen PCT tests no longer cause FIPS provider to enter a unrecoverable error state.", "body": "Reviewed-by: Viktor Dukhovni \nReviewed-by: Paul Dale \nMergeDate: Fri Aug 7 09:57:46 2026\n(Merged from https://github.com/openssl/openssl/pull/32191)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1aaca5ab8ffb7f9387dce2a2de128eff006ac1c4", "date": "2026-08-05T11:29:15+09:00", "subject": "doc: Clarify wal_sender_shutdown_timeout values", "body": "Clarify how special wal_sender_shutdown_timeout values affect shutdown\nwaiting. A value of -1 disables the shutdown timeout and lets the\nWAL sender wait for the receiver to catch up, while 0 causes immediate\ntermination without waiting for catch-up. Positive values bound how long\nshutdown waits, so document that they should be high enough for WAL data\nto be replicated under normal circumstances.\n\nUpdate the main documentation, the GUC long description, and\npostgresql.conf.sample consistently. While here, fix nearby paragraph\nindentation and spacing.\n\nBackpatch to v19, where wal_sender_shutdown_timeout was introduced.\n\nAuthor: Chao Li \nReviewed-by: Ian Barwick \nReviewed-by: Daniel Gustafsson \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/AF4FE756-A220-4DA7-87B7-A126F3F307FD@gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "3d4216d85f42d9f011001367d5f962ca8d1d72e2", "date": "2026-08-05T11:34:29+09:00", "subject": "Clarify wraparound warning percentage messages", "body": "Commit e646450e609 added DETAIL messages for XID and MultiXactId\nwraparound warnings that described the reported percentage as the\npercentage of IDs available for use. However, the percentage is actually\ncalculated from the remaining distance to the wraparound limit, not the\nstop limit where new IDs are refused. As a result, the wording could be\nmisinterpreted as meaning that the reported percentage of IDs can still\nbe allocated.\n\nUpdate the wording to clarify that the reported percentage represents the\nremaining transaction ID space or MultiXactId space before wraparound.\n\nAlso update the related XID warning hints to use \"transaction ID\"\nterminology consistently.\n\nBackpatch to v19, where the DETAIL messages were added.\n\nAuthor: Fujii Masao \nReviewed-by: Bharath Rupireddy \nReviewed-by: Kyotaro Horiguchi \nReviewed-by: Nathan Bossart \nReviewed-by: Yugo Nagata \nDiscussion: https://postgr.es/m/CAHGQGwHWZTsR6bjdfpa+pOkBPjoXXm2LuJ+5nh+CvdyqEASHcQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9484169ba070e1a7fe5b9b1d7a5f6035a37af0a3", "date": "2026-08-05T11:36:43+09:00", "subject": "doc: Update XID wraparound error example", "body": "Commit edee0c621de, and the equivalent v17 commit f2353dd71724,\nchanged the runtime XID wraparound messages to use \"transaction IDs\"\nterminology, but one corresponding error example in maintenance.sgml\nstill used the older XID wording.\n\nUpdate that documentation example in line with the current runtime\nmessage.\n\nBackpatch to v17, where the runtime messages were changed.\n\nAuthor: Fujii Masao \nReviewed-by: Yugo Nagata \nDiscussion: https://postgr.es/m/CAHGQGwHTN-Xc5iDtbzNSjfxuab5Y9qAArw8cB4PrrDJpZ+1fgA@mail.gmail.com\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "eac0cc84aa2472a85a19ee84561c1ba71e381a55", "date": "2026-08-05T10:50:46+08:00", "subject": "fix(bundled-dev): inject client script tag before chunk scripts", "body": "", "style": "conventional", "type": "fix", "scope": "bundled-dev", "breaking": false} {"repo": "vite", "sha": "ad55e344217ad86c1572567e10ba5f40002a13a3", "date": "2026-08-05T11:11:08+08:00", "subject": "test(bundled-dev): enable html playground", "body": "", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "vue-core", "sha": "22b53eafecf56a9fb97f00b8e8b69c90a073d2d4", "date": "2026-08-05T11:29:54+08:00", "subject": "fix(custom-element): warn when props override native properties", "body": "close #12124", "style": "conventional", "type": "fix", "scope": "custom-element", "breaking": false} {"repo": "ant-design", "sha": "1a4b06c135d94077775f404745f0020b22efe9ec", "date": "2026-08-04T21:38:12-07:00", "subject": "docs: subject-verb agreement in ComponentTokenTable locale", "body": "\"here is your component/global tokens\" doesn't agree with the plural\n\"tokens\". This copy is rendered inside the Component Token / Global\nToken demo shown on every component's docs page via ComponentTokenTable.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "07f58006c5b9fb719317196371642abc86775435", "date": "2026-08-05T06:06:43Z", "subject": "refactor(devtools): rely on esbuild to downlevel async generators", "body": "Configure esbuild with supported 'async-await': false in esbuild-base.config.mts so esbuild downlevels async functions and async generators natively.\n\nThis allows removing the custom Babel plugin @babel/plugin-transform-async-generator-functions and downlevelAsyncGeneratorsIfPresent option.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "angular", "sha": "573145afb5001de9bc68e373f0e55b68b84f4e59", "date": "2026-08-05T06:30:10Z", "subject": "refactor(devtools): remove async-await downleveling as devtools is zoneless", "body": "Remove the supported 'async-await': false option in esbuild-base.config.mts as Angular DevTools is zoneless and does not require downleveling async/await syntax.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "postgres", "sha": "c5e11de2f378b8e52893bfcd3f6ffd583fffc160", "date": "2026-08-05T08:35:10+02:00", "subject": "More tab-completion for DROP PROPERTY GRAPH", "body": "This adds completion for CASCADE and RESTRICT, similar to what\ncompletion of other DROP commands provides.\n\nAuthor: Peter Smith \nReviewed-by: Chao Li \nDiscussion: https://www.postgresql.org/message-id/flat/CAHut+PuHsLupVWZ9SUeKm=jKR34wYP6qAZPoOHFhU3uDbx34wQ@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "9b293c1f7be32f85fadddf1fd716a924f0a2b4b1", "date": "2026-08-05T15:46:08+09:00", "subject": "feat: register references in HTML to safe modules", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "4e467d7ae2cd7d0af031fac8865dcff73cb1cf16", "date": "2026-08-05T14:50:07+08:00", "subject": "fix(ssr): normalize hidden states during hydration", "body": "", "style": "conventional", "type": "fix", "scope": "ssr", "breaking": false} {"repo": "tmux", "sha": "72e705cbd6c3d92b48c56fa2f687a12b838a6b66", "date": "2026-08-05T08:26:10+01:00", "subject": "Tweak times for macOS.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0198ee9e51bd47c7947fde3d9f5ebdc13ae82030", "date": "2026-08-05T07:27:45Z", "subject": "Do not log when logging is off, from Michael Grant.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "fe384e06bc9fa83c589c0294ab3f9489fe848c99", "date": "2026-08-05T07:31:08Z", "subject": "Add pane_private_modes format with list of DEC private modes, from George Nachman.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "1995b5559d749bd804ac938500814e2c0e049bfd", "date": "2026-08-05T07:35:35Z", "subject": "Do not crash when synchronizing groups and there is no current window. GitHub issue 5467 from Ju-an Zhang.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "9ea48811fed455d4a869eb100b2216f039f8677a", "date": "2026-08-05T09:43:57+02:00", "subject": "servers: drop duplicate (and interacting) ctrl handlers on Windows, add exit message", "body": "On Windows, the init code calls `SetConsoleCtrlHandler()`, and before\nthis patch also set handlers for all Unixy signals. Of these, `SIGBREAK`\n(used on Windows-only), `SIGINT`, `SIGABRT` and `SIGTERM` were also\nsetting up a `SetConsoleCtrlHandler()`, in addition to the call made\ndirectly. (The rest, `SIGHUP`, `SIGPIPE`, `SIGALRM` are either missing\nthe macros, or ignored by `signal()` on Windows.)\n\nAs per WINE sources, `SetConsolCtrlHandler(, TRUE)` calls are\nadditive, which means the test server set up two console ctrl handlers.\n\nThen the ctrl handler set directly (`ctrl_event_handler()`), was\ntriggering the other signal handler via `raise()`, for the 'initiate\nexit' logic, which in turn triggered exiting a wait within `select_ws()`\nand other loops. The Windows window handler also made use of the\n`SIGTERM` event to initiate exit via `raise()` and the second signal\nhandler.\n\nTo simplify, de-duplicate the ctrl handlers by dropping `signal()` calls\nand keeping the direct Win32 call with `ctrl_event_handler()` doing all\nthe signal handling on Windows. Break out the 'initiate exit' logic into\na function and call it from both Unix and Windows signal/ctrl/window\nhandlers. Also drop calling `raise()` on exit, because it's a no-op\nwithout a `signal()` pair.\n\nAlso:\n- drop logging the actual ctrl type number, replace with just logging\n whether we handled the event, in `ctrl_event_handler()`. To avoid\n using non-signal-safe functions (e.g. `fprintf()`) from the handler.\n- also replace `logmsg()` with `WriteFile()` to prevent regressions.\n Ref: #22045\n- replace `logmsg()` with `WriteFile()` in `main_window_proc()`.\n- fix to forward ctrl handling to the OS in the rare case of failed\n `exit_event` initialization on startup. To swap a possible hang\n (within `WaitForMultipleObjectsEx()`) with an ungraceful shutdown.\n- add support for an 'exit message' string, set by signal/ctrl handlers,\n and log it on app exit. To avoid the need to deal with logging within\n the handlers,", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "e56a04215313b42326e91971cb83759d410d61a2", "date": "2026-08-05T07:50:21Z", "subject": "Use ACS arrows for tree mode.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0e2fb2c8c307aaee5ab47aa92055b53ca146706b", "date": "2026-08-05T16:55:26+09:00", "subject": "Initialize read stream before fetching metapage in hash bulk-deletion", "body": "hashbulkdelete() fetched the relcache's cached hash metapage before\ncalling read_stream_begin_relation(). During the read stream\ninitialization, relation lookups may process pending relcache\ninvalidation messages, which could cause the cached metapage to be\ninvalidated before the read stream uses it, leading to the failure of a\nVACUUM bulk-deletion for a hash index.\n\nThis commit reworks the order of hashbulkdelete() so as its read stream\nis initialized before fetching the cached metapage, so as pending\ninvalidation messages do not interfere with the relation scan.\n\nIssue introduced by bfa3c4f106b1.\n\nAuthor: Mikhail Nikalayeu \nReviewed-by: Bertrand Drouvot \nReviewed-by: Nazir Bilal Yavuz \nReviewed-by: Xuneng Zhou \nDiscussion: https://postgr.es/m/CADzfLwVEJ2_7ioX3ZSB1P7qXW40ghUy_ZCzeFvhDLoa_3Muztg@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "369db905fe3dc806b24ecfcb8402a2c8072aebce", "date": "2026-08-05T10:02:53+02:00", "subject": "CVE-2026-18924: HTTP/2 server push UAF", "body": "http2: make server push transfers inherit share from parent\n\nReported-by: Stephan Zeisberg\nCloses #22488", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "6ed70f61f643ff84894c7d17b590a925509c4b09", "date": "2026-08-05T10:02:53+02:00", "subject": "CVE-2026-18924: HTTP/2 server push UAF", "body": "Cherry-picked from 90325ff0444cbdff368bda5d2\n\nhttp2: make server push transfers inherit share from parent\n\nReported-by: Stephan Zeisberg\nCloses #22488", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "90325ff0444cbdff368bda5d26d6405a0bb6ee43", "date": "2026-08-05T10:02:53+02:00", "subject": "http2: make server push transfers inherit share from parent", "body": "Reported-by: Stephan Zeisberg\nCloses #22488", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "e1f247d36f5ad5f30641500e5aa51b15e9e14fa8", "date": "2026-08-05T13:36:53+05:00", "subject": "fix(Tour): remove unusable per-step indicatorsRender/actionsRender types", "body": "* fix(Tour): remove unusable per-step indicatorsRender/actionsRender types\n\n* test(Tour): drop type-only assertion, already enforced by tsc", "style": "conventional", "type": "fix", "scope": "Tour", "breaking": false} {"repo": "postgres", "sha": "f585671055d1c56d6fba0bc5835d28e68248ccfe", "date": "2026-08-05T10:44:10+02:00", "subject": "Disallow aggregates, window functions, and SRFs in GRAPH_TABLE COLUMNS", "body": "The COLUMNS list of a GRAPH_TABLE query is parsed as an ordinary select\ntarget list, which permits aggregate functions, window functions, and\nset-returning functions. GRAPH_TABLE has no machinery to evaluate them,\nthough: the rewriter copies the COLUMNS target list verbatim into a\nfreshly built subquery whose hasAggs/hasWindowFuncs/hasTargetSRFs flags\nare never set, so the planner builds no Agg/WindowAgg node (and no SRF\nexpansion) and the Aggref/WindowFunc/SRF reaches the executor. This\ntriggers an assertion failure (\"ecxt_aggvalues != NULL\"), or \"Aggref\nfound in non-Agg plan node\" on a non-assert build, for otherwise\nparser-accepted SQL such as\n\n SELECT max(c) FROM GRAPH_TABLE\n (g MATCH (x IS v) COLUMNS (count(*) AS c));\n\nReject these constructs in transformRangeGraphTable() the same way\nsubqueries are already handled: save and clear pstate->p_hasAggs,\np_hasWindowFuncs, and p_hasTargetSRFs around the transformation of the\nCOLUMNS list, and raise a \"not supported\" error if any of them got set.\n\nThis is deliberately a blanket prohibition for now. Once quantified\nelement patterns such as (a)->{1,5} are supported, aggregates over\nproperty references of higher degree (e.g. count(a) or sum(a.val)) can\nbe allowed; at that point the check will need to inspect the aggregate\narguments rather than reject all aggregates outright.\n\nAuthor: Ewan Young \nReviewed-by: Ashutosh Bapat \nDiscussion: https://www.postgresql.org/message-id/flat/CAON2xHOYAmYLkB2jGi6g77d6Fqv8YgOrfV-riQVf0K_7AdxD3w@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "bd483a63cfa6714f765ef396bdcb35a0729149b0", "date": "2026-08-05T08:54:56Z", "subject": "Ignore Ms if it is invalid (for example no %p1) and report in client mode, GitHub issue 5460.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "1c1dc0558f784a66fb6d0f7dd3bae8d89274ad64", "date": "2026-08-05T11:20:05+02:00", "subject": "feat(cwd): support explicit chdir (:bcd/:tcd/…) in temp context", "body": "Problem:\n- Explicit `:bcd` (etc.) persists from `nvim_buf_call()` but not from an\n autocmd handler targeting a hidden buf (`LspAttach`, `TermRequest`, …),\n which needs a `vim.schedule()` workaround.\n- `vim._with()` is supposed to work as a \"sandbox\", discarding\n side-effects, but it leaks CWD changes: `:lcd` from a `win` context,\n any chdir from a visible-buffer context.\n\nSolution:\n- Explicit :cd/:tcd/:bcd during a temp context persists by default.\n - \"Ambient\" directory changes ('autochdir', existing win-local CWD,\n etc.) are still undone, as before.\n- Add `kCtxKeepDirs`: snapshot/restore the target's full CWD state\n (w/b/tp-local, global, cwd). Used by `vim._with()` and `'inccommand'`,\n which must not leak state.", "style": "conventional", "type": "feat", "scope": "cwd", "breaking": false} {"repo": "angular", "sha": "cf1844bcce4716a300c73a38a1e07e39e09e2701", "date": "2026-08-05T09:31:47Z", "subject": "refactor(devtools): use JavaScriptTransformer for Angular optimization esbuild plugin", "body": "Refactor the devtools optimization esbuild plugin to use JavaScriptTransformer from @angular/build/private.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "postgres", "sha": "49a712f32ceebfeecac9017a49f5b1fe89ae399f", "date": "2026-08-05T11:40:39+02:00", "subject": "Fix calculating length of match to localized month/weekday names", "body": "seq_search_localized() returns the length of the matching prefix in\n*len, but because it internally case-folds the inputs, it gets\nconfused on the length. The caller expects to get the length of the\nprefix in the original string, but what it actually returns is the\nlength of the prefix after case-folding, which can be different if the\ncase-folded characters have different byte-length than the original,\nor with ICU, if the case-folding changes the number of characters\n(e.g. \"ß\", the German double s).\n\nTo fix, once we have determined that we have a match, work harder to\nfind the match's length in the original string. This adds some\noverhead, but the strings are expected to be short.\n\nThe function does \"case-folding\" by converting a string to upper-case,\nthen to lower-case, which is a little ugly given that we have\ndedicated functions for case-folding nowadays. But switching to that\ndoesn't seem appropriate to backpatch in a security fix, and that's\nnot available in older stable versions, anyway.\n\nAuthor: Heikki Linnakangas ", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d23507a0f61a28f6d56d5d2b43594a119ee78deb", "date": "2026-08-05T05:37:02-05:00", "subject": "fix(terminal): OSC 52 multiline copy replaces newlines with NUL #41097", "body": "Problem: An OSC 52 sequence from a :terminal job passes the decoded\npayload to the clipboard provider as a single list item. Command-line\nproviders (pbcopy, xclip, ...) receive it with channel semantics, where\na newline inside an item is sent as NUL (:h chansend()), so multiline\ncopies arrive with NUL bytes instead of newlines.\n\nSolution: Split the payload on newlines into a proper list of lines.\nA trailing newline yields a final empty item, which chansend() turns\nback into a newline, so payloads round-trip exactly.\n\n(cherry picked from commit e58f29ca3e6935af5b14879db1e8f2b959555c10)", "style": "conventional", "type": "fix", "scope": "terminal", "breaking": false} {"repo": "neovim", "sha": "e58f29ca3e6935af5b14879db1e8f2b959555c10", "date": "2026-08-05T05:37:02-05:00", "subject": "fix(terminal): OSC 52 multiline copy replaces newlines with NUL #41097", "body": "Problem: An OSC 52 sequence from a :terminal job passes the decoded\npayload to the clipboard provider as a single list item. Command-line\nproviders (pbcopy, xclip, ...) receive it with channel semantics, where\na newline inside an item is sent as NUL (:h chansend()), so multiline\ncopies arrive with NUL bytes instead of newlines.\n\nSolution: Split the payload on newlines into a proper list of lines.\nA trailing newline yields a final empty item, which chansend() turns\nback into a newline, so payloads round-trip exactly.", "style": "conventional", "type": "fix", "scope": "terminal", "breaking": false} {"repo": "curl", "sha": "7babac86904c7ac0b34a22240737c7261cb059eb", "date": "2026-08-05T12:42:15+02:00", "subject": "servers: fix HANDLE leak in UWP builds", "body": "Also: shorten code.\n\nReported by Copilot\nBug: https://github.com/curl/curl/pull/22487#pullrequestreview-4863399903\n\nCloses #22489", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "52e2e2b0c05e556fbd6f24b3b3102a29675f3f33", "date": "2026-08-05T11:54:22+01:00", "subject": "C=1 is parsed and stored as an image-placement flag, suppressing cursor movement and scrolling as required by the Kitty protocol.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "14aa6af4a5f1cde7bf121c2fb0636bf1d0c63897", "date": "2026-08-05T16:01:37+05:00", "subject": "docs: fix misleading `toHaveResolved` JSDoc examples", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "aba0f9b7b9682eb5205e19d725bbd4f535a97288", "date": "2026-08-05T16:02:50+05:00", "subject": "fix: swap replace() arguments so timeout error stack shows the real message", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vitest", "sha": "86d4a9da9d34603a21963d11d1b03ec9f5397854", "date": "2026-08-05T13:04:30+02:00", "subject": "feat!: fail the test when an asynchronous assertion is not awaited", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": true} {"repo": "vite", "sha": "d62b3360ecebdf11c23e99ffeb4b32e77c9a2ec8", "date": "2026-08-05T19:09:55+08:00", "subject": "chore: remove outdated `minimumReleaseAgeExclude` configuration", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "6e96c468d4078f451f889e73106252c9a72dcf8a", "date": "2026-08-05T13:40:18+02:00", "subject": "DEPRECATE.md: HTTP/2 Server Push gets removed in March 2027", "body": "URL: https://curl.se/mail/lib-2026-08/0003.html\nCloses #22490", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b260e9f841dc4e3c04da45e325ce23a11fec514d", "date": "2026-08-05T14:06:57+02:00", "subject": "servers: fix to avoid a non-signal-safe call in signal handler", "body": "`strlen()` is only guaranteed to be signal-safe since POSIX.1-2008.\n\nRef: https://pubs.opengroup.org/onlinepubs/009695399/functions/xsh_chap02_04.html#tag_02_04_03\n\nReported by Copilot\nBug: https://github.com/curl/curl/pull/22487#pullrequestreview-4863961494\nFollow-up to e95f509c66abdd88ae02e3243cdc217f19c4a330 #16852\n\nCloses #22491", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b5716286e9a6064ad202c653c0cc7fb8077418a1", "date": "2026-08-05T14:19:14+02:00", "subject": "tests: keep test names shorter than 70 columns", "body": "- makes test names less complicated\n\n- makes them less likely to wrap lines when using narrow terminals\n\n- runtests now returns error for the test if the name is longer\n\n- replace the \"...\" with a singe space\n\nCloses #22492", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "cc4a76d0be6b333d64812b0c4b1b1bf5cfb48544", "date": "2026-08-05T13:22:15+01:00", "subject": "Removed unconditional Floyd–Steinberg dithering from image-sixel.c:1077. Kitty images now use the adaptive 256-colour median-cut palette with direct nearest-colour mapping.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "4a9dbac43573fe4ba7e9798b159eb718ff8df2d7", "date": "2026-08-05T12:23:25Z", "subject": "Add a refresh-now command and change the r binding back to it in copy mode.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "947ae0e0ccec071e32e2a4f69834e03c209fa518", "date": "2026-08-05T14:40:35+02:00", "subject": "servers: drop redundant guards on Windows", "body": "These `SIG*` macro are provided by all supported Windows toolchains.\n\nCherry-picked from #22487\n\nCloses #22493", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "f81c41929bc9c9921b109bdd5539c0d0fc114d77", "date": "2026-08-05T13:51:07+01:00", "subject": "Fix kitty image resize issue when the kitty image resized smaller than the image, the image was overflowing lines and banding.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "abcb5349e3f38decdb483a22720057036ac77862", "date": "2026-08-05T14:56:40+02:00", "subject": "cf-socket: disable TCP SYN retransmissions for localhost on Windows", "body": "Suggested-by: Marcel Jamin\nURL: https://curl.se/mail/lib-2026-08/0002.html\nURL: https://daniel.haxx.se/blog/2024/08/14/slow-tcp-connect-on-windows/\n\nCloses #22494", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "09860921f0fac6ae3ac563a4121b0182fd89d9aa", "date": "2026-08-05T14:56:40+02:00", "subject": "cf-socket: disable TCP SYN retransmissions for localhost on Windows", "body": "Suggested-by: Marcel Jamin\nURL: https://curl.se/mail/lib-2026-08/0002.html\n\nCloses #22494", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "24071dd53d1eddada3543dc0eeb1eb80e14a0529", "date": "2026-08-05T14:59:42+02:00", "subject": "statem: add missing SSLfatal in compressed certificate construction", "body": "A WPACKET failure in tls_construct_server_compressed_certificate()\nreturned CON_FUNC_ERROR without calling SSLfatal(), unlike the\nidentical failure handling in tls_construct_server_certificate().\nThis trips the check_fatal assertion in write_state_machine() on\ndebug builds and skips the alert and error state otherwise.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Tue Aug 11 07:34:32 2026\n(Merged from https://github.com/openssl/openssl/pull/32197)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d44190def5df1d4d6721956bc22e9ef46173781f", "date": "2026-08-05T14:59:42+02:00", "subject": "statem: add missing SSLfatal in compressed certificate construction", "body": "A WPACKET failure in tls_construct_server_compressed_certificate()\nreturned CON_FUNC_ERROR without calling SSLfatal(), unlike the\nidentical failure handling in tls_construct_server_certificate().\nThis trips the check_fatal assertion in write_state_machine() on\ndebug builds and skips the alert and error state otherwise.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Tue Aug 11 07:34:19 2026\n(Merged from https://github.com/openssl/openssl/pull/32197)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "581983aa5fcfde80f6005ee33c522cae26d5e859", "date": "2026-08-05T14:59:42+02:00", "subject": "statem: add missing SSLfatal in compressed certificate construction", "body": "A WPACKET failure in tls_construct_server_compressed_certificate()\nreturned CON_FUNC_ERROR without calling SSLfatal(), unlike the\nidentical failure handling in tls_construct_server_certificate().\nThis trips the check_fatal assertion in write_state_machine() on\ndebug builds and skips the alert and error state otherwise.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Tue Aug 11 07:34:06 2026\n(Merged from https://github.com/openssl/openssl/pull/32197)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5419b70ba6f70f5f4a565fddfee2edad18924d51", "date": "2026-08-05T14:59:42+02:00", "subject": "statem: add missing SSLfatal in compressed certificate construction", "body": "A WPACKET failure in tls_construct_server_compressed_certificate()\nreturned CON_FUNC_ERROR without calling SSLfatal(), unlike the\nidentical failure handling in tls_construct_server_certificate().\nThis trips the check_fatal assertion in write_state_machine() on\ndebug builds and skips the alert and error state otherwise.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Tue Aug 11 07:33:56 2026\n(Merged from https://github.com/openssl/openssl/pull/32197)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "590e79c6f53625ac61e2622c493e767287321d30", "date": "2026-08-05T14:59:42+02:00", "subject": "statem: add missing SSLfatal in compressed certificate construction", "body": "A WPACKET failure in tls_construct_server_compressed_certificate()\nreturned CON_FUNC_ERROR without calling SSLfatal(), unlike the\nidentical failure handling in tls_construct_server_certificate().\nThis trips the check_fatal assertion in write_state_machine() on\ndebug builds and skips the alert and error state otherwise.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Tue Aug 11 07:33:44 2026\n(Merged from https://github.com/openssl/openssl/pull/32197)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0f6dc3fbaf7646bf46c5a0f75a01c837f4f8c71d", "date": "2026-08-05T15:03:18+02:00", "subject": "ssl: restore missing SSLfatal in keylog output", "body": "Commit e077455e9e dropped the SSLfatal() call together with the\nERR_R_MALLOC_FAILURE cleanup in nss_keylog_int(). Since then an\nallocation failure there propagates a plain zero return through\nssl_log_secret() and ssl_log_rsa_client_key_exchange() callers whose\n\"SSLfatal() already called\" comments no longer hold, tripping the\ncheck_fatal assertion in the state machine on debug builds when a\nkeylog callback is set.\n\nCommit 7d78cd722b papered over one symptom of this by calling SSLfatal()\ndirectly at the early exporter secret call site in\ntls13_change_cipher_state(). With the call restored at the source that\nwould now invoke SSLfatal() twice, pushing a duplicate entry on the\nerror queue, so revert that call site back to the\n\"SSLfatal() already called\" comment used by all the other callers.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nMerge-date: Thu Aug 27 13:54:32 2026\nMerged-from: https://github.com/openssl/openssl/pull/32196\n(cherry picked from commit dde995f2913bf825bbd0e9c25af4c75caf917a10)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dde995f2913bf825bbd0e9c25af4c75caf917a10", "date": "2026-08-05T15:03:18+02:00", "subject": "ssl: restore missing SSLfatal in keylog output", "body": "Commit e077455e9e dropped the SSLfatal() call together with the\nERR_R_MALLOC_FAILURE cleanup in nss_keylog_int(). Since then an\nallocation failure there propagates a plain zero return through\nssl_log_secret() and ssl_log_rsa_client_key_exchange() callers whose\n\"SSLfatal() already called\" comments no longer hold, tripping the\ncheck_fatal assertion in the state machine on debug builds when a\nkeylog callback is set.\n\nCommit 7d78cd722b papered over one symptom of this by calling SSLfatal()\ndirectly at the early exporter secret call site in\ntls13_change_cipher_state(). With the call restored at the source that\nwould now invoke SSLfatal() twice, pushing a duplicate entry on the\nerror queue, so revert that call site back to the\n\"SSLfatal() already called\" comment used by all the other callers.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nMerge-date: Thu Aug 27 13:54:32 2026\nMerged-from: https://github.com/openssl/openssl/pull/32196", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "aad518f2f9373a4ee61a395404e8865b7b4ff359", "date": "2026-08-05T15:04:02+02:00", "subject": "ssl: add missing SSLfatal in PSK premaster secret generation", "body": "An allocation failure for the PSK premaster secret buffer in\nssl_generate_master_secret() returned failure without calling\nSSLfatal(), while its callers on both the client key exchange\nconstruction and processing paths assume the fatal state was\nalready set. Also cover the unreachable no-PSK fallback branch.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Matt Caswell \nMergeDate: Tue Aug 11 07:29:11 2026\n(Merged from https://github.com/openssl/openssl/pull/32195)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d43a0ec947354700bf604acf6f2381378199c7e0", "date": "2026-08-05T15:04:02+02:00", "subject": "ssl: add missing SSLfatal in PSK premaster secret generation", "body": "An allocation failure for the PSK premaster secret buffer in\nssl_generate_master_secret() returned failure without calling\nSSLfatal(), while its callers on both the client key exchange\nconstruction and processing paths assume the fatal state was\nalready set. Also cover the unreachable no-PSK fallback branch.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Matt Caswell \nMergeDate: Tue Aug 11 07:28:59 2026\n(Merged from https://github.com/openssl/openssl/pull/32195)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f3b94dc87b18ab1371e2b1abf87be6a05ef2f34c", "date": "2026-08-05T15:04:02+02:00", "subject": "ssl: add missing SSLfatal in PSK premaster secret generation", "body": "An allocation failure for the PSK premaster secret buffer in\nssl_generate_master_secret() returned failure without calling\nSSLfatal(), while its callers on both the client key exchange\nconstruction and processing paths assume the fatal state was\nalready set. Also cover the unreachable no-PSK fallback branch.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Matt Caswell \nMergeDate: Tue Aug 11 07:28:48 2026\n(Merged from https://github.com/openssl/openssl/pull/32195)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5c1a240787586e477a01d14a833be2f242857928", "date": "2026-08-05T15:04:02+02:00", "subject": "ssl: add missing SSLfatal in PSK premaster secret generation", "body": "An allocation failure for the PSK premaster secret buffer in\nssl_generate_master_secret() returned failure without calling\nSSLfatal(), while its callers on both the client key exchange\nconstruction and processing paths assume the fatal state was\nalready set. Also cover the unreachable no-PSK fallback branch.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Matt Caswell \nMergeDate: Tue Aug 11 07:28:37 2026\n(Merged from https://github.com/openssl/openssl/pull/32195)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "51b82ebb0c3c1153d59cd57b9c18800240acb645", "date": "2026-08-05T15:04:02+02:00", "subject": "ssl: add missing SSLfatal in PSK premaster secret generation", "body": "An allocation failure for the PSK premaster secret buffer in\nssl_generate_master_secret() returned failure without calling\nSSLfatal(), while its callers on both the client key exchange\nconstruction and processing paths assume the fatal state was\nalready set. Also cover the unreachable no-PSK fallback branch.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Matt Caswell \nMergeDate: Tue Aug 11 07:28:26 2026\n(Merged from https://github.com/openssl/openssl/pull/32195)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ba627dbcde3037dd93660a676d0b98e40a787720", "date": "2026-08-05T15:04:02+02:00", "subject": "ssl: add missing SSLfatal in PSK premaster secret generation", "body": "An allocation failure for the PSK premaster secret buffer in\nssl_generate_master_secret() returned failure without calling\nSSLfatal(), while its callers on both the client key exchange\nconstruction and processing paths assume the fatal state was\nalready set. Also cover the unreachable no-PSK fallback branch.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Matt Caswell \nMergeDate: Tue Aug 11 07:28:05 2026\n(Merged from https://github.com/openssl/openssl/pull/32195)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6cfad0a94827754048ac9647b9ba1031ad8fcf36", "date": "2026-08-05T15:04:30+02:00", "subject": "ssl: add missing SSLfatal on session ID lock failure", "body": "The first CRYPTO_THREAD_read_lock() failure in\nssl_generate_session_id() returned failure without calling SSLfatal()\nwhile the second lock right below does, and callers such as\ntls_construct_new_session_ticket() assume the fatal state was\nalready set. This should currently not happen as the read lock does\nnot realistically fail, so this is just future proofing of the error\nhandling.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Tue Aug 11 07:31:08 2026\n(Merged from https://github.com/openssl/openssl/pull/32200)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "677102073341cdd75c9c4b6e6ed146025dba7a19", "date": "2026-08-05T15:05:35+02:00", "subject": "ssl: add missing SSLfatal in TLSv1.3 finished MAC computation", "body": "tls13_final_finish_mac() returned failure on a NULL handshake digest\nwithout calling SSLfatal() or raising any error, while its callers on\nthe Finished construction and processing paths assume the fatal state\nwas already set. This should currently not happen as the handshake\ndigest must have been fetched successfully earlier in the handshake,\nso this is just future proofing of the error handling.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Tue Aug 11 07:49:36 2026\n(Merged from https://github.com/openssl/openssl/pull/32198)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d3961e8a44c8ad873797bc5632f2ab55cad4b608", "date": "2026-08-05T15:06:10+02:00", "subject": "statem: add missing SSLfatal in padding extension construction", "body": "tls_construct_ctos_padding() returned EXT_RETURN_FAIL on a\nnon-positive digest size without calling SSLfatal(), propagating to\nCON_FUNC_ERROR in the ClientHello construction without the fatal\nstate set. This should currently not happen as ssl_md() only returns\ndigests with a positive size, so this is just future proofing of the\nerror handling.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Matt Caswell \nReviewed-by: Frederik Wedel-Heinen \nMergeDate: Tue Aug 11 08:45:51 2026\n(Merged from https://github.com/openssl/openssl/pull/32199)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "68b315d3b88849a82d8c9a8e2941a7e1a5c13f2e", "date": "2026-08-05T15:38:17+02:00", "subject": "limit it to TCP only", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "9fb2f0b69773b13960b2149dec7d1139628ef305", "date": "2026-08-05T15:38:48+02:00", "subject": "(void) the call", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "7dd563bd9bd28d872a22b705dbe21074de919d15", "date": "2026-08-05T15:40:10+02:00", "subject": "'TCP_INITIAL_RTO_NO_SYN_RETRANSMISSIONS': undeclared", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b11e0026f71aa33b47d3b31f7c6a2aee4f929b12", "date": "2026-08-05T15:59:30+02:00", "subject": "GHA/windows: re-enable taskkill in torture jobs", "body": "Torture jobs are arguably the most flaky nowadays. Make a blind try to\nsee if re-enabling taskkill makes an observable improvement for torture.\n\nFollow-up to 208b87744e8ce326e2569c2efd735da43ce74a0b #21039\nFollow-up to f450f3801b6b9dff0ea280f5fb4bf28203f7b313 #19897\nFollow-up to 2701ac6a4d16a62130dad05be1c484903b8545c7 #19421\n\nCloses #22495", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "557ba5e8d3c41a0849ce0bfa24630956e4be27ae", "date": "2026-08-05T16:04:54+02:00", "subject": "statem: fail ClientHello construction on ECH GREASE failure", "body": "tls_construct_ctos_ech() called SSLfatal() on an ECH GREASE\nconstruction failure but returned EXT_RETURN_NOT_SENT, so extension\nconstruction reported success while the connection was already in the\nfatal error state. The ClientHello would still be sent together with\nthe fatal alert and the state machine kept running until it tripped\nover the error state later. Return EXT_RETURN_FAIL to fail the\nconstruction immediately.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Matt Caswell \nReviewed-by: Tomas Mraz \nMerge-date: Wed Sep 16 09:32:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32201", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1ceffb083f1fb5da8f7868547273416cd60bcc6c", "date": "2026-08-05T16:04:54+02:00", "subject": "statem: fail ClientHello construction on ECH GREASE failure", "body": "tls_construct_ctos_ech() called SSLfatal() on an ECH GREASE\nconstruction failure but returned EXT_RETURN_NOT_SENT, so extension\nconstruction reported success while the connection was already in the\nfatal error state. The ClientHello would still be sent together with\nthe fatal alert and the state machine kept running until it tripped\nover the error state later. Return EXT_RETURN_FAIL to fail the\nconstruction immediately.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Matt Caswell \nReviewed-by: Tomas Mraz \nMerge-date: Wed Sep 16 09:32:01 2026\nMerged-from: https://github.com/openssl/openssl/pull/32201", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "bcf6c886dd5336d972444c9c2820cf4d293a0886", "date": "2026-08-05T16:04:54+02:00", "subject": "statem: fail ClientHello construction on ECH GREASE failure", "body": "tls_construct_ctos_ech() called SSLfatal() on an ECH GREASE\nconstruction failure but returned EXT_RETURN_NOT_SENT, so extension\nconstruction reported success while the connection was already in the\nfatal error state. The ClientHello would still be sent together with\nthe fatal alert and the state machine kept running until it tripped\nover the error state later. Return EXT_RETURN_FAIL to fail the\nconstruction immediately.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Matt Caswell \nReviewed-by: Tomas Mraz \nMerge-date: Wed Sep 16 09:31:48 2026\nMerged-from: https://github.com/openssl/openssl/pull/32201", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "76dff307b4c34bc4ceefcb1c7aa86c39daa5ea45", "date": "2026-08-05T16:57:14+02:00", "subject": "fix(compiler): Generate correct expression for optional chaning.", "body": "Optional chaining was generating expressions with included an extra pair of parenthesis which changed the semantics of the expression and threw an unexpected error from the optional chain non nullable extended diagnostic.\n\nfixes #70085\n\n(cherry picked from commit d7b03f5523d220c7935426d23ff5f103c96b00d9)", "style": "conventional", "type": "fix", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "d7b03f5523d220c7935426d23ff5f103c96b00d9", "date": "2026-08-05T16:57:14+02:00", "subject": "fix(compiler): Generate correct expression for optional chaning.", "body": "Optional chaining was generating expressions with included an extra pair of parenthesis which changed the semantics of the expression and threw an unexpected error from the optional chain non nullable extended diagnostic.\n\nfixes #70085", "style": "conventional", "type": "fix", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "f51080b07bc083dd2f41ce72ca9ea08ec1a1ee60", "date": "2026-08-05T10:15:40-05:00", "subject": "refactor(platform-browser): modernize DOM renderer tests", "body": "Migrate fixtures to standalone components and await zoneless updates\n\n(cherry picked from commit ec372d60294ff9b6137bf253201ce8341bd80f42)", "style": "conventional", "type": "refactor", "scope": "platform-browser", "breaking": false} {"repo": "angular", "sha": "ec372d60294ff9b6137bf253201ce8341bd80f42", "date": "2026-08-05T10:15:40-05:00", "subject": "refactor(platform-browser): modernize DOM renderer tests", "body": "Migrate fixtures to standalone components and await zoneless updates", "style": "conventional", "type": "refactor", "scope": "platform-browser", "breaking": false} {"repo": "openssl", "sha": "81e0cf8ea3d2c29073291ead904203634778b32a", "date": "2026-08-05T09:39:37-06:00", "subject": "Don't read an ASN1_STRING data raw and assume it's 0 byte terminated", "body": "Fix latent app bug in policy error diagnostics.\n\nFixes: #32176\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:39 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ed8050370b7cf117dfb0de218695b1e003a51f36", "date": "2026-08-05T11:44:01-04:00", "subject": "Make local buffers pin limit more conservative", "body": "GetLocalPinLimit() and GetAdditionalLocalPinLimit(), currently in use\nonly by the read stream, previously allowed a backend to pin all\nnum_temp_buffers local buffers. This meant that the read stream could\nuse every available local buffer for read-ahead, leaving none for other\nconcurrent pin-holders.\n\nIn 18, this was reported as a sequential scan of a temporary table\nfailing with \"no empty local buffer available\": with\neffective_io_concurrency >= 64 and default io_combine_limit and\ntemp_buffers, the scan's read stream budget,\n(effective_io_concurrency + 1) * io_combine_limit, meets or exceeds\nnum_temp_buffers, so look-ahead could pin the entire local buffer pool.\nAny additional pin request made while returning tuples -- such as\nfetching TOASTed values from the table's (also temporary) TOAST table --\nthen found no unpinned buffer to evict.\n\nCap the local pin limit to num_temp_buffers / 4, providing some\nheadroom. This doesn't guarantee that all needed pins will be available\n-- for example, a backend can still open more cursors than there are\nbuffers -- but it makes it less likely that read-ahead will exhaust the\npool.\n\nThis is a backpatch of commit da6874635db, which was applied to master\nonly during the 19 development cycle, where the issue surfaced as a\nregression test failure after on-access pruning began setting the\nvisibility map. Reports of the failure above on 18 prompted\nbackpatching it now.\n\nNo backpatch to 17, even though its read stream also allows a single\nstream to pin the whole pool in principle. There, sequential scans pass\nREAD_STREAM_SEQUENTIAL, which disables fadvise-based look-ahead, so the\nlook-ahead distance cannot grow past io_combine_limit (at most 32\nbuffers, versus a minimum temp_buffers of 100) and the reported failure\nis unreachable. The remaining theoretical paths require adversarial\nsettings and have never been reported. Moreover, 17 predates these\nfunctions; the fix would have to be reimplemented in\nLimitAdditionalLocalPins(), making the ", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "52a9a4f24750ef15a23343729acad71942c5bc55", "date": "2026-08-05T09:54:47-06:00", "subject": "Don't call strlcpy on a non-string.", "body": "Again, ASN1_STRING_get0_data has never been guaranteed to be nul terminated.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:40 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7acab0ef09e15c56d9159f6bc326285fc0b0e2f3", "date": "2026-08-05T10:19:32-06:00", "subject": "Fix the tests to not pretend ASN1_STRING's are nul terminated.", "body": "Reviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "5eb3570e83292b08c4ab23a0709e3120615b0fb8", "date": "2026-08-05T12:48:02-04:00", "subject": "fix(cache): improve `fsModuleCache` resiliency during external modifications", "body": "", "style": "conventional", "type": "fix", "scope": "cache", "breaking": false} {"repo": "openssl", "sha": "f3ad4106b8b81a18616131fe09f8332b38e8bfd3", "date": "2026-08-05T10:48:04-06:00", "subject": "Make ASN1_STRING_to_UTF8() return a NUL terminated buffer", "body": "While it's really pretty foolish to do so (since it can\ncontain legitimate 0 bytes in the output) Some callers were\ntreating the returned buffer as a C string, in spite of the fact\nthat the documentation does not say that it is and a length\nis returned.\n\nUnfortunately, this probably stems from the bad habit in\nOpenSSL of trying to make things C strings anyway, - The conversion\npath already allocated the extra byte and terminated it as if it\nis a usable C string. The same-format path stored exact-length data\nfrom an ASN1_STRING with no NUL, so termination depended on the input\ntype and how it was constructed.\n\nSo be consistent, and NUL terminate the returned buffer in both cases.\n\nWe then document that the retured buffer *IS* nul terminated, add\nunit tests because of the Beyonce rule, and document in the\nwarnings that in spite of the output buffer being nul terminated\nit is potentially a bad idea to use C string functions on it as\nit can contain internal NUL bytes.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:42 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "9f23f8ec3413beff8072b837f0ed09329661dc5d", "date": "2026-08-05T12:49:26-04:00", "subject": "perf: lowers peak memory usage when using `--changed` on a large graph", "body": "", "style": "conventional", "type": "perf", "scope": "", "breaking": false} {"repo": "vitest", "sha": "8e2108ddaec3c58501621fdd2d78929d87c383c8", "date": "2026-08-05T12:56:26-04:00", "subject": "fix: stale mock metadata breaks automocking with isolate:false (fix #10145)", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "133785b15936d9417fdc41f77b87304b6bf458d5", "date": "2026-08-05T19:10:52+02:00", "subject": "servers: drop re-registering the signal handler on modern systems", "body": "Before this patch modern systems used `sigaction()` and `SA_RESTART` to\ninstall signal handlers, but the signal handler function itself still\nmade a call to the legacy `signal()` function to re-register itself\nbefore returning.\n\nRe-registering the handler is not necessary with `sigaction()`. It's\nalso undesired to use the legacy API when the modern one is available.\n\nFix by guarding off this call in builds that support the modern API.\n\nFollow-up to 3fb6e5a01001b8c7dfdc33a89041178aac381a27 #6529\nFollow-up to 18cbb4d7d645016d1a9c0ee772c724d8c41db3bd\n\nCloses #22497", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "5adcd7a05437bcd00a682ec161025b1d6dbbf0b4", "date": "2026-08-05T19:15:00+02:00", "subject": "chore: apply function decorator to middleware functions", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c9f0103487d39358a6aae76df7a1f4a38212b8ba", "date": "2026-08-05T11:21:19-06:00", "subject": "Fix one byte overrun in UNIVERSALSTRING_to_string", "body": "This function assumed there was one byte more than the\nlength of the string to add NUL byte termination. Don't\ndo that.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:43 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0d8e41fe1c80d1310e0eed0f66af82f60e506c69", "date": "2026-08-05T19:26:30+02:00", "subject": "Fix MSVC warnings from new timezone code", "body": "The new timezone code from commit aeb07c55fab introduced a few new\ncompiler warnings from MSVC:\n\n ../src/timezone/zic.c(353): warning C5287: operands are different enum types '' and ''; use an explicit cast to silence this warning\n ../src/timezone/zic.c(353): warning C5287: operands are different enum types '' and ''; use an explicit cast to silence this warning\n ../src/timezone/zic.c(1654): warning C4146: unary minus operator applied to unsigned type, result still unsigned\n\nSilence these warnings with pragmas.\n\nIn order not to make future code merges more complicated, the pragmas\napply to the entire file, not only to the specific code sections where\nthey are needed. This also prevents further warnings of the same\nkinds creeping in.\n\nReviewed-by: Tom Lane \nDiscussion: https://www.postgresql.org/message-id/flat/2294297.1780270682%40sss.pgh.pa.us", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "b147e64e0c10b4f5589b56c96718880288e5e034", "date": "2026-08-05T19:28:08+02:00", "subject": "chore: narrow type down to a random tuple", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e2c0c634522b7742ef0fb0018bca8b373cf35417", "date": "2026-08-05T23:05:09+05:30", "subject": "fix(statusline): don't clobber global statusline during autocmd #41169", "body": "Problem:\nWith `laststatus=3`, a pager float shares the main grid's statusline\nrow. Setting a diagnostic fires `DiagnosticChanged`, whose handler calls\n`nvim__redraw({ statusline = true })`.\n\nAnalysis:\nInside the autocmd, `curwin` is temporarily switched to the tiled window\nshowing that buffer, so `win_redr_status()` paints its statusline over\nthe pager's `[Pager]` statusline on the shared row. After the autocmd,\n`curwin` is restored but the pager statusline is never repainted.\n\nSolution:\nUse `ctx_saved_curwin()` decide whether to draw the global statusline,\nmatching `win_redr_stl_expr()` and `update_screen()`. No behavior change\nif no buffer-context switch is active.", "style": "conventional", "type": "fix", "scope": "statusline", "breaking": false} {"repo": "neovim", "sha": "9127ed41a5bf4155ae8276a143acf6c0b6568ca3", "date": "2026-08-05T12:36:58-05:00", "subject": "fix(dir): preserve alternate file #41177", "body": "", "style": "conventional", "type": "fix", "scope": "dir", "breaking": false} {"repo": "neovim", "sha": "69664a0aca6981cb93b1a1f01c44e70ad530f57d", "date": "2026-08-05T12:37:24-05:00", "subject": "fix(zip): preserve alternate file #41178", "body": "", "style": "conventional", "type": "fix", "scope": "zip", "breaking": false} {"repo": "openssl", "sha": "62570ca6e93af540907330d50ee581fe82cd18fd", "date": "2026-08-05T11:39:56-06:00", "subject": "Don't strncpy() counted ASN1_STRING data in ossl_sk_ASN1_UTF8STRING2text()", "body": "The element data is counted, not a C string, so strncpy() was wrong two\nways: an empty element yields strncpy(p, NULL, 0), which is undefined, and a\nvalue containing an embedded NUL was truncated and zero padded instead of\ncopied verbatim.\n\nUse memcpy() bounded by the element length, skipping empty elements. The\nseparator copy becomes a plain memcpy() as well, dropping the \"+ 1 to\nsilence gcc\" strncpy() hack. The computed buffer length and terminator are\nunchanged.\n\nNote in the documentation that using utf8 values in C strings is a\nfootgun if they contain NUL bytes internally\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:44 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9a93ac6533e6ce68b0bc66af999408bc74662891", "date": "2026-08-05T19:49:08+02:00", "subject": "refactor(vim.fs): slug() minor cleanup", "body": "", "style": "conventional", "type": "refactor", "scope": "vim.fs", "breaking": false} {"repo": "angular", "sha": "72e766456ef9c9a1d4f64919b634a3b8f2aa96e8", "date": "2026-08-05T13:01:54-05:00", "subject": "refactor(forms): modernize Signal Forms tests", "body": "replace change detection calls with whenStable for async tests\n\n(cherry picked from commit 4631a6e47923dc72cba4fe5e6075873ccefe8012)", "style": "conventional", "type": "refactor", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "4631a6e47923dc72cba4fe5e6075873ccefe8012", "date": "2026-08-05T13:01:54-05:00", "subject": "refactor(forms): modernize Signal Forms tests", "body": "replace change detection calls with whenStable for async tests", "style": "conventional", "type": "refactor", "scope": "forms", "breaking": false} {"repo": "neovim", "sha": "23525dd4e3d7daaf619ab5a484ad45693a885975", "date": "2026-08-05T20:06:28+02:00", "subject": "fix(editorconfig): avoid trim_trailing_whitespace in insert-mode #41175", "body": "Problem:\nDuring insert-mode / replace-mode, `autowrite` may trigger. If it does, the\ncursor position can shift due to the automatic removal of trailing spaces on the\ncurrent line. When I resume typing, the space between the last word and the new\nword is suddenly gone.\n\nSolution:\nDisable the \"remove trailing spaces\" handler during Insert (or a similar) mode.\nAutosave logic is not affected.", "style": "conventional", "type": "fix", "scope": "editorconfig", "breaking": false} {"repo": "neovim", "sha": "a29a9130a9828f32cc3b57bca39c25214118ddbf", "date": "2026-08-06T02:07:59+08:00", "subject": "fix(ui2): do not steal focus when consecutive cmds emit messages #41062", "body": "Problem: A message emitted while a previous expanded message is still\n visible opens the pager and enters it, moving focus away from\n the buffer window without an explicit request (#41061).\nSolution: Only enter the pager when it was explicitly requested (\"g<\",\n :messages, or entered from the expanded cmdline). An unfocused\n pager is dismissed by the cmdline key handler, which stays armed\n across the cmdline and no longer dismisses on non-typed keys\n (#39221).", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "eslint", "sha": "b7208f6ed848d350a8bf2ed08c28ab1e388fc70d", "date": "2026-08-05T23:51:45+05:30", "subject": "Update lib/rule-tester/rule-tester.js", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "a263b0bb0e8a10231fa94f90a8e2665dbfbe9deb", "date": "2026-08-05T23:52:07+05:30", "subject": "Update docs/src/integrate/nodejs-api.md", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "6d04cd2b84496d727344be1acb6287452982c9bd", "date": "2026-08-05T11:36:37-07:00", "subject": "Remove Subscription conninfo field; generate in caller.", "body": "After server-based subscriptions, conninfo became more than just a\ncatalog field. It has its own error paths, and it's important that\ncallers that don't need conninfo don't encounter errors related to it.\n\nReviewed-by: Amit Kapila \nReviewed-by: Hayato Kuroda (Fujitsu) \nDiscussion: https://postgr.es/m/20260710195902.4f.noahmisch%40microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "bbba199848c0acee5df421c4a49a8124f121a395", "date": "2026-08-06T00:13:36+05:30", "subject": "fix: address feedback", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "eslint", "sha": "7364a5179adc221ab112755dae1611339ef60f71", "date": "2026-08-06T00:20:53+05:30", "subject": "Apply suggestion from @snitin315", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f567e066d6e3d09b1db740d0a891b9464c64caa2", "date": "2026-08-05T11:57:36-07:00", "subject": "Build subscription conninfo after checking that it's enabled.", "body": "If a subscription is disabled, don't try to build conninfo because\nthat may generate a confusing error and try to disable an\nalready-disabled subscription.\n\nPartially addresses finding 5 in report from linked discussion.\n\nReported-by: Noah Misch \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/20260710195902.4f.noahmisch%40microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b8424c0d848ac2526fc3c7b9a325be000b119fd4", "date": "2026-08-05T13:02:57-06:00", "subject": "Fix unguarded memcpy's asn1_type_get_int_oct()and X509_NAME_get_text_by_OBJ()", "body": "Reviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:45 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "e872ad54f96af9d9a917c8a6fef5df84235fe8a7", "date": "2026-08-06T00:35:29+05:30", "subject": "fix: update tests", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "neovim", "sha": "ab80ea92cc2b1f0a1e706fa09454b3270b4ca244", "date": "2026-08-05T15:07:19-04:00", "subject": "fix(ui2): pager handling #41179", "body": "- Avoid shared state. Pass `focus` to set_pos()/expand_msg() instead of\n a shared `pager_focus` flag: the flag is only cleared when set_pos()\n actually enters the pager, so \":messages\" from inside the pager left\n it set.\n- pager_shown(): the pager window is invalid after leaving it with \"q\".\n- Reuse pager_shown() in expand_msg().", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "openssl", "sha": "b5570727c8062fe996b0d0afb5a27806a3c23d56", "date": "2026-08-05T13:30:19-06:00", "subject": "Add OPENSSL_NONSTRING and apply it to ASN1_STRING's data field", "body": "OPENSSL_NONSTRING expands to __attribute__((nonstring)) where the compiler\nsupports it and to nothing otherwise. ASN1_STRING data has an explicit\nlength and is not necessarily NUL terminated, so marking the field lets GCC\ndiagnose direct use of it with C string functions such as strlen(), strcpy()\nor \"%s\".\n\nThe diagnostic fires only on GCC under an optimised build; other compilers,\nincluding current Clang, accept the attribute but do not warn.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:51 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "bb7f05163193e70a28dd0a40a4ba62135f3e2527", "date": "2026-08-05T21:32:59+02:00", "subject": "refactor(compiler): remove explicit `strict: true`", "body": "This flag is set by default in TS 6.0\n\n(cherry picked from commit f1c0c405c9d191fc42beec843c43d54392e53498)", "style": "conventional", "type": "refactor", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "f1c0c405c9d191fc42beec843c43d54392e53498", "date": "2026-08-05T21:32:59+02:00", "subject": "refactor(compiler): remove explicit `strict: true`", "body": "This flag is set by default in TS 6.0", "style": "conventional", "type": "refactor", "scope": "compiler", "breaking": false} {"repo": "postgres", "sha": "ae420d66e72051b7ceab49d14f990094f1c8c6b4", "date": "2026-08-05T13:11:04-07:00", "subject": "Be precise about when ALTER SUBSCRIPTION needs conninfo.", "body": "Decide early whether the original conninfo is needed so that errors\nhappen consistently.\n\nAddresses finding 12 in report from linked discussion.\n\nCo-authored-by: Shlok Kyal \nReported-by: Noah Misch \nReviewed-by: Hayato Kuroda (Fujitsu) \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/20260710195902.4f.noahmisch%40microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "6107629c5b0a00b357f377e3ddc1cad34972ad49", "date": "2026-08-05T15:11:13-05:00", "subject": "feat(fs): vim.fs.normalize{plain:boolean} #41127", "body": "`opts.plain=true` does not expand tildes in addition to environment\nvariables, unlike `opts.expand_env=false`.\n\n`opts.expand_env=false` is soft-deprecated.", "style": "conventional", "type": "feat", "scope": "fs", "breaking": false} {"repo": "postgres", "sha": "5ba18183c4ca34beeedf7023d1f260ba1f5e9f87", "date": "2026-08-05T13:18:34-07:00", "subject": "Always check foreign-server USAGE when resolving subscription conninfo.", "body": "Previously, this was skipped in some cases to avoid raising errors\nwhen conninfo wasn't even needed. That was wrong in cases where\nconninfo was needed.\n\nNow that we only build conninfo when needed, always perform the USAGE\ncheck.\n\nAddresses finding 7 in report from linked discussion.\n\nCo-authored-by: Shlok Kyal \nReported-by: Noah Misch \nReviewed-by: Shlok Kyal \nReviewed-by: Hayato Kuroda (Fujitsu) \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/20260710195902.4f.noahmisch%40microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "caadb9e97f571e19f359a42ad186df877ea834c7", "date": "2026-08-05T13:18:44-07:00", "subject": "For subscription DDL, demote user mapping checks to WARNING.", "body": "The checks are useful to report to the user, but there's no reason to\nraise an error. If needed while constructing conninfo, fdwconnection\nwill raise an error then.\n\nPartially addresses finding 1, and addresses finding 13 in report from\nthe linked discussion.\n\nReported-by: Noah Misch \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/20260710195902.4f.noahmisch%40microsoft.com\nDiscussion: https://postgr.es/m/e103ae8daf74485e0c0ebde297fae735d38f54d1.camel@j-davis.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1718a16433a3f6572e0ff9c64ef394ae8fec3765", "date": "2026-08-05T13:19:02-07:00", "subject": "CREATE SUBSCRIPTION: do not construct conninfo unnecessarily.", "body": "Still check that the creating user has USAGE privileges on the server,\nand that the FDW supports subscription connections.\n\nAddresses finding 1 in the report from the linked discussion.\n\nReported-by: Noah Misch \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/20260710195902.4f.noahmisch%40microsoft.com\nDiscussion: https://postgr.es/m/e103ae8daf74485e0c0ebde297fae735d38f54d1.camel@j-davis.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a1e39c501d7a89a01c2738fecf7b66927e4e8bb7", "date": "2026-08-05T13:19:21-07:00", "subject": "When changing owner of a subscription, do not throw an error.", "body": "Errors will be caught when the connection is actually used.\n\nRestore uses multiple DDL commands to restore a subscription, so\nchecks of the intermediate state risk restore errors. In the future we\ncould address this with a more careful restoration order, but the\nDDL-time errors are merely for convenience.\n\nAddresses finding 2 in the report from the linked discussion.\n\nReported-by: Noah Misch \nReviewed-by: Shlok Kyal \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/20260710195902.4f.noahmisch%40microsoft.com\nDiscussion: https://postgr.es/m/e103ae8daf74485e0c0ebde297fae735d38f54d1.camel@j-davis.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "efbe9f0317b11d11f09af4034e88205d1d5edd85", "date": "2026-08-05T22:22:22+02:00", "subject": "Apply suggestions from code review", "body": "Viktor's proposals", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "f5664beaf333238a34a25f6d06d8b128688ba55c", "date": "2026-08-06T01:54:10+05:30", "subject": "test: add more cases", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b844df54f1f8147ac7fe90ac83613a55c5656682", "date": "2026-08-05T14:24:58-06:00", "subject": "Don't assume an ASN1_STRING is nul byte terminated in the test", "body": "Reviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:46 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "73b35687f6f3f34fa1e0c7c93a92011968033212", "date": "2026-08-05T22:25:36+02:00", "subject": "pfff, DOS newlines!?", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c2bccf040e772a4b4db52489f99839f7c4e01426", "date": "2026-08-05T14:29:19-06:00", "subject": "Don't hit the legacy strlen case in ASN1_STRING_copy", "body": "This should avoid the warning for calling strlen() on data in gcc\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:47 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "951056d0f8ce2b0fb47daec84bf4ccd3af15cd99", "date": "2026-08-06T02:02:33+05:30", "subject": "fix: report on enum redeclarations", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "eslint", "sha": "c63d1e0b800f61a75e8b84a90212ca751cb39799", "date": "2026-08-06T02:09:27+05:30", "subject": "fix: not report declarations that use the declare keyword", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "tmux", "sha": "8b44703b41d88ea626b91c3c42e065ae207be08f", "date": "2026-08-05T21:44:25+01:00", "subject": "Fix SIXEL median-cut palette generation", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "8e14c0ae504ffb917ad375f0f63dc0ea27ad514b", "date": "2026-08-05T17:22:10-04:00", "subject": "build(vim-patch): mch_create_pty_channel() is n/a", "body": "Target v8.2.1091.\nFrom v8.0.1000 for 'terminal'.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "curl", "sha": "26b9f3aa9b2531d5da74ce5472bc8f59ea01bd6d", "date": "2026-08-05T23:40:18+02:00", "subject": "rtsp: refactor method handling and improve error checks", "body": "- convert the method switch() to a simple table\n\n- avoid converting the methods from external to internal numbers, they were\n the same anyway so keep the external ones, just use the old defines.\n\n- fix range check. It wrongly used the method numbers as bitmask, which made\n the check not work previously. Also error on OOM.\n\n- Dropped the session-id check. It too wrongly did a bitmask check which was\n wrong and never worked. When fixed, it broke test cases so I dropped the\n entire check.\n\n- split out rtsp_setup_request() from rtsp_do()\n\n- replace the httpversion variable with a define\n\nCloses #22505", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "4ea4c80ae63d0514575ba0f1ea8e47dd0cc44cf4", "date": "2026-08-05T17:52:53-04:00", "subject": "build(vim-patch): python2 test is n/a", "body": "Target v8.2.1120 .\nRemoved since https://github.com/neovim/neovim/pull/3074.\nNvim removed EOL versions in\nhttps://github.com/neovim/neovim/pull/17222 .", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "4f183765aab81df08aa53baf483b26d02ad677bb", "date": "2026-08-05T15:54:01-06:00", "subject": "Sweep for printf UB - don't printf %s NULL", "body": "As an asn1_string of length 0 will have NULL data, don't\nprintf %s NULL.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:48 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "708add2d9fb439f110fd73f943d8a64bbdf10cab", "date": "2026-08-05T17:56:02-04:00", "subject": "Fix DTLS 1.3 early data transcript hash stripping", "body": "When the client installs early write keys, negotiated_version is not\nyet set to DTLS1_3_VERSION, so the RFC 9147 §5.2 stripping of\nmsg_seq, fragment_offset, and fragment_length was skipped in\ntls13_change_cipher_state. Fix by conditioning on\nSSL_CONNECTION_IS_DTLS() alone.\n\nFixes: openssl/project#2043\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Matt Caswell \nMerge-date: Mon Aug 17 08:34:23 2026\nMerged-from: https://github.com/openssl/openssl/pull/32276", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "430aadec30167839410b9887862b077646cfeab0", "date": "2026-08-05T16:11:00-06:00", "subject": "Fix the kerberos external test patch", "body": "Make the patch we use here length safe.\n\n(the original kerb code is not)\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:49 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "d674f34f253583a6fa91579993d6f41738cc5006", "date": "2026-08-05T23:26:11+01:00", "subject": "Support Kitty Unicode placeholder image transmission", "body": "- Parse Kitty graphics commands wrapped in tmux passthrough sequences.\n - Accept unpadded base64 payloads emitted by icat.\n - Handle chunked transfers whose final chunk omits m.\n - Track virtual image placements.\n - Convert Unicode placeholders into shared image cells for Kitty and SIXEL clients.\n - Add regression coverage for wrapped, chunked, unpadded input.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "8d23b9511918560b0713404a1d78879b7434b580", "date": "2026-08-06T00:26:40+02:00", "subject": "build: remove explicit strict options", "body": "Those options are enabled by default, they don't need to be explicit.\n\n(cherry picked from commit 9cf7b44f7c08082eaade725edf05cabe51f1a0b4)", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "9cf7b44f7c08082eaade725edf05cabe51f1a0b4", "date": "2026-08-06T00:26:40+02:00", "subject": "build: remove explicit strict options", "body": "Those options are enabled by default, they don't need to be explicit.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "tmux", "sha": "75b606c3df81a458051879d683fb32fe37ac008f", "date": "2026-08-06T00:06:59+01:00", "subject": "The two Chawan symptoms now have matching fixes:", "body": "- Layout corruption: C=1 is respected.\n - Lingering remnants: tmux now tracks Kitty’s (image ID, placement ID) identity, replaces moved placements, and honors\n placement-specific d=i,...,p=... deletion.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1200dfd60c367e97a5e1f31100d82072eb0178b6", "date": "2026-08-06T08:19:13+09:00", "subject": "Remove redundant unlink() in KeepFileRestoredFromArchive() for non-WIN32", "body": "KeepFileRestoredFromArchive() called unlink() on an existing WAL segment\nbefore durable_rename() replaced it. rename(2) atomically replaces a\ntarget on POSIX systems, and the extra unlink() created a small window\nwhere it would be possible for a concurrent WAL sender to see a segment\nas missing.\n\nNote that the window still exists on WIN32, as we lack a safe concurrent\nalternative. Perhaps something like ReplaceFile() could be looked at\nfor this purpose.\n\nThis problem is hard to reach in practice, so no backpatch is done. I\nalso tend to be conservative regarding recovery changes, even small, in\nstable branches.\n\nAuthor: Stepan Neretin \nReviewed-by: Andrey Borodin \nReviewed-by: Neil Chen \nDiscussion: https://postgr.es/m/CA+Yyo5S0C25jS_pRWQiGgoiy+X=GyJS7ZJkEq=_SW1WWpkL3Sg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5f1c7bb0cc37d18b2fc656e1148977e3e17fe3fe", "date": "2026-08-06T01:52:45+02:00", "subject": "Fix ASN1_TYPE memory leak in asn1_multi()", "body": "In asn1_gen.c `asn1_multi()` function variable allocated as\n`ASN1_TYPE *typ = generate_v3()` is not freed if\n`if (!sk_ASN1_TYPE_push(sk, typ))` branch fails.\n\nBug found by x509v3 fuzzer MFAIL test.\nExample input:\n```\n[default]\n1.2.3.4 = ASN1:SEQUENCE:items\n\n[items]\nvalue = INTEGER:1\n```\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 14 10:29:08 2026\nMerged-from: https://github.com/openssl/openssl/pull/32206", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "05bbcc9491193aaa9f0e60c3385d849b2367111b", "date": "2026-08-06T01:52:45+02:00", "subject": "Fix ASN1_TYPE memory leak in asn1_multi()", "body": "In asn1_gen.c `asn1_multi()` function variable allocated as\n`ASN1_TYPE *typ = generate_v3()` is not freed if\n`if (!sk_ASN1_TYPE_push(sk, typ))` branch fails.\n\nBug found by x509v3 fuzzer MFAIL test.\nExample input:\n```\n[default]\n1.2.3.4 = ASN1:SEQUENCE:items\n\n[items]\nvalue = INTEGER:1\n```\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 14 10:25:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/32206", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0c1e2d98eab6d6409f46b67cf43fa1a4555bb5c8", "date": "2026-08-06T01:52:45+02:00", "subject": "Fix ASN1_TYPE memory leak in asn1_multi()", "body": "In asn1_gen.c `asn1_multi()` function variable allocated as\n`ASN1_TYPE *typ = generate_v3()` is not freed if\n`if (!sk_ASN1_TYPE_push(sk, typ))` branch fails.\n\nBug found by x509v3 fuzzer MFAIL test.\nExample input:\n```\n[default]\n1.2.3.4 = ASN1:SEQUENCE:items\n\n[items]\nvalue = INTEGER:1\n```\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 14 10:15:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32206", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "284b863be865e4c539775e79cdd9b76f7f9cb886", "date": "2026-08-06T01:52:45+02:00", "subject": "Fix ASN1_TYPE memory leak in asn1_multi()", "body": "In asn1_gen.c `asn1_multi()` function variable allocated as\n`ASN1_TYPE *typ = generate_v3()` is not freed if\n`if (!sk_ASN1_TYPE_push(sk, typ))` branch fails.\n\nBug found by x509v3 fuzzer MFAIL test.\nExample input:\n```\n[default]\n1.2.3.4 = ASN1:SEQUENCE:items\n\n[items]\nvalue = INTEGER:1\n```\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 14 09:54:19 2026\nMerged-from: https://github.com/openssl/openssl/pull/32206", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "064c326e96fc03ec2553d2477953cdd86421a67d", "date": "2026-08-06T01:52:45+02:00", "subject": "Fix ASN1_TYPE memory leak in asn1_multi()", "body": "In asn1_gen.c `asn1_multi()` function variable allocated as\n`ASN1_TYPE *typ = generate_v3()` is not freed if\n`if (!sk_ASN1_TYPE_push(sk, typ))` branch fails.\n\nBug found by x509v3 fuzzer MFAIL test.\nExample input:\n```\n[default]\n1.2.3.4 = ASN1:SEQUENCE:items\n\n[items]\nvalue = INTEGER:1\n```\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 14 09:48:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32206", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d9f5cb8965cb56414b97a95a101dc68e08f26476", "date": "2026-08-06T01:52:45+02:00", "subject": "Fix ASN1_TYPE memory leak in asn1_multi()", "body": "In asn1_gen.c `asn1_multi()` function variable allocated as\n`ASN1_TYPE *typ = generate_v3()` is not freed if\n`if (!sk_ASN1_TYPE_push(sk, typ))` branch fails.\n\nBug found by x509v3 fuzzer MFAIL test.\nExample input:\n```\n[default]\n1.2.3.4 = ASN1:SEQUENCE:items\n\n[items]\nvalue = INTEGER:1\n```\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 14 09:42:52 2026\nMerged-from: https://github.com/openssl/openssl/pull/32206", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "014b00a6420d34b80ebf4f6ac9aebd88fe236e5d", "date": "2026-08-06T02:53:00+03:00", "subject": "fix(wasm): fix browser input and runtime exit handling", "body": "Use ev.key directly to preserve Caps Lock casing, implement clipboard\npaste, and prevent :q/:qa from leaving the runtime in a loop.", "style": "conventional", "type": "fix", "scope": "wasm", "breaking": false} {"repo": "ant-design", "sha": "cd2c5c83dec2b1d48be8fc9074d086e127f03cb2", "date": "2026-08-06T09:53:08+08:00", "subject": "refactor(utils): centralize guards function", "body": "* refactor(utils): centralize guards function\n\n* update\n\n* update", "style": "conventional", "type": "refactor", "scope": "utils", "breaking": false} {"repo": "vue-core", "sha": "6945e1426ea9c865b02e92762dce72e3f0e12e0b", "date": "2026-08-06T10:23:23+08:00", "subject": "fix(vapor): mark forwarded root slot outlets", "body": "Propagate exposed validity changes from forwarded root slot outlets so\nparent slot boundaries can re-resolve content and fallback correctly.", "style": "conventional", "type": "fix", "scope": "vapor", "breaking": false} {"repo": "ant-design", "sha": "461afa0011f6661a99aa1162b8aed8ef47f8d0d1", "date": "2026-08-06T11:20:53+08:00", "subject": "feat(Tooltip/Popover/Popconfirm): support global mouseEnterDelay/mouseLeaveDelay via ConfigProvider", "body": "* feat(Tooltip): support global mouseEnterDelay/mouseLeaveDelay via ConfigProvider\n\n- Add mouseEnterDelay and mouseLeaveDelay to TooltipConfig type\n- Consume global config in Tooltip component with component props override\n- Add unit tests for global delay config\n- Update documentation for ConfigProvider tooltip config\n\nCo-Authored-By: Claude \n\n* docs: update sharedProps global config version\n\nCo-Authored-By: Claude \n\n* docs: revert ConfigProvider tooltip section title\n\nCo-Authored-By: Claude \n\n* fix: use merged delay values in mergedProps\n\n* fix: use merged delay values in mergedProps for Tooltip and Popconfirm\n\nCo-Authored-By: Claude \n\n---------", "style": "conventional", "type": "feat", "scope": "Tooltip/Popover/Popconfirm", "breaking": false} {"repo": "vue-core", "sha": "87faca486dba7da4379cc17b163dc654815298b4", "date": "2026-08-06T11:22:08+08:00", "subject": "fix(vapor): handle keyed slot roots", "body": "", "style": "conventional", "type": "fix", "scope": "vapor", "breaking": false} {"repo": "vite", "sha": "ef02435114c57d0422028f0e6987f3df8db72969", "date": "2026-08-06T04:22:17+01:00", "subject": "docs(build): fix incomplete `@default` for build.minify", "body": "", "style": "conventional", "type": "docs", "scope": "build", "breaking": false} {"repo": "postgres", "sha": "4edd59de527780f536fbbf97d5e117524ded2723", "date": "2026-08-06T17:43:26+12:00", "subject": "Fix unlikely incremental tuple deform bug with missing attrs", "body": "The code added in c456e3911 added populate_isnull_array() to bulk\npopulate the slot's tts_isnull array 8 elements at a time. When tuples\ndon't have an exact multiple-of-eight number of attributes, this will\nlead to populating the tts_isnull elements for attributes that don't\nexist in the tuple. This is ok as the array is large enough. However,\nif we perform tuple deforming in two passes, and on the first pass\ndeform *some* of the attributes with slot_getmissingattrs() then later\nwhen we deform the remaining missing attributes, the subsequent call to\npopulate_isnull_array() would overwrite the tts_isnull values previously\nset by slot_getmissingattrs(), and since that function only continues\nwhere it left off, it wouldn't reapply the previously set values and those\nwould be left as NULLs, as populate_isnull_array() would have set them.\n\nHere, we fix by passing the tuple's natts to slot_getmissingattrs()\nrather than the attnum we're deforming from. This means we apply all\nthe missing attribute values each deform pass, so slightly more work,\nbut deforming several missing values in different deform passes is\nlikely exceedingly rare. Doing that seems much better than adding\noverhead in the happy path to check for this and skip the subsequent call\nto populate_isnull_array().\n\nAuthor: David Rowley \nReported-by: Peter Geoghegan \nDiscussion: https://postgr.es/m/CAH2-WznHo4b+6AmAj0GZ0jXqDSK69MfHe8fAQwuY_01y7cVNdw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "3822fa865894da8c3a94aee23f14551c60af05c2", "date": "2026-08-06T07:45:14+02:00", "subject": "servers: de-duplicate shutdown signal/logging code", "body": "To share more common code between servers, and to log the same set of\ninformation for all of them.\n\nAlso:\n- move server unix socket to global variable.\n\nFollow-up to d1eca3861c9cf8f3cd03274ce0db3c500bd246ae #22501\nFollow-up to 1637bbc9ced4213b902703b9223273ab2dcd2c6e #22498\n\nCloses #22503", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "1637bbc9ced4213b902703b9223273ab2dcd2c6e", "date": "2026-08-06T07:59:29+02:00", "subject": "servers: sync socket type global variables", "body": "Replace `ipv_inuse` and `use_ipv6` with `socket_type` and\n`socket_domain` (where missing) to avoid dupliicate globals with\noverlapping purposes. The replacement variables also support Unix\nsockets.\n\nAlso:\n- simplify/reduce IPv6 guards.\n- socksd: fix to reset `socket_domain` for `--ipv4` option.\n\nCloses #22498", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "2f82eafc2112e77ae2d4b85aa7cce345e4b4db36", "date": "2026-08-06T14:07:53+08:00", "subject": "fix(vapor): resolve fallback across multiple slot roots", "body": "", "style": "conventional", "type": "fix", "scope": "vapor", "breaking": false} {"repo": "postgres", "sha": "4790abd112fcb889bac3e1e45b2e7238d48184e5", "date": "2026-08-06T11:41:18+05:30", "subject": "Fix race condition in subscription TAP test 023_twophase_stream.", "body": "Buildfarm member olingo intermittently failed this test, timing out while\nwaiting for the subscriber log to report an ERROR because\nmax_prepared_transactions is zero there. The test captured the log offset\nonly after issuing the publisher's\nBEGIN/INSERT/PREPARE TRANSACTION/COMMIT PREPARED sequence.\n\nSince streaming is enabled, the subscriber can receive and apply the\ntransaction, and log the expected ERROR, before that publisher SQL command\neven returns, i.e. before the test captures the offset. The subsequent\nwait_for_log() calls then searched only from a point after the message had\nalready been written, and timed out waiting for it.\n\nFix by moving the offset capture to before the publisher's transaction is\nissued, ensuring it always precedes the point where the ERROR can appear\nin the subscriber log.\n\nReported-by: Alexander Lakhin \nAuthor: Zhijie Hou \nReviewed-by: Amit Kapila \nBackpatch-through: 16, where test was introduced\nDiscussion: https://postgr.es/m/c43753d8-5265-4f77-83ff-9b1167276ec5@gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "58ff4a0a08679356e502dfcbbc50e9bec94b77b1", "date": "2026-08-06T15:39:56+09:00", "subject": "Tighten tid input parsing with strtoul() to reject empty fields", "body": "strtoul() informs that in the case of no conversion, \"endptr\" is set\nequal to the location of the input pointer, which is something that\ntidin() has never checked.\n\nThis allowed inputs like \"(,)\", \"(1,)\" or \"(,1)\" to be accepted on some\nplatforms, like Linux where \"(1,)\" implied \"(1,0)\", but rejected on\nothers like macos where \"(1,)\" is rejected, due to the fact that errno\nmay not be set.\n\nThis change may cause some issues as some tid values will now be\nrejected, but at least we get the consistent strtoul() experience.\nIf it proves to be a pain for some, it could always be reverted.\nAnother point of this change is that we get one step closer to a\npossible common strtoul() where the error handling could be unified in\ncore, with all the errors handled the same way for errno and no input\nparsed, at least.\n\nReported-by: Michael Malis \nAuthor: Tristan Partin \nDiscussion: https://postgr.es/m/19584-e60c446ba6f57c9c@postgresql.org\nDiscussion: https://postgr.es/m/DKBS2Z9CGARC.2T07O6TJYSE8B@partin.io", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "fd24c4bcf4880b0b8ca02e48075abd3adb02af18", "date": "2026-08-06T08:58:22+02:00", "subject": "tests/server/dnsd: fix to install signal handlers on startup", "body": "To sync with the rest of code which already handled `got_exit_signal`\nand did the restoration on exit. Also syncing with rest of servers.\n\nAlso: turn off `keep_sigalrm` to sync with most servers.\n\nCloses #22499", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c04189523cb848922d7ea03898b525cfffe3bc9c", "date": "2026-08-06T09:13:30+02:00", "subject": "cookie: refuse to load cookies set against a PSL domain", "body": "Verified by test 409\n\nReported-by: 1rhino2 on hackerone\n\nCloses #22500", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "d1eca3861c9cf8f3cd03274ce0db3c500bd246ae", "date": "2026-08-06T09:27:41+02:00", "subject": "servers: sync server port global variable across servers, make port variables `uint16_t`", "body": "Follow-up to 1637bbc9ced4213b902703b9223273ab2dcd2c6e #22498\n\nCloses #22501", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "8b0f33a1abe70266119ad7dd97fda007e86f09f5", "date": "2026-08-06T02:39:53-05:00", "subject": "feat(dir): set buffer-local CWD #41174", "body": "", "style": "conventional", "type": "feat", "scope": "dir", "breaking": false} {"repo": "postgres", "sha": "290bb8afe6b47adb8b080688bd97a710af351cb1", "date": "2026-08-06T16:46:03+09:00", "subject": "Tighten proto_version parsing in pgoutput", "body": "This is similar to 58ff4a0a0867, but this time for the parameter\n\"proto_version\", when given as a parameter to pgoutput for logical\nreplication. strtoul() lacked a check for an empty string, leading to\nan inconsistent behavior depending on the platform.\n\nThe author has originally proposed a TAP test to check this empty value\npattern. I have added a cheaper SQL query, instead, test_decoding\nincluding a test area where this is possible.\n\nAuthor: Tristan Partin \nDiscussion: https://postgr.es/m/DKBS2Z9CGARC.2T07O6TJYSE8B@partin.io", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a478393759f302d3a55e62094ceb61e64ca610f4", "date": "2026-08-06T10:03:56+02:00", "subject": "psl: update a comment to understandable English", "body": "Closes #22502", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "674cdd94f58a83d99e5d21116f5e483bb49bc4f6", "date": "2026-08-06T09:11:45+01:00", "subject": "Add bptato's test to the regression test script.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9b917a93116e6b8aac7ed837ded27696f7232505", "date": "2026-08-06T10:23:44+02:00", "subject": "Report duplicate property and label names with a proper error", "body": "Adding a property with the same name multiple times to a label on an\nelement, either within a single PROPERTIES clause or across statements\nvia ALTER PROPERTY GRAPH ... ADD PROPERTIES, previously failed with a\nunique-index violation on pg_propgraph_label_property. The same class\nof bug existed for labels: listing the same label multiple times on\none element in CREATE PROPERTY GRAPH, or adding a label to an element\nthat already has it via ALTER PROPERTY GRAPH ... ADD LABEL, failed\nwith a unique-index violation on pg_propgraph_element_label. Detect\nthe duplicates up front and raise a friendlier error in both cases.\n\nFor properties, the cross-statement duplicate is caught by a syscache\nprobe before insertion. The in-clause duplicate could have been\ncaught by the same probe if we issued a CommandCounterIncrement()\nbetween property inserts, but that forces a catalog invalidation per\nproperty purely to detect a condition we can check for free on the\nin-memory target list. The list-based check is only needed when the\nproperties are listed explicitly; when they are derived from the\ntable's attributes, names are already unique.\n\nFor labels, a single syscache probe on pg_propgraph_element_label\nsuffices for both the in-clause and cross-statement cases because\ninsert_element_record() already issues CommandCounterIncrement()\nbetween successive label inserts.\n\nAuthor: Ashutosh Bapat \nReported-by: Noah Misch \nDiscussion: https://www.postgresql.org/message-id/flat/20260630173053.51.noahmisch%40microsoft.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "05adcb13542061dc7fadde5cd58d23d405a96d93", "date": "2026-08-06T13:49:34+05:00", "subject": "test: fix failing ecosystem test for `eslint-plugin-unicorn`", "body": "* test: fix failing ecosystem test for `eslint-plugin-unicorn`\n\n* test: don't use `--save-dev` for `core-js-compat` and unpin `eslint-plugin-eslint-plugin` in ecosystem test config", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "curl", "sha": "de9919f38a0504d3eba66da35aa9d93c49f34e4a", "date": "2026-08-06T11:05:01+02:00", "subject": "TODO: ECH for QUIC", "body": "And drop:\n\n- Consider OCSP stapling by default\n\nIt is a practice that is going out-of-style, so doing this by default now\nseems wrong.\n\n- Provide callback for cert verification\n\nWe have lots of options already. Let's not do this.\n\nCloses #22504", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "641e75e6e251de581fa3154fd2d1523d17d76395", "date": "2026-08-06T09:05:04Z", "subject": "Add T key to change pane title, from Fouad Wahabi in GitHub issue 5461.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "7d2d187db4990a93e057fee0f7704e3b99f7036e", "date": "2026-08-06T10:25:29+01:00", "subject": "Inline basic image fallback renderer", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "3aae64e4fbee7c1fa408c54df18d3f631781c283", "date": "2026-08-06T11:54:04+02:00", "subject": "servers: drop complex and redundant signal handler output", "body": "In year 2020 the Unixy signal handler received a `logmsg()` call to log\nthe signal number, but at the same time it already saved it to a global\nvariable and logged it on exit, meaning this extra `logmsg()` was\nredundant. Because `logmsg()` is not signal-safe, this call was replaced\nin 2025 with signal-safe logging, but without the signal number, while\nalso adding complexity, spent on trying to open the log file and handle\nerrors. All for nothing, because the signal number was logged all along.\n\nThis patch removes all this, and simplifies it to a single, signal-safe\n`write()` to STDERR to say that the signal handler triggered. This is\nalso non-critical, but may help debugging.\n\nAlso: point the POSIX documentation to the 2004 revision, which has a\nshorter list of safe functions. (was: 2018)\n\nFollow-up to e95f509c66abdd88ae02e3243cdc217f19c4a330 #16852\nFollow-up to 9869f6dc5af85caf2e0fd4c56713a7f2049ecfff #5218\n\nCloses #22507", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "eb19a52b7c235ad5dd86167a6268b1beaf560df8", "date": "2026-08-06T13:16:48+02:00", "subject": "feat(vim._with): `keepcwd`", "body": "", "style": "conventional", "type": "feat", "scope": "vim._with", "breaking": false} {"repo": "tmux", "sha": "596de59d118bfa421f9c0e1814b8675fb09bcee2", "date": "2026-08-06T12:24:10+01:00", "subject": "Make image types opaque", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "c2155fe4d5c8d25fba3a7366d367e3296ae669fa", "date": "2026-08-06T19:24:34+08:00", "subject": "test(bundled-dev): enable sourcemap playgrounds", "body": "", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "curl", "sha": "7f6a75664f9fb45390193e492ee3361fb795d098", "date": "2026-08-06T13:46:41+02:00", "subject": "dnsd: add support for DoH", "body": "dnsd now opens UDP+TCP sockets and accepts http: DoH requests to obtain\nthe same, configured answers (records, delays, error codes) as over UDP.\n\nDoH: use `async->queries_ongoing` like all other resolvers instead of\nthe internal `pending` counter. Fixes waiting for results.\n\nTests: in pytest, parameterize dnsd tests to use both DNS and DoH.\n\nCloses #22506", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "3b0cb22eb6246b0da2b944e713f801339443303e", "date": "2026-08-06T14:07:15+02:00", "subject": "servers: add option to build without window handler (Windows)", "body": "curl already built like this in UWP mode; add a macro to allow building\nregular Win32 targets too, meaning without the hidden GUI window, window\nhandler, and thus without support for shutting down `servers.exe`\ngracefully. To test the server in CI as a plain console app and remove\none component/variable from the system, in an effort chasing flaky runs.\n\nThe graceful option was used by `taskkill` (without `-f`) in runtests,\nand it has been disabled (in favor of always `-f` (forced) kill), in an\nattempt to avoid taskkill hanging. Those resolved, but general CI\nflakiness remained. We see `taskkill` invoked seldom per job, and in\nmost of these cases it doesn't find the PID requested. It means the\ngraceful option is not actually used since 2025-04, and rather rarely\nbefore that. Also just today, the callback and main loop received a few\nupdates to avoid potentially unsafe calls and interaction with the ctrl\nhandler.\n\nNot expecting anything breakthrough here. Also graceful shutdown\nis a useful feature to cleanup properly and to be in sync with other\nplatforms.\n\nAlso: GHA/windows: enable in Windows CI jobs.\n\nRef: https://learn.microsoft.com/windows/console/registering-a-control-handler-function#listen-with-hidden-window-example\n\nFollow-up to 7db9947fcf5d3deebe98e7992ebedc9a2f624bf5 #22045\nFollow-up to 9ea48811fed455d4a869eb100b2216f039f8677a #22487\nFollow-up to b11e0026f71aa33b47d3b31f7c6a2aee4f929b12 #22495\nFollow-up to 208b87744e8ce326e2569c2efd735da43ce74a0b #21039\nFollow-up to f450f3801b6b9dff0ea280f5fb4bf28203f7b313 #19897\nFollow-up to 2701ac6a4d16a62130dad05be1c484903b8545c7 #19421\nFollow-up to 4e203f65a1e5e448daf1da2b8c3aaa33654dd080 #17054\n\nCloses #22496", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f0178deecbb56283fda8f4c504c8fe16ce23e1a6", "date": "2026-08-06T14:07:57+02:00", "subject": "doc: Define MXID as an acronym", "body": "XID is defined as an acronym in the documentation but MXID (Multixact\nID) was missing from the acronym list. This adds MXID as an acronym\nand links it to the \"Multixacts and Wraparound\" section.\n\nAuthor: Koshino Taiki \nDiscussion: https://postgr.es/m/OS9PR01MB17364A940ACFCADBB7593498A94D32@OS9PR01MB17364.jpnprd01.prod.outlook.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "344ea602f655a7d22ab3567536fef0dda167d2ae", "date": "2026-08-06T14:18:57+02:00", "subject": "refactor(ex_cmds): use function arguments for magic behavior", "body": "Now this was a cargo cult anti-pattern to write home about.\nDoing painful save-and-restore bookkeeping around a separate\n`magic_overruled` decoy global is just as messy as doing painful\nsave-and-restore logic around `p_magic` itself. only that now you need\nto wrap every access to the effective value in a function call.\n\nThis replaces this with a marvellous new Clean Code technique™:\npassing in the intended behavior as a function parameter to functions\nwhere either the option or an explicit value might be used.", "style": "conventional", "type": "refactor", "scope": "ex_cmds", "breaking": false} {"repo": "vite", "sha": "15f03073c915d6ffb9a1fda447ef66b02bf5cde8", "date": "2026-08-06T21:37:09+09:00", "subject": "fix(build): make client chunkImportMap work with `sharedPlugins: true`", "body": "", "style": "conventional", "type": "fix", "scope": "build", "breaking": false} {"repo": "vite", "sha": "de041a79b05a0be965c874592fe2c1505bcd48df", "date": "2026-08-06T17:38:08+05:00", "subject": "fix(css): don't re-run lightningcss visitor during minify (fix #23146)", "body": "", "style": "conventional", "type": "fix", "scope": "css", "breaking": false} {"repo": "neovim", "sha": "3a396466930f22d279934023e1459225a6de9389", "date": "2026-08-06T07:39:59-05:00", "subject": "fix(window): 'laststatus' change pushes win past last row #41189", "body": "Problem:\nGiving a window a status line when `'laststatus'` starts requiring one\ntakes the row from a resizable frame found by walking up the frame\ntree, but only the window's own leaf frame is grown back. Every frame\nbetween the leaf and the donor keeps its old height, so a later resize\nhands out a row that does not exist and a window's status line ends up\non the command line.\n\nSolution:\nGrow the window's frame with `frame_setheight()`, which takes the row\nfrom a neighbouring frame and keeps every enclosing frame consistent.", "style": "conventional", "type": "fix", "scope": "window", "breaking": false} {"repo": "vite", "sha": "16944824fd5fae014101737ec0029ab6a58e774b", "date": "2026-08-06T20:49:56+08:00", "subject": "test: poll after editFile in playground tests", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vite", "sha": "fddf4ea41de5f7889037a2f957438857ac12a260", "date": "2026-08-06T18:31:23+05:30", "subject": "fix(server): use a random port when port is 0", "body": "", "style": "conventional", "type": "fix", "scope": "server", "breaking": false} {"repo": "tmux", "sha": "0b2739c62c520a48d6d821203f60bae21092e7ab", "date": "2026-08-06T14:27:07+01:00", "subject": "Use one image support switch", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1b25e3fca5772fce2edd71a6adfc306dbe254c3e", "date": "2026-08-06T15:33:27+02:00", "subject": "Reject duplicate fields in X509v3 config", "body": "Reject repeated fields in policyConstraints, basicConstraints\nand basicAttConstraints instead of silently accepting the last value.\n\nAssisted-by: ChatGPT:gpt-5.6\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nMergeDate: Sun Aug 9 04:06:37 2026\n(Merged from https://github.com/openssl/openssl/pull/32181)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dfb78caa5711d439e66bcaf3090f88d6aae47413", "date": "2026-08-06T16:05:37+02:00", "subject": "Add x509v3 duplicate field config tests", "body": "Reviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nMergeDate: Sun Aug 9 04:06:42 2026\n(Merged from https://github.com/openssl/openssl/pull/32181)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "35ab6600429881b7a594548ca23d7c0d5006c2a1", "date": "2026-08-06T10:23:46-04:00", "subject": "ci: only run `lock` and `regress` workflows on the main repo", "body": "Forks are just wasting CI cycles spinning actions which are probably not\nlooked at. Also avoid locking issues and PRs on any forks with\nupstream's policies.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "17e4b50a241d2f9f7a862784396ee97552f1011c", "date": "2026-08-06T16:42:37+02:00", "subject": "Run clang-format - chore", "body": "Reviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nMergeDate: Sun Aug 9 04:06:46 2026\n(Merged from https://github.com/openssl/openssl/pull/32181)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "95b6adbd61ec37f064beffcf85457cdae8d7db08", "date": "2026-08-06T16:17:21+01:00", "subject": "Simplify image support terminology", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "24ff74fa8db2dd84b2acd190a1bcc04b7d69e1e7", "date": "2026-08-06T08:24:02-07:00", "subject": "CI: fix labeler matches for vdns file move", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "d3d46c947516dfd0e64f2808311b99aa9d5e08c0", "date": "2026-08-06T18:37:56+02:00", "subject": "fix(devtools): fix regression for node serialization.", "body": "This fixes a regression introduced by #69309 where the format of `serializedId` from a coma separated string, to a JSON representation of an array.\n\nfixes #70104\n\n(cherry picked from commit 03f1758e2ae7e033c14360a72583d51825993217)", "style": "conventional", "type": "fix", "scope": "devtools", "breaking": false} {"repo": "angular", "sha": "03f1758e2ae7e033c14360a72583d51825993217", "date": "2026-08-06T18:37:56+02:00", "subject": "fix(devtools): fix regression for node serialization.", "body": "This fixes a regression introduced by #69309 where the format of `serializedId` from a coma separated string, to a JSON representation of an array.\n\nfixes #70104", "style": "conventional", "type": "fix", "scope": "devtools", "breaking": false} {"repo": "neovim", "sha": "693de029bc0377dc89291ab3dedeec1f01051a73", "date": "2026-08-06T12:53:45-04:00", "subject": "build(vim-patch): vim9 skip_type() is n/a", "body": "v8.2.1300", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "4b581a4666c3e470a01a7323801b2ba8ccfa478c", "date": "2026-08-06T11:05:43-06:00", "subject": "Add a migration entry for ASN1_STRINGs", "body": "Re-document that ASN1_STRINGs are not NUL byte terminated, and\nhave been documented as such for over 20 years.\n\nASN1_STRING data has never been guaranteed to be NUL byte terminated.\nASN1_STRING_set() added one anyway; its replacements do not, so code\nrelying on UB, and treating it as a C string is more likely to break.\n\nReviewed-by: Neil Horman \nReviewed-by: Mounir Idrassi \nReviewed-by: Norbert Pocs \nReviewed-by: Tomas Mraz \nMerge-date: Fri Sep 4 17:35:40 2026\nMerged-from: https://github.com/openssl/openssl/pull/32233", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "12d9f33c155dcd06c3920defa72db081db4c5774", "date": "2026-08-06T13:16:35-04:00", "subject": "build(vim-patch): vim9 error 1103 is n/a", "body": "v8.2.1341", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "tmux", "sha": "628acfd802ba2422c98a343ff96587fee75489b9", "date": "2026-08-06T19:08:59+01:00", "subject": "Pass through wrapped Kitty graphics", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5593aca0c62e08f5827d706d03be99481ef9bcb0", "date": "2026-08-06T13:10:00-05:00", "subject": "Remove lo_hton64() and lo_ntoh64().", "body": "These functions date back to commit 461ef73f09, which predates the\n64-bit byte-swapping routines in pg_bswap.h (added in commit\n510b8cbff1). Commit 0ba99c84e8 converted them to use pg_hton32()\nand pg_ntoh32() but stopped short of removing them, so let's do\nthat now.\n\nDiscussion: https://postgr.es/m/ahXE28klgxIJXBLq%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "92fc04afc613860cb5f1c9df956370aa0390d60b", "date": "2026-08-06T13:29:45-06:00", "subject": "Remove dead code from pkcs7", "body": "The \"state\" variable is dead, nothing ever reads it. We will\nbe repurposing it for something else.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Viktor Dukhovni \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 13:56:39 2026\nMerged-from: https://github.com/openssl/openssl/pull/32242", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "031dddc69998b3d76446e41680c7917216b4f772", "date": "2026-08-06T13:39:50-06:00", "subject": "Store the CMS created-content state in CMS_ContentInfo", "body": "Whether the content octet string was created locally (to be filled in)\nor read in via d2i is CMS's own state; it was being kept in the\nASN1_STRING via ASN1_STRING_FLAG_CONT, of which the only user was\nCMS. Track it in a CMS_ContentInfo field instead of using ASN1_STRING\nas a convenient internal state toilet. No change in behaviour.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Viktor Dukhovni \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 13:56:40 2026\nMerged-from: https://github.com/openssl/openssl/pull/32242", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "636bb3e45f8f97b1916f1bd9b190f6645dfc259b", "date": "2026-08-06T13:59:38-06:00", "subject": "Track the PKCS7 streaming state in the state field, not the ASN1_STRING", "body": "Whether the content is an indefinite-length (NDEF) streaming placeholder\nis PKCS7's own state, but it was read back from ASN1_STRING_FLAG_NDEF on\nthe content octet string. Record it in a spare bit of the PKCS7 state\nfield instead, set in PKCS7_stream and read where the content BIO is set\nup and finalized. PKCS7_stream still sets the NDEF flag for the ASN.1\nencoder; that move follows separately. No change in behaviour.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Viktor Dukhovni \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 13:56:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/32242", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2ebd8e127d91f467f3dc868bb4a68f3f9b9c1883", "date": "2026-08-06T14:31:16-06:00", "subject": "Do the NDEF streaming setup in the ASN.1 layer, not in CMS and PKCS7", "body": "CMS_stream() and PKCS7_stream() marked the content octet string with\nASN1_STRING_FLAG_NDEF and handed BIO_new_NDEF() the address of its data\nfield, where the encoder records the position of the content. Add an\nASN1_OP_STREAM_CONTENT callback operation through which BIO_new_NDEF()\nobtains that string, sets the flag, and reads the position back out, so\nboth are manipulated only by the ASN.1 code that uses them. What is left\nof the two functions becomes ossl_cms_stream() and ossl_pkcs7_stream();\nthe public ones remain as wrappers but are unused.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Viktor Dukhovni \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 13:56:42 2026\nMerged-from: https://github.com/openssl/openssl/pull/32242", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "48c6a6f83951da7367e5ea847e240a241ffaf6a7", "date": "2026-08-06T22:36:43+02:00", "subject": "docs: v9.x EOL", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "1ea07d6e0a1263aaa549a8253db619e87af1da34", "date": "2026-08-06T22:52:45+02:00", "subject": "CI: filter all .[ch] files before doing typo checks", "body": "The filter only leaves comments and strings in all files using .c or .h\nextensions.\n\nIt is important that the filter runs after all the other checks are done\non these files as this process destroys the files.\n\nFixes #22508\nCloses #22509", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "curl", "sha": "a385dd06c337cffdf3f12c9048e34d46286dab80", "date": "2026-08-06T22:52:45+02:00", "subject": "CI: filter all .[ch] files before doing typo checks", "body": "The filter only leaves comments and strings in all files using .c or .h\nextensions.\n\nIt is important that the filter runs after all the other checks are done\non these files as this process destroys the files.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "curl", "sha": "ff4bab0a68fcb1d3ffc31b61b71c877473dc04f4", "date": "2026-08-06T23:14:25+02:00", "subject": "ldap_do: refactor error handling and simplify show_vals function", "body": "- (much) less code repetition\n\n- simplifies ldap_do somewhat\n\nCloses #22510", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "585181e077486324e12f6071a70955a0a48906aa", "date": "2026-08-06T17:16:23-04:00", "subject": "Restore vacuum failsafe abandonment of buffer access strategy", "body": "VACUUM's wraparound failsafe mode exists to reclaim transaction IDs as\nquickly as possible. 4830f1024325 made the failsafe stop using the\nBAS_VACUUM buffer access strategy so that the rest of the vacuum could\nmake use of all of shared buffers rather than being confined to the\nsmall strategy ring.\n\nHowever, when 9256822608f3 made vacuum's first heap pass use the read\nstream, this was accidentally disabled. The read stream keeps its own\nreferences to the buffer access strategy, so clearing vacrel->bstrategy\nin lazy_check_wraparound_failsafe() no longer had any effect on the\nreads issued by the first pass.\n\nFix this by adding clearing the BufferAccessStrategy reference actually\nbeing used by the ongoing scan -- those in the ReadBuffersOperations\nstructs themselves.\n\nTwo things we accept rather than fix, as neither is worth the added\ncomplexity given how rarely failsafe mode is reached:\n\n- A small amount of read time for IOs that were already in progress when\n the strategy was cleared may be attributed to IOCONTEXT_NORMAL instead\n of IOCONTEXT_VACUUM. WaitReadBuffers() derives the IOContext from the\n (now cleared) strategy, so the wait time of these in-flight IOs is\n misattributed. This is bounded by the stream's look-ahead window and\n happens at most once per vacuum, when the strategy is first cleared.\n\n- The stream's buffer pin limit stays lower than it would have been had\n no strategy been used at all. max_pinned_buffers is capped by the\n strategy's pin limit when the stream is created and is not recomputed\n when the strategy is cleared. Raising it would mean building a new,\n larger ring, which would require first waiting for all in-progress IOs\n to complete. That didn't seem worth it.\n\nReported-by: Jingtang Zhang \nDiscussion: https://postgr.es/m/CAPsk3_APRYVLhAJ5TMwdmpSx8W_%3DPHMm%3DPmKAvnC3gBrfNommQ%40mail.gmail.com\nBackpatch-to: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "ba2dc2cfb6f5c716e2670be84e815959079b7062", "date": "2026-08-06T22:17:19+01:00", "subject": "Clear text beneath graphical images", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "db0c984b1d101b7519df1922c8a880d14f19de82", "date": "2026-08-06T23:27:59+02:00", "subject": "Remove scram_client_key_len and scram_server_key_len from PGconn", "body": "Commit 761c79508e7 added scram_client_key_len and scram_server_key_len\nto PGconn, but no consumer read the fields and there were no accessor\nfunctions defined. Code which need the length use SCRAM_MAX_KEY_LEN\nwhich the new fields were also verified to match, so remove the fields\nas they are not used.\n\nAuthor: Aleksander Alekseev \nReviewed-by: Matheus Alcantara \nDiscussion: https://postgr.es/m/CAJ7c6TMx3pyuW86zEjGs4kua0vQPzZEg2sCUe+NnP08fBKuVoQ@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "fb931f167619ce6c79380bdfa28c1b35860f7af9", "date": "2026-08-06T22:34:49+01:00", "subject": "Use compact image references in grid cells", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "6f9a64e6f5b1e3faada6514e8b090f9d61aa8e54", "date": "2026-08-07T00:00:04+02:00", "subject": "fix(compiler): remove namespaced MathML script elements", "body": "Treat MathML-namespaced script elements as scripts during template preprocessing. This prevents scripts nested in MathML HTML integration points from surviving template compilation.\n\n(cherry picked from commit 107f6fa49d00ec250b94a8d5f484bae2fc3d71c9)", "style": "conventional", "type": "fix", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "107f6fa49d00ec250b94a8d5f484bae2fc3d71c9", "date": "2026-08-07T00:00:04+02:00", "subject": "fix(compiler): remove namespaced MathML script elements", "body": "Treat MathML-namespaced script elements as scripts during template preprocessing. This prevents scripts nested in MathML HTML integration points from surviving template compilation.", "style": "conventional", "type": "fix", "scope": "compiler", "breaking": false} {"repo": "neovim", "sha": "2b7c53be421b6d733a161bb5b11cfa54090db519", "date": "2026-08-06T19:33:43-04:00", "subject": "build(vim-patch): runnvim should make redir_test_to_null n/a", "body": "Prior patches for it are marked n/a.\nTarget v8.2.2630 .", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "angular", "sha": "12b0acd498b01a743de24eaefd99b33e311bba96", "date": "2026-08-06T23:42:02Z", "subject": "feat(devtools): implement persistent breakpoints on signal consumers using CDP", "body": "Implements persistent breakpoints for Angular signal consumers (computeds and effects) in Angular DevTools using Chrome DevTools Protocol (CDP).\n\n- Adds debugger permission to Chrome and Firefox extension manifests.\n- Implements CDP breakpoint orchestration in background script.\n- Updates SignalDetailsComponent and SignalGraphPaneComponent with label icon toggle and persistent state sync across reloads.", "style": "conventional", "type": "feat", "scope": "devtools", "breaking": false} {"repo": "nestjs", "sha": "7637d934b9eef57a20999319dab211b58c80a884", "date": "2026-08-07T05:46:27+06:00", "subject": "fix(core): share existing singleton providers with lazily loaded modules", "body": "Fixes #17428", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "ec6deea51329b0c8410c75943d90ce823081e216", "date": "2026-08-06T17:09:48-07:00", "subject": "fix(compiler-cli): record class extends clause references in DeferredSymbolTracker", "body": "`DeferredSymbolTracker.lookupIdentifiersInSourceFile` prunes `ts.isTypeNode`\nsubtrees so that references appearing exclusively inside type annotations\ndo not keep static import declarations in the emitted JavaScript.\n\nHowever, `ts.isTypeNode` returns `true` for `ts.ExpressionWithTypeArguments`,\nwhich TypeScript uses to represent both `extends` and `implements` heritage\nclauses. An `extends` clause on a class declaration or class expression is a\nvalue position that survives in the emitted JavaScript output.\n\nBecause `isTypeNode` returned `true`, references to base classes imported\nalongside deferred dependencies were ignored. As a result, the compiler\nerroneously marked the static import statement as deferrable and deleted it\nfrom the emitted JavaScript, leaving the `extends ` clause referencing\nan undeclared identifier and causing a runtime `ReferenceError`.\n\nThis commit ensures that `ExpressionWithTypeArguments` under a class `extends`\nclause is not treated as an erasable type node, preserving the static import\nwhenever a base class is referenced.\n\n(cherry picked from commit 8d6c925392c7230fbe6b39abdad05381672631b1)", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "8d6c925392c7230fbe6b39abdad05381672631b1", "date": "2026-08-06T17:09:48-07:00", "subject": "fix(compiler-cli): record class extends clause references in DeferredSymbolTracker", "body": "`DeferredSymbolTracker.lookupIdentifiersInSourceFile` prunes `ts.isTypeNode`\nsubtrees so that references appearing exclusively inside type annotations\ndo not keep static import declarations in the emitted JavaScript.\n\nHowever, `ts.isTypeNode` returns `true` for `ts.ExpressionWithTypeArguments`,\nwhich TypeScript uses to represent both `extends` and `implements` heritage\nclauses. An `extends` clause on a class declaration or class expression is a\nvalue position that survives in the emitted JavaScript output.\n\nBecause `isTypeNode` returned `true`, references to base classes imported\nalongside deferred dependencies were ignored. As a result, the compiler\nerroneously marked the static import statement as deferrable and deleted it\nfrom the emitted JavaScript, leaving the `extends ` clause referencing\nan undeclared identifier and causing a runtime `ReferenceError`.\n\nThis commit ensures that `ExpressionWithTypeArguments` under a class `extends`\nclause is not treated as an erasable type node, preserving the static import\nwhenever a base class is referenced.", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "vue-core", "sha": "1452e5c49f445aad7036313a4012b8b9386045d0", "date": "2026-08-07T08:18:36+08:00", "subject": "refactor(vapor): replace logicalIndex with block placeholders", "body": "Give each anchored dynamic block its own template placeholder so\nCSR element indexes align with SSR logical units. This removes the\nseparate hydration index bookkeeping and unifies compiler and runtime\ninsertion behavior.\n\nReuse eligible placeholders as fragment anchors, remove the\nprepend-specific runtime paths, and fix invalid-nesting insertion\ncodegen.\n\nThis changes the internal Vapor compiler/runtime codegen contract and\nthe CSR DOM anchor shape. Output from older Vapor compiler versions is\nnot compatible.", "style": "conventional", "type": "refactor", "scope": "vapor", "breaking": false} {"repo": "ant-design", "sha": "0d0b6b73c82a893709361f6cd4d92fe7d9b2de3d", "date": "2026-08-07T13:04:33+12:00", "subject": "docs: fix typos in code comment and docs", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "7e221fc61e3f4c57371ce7dc24c51e887ae33157", "date": "2026-08-07T10:12:20+09:00", "subject": "docs: explain browser target using words used by baseline-browser-mapping", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "b24346fe31ad45a91de7d9be0698c8b2c0954d3d", "date": "2026-08-06T22:00:42-04:00", "subject": "api: fix printf format specifier", "body": "- Use %hu for unsigned short instead of %u.\n\nPrior to this change some compilers could warn of an argument type\nmismatch. C argument promotion rules promote the unsigned short argument\nto an int, which does not match %u (unsigned int) but does match %hu\n(unsigned short).\n\nAssisted-by: Viktor Szakats\n\nCloses https://github.com/curl/curl/pull/22511", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "01055a7a83bedb6a183f279670d1e2fac0eb51a6", "date": "2026-08-07T10:58:44+08:00", "subject": "fix(DatePicker): pass props to RangePicker semantic callbacks", "body": "", "style": "conventional", "type": "fix", "scope": "DatePicker", "breaking": false} {"repo": "vite", "sha": "e2d47490c931f25fa79fb41f50c23e85b0254e9b", "date": "2026-08-07T11:03:13+08:00", "subject": "docs(migration): import swc in the decorator workaround", "body": "", "style": "conventional", "type": "docs", "scope": "migration", "breaking": false} {"repo": "vite", "sha": "6a3339811d0f9cd5baed22a34571dff2596be1a1", "date": "2026-08-07T11:03:23+08:00", "subject": "docs(api-javascript): fix typo and update configuration reference", "body": "", "style": "conventional", "type": "docs", "scope": "api-javascript", "breaking": false} {"repo": "neovim", "sha": "107b806bbe6461f76f5167f1cf5c56def705bdd7", "date": "2026-08-06T23:15:57-04:00", "subject": "build(vim-patch): os makefile for tests are n/a", "body": "Following remain relevant:\n\n- src/testdir/Makefile (default setup in Unix-like OS)\n- src/testdir/Make_all.mak (define old/new style Vim tests)\n\nRest are n/a because Nvim runs Vim's tests on GHA Windows runner.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "postgres", "sha": "cacba321c2ef9dfd1c1b6924b6c7291a1182d2d6", "date": "2026-08-07T12:28:44+09:00", "subject": "Make 030_pg_recvlogical robust against PID reuse", "body": "The reconnection test assumed that a restarted walsender would always\nhave a different PID from the previous one. On some platforms, however,\nPIDs can be reused quickly, causing the test to time out even though\npg_recvlogical has successfully reconnected.\n\nThis issue was reported by buildfarm member fairywren:\nhttps://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=fairywren&dt=2026-07-30%2009%3A58%3A12\n\nFix this by waiting for the server log message indicating that the\nlogical replication slot has been acquired again, rather than relying on\na PID change.\n\nBackpatch to v19, where the affected reconnection test was introduced by\ncommit d89b1d817513.\n\nAuthor: Hayato Kuroda \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/OS9PR01MB121497CF10DFF54F745E37204F5C92@OS9PR01MB12149.jpnprd01.prod.outlook.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a15cf1ea07963ab9d92803e23a96ac5541c6ba0f", "date": "2026-08-06T23:30:50-04:00", "subject": "configure: clarify --enable-debug option", "body": "- Change --enable-debug help text to say it is for developing curl.\n\n- Warn when --enable-debug is used.\n\nThis change copies the help text and warning from cmake ENABLE_DEBUG.\nAlso, it shortens the \"for debugging curl itself\" to just say \"for\ndebugging curl\".\n\nNow it looks like this:\n\n --enable-debug Enable curl debug features (for developing\n curl)\n --disable-debug Disable curl debug features\n\nRef: https://github.com/curl/curl/issues/22481#issuecomment-5207220359\n\nReported-by: Viktor Szakats\n\nCloses https://github.com/curl/curl/pull/22512", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "b8ac44e8c0df01462e6124fc11e969f0ddb5a9ea", "date": "2026-08-07T12:01:38+08:00", "subject": "fix(drawline): 'statuscolumn' breaks unprintable char wrapping", "body": "Problem:\nEvaluating 'statuscolumn' overwrites transchar_charbuf[], which breaks\nthe drawing of an unprintable char if p_extra points there.\n\nSolution:\nMake a copy in wlv.extra so that it won't be overwritten.", "style": "conventional", "type": "fix", "scope": "drawline", "breaking": false} {"repo": "postgres", "sha": "10e20e59e9fdf4458c23570bc813333acc328e6c", "date": "2026-08-07T14:23:39+09:00", "subject": "Fix local pgstat entry leak on OOM during entry creation", "body": "When pgstat_init_entry() fails due to an OOM in the DSA allocation,\npgstat_get_entry_ref() cleaned up the shared hashtable but forgot to\nremove the local reference that pgstat_get_entry_ref_cached() had\nalready inserted into pgStatEntryRefHash.\n\nMissing this cleanup would leave a backend with a stale local cache\nentry whose entry_ref points to a NULL shared_stats. If\npgstat_gc_entry_refs() runs with this reference still around, it would\ncrash due to a pointer dereference.\n\nThe local reference is now removed before removing the shared entry,\nthe order being sensitive to pending interrupts.\n\nOversight in 8191e0c16a03.\n\nAuthor: Niall Newman \nDiscussion: https://postgr.es/m/2FDAA194-9CF3-4FD7-A450-F1A4BEB125F6@turacolabs.com\nBackpatch-through: 15", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "5ade9944d67cb53bc0e7eab3b4edd986625e8c76", "date": "2026-08-07T01:26:40-04:00", "subject": "docs: fix grammar in detail page spec", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "bf80a4c2d238834073b09c231294c4479157f16e", "date": "2026-08-07T14:52:09+09:00", "subject": "Fix shared refcount leak in pgstat entry acquisition", "body": "pgstat_acquire_entry_ref() does an entry acquisition with the following\nactions:\n- Increment of a shared ebtr's refcount.\n- Release dshash partition lock.\n- Update local entry_ref fields.\n\nWith a pending interrupt, the LWLock acquired by dshash_release_lock()\nin the second step could fail, leaving around a local entry_ref with\nshared_stats set to NULL and the refcount permanently inflated as\npgstat_release_entry_ref() skips the refcount decrement is shared_stats\nis not set.\n\nThe actions are now reordered so as the lock is released after the local\nentry_ref fields are populated.\n\nThe consequence of this problem is a minor leak. It can only happen if\nan interrupt is processed when a dshash lock is released, which is a\nvery narrow window (very unlikely), so no backpatch is done.\n\nAuthor: Niall Newman \nDiscussion: https://postgr.es/m/2FDAA194-9CF3-4FD7-A450-F1A4BEB125F6@turacolabs.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "57f3847734df908c0dfcdc4b51da7aa448f51d76", "date": "2026-08-07T14:22:59+08:00", "subject": "types(runtime-vapor): improve custom directive types", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ccfa6c9644132d95c09dd6512643493fb5081c6a", "date": "2026-08-07T16:12:47+09:00", "subject": "Remove stale comment on ri_FastPathCheck()", "body": "The note claiming ri_FastPathCheck() is only used by the ALTER TABLE\nvalidation path is no longer accurate. Rather than reword it, remove\nit; the block comment above RI_FKey_check()'s call to the function\nalready documents when each path is taken, and a duplicated\ncross-reference here would only drift out of date again.\n\nReported-by: Noah Misch \nReviewed-by: Ayush Tiwari \nDiscussion: https://postgr.es/m/20260705210533.ee.noahmisch@microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "467ad033bff337a2d980921c640509b6ff2191fe", "date": "2026-08-07T15:19:23+08:00", "subject": "fix(compiler-vapor): fold falsy hidden bindings into static templates", "body": "", "style": "conventional", "type": "fix", "scope": "compiler-vapor", "breaking": false} {"repo": "vue-core", "sha": "73496879247db88c8a801bd08006f0943994b505", "date": "2026-08-07T15:20:49+08:00", "subject": "chore(types): use bundler resolution for vapor dts tests", "body": "", "style": "conventional", "type": "chore", "scope": "types", "breaking": false} {"repo": "ant-design", "sha": "790e7625ef565d0df2b36d2ce5787cb65531642f", "date": "2026-08-07T15:26:24+08:00", "subject": "feat: support custom components in Pagination", "body": "* feat(Pagination): support custom size changer render\n\n* refactor(Pagination): use components for custom size changer\n\n* refactor(Pagination): refine custom size changer API", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "9952224fd5c8cac548f3b2ff05d1d66944743838", "date": "2026-08-07T15:27:14+08:00", "subject": "chore(compiler-sfc): satisfy lint rules for template imports", "body": "", "style": "conventional", "type": "chore", "scope": "compiler-sfc", "breaking": false} {"repo": "postgres", "sha": "edc92e0ff9572e41dc2d2fc14fbab313fa9d7bc5", "date": "2026-08-07T16:30:21+09:00", "subject": "Restrict RI fast-path FK check to btree referenced indexes", "body": "The RI fast-path check probes the referenced index directly and, for\nsingle-column keys, uses SK_SEARCHARRAY. Both assume the index is a\nbtree. A comment claimed \"PK indexes are always btree\", but a foreign\nkey's referenced index need not be a primary key:\ntransformFkeyCheckAttrs() accepts any unique (or, for temporal keys,\nexclusion) index, so an out-of-tree access method advertising\namcanunique could supply a non-btree index reachable by the fast path.\n\nAdd pk_index_is_btree to RI_ConstraintInfo, set from the referenced\nindex's access method when the constraint is loaded, and make\nri_fastpath_is_applicable() return false for non-btree indexes so such\nconstraints fall back to the SPI path.\n\nReported-by: Noah Misch \nReviewed-by: Ayush Tiwari \nDiscussion: https://postgr.es/m/20260705210533.ee.noahmisch@microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f68f47fe7d66a475d934823624013bcb7fae6f01", "date": "2026-08-07T16:34:00+09:00", "subject": "Begin RI fast-path index scan under the switched user id", "body": "ri_FastPathCheck() called index_beginscan() before switching to the\nreferenced relation's owner for the permission check and probe. For\nbtree this has no consequence, but starting the scan before the user id\nswitch sets a poor example for code dealing with out-of-tree access\nmethods, whose beginscan could observe the wrong user id. Move\nindex_beginscan() after the SetUserIdAndSecContext() call.\n\nReported-by: Noah Misch \nReviewed-by: Ayush Tiwari \nDiscussion: https://postgr.es/m/20260705210533.ee.noahmisch@microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "57fea001d154e7dd8d5d74d3082731f1dcfd31be", "date": "2026-08-07T15:37:01+08:00", "subject": "docs(static-deploy): update new platform rules", "body": "", "style": "conventional", "type": "docs", "scope": "static-deploy", "breaking": false} {"repo": "angular", "sha": "4d985a179e66428d46e60b4622435f88c39e1d2b", "date": "2026-08-07T09:58:22+02:00", "subject": "fix(core): incorrect loop in defer blocks", "body": "Fixes that we were running a `for in` loop on an array and silently doing the wrong thing.\n\nFixes #70113.\n\n(cherry picked from commit 565dbb2fe5406f2a86d43a5b3c6993434b7124a7)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "565dbb2fe5406f2a86d43a5b3c6993434b7124a7", "date": "2026-08-07T09:58:22+02:00", "subject": "fix(core): incorrect loop in defer blocks", "body": "Fixes that we were running a `for in` loop on an array and silently doing the wrong thing.\n\nFixes #70113.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "tmux", "sha": "10048f6579209fd1ca18ff6c5436628dec012371", "date": "2026-08-07T08:10:53Z", "subject": "Only unzoom in the floating-split case, modal panes take care of themselves.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "7d9d43db16260bf2cb90119161bccaddc3acfd62", "date": "2026-08-07T16:24:13+08:00", "subject": "fix(vapor): handle markerless slot hydration anchors", "body": "", "style": "conventional", "type": "fix", "scope": "vapor", "breaking": false} {"repo": "postgres", "sha": "acdfaee1b947776ec381a4756f01979d31316774", "date": "2026-08-07T17:25:42+09:00", "subject": "Fire fast-path FK batches inside the deferred trigger loop", "body": "AfterTriggerFireDeferred() drained its queued events in a loop, then\ncalled FireAfterTriggerBatchCallbacks() once after the loop to flush\nthe RI fast-path FK-check batches the fired triggers accumulated.\n\nA batch callback runs user-supplied cast or equality functions, whose\nDML can queue further deferred trigger events. Because the flush ran\nafter the loop had exited, such an event was left in\nafterTriggers.events with nothing to fire it, since\nAfterTriggerFireDeferred() is the last drainer at commit. The deferred\ncheck was skipped and a row violating the constraint committed.\n\nMove the flush inside the loop, after afterTriggerInvokeEvents(), and\ndrop the \"all fired\" break, so afterTriggerMarkEvents() re-checks after\neach flush and fires events a flush queued at the correct time. The\nother FireAfterTriggerBatchCallbacks() callers leave any queued event\nfor the eventual commit-time AfterTriggerFireDeferred(), so only the\ncommit-time firing, which has no later drainer, lost events.\n\nReported-by: Noah Misch \nReviewed-by: Ayush Tiwari \nDiscussion: https://postgr.es/m/20260705210533.ee.noahmisch@microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "cec48686a0d781806a262c548adb7d0fb06868fc", "date": "2026-08-07T17:30:06+09:00", "subject": "Drain pending asynchronous requests during ExecReScanAppend.", "body": "The logic for asynchronous Append assumes that pending requests made for\nsubplans of an Append are drained during ExecReScanAppend. To ensure\nthat, commit 9e283fc85 modified postgresReScanForeignScan to drain such\na request if any, but failed to take into account that if such a request\nwas made for a subplan that is re-scanned with parameter changes or\npruned in the next round by runtime pruning, the postgres_fdw callback\nfunction is called after ExecReScanAppend or never called, respectively.\nThis would cause such a request to remain even after ExecReScanAppend,\nleading to incorrect results, an infinite loop, or an assertion failure.\n\nTo fix, modify ExecReScanAppend to, for each of the pending requests,\ngive the FDW a chance to drain that request using the existing\nForeignAsyncConfigureWait/ForeignAsyncNotify callback functions. This\nmakes the change made to postgresReScanForeignScan useless, so remove it\nas well.\n\nBack-patch to v14 where asynchronous Append was added.\n\nReported-by: Alexander Korotkov \nCo-authored-by: Alexander Korotkov \nCo-authored-by: Gleb Kashkin \nCo-authored-by: Etsuro Fujita \nReviewed-by: Alexander Pyhalov \nReviewed-by: Gleb Kashkin \nDiscussion: https://postgr.es/m/CAPpHfduMOTnV5Zj2KGJ7zanL_10QvccZHtPUaDfJvBhsh9axnQ%40mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "dea1702503cbc2ac91d1f8235b1912cb119afe16", "date": "2026-08-07T14:31:54+06:00", "subject": "fix(core): correctly scan providers when overriding a lazy-loaded module", "body": "The lazy-load short-circuit in scanner.ts relies on the inserted flag\nto skip re-scanning already-loaded modules. NestContainer.replaceModule()\nwas incorrectly returning inserted: false, causing overridden lazy\nmodules (via testing module builder) to be skipped entirely and their\nproviders never scanned. replaceModule() always creates a new module\nreference that must be scanned, so it now correctly returns\ninserted: true.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "d34f9505353dbf3888b3f7000ce2b666bac928d7", "date": "2026-08-07T17:34:08+09:00", "subject": "Fix indentation issue introduced by commit 291a4bd2ca", "body": "Noticed before the commit reached koel.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0ec3f048bfc15c8eb9933e8228b847593389da1b", "date": "2026-08-07T17:39:24+09:00", "subject": "Handle nullable referenced key in RI fast-path check", "body": "The RI fast-path FK check asserted that the referenced key is never\nNULL, in ri_FastPathFlushArray() and in recheck_matched_pk_tuple().\nThat holds for a primary key, but a foreign key may reference any\nunique column, and a UNIQUE column is nullable.\n\nThe assertion is reachable under READ COMMITTED. ri_LockPKTuple()\nlocks the matched PK tuple with TUPLE_LOCK_FLAG_FIND_LAST_VERSION, so\nwhen a concurrent transaction commits a key-changing UPDATE while the\ncheck waits, the lock follows the update chain to the latest version.\nIf that version now has NULL in the referenced column, the fast path\nreaches the assert; in a non-assert build it would compare against the\nNULL and treat it as a match.\n\nA NULL referenced key cannot equal any (non-null) FK value, so treat it\nas no match and let the ordinary foreign-key violation be raised. This\nmatches the SPI path, whose requalifying \"pkatt = $n\" evaluates to NULL\nfor such a row, so the row is not returned and the check reports a\nviolation.\n\nReported-by: Noah Misch \nReviewed-by: Ayush Tiwari \nDiscussion: https://postgr.es/m/20260705210533.ee.noahmisch@microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a4a544032a0ba7a3bb0aea710f9f47b6cd2885db", "date": "2026-08-07T04:41:37-04:00", "subject": "feat(cwd)!: :lcd! (bang), rearrange :bcd/:lcd/… scope precedence #41194", "body": "Problem:\n- buf-local CWD scope is lower priority than :lcd, which is weird.\n ```\n win > buf > tab > global\n ```\n- No way to clear current CWD at a given scope.\n\nSolution:\n- Rerrange scope precedence to:\n ```\n buf > win > tab > global\n ```\n- Introduce \"bang\" variants (`:bcd!`/`:lcd!`/`:tcd!`) which clears the\n local CWD for the given scope.", "style": "conventional", "type": "feat", "scope": "cwd", "breaking": true} {"repo": "vue-core", "sha": "3c4e5b7916efc08268687c03f8fb305d23c1370e", "date": "2026-08-07T17:00:46+08:00", "subject": "fix(vapor): preserve vdom slot fallback position", "body": "", "style": "conventional", "type": "fix", "scope": "vapor", "breaking": false} {"repo": "neovim", "sha": "a8da01f8e610aebcf33df460174de78be222bc25", "date": "2026-08-07T02:09:48-07:00", "subject": "fix(decor): highlight without 'hl_eol' bleeds into wrap gaps #41160", "body": "Problem:\n'linebreak' filler and 'breakindent'/'showbreak' padding are screen\ncells with no buffer character behind them, yet a decoration draws over\nthem whether or not it asked to cover such cells. A highlight bounded to\nits text then paints a tail out to the edge of the row, most visible on\ninline code spans from plugins. That same highlight already leaves the\ncells past the end of a line alone, so it treats identical cells two\ndifferent ways.\n\nSolution:\nOnly a decoration with 'hl_eol' draws the gaps, which is what the flag\nalready means at the end of a line. A full-width background such as a\nfenced code block sets it and still covers them. Classic :syntax has no\nsuch flag and is unchanged.", "style": "conventional", "type": "fix", "scope": "decor", "breaking": false} {"repo": "vue-core", "sha": "8f89be88faaadd9e1025f7dc1430b3854341fb72", "date": "2026-08-07T17:20:56+08:00", "subject": "chore(playground): default to TypeScript 6", "body": "fix #15217", "style": "conventional", "type": "chore", "scope": "playground", "breaking": false} {"repo": "curl", "sha": "656e371fcd39ed6d9aed4746e1a8cc6522722e6b", "date": "2026-08-07T11:25:15+02:00", "subject": "fixup c-comments", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "18eb0a7e787b9fac3049ef3dad0e845d2bd940a4", "date": "2026-08-07T18:27:49+09:00", "subject": "fix: prevent ASI hazard in `no-unused-labels` autofix", "body": "* fix: prevent ASI hazard in `no-unused-labels` autofix\n\nWhen removing an unused label whose body starts with `(`, `[`, `` ` ``,\n`/`, `+`, or `-`, the autofix could silently change program semantics by\nmerging the body with the preceding statement via ASI.\n\ne.g. `foo()\\nLABEL: [1,2,3].forEach(x => x)` was incorrectly fixed to\n`foo()\\n[1,2,3].forEach(x => x)`, which parses as `foo()[1,2,3].forEach(...)`.\n\nThe fix skips autofixing in those cases when the token before the\nLabeledStatement is not a safe ASI boundary (`;`, `{`, `}`, `:`).\n\n* fix: remove } from safe ASI boundary tokens in no-unused-labels\n\n} is not always a safe ASI boundary — e.g. a function expression ending with } can still be continued by [ on the next line:\n\n const foo = function() {}\n LABEL: [1, 2, 3].forEach(x => x)\n\nRemoving LABEL: here would produce `function() {}[1, 2, 3]` which is a member access, not two separate statements.\n\n* Apply suggestion from @mdjermanovic\n\ngetFirstToken(node.body) always returns a token.\n\nCo-authored-by: Milos Djermanovic \n\n* Update tests/lib/rules/no-unused-labels.js\r\n\r\nThis test was already being caught by the isStaticTemplateLiteral check above\n\nCo-authored-by: Milos Djermanovic \n\n* Update tests/lib/rules/no-unused-labels.js\n\nCo-authored-by: Milos Djermanovic \n\n---------", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "2f4059a86147c51f4a5ce71f0ee3e819f2e48f99", "date": "2026-08-07T17:29:32+08:00", "subject": "feat: support progress ring for FloatButton.BackTop", "body": "* feat: add BackTop progress ring\n\n* fix(BackTop): sync progress ring with target scroll state\n\n* test(BackTop): cover null progress target\n\n* fix(BackTop): handle document targets in isolated environments\n\n* fix(FloatButton): correct BackTop progress ring rendering\n\n* test(FloatButton): update progress ring demo snapshot\n\n* fix lint error\n\n* test(utils): remove obsolete getScroll regression test\n\n* fix(utils): guard missing document constructor\n\n* docs(FloatButton): improve progress ring demo\n\n* docs(FloatButton): show progress ring immediately in demo\n\n* demo: update demo\n\n* add useScroll hook\n\n* fix(FloatButton): render progress ring on button border\n\n* refactor: address BackTop progress review feedback\n\n* fix(utils): guard missing document constructor\n\n* test(FloatButton): ignore progress ring demo snapshots\n\n---------", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "babel", "sha": "7b8e231d4e41b69f3551a2c2253b74794675624a", "date": "2026-08-07T17:30:21+08:00", "subject": "Clean up Prettier E2E test script", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "d8b12623a364be085b5cdddfdc8b13af3a6fef3f", "date": "2026-08-07T17:30:59+08:00", "subject": "fix: Define child node types correctly", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "openssl", "sha": "723e67bb0d66208730344e9d56db32b75a44cf06", "date": "2026-08-07T18:34:00+09:00", "subject": "Fix X509_ATTRIBUTE_set1_data() for BIT STRING attribute values", "body": "Commit 9044e5f42556 (\"Convert internal use of ASN1_STRING_set and\nASN1_STRING_length\") replaced ASN1_STRING_set() with the new\nASN1_STRING_set1_data() in X509_ATTRIBUTE_set1_data(). The new setter\nintentionally rejects strings of type V_ASN1_BIT_STRING, so creating an\nattribute value of type BIT STRING from raw bytes with an explicit\nlength now fails with ASN1_R_ILLEGAL_BITSTRING_FORMAT.\n\nThis broke PKCS8_add_keyusage() and therefore PKCS#12 creation with a\nnon-zero keytype (e.g. 'openssl pkcs12 -export -keyex' or '-keysig'),\nas well as the public X509_ATTRIBUTE_create_by_*() and\nX509at_add1_attr_by_*() APIs when used with BIT STRING values supplied\nas bytes plus an explicit length.\n\nRestore the previous behaviour by dispatching to ASN1_BIT_STRING_set1()\nwith zero unused bits when attrtype is V_ASN1_BIT_STRING, keeping\nASN1_STRING_set1_data() for the other types: the signature of this\nexplicit-length raw-data API provides no way to specify another\nunused-bit count, so the data is taken as complete octets, which\npreserves the previous encoding for supplied raw-byte values. Document\nthis in the X509_ATTRIBUTE manual page.\n\nFixes #32234\n\nAssisted-by: Cline:Qwen3.8 Max\nReviewed-by: Richard Levitte \nReviewed-by: Bob Beck \nMerge-date: Wed Aug 19 15:44:48 2026\nMerged-from: https://github.com/openssl/openssl/pull/32238", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "bd482e23d0001805ae7c6a9a7f65cc0aa0fb6ced", "date": "2026-08-07T18:34:11+09:00", "subject": "Add regression tests for BIT STRING X509 attributes", "body": "Test that X509_ATTRIBUTE_create_by_NID() accepts a BIT STRING value\nsupplied as raw bytes with an explicit length, as done for the keyUsage\nattribute by openssl pkcs12 -export -keyex and -keysig via\nPKCS8_add_keyusage(). Both KEY_EX (0x10) and KEY_SIG (0x80) are exercised\nat this common boundary, and the resulting attribute is checked for the\nexpected type, value, length, and zero unused bits.\n\nAlso pass -keyex to an existing PKCS12 export test so the reported CLI\npath is covered end to end.\n\nRegression test for https://github.com/openssl/openssl/issues/32234\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Richard Levitte \nReviewed-by: Bob Beck \nMerge-date: Wed Aug 19 15:44:50 2026\nMerged-from: https://github.com/openssl/openssl/pull/32238", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "2a5d6a5502135a0dc8cc2c074a3913ba34c84440", "date": "2026-08-07T11:41:16+02:00", "subject": "curl_url_set.md: expand the CURLU_NO_AUTHORITY description", "body": "Closes #22515", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "6e783750e2d4080fc707b38cd10d92edf2fa2d5f", "date": "2026-08-07T11:41:16+02:00", "subject": "curl_url_set.md: expand the CURLU_NO_AUTHORITY description", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "bae17e3191980eba5288266d54e3fdbced5f49a4", "date": "2026-08-07T17:41:30+08:00", "subject": "fix: Load ESM fixture options default exports", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "07457549a3c545514ea310646960f5878966718e", "date": "2026-08-07T11:43:04+02:00", "subject": "hostname", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "0e3ec51574dad14c4db5478b726623e7601a180e", "date": "2026-08-07T15:53:57+06:00", "subject": "refactor: replace non-null assertions with default values", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "0b547d25809b5f2ee9d5ae3383bd2ec690ea573c", "date": "2026-08-07T17:56:27+08:00", "subject": "fix(runtime-vapor): preserve hydrated vdom slot fallback position", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "366227d2482f8a83be6f2920a63072c1fac05e79", "date": "2026-08-07T12:02:26+02:00", "subject": "smoother wording", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "14ef8d13f4e1756800a54b9e9221a7265c59916c", "date": "2026-08-07T18:15:02+08:00", "subject": "fix(compiler-vapor): preserve optional chaining in cached member expressions", "body": "close #15226", "style": "conventional", "type": "fix", "scope": "compiler-vapor", "breaking": false} {"repo": "nestjs", "sha": "07410071df6951454e9385d035188178273868ed", "date": "2026-08-07T16:15:15+06:00", "subject": "style: format union types", "body": "", "style": "conventional", "type": "style", "scope": "", "breaking": false} {"repo": "tmux", "sha": "b6c709307ee363e259300a04c50af6fea0abdf4d", "date": "2026-08-07T11:21:10+01:00", "subject": "Dither alpha in SIXEL output", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "cb3db5dd752871f0cdd7854d4083c3e78b08e000", "date": "2026-08-07T18:29:55+08:00", "subject": "site: homepage tweaks — dynamic contributor avatars, email fix & theme keyboard a11y", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "3442c1f26f3113a3df29da871e7ec1bc2a4945cf", "date": "2026-08-07T11:34:16+01:00", "subject": "Fix underlay bookkeeping bug.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "0105f15201743fe80a9af105c2117510d3b6fdf8", "date": "2026-08-07T05:39:13-05:00", "subject": "fix(window): statusline can silently fail to make room #41201", "body": "", "style": "conventional", "type": "fix", "scope": "window", "breaking": false} {"repo": "ant-design", "sha": "60e6bf4867ed9c81c614b29ce046d7104c786188", "date": "2026-08-07T18:44:13+08:00", "subject": "test: fix missing DOM constructors in image test environment", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "bc9d27b0bc04307a1e1c2b81713ac7372e46da95", "date": "2026-08-07T06:06:23-05:00", "subject": "fix(window): 'winbar' in single-row win overlaps global 'statusline' #41188", "body": "", "style": "conventional", "type": "fix", "scope": "window", "breaking": false} {"repo": "angular", "sha": "f746463caa5d832a35b3ce389fc3f7af2af62082", "date": "2026-08-07T14:09:22+03:00", "subject": "fix(devtools): breadcrumbs missing nav arrows", "body": "Show the missing nav arrows when you select a deeply nested component where\nthe breadcrumbs path is longer than the container. The nav arrows used to\nappear only when the split is resized. Add some other minor improvements.", "style": "conventional", "type": "fix", "scope": "devtools", "breaking": false} {"repo": "neovim", "sha": "2e0a5a596aeb3d3f31a106604bf9274288c92989", "date": "2026-08-07T07:44:19-04:00", "subject": "fix(terminal): spawn in effective CWD #41211", "body": "Problem:\n`:terminal` does not respect the invocation-time CWD.\nThis wasn't noticeable with `:lcd` because the window-local CWD gets\napplied to the new terminal buffer. But it is noticeable with `:bcd`.\n\nSolution:\nSpecify `cwd` in the job spec.", "style": "conventional", "type": "fix", "scope": "terminal", "breaking": false} {"repo": "vitest", "sha": "62c795d63f333dd68eedd5f050450da01906a48b", "date": "2026-08-07T16:52:23+05:00", "subject": "fix(browser): resolve `connectTimeout` from the project config (fix #10879)", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "vitest", "sha": "3850323e2123bd5cc8972436de79a9e80f395340", "date": "2026-08-07T13:52:56+02:00", "subject": "fix(vm): stop retaining every finished test file in vm pool workers", "body": "", "style": "conventional", "type": "fix", "scope": "vm", "breaking": false} {"repo": "angular", "sha": "9711e86f2aae459d0395943eb112ead229952372", "date": "2026-08-07T14:56:35+03:00", "subject": "refactor(devtools): collapse injected services section in the side pane by default", "body": "Collapse the section since it takes a significant portion of the vertical space,\nwhereas most of the times users are exploring the properties leading them to scroll.\n\nAddiotionally, introduce some other minor UI improvements.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "vitest", "sha": "01298e5b355df8c405e656e8a9b07e1d57cd05cf", "date": "2026-08-07T14:11:47+02:00", "subject": "feat(vm): support `require(esm)` in vm pools", "body": "", "style": "conventional", "type": "feat", "scope": "vm", "breaking": false} {"repo": "neovim", "sha": "87a8bbabcf7bb8b1c33de21df1529309697dd698", "date": "2026-08-07T15:33:14+03:00", "subject": "fix(highlight): explicitly define `TabLineSel` bg/fg #41196", "body": "Problem: `TabLineSel` by default is defined with only bold attribute.\n This makes selected tabpage stand out relatively well, but when used\n in custom tabline it will be combined with `TabLineFill` (which is\n reserved for the \"non tabpage\" part of the tabline).\n Plus the behavior is not consistent with the default 'tabline' which\n combines it with `Normal`.\n\nSolution: Explicitly define background and foreground colors to come\n from `Normal` to ensure the current default 'tabline' behavior even in\n custom tabline.\n The more semantically driven solution here is to link `TabLineSel` to\n `StatusLine`. This makes it more aligned with how `TabLine` is linked\n to `StatusLineNC`. But the current \"bold text with `Normal` colors\" is\n argued to look better as a default.\n\n Also make sure that 'notermguicolors' highlighting in default and\n custom tablines is the same: the current tabpage does not have\n underline and is not bold.", "style": "conventional", "type": "fix", "scope": "highlight", "breaking": false} {"repo": "vitest", "sha": "d487659cda2d7eaa243243807137452236c1eb50", "date": "2026-08-07T14:34:16+02:00", "subject": "test: deflake tests sharing the watch fixture", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vitest", "sha": "1216b5af96065b300df051227d1641f80b4106a6", "date": "2026-08-07T14:34:48+02:00", "subject": "fix: don't lose worker output on teardown, deflake timing-sensitive tests", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vite", "sha": "e26053b9cb51027592eb7f5d3d5d04d0e56a0b47", "date": "2026-08-07T20:52:10+08:00", "subject": "test: repro for css sourcemap sources resolved against the wrong base", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "babel", "sha": "1e5eb959ef9b6cc519978f8144a50ac3c2a4884f", "date": "2026-08-07T09:08:10-04:00", "subject": "Handle null in `getAll{Prev,Next}Siblings`", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "7af8755aed8b5013e0e81350e396553b5db6a0cd", "date": "2026-08-07T15:15:37+02:00", "subject": "perf(browser): serve framework assets as immutable", "body": "", "style": "conventional", "type": "perf", "scope": "browser", "breaking": false} {"repo": "vitest", "sha": "c67d296f42f93ec888ff148e821877194969cea9", "date": "2026-08-07T15:21:55+02:00", "subject": "feat: report the duration breakdown as percentages", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "babel", "sha": "d3d4651caef80c2b01c45e1e7689ac6121d7e2d5", "date": "2026-08-07T15:47:16+02:00", "subject": "Update node flags list", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7c25cdb1ebf64b1e1c9053ca27f9552c9a528a12", "date": "2026-08-07T09:59:55-04:00", "subject": "Only clear VACUUM's read stream strategy once in failsafe mode", "body": "112c2683807b4d690 restored failsafe vacuum's abandonment of a buffer\naccess strategy by clearing the ReadBuffersOperations' strategy\nreferences. But it did so in lazy_scan_heap()'s main loop, meaning it\nlooped through all the ReadBuffersOperations once per block after\nfailsafe was engaged. Track it with a local flag and clear the strategy\nonly once.\n\nReported-by: Melanie Plageman \nDiscussion: https://postgr.es/m/CAAKRu_Zse14nSNeCgtnE1LUAH8Of7OmYR%2BCc3O_DAzxt3m6T-g%40mail.gmail.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "7e53d3ba4df39003ca3c8a98637943348fec76ae", "date": "2026-08-07T10:10:21-04:00", "subject": "fix(cwd): keep buffer-local dir when re-editing #41215", "body": "Problem:\nRe-editing a buffer (`:edit!`, re-reading a dir.lua buffer, etc.) drops\nits `:bcd` directory, so the CWD falls back to the global one. Whereas\nother buffer-local state (`b:` vars, local options) survives a reload.\n\nSolution:\nDon't clear buf dir in `buf_freeall()`; `do_ecmd()` calls that when\nreloading/re-editing. `free_buffer_stuff()` still clears them when\na buffer is freed or reused for another file.", "style": "conventional", "type": "fix", "scope": "cwd", "breaking": false} {"repo": "babel", "sha": "36057c706ce82a0d0a1c8a5f7721e6fcae4ab1b6", "date": "2026-08-07T19:49:57+05:30", "subject": "fix(babel-register): preserve app-managed graceful shutdown", "body": "", "style": "conventional", "type": "fix", "scope": "babel-register", "breaking": false} {"repo": "ant-design", "sha": "ee37db23e916742685091efa0c87546ede54d666", "date": "2026-08-07T22:21:37+08:00", "subject": "refactor(DatePicker): avoid semantic props type assertion", "body": "", "style": "conventional", "type": "refactor", "scope": "DatePicker", "breaking": false} {"repo": "tmux", "sha": "ae0fa8068c3d729531dd53a15ed721ea0894bc8c", "date": "2026-08-07T16:09:00+01:00", "subject": "Preserve Kitty image crop width after cell rounding.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "311e66df9cc857dcbb02024270f5130815682d63", "date": "2026-08-08T00:09:05+09:00", "subject": "Fix hot standby accepting connections too early after a crash reset", "body": "Commit b53b88109f9 made the postmaster maintain reachedConsistency in\naddition to the startup process. Since the startup process is forked\nfrom the postmaster, it begins life holding whatever value the\npostmaster last set.\n\nOn a crash reset the postmaster re-forks the startup process while its\nown copy still says true: it clears that copy only on receipt of\nPMSIGNAL_RECOVERY_STARTED, which the replacement process cannot send\nbefore it exists. The replacement therefore starts out believing the\ndatabase is already consistent.\n\nCheckRecoveryConsistency() then skips the minRecoveryPoint comparison\naltogether, so hot standby is announced at redo start while replay may\nbe arbitrarily far behind minRecoveryPoint. Read-only connections are\naccepted and answer from heap pages that were flushed ahead of the\nreplay position, returning wrong results with no error raised. The\nsame branch also runs XLogCheckInvalidPages() and\nCheckTablespaceDirectory(), which are skipped as well, and\nlog_invalid_page() treats page references that are normal before\nconsistency as a PANIC.\n\nFix by clearing reachedConsistency in InitWalRecovery(), so that a\nstartup process never depends on the value it inherited. The\npostmaster's own copy is deliberately left alone: forked backends read\nit to choose the \"not yet accepting connections\" errdetail, and it\nconverges once the new startup process sends PMSIGNAL_RECOVERY_STARTED\nand, on reaching minRecoveryPoint, PMSIGNAL_RECOVERY_CONSISTENT.\n\nSuccessive crash resets alternate. A startup process that skips the\nbranch never sends PMSIGNAL_RECOVERY_CONSISTENT, so the postmaster's\ncopy stays false and the next reset forks a process holding the correct\nvalue; that pass reaches consistency properly, which sets the\npostmaster's copy back to true and re-arms the problem for the reset\nafter it. Roughly every other crash reset is therefore affected, not\njust the first one. EXEC_BACKEND builds are unaffected, as\nreachedConsistency is not carried in BackendPa", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "69ee9cf7f4b8bdb984eca8b52dafa15af8cd6a3c", "date": "2026-08-07T16:13:07+01:00", "subject": "Ordinary text writes replaced the entire grid cell, dropping GRID_FLAG_IMAGE and releasing the mage reference. Deferred pane redraws explain the intermittent “image never appears” behavior.", "body": "Fixed:\n - Preserves image metadata while updating the text underlay in the collected ASCII path: grid.c:779\n - Handles the single-cell/non-ASCII path and schedules an image redraw: screen-write.c:2631\n - Uses the requested if/else for new_id.\n - Adds a deterministic regression reproducing point (2): regress/image-support.sh:240\n\n This matches Kitty’s requirement that normal text operations must not affect graphics; graphics require\n explicit deletion. Kitty graphics protocol", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "7d32e7f3798815b8d515707940d9809867f3b117", "date": "2026-08-07T16:18:30+01:00", "subject": "Fix bad merge.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "817d75e25bb61b33f8976279c040ffa17f563639", "date": "2026-08-07T17:22:46+02:00", "subject": "tests/server/tftpd: replace duplicate code with `open_udp_sock()` call", "body": "Follow-up to 7f6a75664f9fb45390193e492ee3361fb795d098 #22506\n\nCloses #22517", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8b651cdcc21fa3e1d3715a382961937e8376c7e0", "date": "2026-08-07T16:26:11+01:00", "subject": "DTLS 1.3 Signal the listener notifier on accept queue push", "body": "The listener's notifier was signalled when the demux routed a datagram into a\nconnection's receive queue, but not when dtls_listener_drive_pending() moved\na completed connection onto the accept queue - even though the latter is\nprecisely the readiness event a thread waiting in SSL_poll() for\nSSL_POLL_EVENT_IC is waiting for.\n\nUsually the injection signal covers for this. Every connection reaching the\naccept queue got there because a datagram was demuxed into its receive queue,\nso the waiter is woken by that, ticks the listener during its own readout,\nand finds the connection. The exception is the window in which the injection\nand the queue push straddle a waiter registering, since signalling only\nhappens if there is a waiter at the time:\n\n 1. Accept thread A polls the listener for SSL_POLL_EVENT_IC. Its readout\n ticks the listener, finds nothing, and it decides to block. It is not a\n registered waiter yet.\n 2. Worker thread B polls one of its own connections, which also ticks the\n listener. The pump reads a client's final ClientHello and injects it into\n that pending connection's queue. There are no waiters, so nothing is\n signalled.\n 3. A enters the blocking section. Its re-check runs without ticking, so it\n sees only the accept queue, still empty, and blocks.\n 4. B's tick reaches dtls_listener_drive_pending(), which completes the\n connection against the buffered ClientHello and pushes it onto the accept\n queue.\n\nWith no signal at step 4, A sleeps on while a validated connection sits ready\nto be accepted, until an unrelated datagram makes the socket readable again.\nB consumed the only one in flight and the client is now waiting on the server,\nso on a quiet listener that means waiting for the client to retransmit its\nhandshake.\n\nFactor the signalling out of dtls_listener_packet_handler() into a helper and\ncall it from both places. Both call sites already hold the listener mutex,\nwhich the helper requires.\n\nAssisted-by: Claude Code", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "87e65c6079fab1ba556fbc1d358c2af455ac720c", "date": "2026-08-07T16:26:20+01:00", "subject": "DTLS 1.3 Honour the DTLS retransmit timer in SSL_poll", "body": "When SSL_poll() decides to block it computes a wakeup deadline from the\nper-object event timeout, so that timer driven work is not delayed by the\nwait. This was done for QUIC objects but not for DTLS connections, whose\ntimeout is the handshake retransmission timer.\n\nThe consequence was that a poll with no user timeout would sleep until a\ndatagram arrived, straight through the point at which the connection should\nhave retransmitted. If the peer had itself lost the message we were waiting\nfor, neither side would make progress.\n\nFold SSL_get_event_timeout() into the deadline for DTLS connections as the\nQUIC branch already does. It reports the DTLS timer already, so no new\nplumbing is needed there.\n\nWaking at the deadline is only useful if something then services the timer.\nUnlike the QUIC case, where the readout ticks the reactor and that handles\ntimeouts, the DTLS readout only pumped the listener's demux. Nothing\nretransmitted, and because an expired timer reports zero time remaining the\nrecomputed deadline would be \"now\" on every subsequent wait, turning the\nsleep into a spin. Call SSL_handle_events() from the readout as well, which\nis what a caller polling without SSL_POLL_FLAG_NO_HANDLE_EVENTS is asking\nfor, and which for a DTLS connection services that timer.\n\nBoth of those are the calls the documentation asks new code to use:\nDTLSv1_get_timeout(3) and DTLSv1_handle_timeout(3) each record that\nSSL_get_event_timeout(3) and SSL_handle_events(3) respectively supersede all\nof their use cases. The test asserts its precondition through\nSSL_get_event_timeout() for the same reason, which has the side benefit of\nusing the same call SSL_poll() uses to compute the deadline being tested.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Tomas Mraz \nMerge-date: Fri Aug 14 15:45:37 2026\nMerged-from: https://github.com/openssl/openssl/pull/32239", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "de4a1979169d61d4a45d44b5c4978beceec4a1f8", "date": "2026-08-07T16:26:20+01:00", "subject": "DTLS 1.3 Enter a blocking section when polling a DTLS listener", "body": "A DTLS listener added only the network socket to the poll set. Three things\nfollowed from that. The notifier was not in the poll set, so a thread blocked\nin SSL_poll() on the listener could not be woken by it. The thread never\ncalled ossl_dtls_listener_enter_blocking_section(), so cur_blocking_waiters\nstayed at zero - and since signalling is conditional on there being a waiter,\nno other thread even attempted to signal. And there was no re-check after\nregistering, so readiness arising between the readout deciding nothing was\nready and the wait actually starting was lost.\n\nPolling the socket alone is not sufficient, though not because a wakeup can\nbe missed outright. poll() reports whatever is currently sitting in the\nsocket buffer and returns immediately if there is any, so a thread cannot\nmiss a datagram just by being outside poll() when it arrives. What it can\nmiss is a datagram another thread has already taken. With several threads\npolling the one shared socket that happens constantly: an arriving datagram\nwakes all of them, only one gets it, and the rest find nothing. Any of them\ncan be the one that takes it, because SSL_read() on a connection pumps the\ndemux and SSL_poll() on a connection ticks the whole listener.\n\n 1. Accept thread A polls the listener for SSL_POLL_EVENT_IC. Its readout\n ticks the listener, finds nothing, and it decides to block.\n 2. A client's final ClientHello lands on the shared socket.\n 3. Worker thread B, polling one of its own connections, ticks the listener\n and is the one that takes the datagram. Its tick completes the pending\n connection and pushes it onto the accept queue. Signalling is attempted,\n but A never registered as a waiter, so nothing is signalled.\n 4. A reaches its poll, watching the socket alone. B drained it, so it is\n empty. A sleeps with a validated connection sitting on the accept queue,\n and nothing can wake it until an unrelated datagram arrives or its\n deadline expires.\n\nA's readout, ", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "fe3aa649456472f21921727a99d31761ca333594", "date": "2026-08-07T11:27:19-04:00", "subject": "feat(lsp): pass target buffer to `reuse_client` predicate #41163", "body": "Problem: The reuse_client predicate does not pass the target buffer,\npreventing decisions from being truly made per buffer.\n\nSolution: Pass the target buffer.", "style": "conventional", "type": "feat", "scope": "lsp", "breaking": false} {"repo": "tmux", "sha": "e5bcf59768bc029293a7068b738b574ff017404f", "date": "2026-08-07T16:39:28+01:00", "subject": "Clear text beneath SIXEL image placements.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "59597316a60b11e0cde262a9b55b3e326a754e29", "date": "2026-08-08T00:09:22+08:00", "subject": "fix(cwd): :bcd lost by nvim_win_set_buf, nvim_open_win #41218", "body": "Problem:\n:bcd (buffer-local directory) is not preserved after\n`nvim_open_win` or `nvim_win_set_buf`\n\nAnalysis:\nset_curbuf() ends with update_cwd(), which falls back to\nos_chdir(globaldir) when the target buffer has no b_localdir.\n\nSolution:\nPass kCtxKeepCwd to ctx_switch().", "style": "conventional", "type": "fix", "scope": "cwd", "breaking": false} {"repo": "angular", "sha": "ec5f4e86f0383e6b5f3ab276e65b58c117b1d4d3", "date": "2026-08-07T19:49:22+03:00", "subject": "docs: add code style in Antigravity file reference", "body": "(cherry picked from commit 144b90c907b69f88ce869b8929af0a81aba13b01)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "144b90c907b69f88ce869b8929af0a81aba13b01", "date": "2026-08-07T19:49:22+03:00", "subject": "docs: add code style in Antigravity file reference", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2df3cbc19dfecff99e166c42bdfb0ca2b49ba5be", "date": "2026-08-07T10:51:05-06:00", "subject": "Deprecate CMS_stream() and PKCS7_stream()", "body": "These set up the indefinite length streaming used by BIO_new_NDEF(), which\nnow obtains the content octet string through the ASN.1 callback, so they\nno longer return a streaming boundary and have no use outside the library.\nNeither has ever been documented, so there is no manual page to update.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Viktor Dukhovni \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 13:56:43 2026\nMerged-from: https://github.com/openssl/openssl/pull/32242", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "0a2676e54adc4a27343fe2fce3042f97c371b06c", "date": "2026-08-07T13:21:17-04:00", "subject": "fix(messages): :read starts a \"bufwrite\" progress #41219", "body": "Problem:\nfilemess() treats an empty suffix as \"a buffer write is starting\", but\nreadfile() calls it that way too. So \":read\" (and \":edit\", …) opens a\n`nvim.bufwrite \"\"` progress that is never completed.\n\nUsers of e.g. ghostty will see a stuck \"progress\" spinner.\n\nSolution:\nOnly `buf_write()` starts the progress, via `filemess_progress()`.", "style": "conventional", "type": "fix", "scope": "messages", "breaking": false} {"repo": "angular", "sha": "ed6b7d71d4870b4a782fde80c21fb8be5489b3ae", "date": "2026-08-07T21:00:26+03:00", "subject": "docs: update Angular Aria roadmap status", "body": "Update the roadmap to reflect that Angular Aria became stable in Angular v22 while development of new patterns continues.\n\n(cherry picked from commit 8e41e97dfea09ceba52eca2b7ce4c06196b09b04)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "8e41e97dfea09ceba52eca2b7ce4c06196b09b04", "date": "2026-08-07T21:00:26+03:00", "subject": "docs: update Angular Aria roadmap status", "body": "Update the roadmap to reflect that Angular Aria became stable in Angular v22 while development of new patterns continues.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3dd6e2df85da5bf9f740da497ecf721a3a4574a8", "date": "2026-08-07T11:55:37-07:00", "subject": "doc: FIPS provider is compatible with releases from 3.0 onwards", "body": "State in fips_module(7) that a FIPS provider built from a validated\nversion can be used with libcrypto and libssl from any supported\nrelease from OpenSSL 3.0 onwards, backward and forward, including\nfuture major release series, replacing the narrower wording that\nlimited this to the same major release series.\n\nThe downloads page reference in the same paragraph now points at\nhttps://openssl-library.org/source/, which www.openssl.org/source/\nredirects to.\n\nAligns the documentation with the OpenSSL Library FIPS Module\nSupport Policy.\n\nAssisted-by: Claude Code:claude-fable-5\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nReviewed-by: Richard Levitte \nMerge-date: Tue Aug 18 07:21:22 2026\nMerged-from: https://github.com/openssl/openssl/pull/32105\n(cherry picked from commit 090868518bbff01f8e9590444aed3c06a254a3b2)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "090868518bbff01f8e9590444aed3c06a254a3b2", "date": "2026-08-07T11:55:37-07:00", "subject": "doc: FIPS provider is compatible with releases from 3.0 onwards", "body": "State in fips_module(7) that a FIPS provider built from a validated\nversion can be used with libcrypto and libssl from any supported\nrelease from OpenSSL 3.0 onwards, backward and forward, including\nfuture major release series, replacing the narrower wording that\nlimited this to the same major release series.\n\nThe downloads page reference in the same paragraph now points at\nhttps://openssl-library.org/source/, which www.openssl.org/source/\nredirects to.\n\nAligns the documentation with the OpenSSL Library FIPS Module\nSupport Policy.\n\nAssisted-by: Claude Code:claude-fable-5\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nReviewed-by: Richard Levitte \nMerge-date: Tue Aug 18 07:21:22 2026\nMerged-from: https://github.com/openssl/openssl/pull/32105", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e46f25a535aacf1e4c3f4362a55d91043be09d86", "date": "2026-08-07T11:55:38-07:00", "subject": "README-FIPS: clarify provider and library cross-version compatibility", "body": "State that a FIPS provider built from any validated version may be\nused together with an OpenSSL library built from any supported release\nfrom OpenSSL 3.0 onwards, backward and forward, including future major\nrelease series.\n\nThe downloads page references in the updated paragraph now point at\nhttps://openssl-library.org/source/, which www.openssl.org/source/\nredirects to.\n\nAligns the documentation with the OpenSSL Library FIPS Module\nSupport Policy.\n\nAssisted-by: Claude Code:claude-fable-5\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nReviewed-by: Richard Levitte \nMerge-date: Tue Aug 18 07:21:24 2026\nMerged-from: https://github.com/openssl/openssl/pull/32105\n(cherry picked from commit 93a175f15adfad7f469f98cd2455cd65551e4eb6)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "93a175f15adfad7f469f98cd2455cd65551e4eb6", "date": "2026-08-07T11:55:38-07:00", "subject": "README-FIPS: clarify provider and library cross-version compatibility", "body": "State that a FIPS provider built from any validated version may be\nused together with an OpenSSL library built from any supported release\nfrom OpenSSL 3.0 onwards, backward and forward, including future major\nrelease series.\n\nThe downloads page references in the updated paragraph now point at\nhttps://openssl-library.org/source/, which www.openssl.org/source/\nredirects to.\n\nAligns the documentation with the OpenSSL Library FIPS Module\nSupport Policy.\n\nAssisted-by: Claude Code:claude-fable-5\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nReviewed-by: Richard Levitte \nMerge-date: Tue Aug 18 07:21:24 2026\nMerged-from: https://github.com/openssl/openssl/pull/32105", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "906639e34b13355f23af42e7d548b6ac01f4b591", "date": "2026-08-07T14:59:32-04:00", "subject": "build(vim-patch): vim9 check_typval_type() is n/a", "body": "v8.2.1551", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "neovim", "sha": "33a688f9fe3fd4ee475d819098a9d7a7dede8b5a", "date": "2026-08-07T16:10:02-04:00", "subject": "fix(messages): dangling progress-messages #41222", "body": "Problem:\nSome builtin features emit progress-messages which never \"complete\".\n- On failure, `:write` does not complete the progress-msg it started.\n- ins-completion never ends its \"Scanning...\" message.\n\nSolution:\n- `buf_write()` emits \"failed\" status on failure.\n- `ins_compl_stop()` ends the completion one.", "style": "conventional", "type": "fix", "scope": "messages", "breaking": false} {"repo": "neovim", "sha": "9e283f273b2bece25c1602fb4d26b1dff76ec105", "date": "2026-08-07T22:38:32+02:00", "subject": "test(terminal): unreliable \"spawns in CWD effective at time of invocation\"", "body": "FAILED …/terminal/ex_terminal_spec.lua @ 270: :terminal (fake shell) spawns in CWD effective at time of invocation\n Expected values to differ.\n Value:\n \"~/work/neovim/neovim/build/Xtest_xdg_terminal\"\n stack traceback:\n …/terminal/ex_terminal_spec.lua:276: in function <…/terminal/ex_terminal_spec.lua:270>", "style": "conventional", "type": "test", "scope": "terminal", "breaking": false} {"repo": "neovim", "sha": "eecc4b73ffe614a5fa79992d4890c234c0fe2f0f", "date": "2026-08-07T22:40:32+02:00", "subject": "fix(coverity): UNINIT, FORWARD_NULL", "body": "CID 652778: Uninitialized variables (UNINIT)\n /src/nvim/context.c: 353 in ctx_dirs_save()\n 347 if (curbuf->b_sfname != NULL && curbuf->b_fname == curbuf->b_sfname) {\n 348 cs->cs_save_sfname = xstrdup(curbuf->b_sfname);\n 349 }\n 350 do_autochdir();\n 351 char autocwd[MAXPATHL];\n 352 if (os_dirname(autocwd, MAXPATHL) == OK) {\n >>> CID 652778: Uninitialized variables (UNINIT)\n >>> Using uninitialized value \"*cwd\" when calling \"strcmp\".\n 353 cs->cs_apply_acd = strcmp(cwd, autocwd) == 0;\n 354 }\n 355 }\n 356 }\n 357\n 358 /// Restores the dir state saved by ctx_dirs_save(), undoing any chdir made while switched. The\n\n CID 652777: Null pointer dereferences (FORWARD_NULL)\n /src/nvim/context.c: 556 in ctx_switch()\n 550 cs->cs_target_win = wp->handle;\n 551 cs->cs_target_old_pos = wp->w_cursor;\n 552 }\n 553 // The CWD-state snapshot is only for a real window target; hidden-buffer target is handled by the\n 554 // ctx_win machinery (ctx_win_prep).\n 555 if (buf == NULL || wp != NULL) {\n >>> CID 652777: Null pointer dereferences (FORWARD_NULL)\n >>> Passing null pointer \"wp\" to \"ctx_dirs_save\", which dereferences it.\n 556 ctx_dirs_save(cs, wp, tp == NULL ? curtab : tp, buf);\n 557 }\n 558\n 559 // Save the current state.\n 560 cs->cs_curwin = curwin->handle;\n 561 cs->cs_prevwin = prevwin == NULL ? 0 : prevwin->handle;", "style": "conventional", "type": "fix", "scope": "coverity", "breaking": false} {"repo": "neovim", "sha": "1ab773296292a08d748426669b2f58199e10a822", "date": "2026-08-07T16:49:12-04:00", "subject": "backport: feat(lsp): pass target buffer to `reuse_client` predicate #…", "body": "Problem: The reuse_client predicate does not pass the target buffer,\npreventing decisions from being truly made per buffer.\n\nSolution: Pass the target buffer.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "03efb5e293808ebae346cfc2bb99555592b5ac25", "date": "2026-08-07T22:06:09+01:00", "subject": "Preserve text behind transparent Kitty images.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "d53f8eff9e53a3b7584e38a73938f55e5c0c25cc", "date": "2026-08-07T22:16:16+01:00", "subject": "Add comments to image functions", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "04644c726741ed92aa8a4282eea4cd78ad7ea77f", "date": "2026-08-07T22:41:00+01:00", "subject": "Comment sixel_from_image().", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7fba23f197791b95f72d30ef6adee0f42c43bdaf", "date": "2026-08-08T00:09:15+02:00", "subject": "Add valid single field tests", "body": "Reviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nMergeDate: Sun Aug 9 04:06:52 2026\n(Merged from https://github.com/openssl/openssl/pull/32181)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "b53c00b42599f31c0bd30be75b4dd4be32fd883d", "date": "2026-08-07T18:49:44-04:00", "subject": "fix(progress): ins-compl progress-msg during pum #41226", "body": "Problem:\nThe insert-mode completion progress-message is in \"running\" state while\nthe user is selecting an item. That is noisy and unwanted UX; it was\nonly intended for the \"Scanning...\" task.\n\nSolution:\nEnd the progress-msg just after `ins_compl_show_statusmsg`.", "style": "conventional", "type": "fix", "scope": "progress", "breaking": false} {"repo": "openssl", "sha": "3a9b327813e3d14388138caeeda18450fe9ec2fe", "date": "2026-08-07T17:03:56-06:00", "subject": "Replace generated crypto/objects files only when their content changes", "body": "Reviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a0757394d3f77fd345fdc594ddcc36d2d1c01da0", "date": "2026-08-07T17:03:56-06:00", "subject": "Skip error and ordinal regeneration when nothing has changed", "body": "Reviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:02:59 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ace29c1a8b9ac290e2b4ff8f1fb48aabc2e02079", "date": "2026-08-07T17:03:56-06:00", "subject": "copyright.pm: cope with files that git does not know", "body": "Reviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:02:58 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "eedad09db21202d693fa0eabcedb0ced9297b82b", "date": "2026-08-07T17:13:07-06:00", "subject": "Skip the small generator scripts when their output is current", "body": "Guard the generate_apps, generate_crypto_bn, generate_crypto_conf,\ngenerate_crypto_asn1 and generate_fuzz_oids recipes with test -nt\nchecks against their inputs, so \"make update\" no longer reruns them\nwhen nothing they consume has changed.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:01 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "7b404567924840aabdb955704a9bc04778b14f73", "date": "2026-08-08T01:19:13+02:00", "subject": "fix(core): reject prefixed SVG script hosts", "body": "Use the host element's namespace-aware local name when checking for script\nelements. A prefixed SVG script can expose a qualified tag name such as\n\"x:script\" while its local name remains \"script\".\n\n(cherry picked from commit 9f8e32616b79892785051fab1c3120a770f5f682)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "9f8e32616b79892785051fab1c3120a770f5f682", "date": "2026-08-08T01:19:13+02:00", "subject": "fix(core): reject prefixed SVG script hosts", "body": "Use the host element's namespace-aware local name when checking for script\nelements. A prefixed SVG script can expose a qualified tag name such as\n\"x:script\" while its local name remains \"script\".", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "openssl", "sha": "5c0fe36ea3626a3428e94fe257f917201bc716f9", "date": "2026-08-07T17:38:56-06:00", "subject": "Generate the committed generator-output headers at build time", "body": "Move charmap.h, bn_prime.h, conf_def.h, obj_mac.h, obj_dat.h and\nobj_xref.h from committed generator output to GENERATE rules with\nDEPENDs on their consumers; obj_mac.h joins the mandatory generated\nheaders built before everything else. objects.pl learns -a to append\nthe obj_compat.h aliases itself, replacing the sed step, and\nobj_dat.pl reads only the guarded body of obj_mac.h so the appended\naliases cannot pollute its tables. obj_mac.num remains the curated NID\nregistry in git; the generate_crypto_objects developer target shrinks\nto maintaining it, generate_fuzz_oids now depends on the\nbuild-generated obj_dat.h, and the other generate_* developer targets\nfor these files are retired.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:02 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7a3fa58d9da7a7b60d5a685a6eb96a13c80f5256", "date": "2026-08-07T17:53:57-06:00", "subject": "Generate the error headers and tables at build time", "body": "Teach util/mkerr.pl two new modes. -emit FILE writes one generated\nerror file to standard output as a pure function of openssl.ec and\nopenssl.txt, with no source scan and no state update; every file in\nthe openssl.ec L table becomes a GENERATE rule driven by it, and the\ngenerated error headers join the mandatory generated headers.\n-state scans the sources and maintains openssl.txt but writes no\nother files; \"make errors\" now uses it, so the developer flow only\ncurates the state and the build emits everything else.\n\nReason-code assignment moves out of the header emission into its own\npass, the three emitters become subroutines writing to a filehandle,\nand the up-to-date stamp moves from the statefile's own timestamp to\na separate gitignored stamp file, since the statefile is now a build\ninput whose timestamp must track its content.\n\nRemoving the committed copies also retires 15 that had silently\ndrifted from the generator's output, among them crypto/bio/bio_err.c,\nwhich was missing the reason strings for three BIO error codes that\nexist in the state file.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:03 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f0a01822850bf7324326a3424f6b9a6892e5a768", "date": "2026-08-08T07:54:29+08:00", "subject": "refactor(path): pathcmp() #41035", "body": "Problem:\nRedudant code\n\nSolution:\nAdd `path_fold_char()` to normalize path sep.\nRefactor `pathcmp()` and `path_fnamencmp()` into `path_cmp`.\n\n| feature | pathcmp | path_fnamencmp | path_cmp(now) |\n| ------------------------ | -------------- | -------------- | ------------- |\n| case folding | `mb_toupper()` | `utf_fold()` | `utf_fold()` |\n| consults fileignorecase | ✅ | ✅ | `ic` param |\n| `maxlen` | ✅ | ✅ | ✅ |\n| `/` == `\\` | ✅ | ✅ | ✅ |\n| MSWIN drive letter | ❌ | ✅ | ✅ |\n| sep affects sorting | ✅ | ❌ | ✅ |\n| ignores a trailing slash | ✅ | ❌ | ✅ |\n\nRefactor `path_fnamecmp` and `path_full_compare` into `path_equal`,\nwith flags controlling \"no filesystem\" comparison (i.e. `path_cmp`), env\nvariables expansion, absolute path resolution and filesystem access.", "style": "conventional", "type": "refactor", "scope": "path", "breaking": false} {"repo": "openssl", "sha": "db0d21bad3bc2dd8b9c4c68d9f63a651518a38bc", "date": "2026-08-07T18:06:04-06:00", "subject": "Install man pages a section at a time, not a process per page", "body": "Replace util/write-man-symlinks with util/install-man-pages.pl,\nwhich installs or uninstalls an entire man section in one process:\ncopy, chmod and the NAME-alias symlinks that previously cost a\nbasename, cp, chmod and a full perl interpreter for every one of\nthe ~900 installed pages.\n\nStale alias links are removed before being recreated, with one\nconstraint carried over from the old script's silent symlink\nfailures: on a case-insensitive file system an alias can collide\nwith an installed page (OSSL_TRACE_ENABLED alongside\nOSSL_trace_enabled), so nothing that is not a symlink is ever\nremoved, and an occupied alias name is skipped.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:04 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "274cc16ec3eca1c89807de99130c6649b53dc063", "date": "2026-08-07T18:14:43-06:00", "subject": "Install only changed development and runtime files", "body": "Install the public headers and the pkg-config and cmake exporters\nthrough a new util/install-files.pl, which handles a whole file set\nin one process and replaces a file only when its content differs; an\nunchanged file keeps its timestamp, so builds against the installed\ntree no longer see every header change after each install.\n\nThe static and shared library installs keep their staged copies --\nthe mv is what makes replacing a mapped library safe -- but now\ncompare the staged copy against what is installed and discard it\nwhen identical, and only report an install when one happens.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:05 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "662b6cd817b34334951f2416060369ccae9be7c4", "date": "2026-08-07T18:37:58-06:00", "subject": "Fix the emitted error files for out-of-tree and Windows builds", "body": "Add INCLUDE directives for the objects whose sources include a\ngenerated header by its bare name; in-tree builds masked the missing\nbuild-directory include paths. Make mkerr.pl's config and state\nparsing tolerate CRLF line endings, since it now runs at build time\non Windows checkouts.\n\nAlso repair the error state itself: openssl.txt disagreed with the\npublished bioerr.h on BIO_R_PEER_ADDR_NOT_AVAILABLE (151, not 114)\nand had lost BIO_R_NULL_PARAMETER entirely -- the shipped header is\nthe ABI, so the state now matches it. The out-of-order defines at\nthe bottom of the old header show it had been edited by hand without\nregenerating, which is exactly the failure mode that generating\nthese files at build time removes.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:06 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d5ab5aa63c70b05217444dd9842d7cd42bec5a00", "date": "2026-08-07T18:45:25-06:00", "subject": "Keep the HTTP reason codes visible in no-http builds", "body": "no-http removes the HTTP client but still builds http_lib.c for URL\nparsing, and it raises HTTP reason codes; mkerr must not wrap\nhttperr.h in OPENSSL_NO_HTTP, so HTTP joins SSL, ASYNC and DSO on\nthe list of libraries whose error codes survive disabling.\n\nThe committed httperr.h never had the guard -- it predates HTTP\nbecoming disablable and was never regenerated -- so no-http builds\nhave been quietly relying on the committed copy being stale. The\nemitted header now matches the published one.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8920a41604ba568b8bf3f41cc28589098f75f844", "date": "2026-08-07T18:59:41-06:00", "subject": "Strip the line ending from continuation lines in the mkerr state", "body": "Long reason strings in openssl.txt are stored on a continuation\nline, which was read without stripping its line ending; the greedy\nwhitespace trim kept the carriage return on a CRLF checkout, and\nthe emitted string then carried a bare CR, which compilers treat\nas a line ending inside the literal.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:08 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "3f2b321bac19de096eed5296b6bd0dee9dc5d832", "date": "2026-08-08T10:53:23+08:00", "subject": "fix(site): exclude bot accounts from contributors", "body": "* fix(site): exclude bot accounts from contributors\n\n* update\n\n* Potential fix for pull request finding\n\nCo-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>\nSigned-off-by: lijianan <574980606@qq.com>\n\n* update\n\n* update\n\n---------", "style": "conventional", "type": "fix", "scope": "site", "breaking": false} {"repo": "angular", "sha": "c19a36c2fbc51f41ce974ea1fd87f9f56b3e2444", "date": "2026-08-07T23:10:04-05:00", "subject": "fix(platform-browser): disallow event handler attributes in Meta", "body": "Prevent arbitrary MetaDefinition properties from writing on* handlers directly to meta elements. Browser events can execute these handlers, including on meta elements rendered in the document body.\n\nReject handler attributes case-insensitively before matching or mutating tags.\n\nFixes #70126", "style": "conventional", "type": "fix", "scope": "platform-browser", "breaking": false} {"repo": "curl", "sha": "3f1c033afe008123680306663cc01279e831ac2d", "date": "2026-08-07T21:16:52-07:00", "subject": "libcurl.pc: add the Link.ABI and Source tags", "body": "Link.ABI is a feature of pkgconf 3.0 and indicates the ABI against which\na consumer must link the package. For libcurl, this is hard-coded as C.\nSource is a URL from which to download the package tarball.\n\nCloses #22519", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "b2be2461f1a1228ab7871e8730acbce464ead437", "date": "2026-08-08T13:13:44+08:00", "subject": "fix(runtime-vapor): pause KeepAlive scopes when deactivated", "body": "Vapor effects can continue tracking parent-owned reactive sources after a\ncached branch is moved into storage. This allows render and watcher jobs to\nconsume transient values while the branch is inactive.\n\nCollect the detached component scopes within each KeepAlive boundary and\npause them on deactivation, then resume dirty effects on activation. Ensure\neffects and child scopes created under an already-paused scope inherit its\nstate, and prevent jobs queued before deactivation from running.\n\nAllow unresolved async wrappers to finish resolving while ensuring their\nresolved component effects start paused when the branch is cached.\n\nCloses #15228", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "725fbeaad9532c42acfc9d8e66b48f38425a7697", "date": "2026-08-07T22:32:02-07:00", "subject": "tests: fix typo in assert message in http test", "body": "Found by pyrefly.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "99cbe04815cfcbeacf08ff7b987e2023b037e591", "date": "2026-08-08T14:18:37+08:00", "subject": "fix(runtime-vapor): handle nested KeepAlive boundaries", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "tmux", "sha": "db7ec915eaa6a5c994cdbe1250b7ff467d3f6f43", "date": "2026-08-08T09:14:07+01:00", "subject": "Do not enable sync for SIXEL", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "cce975eb77c0e63c69ba55be562f2cde9756e96f", "date": "2026-08-08T09:19:39+01:00", "subject": "Move RB_GENERATE_STATIC to top of file.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "5c9ae058bfd10eb053fc2368428a07109b59583a", "date": "2026-08-08T10:21:30+01:00", "subject": "Expose image backend flags", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "37b71800b59a34a23983dc8a01c524f90f92998e", "date": "2026-08-08T10:26:50+01:00", "subject": "Move image cell drawing into image.c", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "48b6641d97209bb1cc3eb39c2790e631daf570d3", "date": "2026-08-08T10:33:54+01:00", "subject": "Avoid overflow in image cell size calculation", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "89a7dcc8ab0878336f319f4c69f97d2a0b7461f5", "date": "2026-08-08T10:40:56+01:00", "subject": "Shorten image and SIXEL names", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "d0544db735b1ca5e61c6d1581783de08343c1cbb", "date": "2026-08-08T17:46:17+08:00", "subject": "fix(Drawer): support disabled close button", "body": "", "style": "conventional", "type": "fix", "scope": "Drawer", "breaking": false} {"repo": "tmux", "sha": "c1cab6ea34ecb2d76265782406dcb326cedc7947", "date": "2026-08-08T10:49:13+01:00", "subject": "More cleanup of \"histogram\" name.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "b89f8b062473846e2b6e70e2783996640e82dd7c", "date": "2026-08-08T10:58:15+01:00", "subject": "Update Kitty image names", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0aad77e779ab27ad79532b7cfe1bd8c0766b4c25", "date": "2026-08-08T10:58:58+01:00", "subject": "Update fallback image names", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "1eac4481473df52fbbcb452c4dca8d79039dbb63", "date": "2026-08-08T19:35:07+09:00", "subject": "refactor: replace `semver` with `verkit`", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "curl", "sha": "2d30fd26a060e7c3de3393503fb5ba7e8f3840f8", "date": "2026-08-08T13:20:15+02:00", "subject": "DoH: improvements", "body": "- decode results when individual requests are done\n- makes happy eyeballing start asap\n- remove doh_resp structures as no longer needed\n- remove CURL_DNS_TYPE_NS, CURL_DNS_TYPE_CNAME and\n CURL_DNS_TYPE_DNAME from DoH\n- DoH: do not set PIPEWAIT and SSL OPTS when url starts with http:\n- mark Doh master handle as dirty after every sub-request, not only the last\n- Doh: start probe on AAAA before A (was the other way).\n\ncf-dns: set EXPIRE_HAPPY_EYEBALLS timer when waiting 50ms on AAAA\nresult or progress will not be triggered in time.\n\nAdd debug env var CURL_DBG_HE_AAAA_AWAIT_MS to override the\ndefault 50ms on waiting for the AAAA result to arrive.\n\ntest2100: set CURL_DBG_HE_AAAA_AWAIT_MS to 60 seconds to\nprovide enough time for slow CI runs to sent all DoH requests.\n\ncloses #22514", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c6c9d43917907250daab6f29a143b1d466747f08", "date": "2026-08-08T06:18:31-06:00", "subject": "Install the HTML manual pages through install-files.pl", "body": "The per-page copy loops become one process per section that\nreplaces only a page whose content changed, matching the man page\nand header installs; an unchanged page keeps its timestamp and a\nreinstall is quiet. The image documents get the same treatment.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:09 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "b7745d68b8b3d9635ff85b0705aa435af20670ff", "date": "2026-08-08T20:55:47+08:00", "subject": "test(Form): replace compiler assertion with dist regression test", "body": "* test(Form): replace compiler assertion with dist regression test\n\n* update", "style": "conventional", "type": "test", "scope": "Form", "breaking": false} {"repo": "ant-design", "sha": "ef787d0b8eed12aa5dd3f89233a36d1ce214179c", "date": "2026-08-08T20:55:56+08:00", "subject": "chore: update authors", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "axios", "sha": "f5e83e49ee32b1004db82dad2b14d1e273a99e07", "date": "2026-08-08T19:02:59+02:00", "subject": "docs: refresh sponsor data", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "0eddc16ed0badae1fecc4e005d4e1f317b6d4102", "date": "2026-08-08T22:28:43+03:00", "subject": "ci: add nvim.wasm build and nightly release", "body": "Add a dedicated nvim.wasm build workflow and\ninclude the resulting WASM artifact in nightly releases.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "ca97a3f420e568e9d1ccf19834730a7f7365439d", "date": "2026-08-09T01:13:59+05:30", "subject": "fix(common): match string file types containing regex metacharacters", "body": "`String#match` coerces a string argument into a RegExp, so a `fileType` such as\n`image/svg+xml` or `application/ld+json` was compiled into a pattern that cannot\nmatch the MIME type it was written as, and files were rejected with a message\nnaming the expected type as the type they already had.\n\nCompare a string `fileType` literally before falling back to the existing regex\nmatch, which partial values like `jpeg` still rely on.", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "vue-core", "sha": "644a410ca284cdc53b2ac01ef17701dce272b726", "date": "2026-08-09T10:44:58+08:00", "subject": "fix(runtime-vapor): defer work in paused KeepAlive scopes", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "openssl", "sha": "919b66f56d25f3753c0f160cce37404e908a8d45", "date": "2026-08-09T14:28:39+10:00", "subject": "evp_test: condition CTR-DRBG tests on FIPS version when the df is not used", "body": "Newer versions require the use of the derivation function and non-use is\nan error. This means a number of tests must be inhibited in such situations.\n\nReviewed-by: Neil Horman \nReviewed-by: Nikola Pajkovsky \nMergeDate: Tue Aug 11 05:09:36 2026\n(Merged from https://github.com/openssl/openssl/pull/32250)\n\n(cherry picked from commit 72f063b8403210c876372ab1941c9e8c1b4c5dda)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "72f063b8403210c876372ab1941c9e8c1b4c5dda", "date": "2026-08-09T14:28:39+10:00", "subject": "evp_test: condition CTR-DRBG tests on FIPS version when the df is not used", "body": "Newer versions require the use of the derivation function and non-use is\nan error. This means a number of tests must be inhibited in such situations.\n\nReviewed-by: Neil Horman \nReviewed-by: Nikola Pajkovsky \nMergeDate: Tue Aug 11 05:09:36 2026\n(Merged from https://github.com/openssl/openssl/pull/32250)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "200d1e673f59e5969662a2d23452f50321c7959f", "date": "2026-08-09T14:28:39+10:00", "subject": "evp_test: condition CTR-DRBG tests on FIPS version when the df is not used", "body": "Newer versions require the use of the derivation function and non-use is\nan error. This means a number of tests must be inhibited in such situations.\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Mon Aug 10 14:00:08 2026\n(Merged from https://github.com/openssl/openssl/pull/32249)\n\n(cherry picked from commit e12d71f09a4cd41ca45920fe589b99df3b85532a)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e12d71f09a4cd41ca45920fe589b99df3b85532a", "date": "2026-08-09T14:28:39+10:00", "subject": "evp_test: condition CTR-DRBG tests on FIPS version when the df is not used", "body": "Newer versions require the use of the derivation function and non-use is\nan error. This means a number of tests must be inhibited in such situations.\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Mon Aug 10 14:00:08 2026\n(Merged from https://github.com/openssl/openssl/pull/32249)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5c6119870fc0b855520b69271fd73798f0761070", "date": "2026-08-09T14:28:39+10:00", "subject": "evp_test: condition CTR-DRBG tests on FIPS version when the df is not used", "body": "Newer versions require the use of the derivation function and non-use is\nan error. This means a number of tests must be inhibited in such situations.\n\nReviewed-by: Neil Horman \nReviewed-by: Viktor Dukhovni \nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMergeDate: Mon Aug 10 12:01:44 2026\n(Merged from https://github.com/openssl/openssl/pull/32235)\n\n(cherry picked from commit 5b4df6eac1a7d228ad2a8f08d45c7d0a13992026)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5b4df6eac1a7d228ad2a8f08d45c7d0a13992026", "date": "2026-08-09T14:28:39+10:00", "subject": "evp_test: condition CTR-DRBG tests on FIPS version when the df is not used", "body": "Newer versions require the use of the derivation function and non-use is\nan error. This means a number of tests must be inhibited in such situations.\n\nReviewed-by: Neil Horman \nReviewed-by: Viktor Dukhovni \nReviewed-by: Milan Broz \nReviewed-by: Andrew Dinh \nMergeDate: Mon Aug 10 12:01:44 2026\n(Merged from https://github.com/openssl/openssl/pull/32235)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "4d00ca4064ae0d1a75b604a16c68ab9f386ad388", "date": "2026-08-09T16:15:05+09:00", "subject": "docs: update ESLint peer dependency to `^10.0.0` in shareable configs", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "eslint", "sha": "1a761e1d11b011fcb6bee181231a51010c500e4d", "date": "2026-08-09T12:30:38+05:00", "subject": "docs: update moved JSX specification links", "body": "* docs: update moved JSX specification links\n\nThe JSX specification moved from the `facebook` GitHub organization to\n`react`, so https://facebook.github.io/jsx/ now returns a 404. GitHub\nPages does not redirect across organization renames, so the old URL is\npermanently dead.\n\nPoints the three remaining references at https://react.github.io/jsx/.\n\nThe nearby link to github.com/facebook/jsx/issues/25 is left unchanged\nbecause GitHub does redirect renamed repositories, so it still resolves.\n\n* Update lib/rules/no-useless-escape.js\n\nCo-authored-by: Milos Djermanovic \n\n---------", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "tmux", "sha": "b6dbf10057236eb8bb840fe1f879b496181cf978", "date": "2026-08-09T08:32:25+01:00", "subject": "Redraw SIXEL images before or after text depending on which was placed first in the cell.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "ffc87d6234b2aa4335eec069e5c4d6ac04832b9e", "date": "2026-08-09T17:08:43+09:00", "subject": "docs: fix broken links in Further Reading sections", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "axios", "sha": "ba98559a7f5a18e531b5762387e5957bd281af3d", "date": "2026-08-09T10:34:40+02:00", "subject": "docs: add ScrapingBee sponsor", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "791b6883edc1c02c4b253e8b47237ec13ac66d06", "date": "2026-08-09T11:07:27+02:00", "subject": "build: replace `assert()` with `DEBUGASSERT()`, where missing", "body": "Follow-up to 9a2663322c330ff11275abafd612e9c99407a94a #17572\nFollow-up to c96f982166fbc671b77f84a1730e9b6d11357e8f #10451\n\nCloses #22521", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "260a3b99844ff124f36abd0dda3e294e92adcabe", "date": "2026-08-09T18:20:35+09:00", "subject": "CCM: authenticate empty Final when payload Update is skipped", "body": "CCM Final was routed through an update helper whose pointer-based\ndispatch treats a NULL-input call as a no-op or a length declaration.\nIf an empty message skips payload Update, decryption can therefore skip\ntag verification and encryption does not generate a tag.\n\nProcess an empty payload during Final only when the existing state flags\nshow that no payload operation took place. Route the NULL-input\nEVP_Cipher() form through the same Final path.\n\nFixes #32253\n\nAssisted-by: Codex:gpt-5.6-sol\n\n(cherry picked from commit 493a46c61983195c2b49bfde75ffba0321e5e69d)\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 21 10:04:26 2026\nMerged-from: https://github.com/openssl/openssl/pull/32427", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9fa890d55d49ea7062290450de83248b9cc3d0e0", "date": "2026-08-09T18:20:35+09:00", "subject": "CCM: authenticate empty Final when payload Update is skipped", "body": "CCM Final was routed through an update helper whose pointer-based\ndispatch treats a NULL-input call as a no-op or a length declaration.\nIf an empty message skips payload Update, decryption can therefore skip\ntag verification and encryption does not generate a tag.\n\nProcess an empty payload during Final only when the existing state flags\nshow that no payload operation took place. Route the NULL-input\nEVP_Cipher() form through the same Final path.\n\nFixes #32253\n\nAssisted-by: Codex:gpt-5.6-sol\n\n(cherry picked from commit 493a46c61983195c2b49bfde75ffba0321e5e69d)\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 21 10:00:11 2026\nMerged-from: https://github.com/openssl/openssl/pull/32427", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1edbbe38105b37d816ffeb91988d65953518324d", "date": "2026-08-09T18:20:35+09:00", "subject": "CCM: authenticate empty Final when payload Update is skipped", "body": "CCM Final was routed through an update helper whose pointer-based\ndispatch treats a NULL-input call as a no-op or a length declaration.\nIf an empty message skips payload Update, decryption can therefore skip\ntag verification and encryption does not generate a tag.\n\nProcess an empty payload during Final only when the existing state flags\nshow that no payload operation took place. Route the NULL-input\nEVP_Cipher() form through the same Final path.\n\nFixes #32253\n\nAssisted-by: Codex:gpt-5.6-sol\n\n(cherry picked from commit 493a46c61983195c2b49bfde75ffba0321e5e69d)\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 21 09:57:46 2026\nMerged-from: https://github.com/openssl/openssl/pull/32427", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a94aa9a2c798f050082f67eb5539701e69ec363f", "date": "2026-08-09T18:20:35+09:00", "subject": "CCM: authenticate empty Final when payload Update is skipped", "body": "CCM Final was routed through an update helper whose pointer-based\ndispatch treats a NULL-input call as a no-op or a length declaration.\nIf an empty message skips payload Update, decryption can therefore skip\ntag verification and encryption does not generate a tag.\n\nProcess an empty payload during Final only when the existing state flags\nshow that no payload operation took place. Route the NULL-input\nEVP_Cipher() form through the same Final path.\n\nFixes #32253\n\nAssisted-by: Codex:gpt-5.6-sol\n\n(cherry picked from commit 493a46c61983195c2b49bfde75ffba0321e5e69d)\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 21 09:54:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/32427", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "aef847783abd5e9da789179eb41d243896c83668", "date": "2026-08-09T18:20:35+09:00", "subject": "CCM: authenticate empty Final when payload Update is skipped", "body": "CCM Final was routed through an update helper whose pointer-based\ndispatch treats a NULL-input call as a no-op or a length declaration.\nIf an empty message skips payload Update, decryption can therefore skip\ntag verification and encryption does not generate a tag.\n\nProcess an empty payload during Final only when the existing state flags\nshow that no payload operation took place. Route the NULL-input\nEVP_Cipher() form through the same Final path.\n\nFixes #32253\n\nAssisted-by: Codex:gpt-5.6-sol\n\n(cherry picked from commit 493a46c61983195c2b49bfde75ffba0321e5e69d)\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 21 09:37:17 2026\nMerged-from: https://github.com/openssl/openssl/pull/32427", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "97fada9e5243a5044a073226fe0894b287040525", "date": "2026-08-09T18:20:35+09:00", "subject": "CCM: authenticate empty Final when payload Update is skipped", "body": "CCM Final was routed through an update helper whose pointer-based\ndispatch treats a NULL-input call as a no-op or a length declaration.\nIf an empty message skips payload Update, decryption can therefore skip\ntag verification and encryption does not generate a tag.\n\nProcess an empty payload during Final only when the existing state flags\nshow that no payload operation took place. Route the NULL-input\nEVP_Cipher() form through the same Final path.\n\nEnable CCM in the generic zero-length AEAD test. Declare the zero payload\nlength and supply AAD while deliberately omitting payload Update.\n\nFixes #32253\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Fri Aug 21 08:51:03 2026\nMerged-from: https://github.com/openssl/openssl/pull/32256", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "04f802a2f6387028e3389334e6e04abc28d92587", "date": "2026-08-09T06:49:05-04:00", "subject": "build(vim-patch): n/a patches from comments, ifdefs #41247", "body": "vim-patch.sh fails to detect n/a patches\nbecause of ifdef FEAT_ guards and reserved Vim9script error codes.\n\nIgnore all conditional directives for Vim's \"FEAT_\" guards.\nhttps://cppreference.com/c/preprocessor/conditional", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "4a98c4ade08bf7e1abd3fb1958782dbf8bf4680a", "date": "2026-08-09T09:13:16-04:00", "subject": "Test the constant-time validation helpers", "body": "Suggested by @bbbrumley in review of #31617.\n\nReferences #15076 (add constant time coverage everywhere).\n\nAssisted-by: Claude:claude-sonnet-5\nReviewed-by: Jakub Zelenka \nReviewed-by: Shane Lontis \nMerge-date: Tue Sep 1 14:06:52 2026\nMerged-from: https://github.com/openssl/openssl/pull/32255", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "18285605f351214c0c051ba6464d1477e7e453b7", "date": "2026-08-09T23:52:24+09:00", "subject": "Fix VAES CBC round-key stack cleanup", "body": "The VAES CBC decryptor uses volatile stores to clear its source-level\nround-key array. With GCC 13.3 at -O3, scalar replacement separates\nthat object from the stack slots holding broadcast round keys. The\nstores remain, but the actual key spills survive return.\n\nRoute cleanup through a dedicated perlasm helper. Passing rk across\nthis opaque assembly boundary forces optimized callers to materialize\nthe consumed key array. The helper erases exactly 11, 13, or 15\n64-byte entries and scrubs caller-clobbered vector and general-purpose\nregister state within the System V and Win64 ABI constraints. Gate\nVAES-CBC eligibility on assembler support and include the helper in\nregular and FIPS builds.\n\nThis is a pragmatic boundary rather than a guarantee against every\ncompiler-created spill or abnormal exit. In particular, the Win64 ABI\nrequires the low 128 bits of XMM6-XMM15 to remain intact until the\ncaller restores them. A complete guarantee would require moving the\ndecryptor behind an assembly or custom-ABI boundary.\n\nFixes #32254\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Shane Lontis \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 14:00:40 2026\nMerged-from: https://github.com/openssl/openssl/pull/32257", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "661fc3f81a15179575fb924e3d1afa54d6aef050", "date": "2026-08-09T17:04:04+02:00", "subject": "OSSL_CMP_{exec_certreq,SRV_CTX}_new.pod: enrich doc of CMP client and server", "body": "In particular, clarify that they focus on the Lightweight CMP Profile\nand that the server is single-threaded.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Thu Aug 27 07:24:59 2026\n(Merged from https://github.com/openssl/openssl/pull/28988)\n\n(cherry picked from commit 44635ff58b34e677b1bcf4bd87be07c4698d414c)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "44635ff58b34e677b1bcf4bd87be07c4698d414c", "date": "2026-08-09T17:04:04+02:00", "subject": "OSSL_CMP_{exec_certreq,SRV_CTX}_new.pod: enrich doc of CMP client and server", "body": "In particular, clarify that they focus on the Lightweight CMP Profile\nand that the server is single-threaded.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Thu Aug 27 07:24:59 2026\n(Merged from https://github.com/openssl/openssl/pull/28988)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c0cbc0fa2e12d7c9f8165f43b6271be7588e39e2", "date": "2026-08-09T12:52:48-04:00", "subject": "Release notes for 18.5, 17.11, 16.15, 15.19, 14.24.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "1a8c273683db344c50f7aa883376fec934fa6b34", "date": "2026-08-09T20:07:24+02:00", "subject": "fix(microservices): call on processing end hook", "body": "", "style": "conventional", "type": "fix", "scope": "microservices", "breaking": false} {"repo": "nestjs", "sha": "361ce5fe0a1c211ab5f42cab06ced6e9aaabe3a0", "date": "2026-08-09T15:42:49-04:00", "subject": "feat: narrow body parser options to the selected parser", "body": "`useBodyParser()` never checked its options argument.\n`Options` occurs only inside `Omit`, a non-inferrable position, so it always\nfell back to its default, and that default carries a `[key: string]: unknown` index signature.\n`keyof` on it is `string | number`, `Exclude` removes nothing, and the\nresulting `Pick` collapses to bare index signatures. Every call therefore\naccepted any key, any value type, and even `verify` (the single option the signature claims to\nremove, and the one Nest reserves to capture `rawBody`).\n\nKey the options off the `parser` argument instead. `NestExpressBodyParserOptionsMap` maps each\nparser to the options object its Express factory accepts, derived from `express` itself so the two\ncannot drift, and `NestExpressBodyParserType` is now `keyof` that map. The type parameter is\ninferred from `parser`, a real inference site, so `useBodyParser('json', ...)` accepts only\n`express.json` options and now rejects typos, wrong value types, options belonging to a different\nparser, and `verify`.\n\nThis is a types-only change; no emitted JavaScript differs.\n\n`NestExpressBodyParserOptions` is\ndeprecated rather than removed, and calls that pass no explicit type argument are unaffected.\n\nBREAKING CHANGE: `useBodyParser` now takes the parser type as its first type argument.\nCalls that\npass an explicit options type, such as `app.useBodyParser('urlencoded', {...})`, no longer compile.\nDrop the type argument and let it be inferred from the parser argument.", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "tmux", "sha": "cf80845be7426486767f4c8934ad4aa115bca5f9", "date": "2026-08-09T22:13:15+01:00", "subject": "Change some names of things for clarity.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "beea19dc97213e69174426b18ea3e8cd4297789f", "date": "2026-08-09T22:15:37+01:00", "subject": "Update image size accessor call sites", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "043e9517256b903816e6ec63dfb8b174db81d331", "date": "2026-08-09T17:22:22-04:00", "subject": "fix(platform-express): address body parser review comments", "body": "", "style": "conventional", "type": "fix", "scope": "platform-express", "breaking": false} {"repo": "neovim", "sha": "abc271d8f3db2d6f7c792cdb6868cbb0cc80c3a1", "date": "2026-08-09T17:33:53-04:00", "subject": "fix(logging): vimscript API calls #41253", "body": "Problem:\n\":call nvim_get_mode()\" logs a nonsense channel-id:\n\n RPC: ch 9223372036854775808: invoke nvim_get_mode\n\nSolution:\nCheck `is_internal_call`.\nAlso, skip this logging for RPC calls, because it's redundant with\n`log_request`/`log_notify`.\n\n API: vim -> nvim_get_mode", "style": "conventional", "type": "fix", "scope": "logging", "breaking": false} {"repo": "curl", "sha": "05ddf5511ac765c8fc4eb44507ab7c585116f665", "date": "2026-08-10T00:48:09+03:00", "subject": "headers: name the arguments the way the definitions name them", "body": "Eleven arguments are named one way in the header and another in the\ndefinition. the compiler only checks types so nothing is broken, but the\nheader is what people read first, and then the name changes under them\nin the source:\n\n```\ncurl_dbg_malloc size -> wantedsize\ncurl_dbg_calloc n, size -> wanted_elements, wanted_size\ncurl_dbg_realloc size -> wantedsize\ncurl_dbg_strdup src -> source\nr_freeaddrinfo res -> cahead\nmy_get_line fp, db -> input, buf\ncurl_slist_append_ccsid l -> list\ncurl_version_info_ccsid cid -> ccsid\ncurl_easy_setopt_ccsid curl -> easy\n```\n\nFollow-up to df6014894b789768b139fff27e8aed9dbc46ed79 #20794\n\nCloses #22550", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "127ce776983f19e81ba8abda1984f6f0e229c1b5", "date": "2026-08-10T08:37:32+09:00", "subject": "Use OOM-safe routine for pgstats shared hashtable insert", "body": "pgstat_get_entry_ref() called dshash_find_or_insert() to insert a new\nentry into the shared pgstats hashtable. If the allocation required for\nthe new entry fails, a local cache entry would be left in\npgStatEntryRefHash with a NULL shared_entry. This code is switched to\nuse dshash_find_or_insert_extended(), so as local hash entries can be\ncleaned up before failuring. This change makes pgstat_get_entry_ref()\nmore robust on re-entry.\n\nTaken in isolation within a single backend, local entries are handled\nfine on HEAD as pgstat_get_entry_ref() is able to handle the case of a\nNULL shared_entry on re-entry. However it is possible to finish with\nsome NULL pointer dereference in more complex scenarios, like:\n- First backend fails insert into the shared hashtable on OOM.\n- Second backend bumps the same entry's refcount (due to\nreinitialization of the same entry, for example).\n- First backend calls pgstat_gc_entry_refs(), uses its local entry\nreference with the NULL shared_entry, crashes when checking if the\nshared entry has been dropped.\n\n6f0738ddec85 has introduced dshash_find_or_insert_extended() for the\nsame reason as what we are dealing with here: the backend could be left\nin an inconsistent stat, and clean up actions need to be taken before\nissuing an error. The failure is unlikely going to show up in practice,\nso no backpatch is done. This hash insert extended routine is new as of\nv19.\n\nAuthor: Michael Paquier \nDiscussion: https://postgr.es/m/anV8ftNMW3xyKSMG@paquier.xyz", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b91f79cd08abba7a800d3faf79bb96e8adc8752e", "date": "2026-08-10T09:00:38+09:00", "subject": "Use explicit fetching for digests in cryptohash (OpenSSL >= 3.0)", "body": "cryptohash_openssl.c initialized the EVP_MD_CTX with the implicit static\ndigest objects (EVP_sha256() and friends). These do not dispatch\nthrough a loaded provider. OpenSSL 3.0 and newer versions recommend to\nswitch to an explicit fetch, using EVP_MD_fetch(), the older routines\nbeing deprecated, available for backward-compatibility purposes.\n\nNow, when building with OpenSSL 3.0 or a newer bersion, we fetch the\ndigest by name with EVP_MD_fetch(), cache it in the context, and free it\non teardown. This has as consequence to feed the hash from a provider.\nThe implicit path is kept for older OpenSSL, 1.1.1 being the oldest\nversion still supported on HEAD, and for LibreSSL.\n\nThis is a slight change in the hashing logic. No backpatch is done as\nthat looks safer, and the code in the back-branches is also able to\nwork. This has been locally tested across OpenSSL 1.1.1 up to 3.5, so\nhopefully nothing is broken.\n\nNote: The submitted code was clearly AI-generated. I have spend a bit\nof time cleaning it up, making it more consistent with the project style\nin terms of coding, adjusting a few more things on the way.\n\nAuthor: Mark Atwood \nCo-authored-by: Michael Paquier \nDiscussion: https://postgr.es/m/178596055358.1584287.8485463954311014881@reviewcommit.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "b527b64be31f720ea01c3df5b0db1fb10881ac81", "date": "2026-08-10T08:14:55+08:00", "subject": "chore(playground): disable unsupported ts7 and eliminate vite warn", "body": "", "style": "conventional", "type": "chore", "scope": "playground", "breaking": false} {"repo": "vue-core", "sha": "40423896796ff7438a1aa70ee3cf22fe5b0ac224", "date": "2026-08-10T08:16:38+08:00", "subject": "chore: remove outdated configuration", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "88b11512b8a549fe3c58884a646aa9f4dd0d3edc", "date": "2026-08-10T08:19:24+08:00", "subject": "fix(types): prevent infer Exposed as JSX.Element for defineVaporComponent", "body": "", "style": "conventional", "type": "fix", "scope": "types", "breaking": false} {"repo": "vue-core", "sha": "25dcb3720ecb1015f92b1c6895bed89aa43d0398", "date": "2026-08-10T08:24:13+08:00", "subject": "fix(runtime-vapor): clean up teleport targets on scope disposal", "body": "Closes #15236", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "a61f2f9219bc121eea41a870e181e93a79623743", "date": "2026-08-10T00:31:24Z", "subject": "ldap: reject control characters in URL-decoded filter values", "body": "Closes #22524", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "35ba88ac0893b5c89fac563a36dad6152d4ca31d", "date": "2026-08-10T05:50:24+05:00", "subject": "fix(runtime-vapor): normalize class prop passed to components", "body": "close #15227", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "openssl", "sha": "2d17f2d7c0aa02e4e984dbf47d754d5d3defdf54", "date": "2026-08-10T12:44:32+09:00", "subject": "Add tests for empty AEAD EVP_Cipher() finalization", "body": "Generate an empty-message tag with the streaming path, then check that\nEVP_Cipher() decryption accepts the correct tag and rejects a corrupted\none. Also verify that EVP_Cipher() encryption produces the same tag.\n\nCover AES-OCB and ChaCha20-Poly1305, with and without AAD.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Bob Beck \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 19 17:43:25 2026\nMerged-from: https://github.com/openssl/openssl/pull/32417", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "17aee36db64f351a62c6b3ff9f1e9560996a67e7", "date": "2026-08-10T12:44:32+09:00", "subject": "Add tests for empty AEAD EVP_Cipher() finalization", "body": "Generate an empty-message tag with the streaming path, then check that\nEVP_Cipher() decryption accepts the correct tag and rejects a corrupted\none. Also verify that EVP_Cipher() encryption produces the same tag.\n\nCover AES-OCB, ChaCha20-Poly1305, and AES-GCM-SIV, with and without AAD.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Bob Beck \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 19 17:41:17 2026\nMerged-from: https://github.com/openssl/openssl/pull/32416\n(cherry picked from commit e9fb43aee8654b492ed8a239d3f2bacf13bb9081)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e9fb43aee8654b492ed8a239d3f2bacf13bb9081", "date": "2026-08-10T12:44:32+09:00", "subject": "Add tests for empty AEAD EVP_Cipher() finalization", "body": "Generate an empty-message tag with the streaming path, then check that\nEVP_Cipher() decryption accepts the correct tag and rejects a corrupted\none. Also verify that EVP_Cipher() encryption produces the same tag.\n\nCover AES-OCB, ChaCha20-Poly1305, and AES-GCM-SIV, with and without AAD.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Bob Beck \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 19 17:41:17 2026\nMerged-from: https://github.com/openssl/openssl/pull/32416", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b51c9af53180a33959d847de610eb5af116892cd", "date": "2026-08-10T12:44:32+09:00", "subject": "Add tests for empty AEAD EVP_Cipher() finalization", "body": "Generate an empty-message tag with the streaming path, then check that\nEVP_Cipher() decryption accepts the correct tag and rejects a corrupted\none. Also verify that EVP_Cipher() encryption produces the same tag.\n\nCover AES-OCB, ChaCha20-Poly1305, and AES-GCM-SIV, with and without AAD.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 19 15:58:28 2026\nMerged-from: https://github.com/openssl/openssl/pull/32259", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "14d058f2874d38969aa934d23d5d51050c2f2dca", "date": "2026-08-10T12:44:32+09:00", "subject": "Add tests for empty AEAD EVP_Cipher() finalization", "body": "Generate an empty-message tag with the streaming path, then check that\nEVP_Cipher() decryption accepts the correct tag and rejects a corrupted\none. Also verify that EVP_Cipher() encryption produces the same tag.\n\nCover AES-OCB, ChaCha20-Poly1305, and AES-GCM-SIV, with and without AAD.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 19 15:56:03 2026\nMerged-from: https://github.com/openssl/openssl/pull/32300", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "2450ae797c5adee90c267cf28baaab54eebf9796", "date": "2026-08-10T00:29:42-04:00", "subject": "build(vim-patch): sound feature is n/a", "body": "Vim core did not leverage it to override/customize bell/beep.\nIt wasn't used for custom sounds for system/user (autocmd) events.\nIt should be in-scope for GUI, unlike terminal, even as a plugin\nby leveraging some internal option similar to `set guioptions+=!`.\n\nNo progress as of Vim 9.2 so I quit.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "postgres", "sha": "1f3b9bb109b8b1a7bb60d2d7343c1759679e501b", "date": "2026-08-10T14:53:03+09:00", "subject": "Use EVP_MAC for HMAC (OpenSSL >= 3.0)", "body": "The legacy HMAC_CTX interface (HMAC_CTX_new(), HMAC_Init_ex(), etc.) has\nbeen deprecated since OpenSSL 3.0. The legacy routines do not dispatch\nthrough the provider framework, so a loaded provider's HMAC\nimplementation could be bypassed.\n\nThis commit updates hmac_openssl.c to use EVP_MAC when building with\nOpenSSL 3.0 or newer versions. OpenSSL 1.1.1 is the oldest version\nsupported on HEAD, so the legacy routines need to be kept. Note that\nLibreSSL uses the legacy routines.\n\nLike b91f79cd08ab, no backpatch is done.\n\nAuthor: Mark Atwood \nCo-authored-by: Michael Paquier \nDiscussion: https://postgr.es/m/178596055149.1584250.13974609482797470185@reviewcommit.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "5ef4e1cabca04ed60159597e8d32d7349195648b", "date": "2026-08-10T08:02:37+02:00", "subject": "checksrc: check for `assert`, replace in tests where missing, drop from examples", "body": "Follow-up to 791b6883edc1c02c4b253e8b47237ec13ac66d06 #22521\n\nCloses #22527", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "39beb680714add8a4f066bb591179030ea4c2380", "date": "2026-08-10T14:17:06+08:00", "subject": "fix(Select): remove native clear button background", "body": "* fix(Select): reset clear button background\n\n* test(Select): remove redundant style test", "style": "conventional", "type": "fix", "scope": "Select", "breaking": false} {"repo": "vite", "sha": "6452703914bb44e4f428864f15dac7df8e31c9b9", "date": "2026-08-10T14:17:17+08:00", "subject": "docs(build): correct 'rollup' to 'Rolldown'", "body": "", "style": "conventional", "type": "docs", "scope": "build", "breaking": false} {"repo": "postgres", "sha": "6e89028ec487ffe6f2150b3e87af7e983f7295b2", "date": "2026-08-10T08:44:49+02:00", "subject": "Fix errmsg grammar for publication membership", "body": "Using \"a member of\" instead of \"member of\" is consistent with other\nsimilar error messages.\n\nAuthor: Peter Smith \nReviewed-by: Ewan Young \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/CAHut+Pvyj3ad7g-Sb98LDsf1i4c+92rqr+OVG9Lb3fikc_POLg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "39bf0422960291b593b6661cdfd273620006f7b0", "date": "2026-08-10T09:42:01+02:00", "subject": "ci: trigger backports on ready_for_review", "body": "The intended sequence:\n 1. Convert the PR to draft.\n 2. Update branch labels.\n 3. Mark it ready for review.\n 4. Backports CI runs with the updated labels.\n\nSigned-off-by: Nikola Pajkovsky \nReviewed-by: Richard Levitte \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:08:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32260\n(cherry picked from commit 0eb3b1575caed6c622f63bb380b5257618a1fa3b)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0eb3b1575caed6c622f63bb380b5257618a1fa3b", "date": "2026-08-10T09:42:01+02:00", "subject": "ci: trigger backports on ready_for_review", "body": "The intended sequence:\n 1. Convert the PR to draft.\n 2. Update branch labels.\n 3. Mark it ready for review.\n 4. Backports CI runs with the updated labels.\n\nSigned-off-by: Nikola Pajkovsky \nReviewed-by: Richard Levitte \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:08:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32260", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "619c95a8c9d127c26f2f2afa9130eb7a298604ad", "date": "2026-08-10T08:58:55+01:00", "subject": "Stop the DTLS timer when the retransmit budget is exhausted", "body": "dtls1_handle_timeout() fails the connection with SSLfatal() once\ndtls1_check_timeout_num() reports that DTLS1_TMO_ALERT_COUNT unanswered\nretransmissions have been sent. It returns at that point without reaching\ndtls1_start_timer(), so next_timeout is left holding a time in the past.\n\nNothing re-arms or clears it afterwards, the connection being finished, so\nevery later dtls1_get_timeout() computes a negative remaining time, clamps it\nto zero and reports the timeout as due immediately. A caller which waits on\nthat timeout therefore never waits at all. The documented\nDTLSv1_get_timeout() plus select() loop turns into a busy loop on any\nconnection which has given up retransmitting, consuming a core until the\napplication gives up on the connection itself.\n\nStop the timer instead. dtls1_stop_timer() zeroes next_timeout, which\ndtls1_get_timeout() already reports as \"no timer running\", so callers see no\npending timeout and wait on whatever other deadline they have.\n\nThe test drives this without waiting by forcing a retransmission timeout of\n1ms, which is below the 15ms floor at which dtls1_get_timeout() treats a\ntimer as already expired. The server therefore retransmits on every read\nattempt and exhausts its budget inside SSL_accept().\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Tomas Mraz \nMerge-date: Fri Aug 14 15:45:40 2026\nMerged-from: https://github.com/openssl/openssl/pull/32239", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0670df4cbb543f5b23a4ecb6ff05fbbb5adb8e5f", "date": "2026-08-10T08:59:06+01:00", "subject": "Wait consistently across platforms when there is nothing to poll", "body": "poll_translate_ssl_dtls_conn() adds no file descriptor to the poll set when\nthe connection's BIO cannot supply one, which is the case for the memory and\ndatagram-pair BIOs the test suite is built on. Unlike the QUIC and DTLS\nlistener paths it does not report that as an error, and rightly so: a DTLS\nconnection with no pollable BIO has no readiness to wait for, but it still\nhas a retransmission deadline to wake for, so a wait with nothing to watch is\nmeaningful.\n\nThe resulting empty descriptor set was then handed to the operating system,\nwhere the behaviour differs. poll() treats it as a plain sleep, so this\nworked on Unix. Windows' select() rejects a call with no descriptors, so\nossl_rio_poll_builder_poll() returned failure and SSL_poll() failed with it.\n\nSleep out the deadline explicitly rather than relying on the platform. With\nno descriptors and no deadline nothing could ever provide a wakeup, so that\ncombination is reported as a failure instead of sleeping for ever.\n\nThis has no test of its own: on any platform whose poll() already sleeps, the\nbehaviour before and after is identical, so only the Windows CI builds\ndistinguish them.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Tomas Mraz \nMerge-date: Fri Aug 14 15:45:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/32239", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ad0c0af59157b51f5e7eae7f6e69fd7f905b3026", "date": "2026-08-10T10:00:08+02:00", "subject": "move the test code outside of the #ifdef", "body": "during the test backportg, the test was accidentally placed into\nifdef\n\nFixes: 69de45cd12e5 (\"Reject clearly degenerate RSASVE parameters.\")\nSigned-off-by: Nikola Pajkovsky \n\nReviewed-by: Tomas Mraz \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Mon Aug 10 10:34:32 2026\n(Merged from https://github.com/openssl/openssl/pull/32261)\n\n(cherry picked from commit 190a56c010a306d8e03216a1db5c4d495f52cb94)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "190a56c010a306d8e03216a1db5c4d495f52cb94", "date": "2026-08-10T10:00:08+02:00", "subject": "move the test code outside of the #ifdef", "body": "during the test backportg, the test was accidentally placed into\nifdef\n\nFixes: 69de45cd12e5 (\"Reject clearly degenerate RSASVE parameters.\")\nSigned-off-by: Nikola Pajkovsky \n\nReviewed-by: Tomas Mraz \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Mon Aug 10 10:34:32 2026\n(Merged from https://github.com/openssl/openssl/pull/32261)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "92aa95bd47d54d11da66aad9ee0a2900b4eb16f8", "date": "2026-08-10T16:11:27+08:00", "subject": "fix(channel): handle socket accept failure #41240", "body": "(cherry picked from commit 4b8c4dfc680fe0b6c6883ff0e641c1ee7f835711)", "style": "conventional", "type": "fix", "scope": "channel", "breaking": false} {"repo": "neovim", "sha": "4b8c4dfc680fe0b6c6883ff0e641c1ee7f835711", "date": "2026-08-10T16:11:27+08:00", "subject": "fix(channel): handle socket accept failure #41240", "body": "", "style": "conventional", "type": "fix", "scope": "channel", "breaking": false} {"repo": "curl", "sha": "1a17959fc7a80864c9c9b37097856bdb4a465cf8", "date": "2026-08-10T10:17:06+02:00", "subject": "url: fix negotiate/ntlm connection reuse", "body": "Reorder logic when looking for matching connections. Check candidate\nNegotiate/NTLM state first. Require same \"input\" parameters when\nconnection is already authenticated. Same for proxy.\n\nDeny connection reuse for empty usernames for NTLM/Negotiate using\nconnections.\n\nReported-by: Martin Dukek\nCloses #22528", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "8d406ed2acac032d2d7ff560ec26c07d5e284c27", "date": "2026-08-10T03:17:51-05:00", "subject": "fix(defaults): emit events on automatic background change #41242", "body": "Problem:\nAfter #40270, events are no longer emitted from the automatic background\ndetection. This applies not just during startup, but also if the user\nmanually changes the background of their terminal.\n\nSolution:\nSet the background as normal, assuming that a normal terminal will\nrespond within 100 ms. Change test to match expected behavior:\n- BG set during startup won't trigger user autocmds since it runs before\n any user config\n- If the terminal takes longer than 100 ms to respond to initial OSC 11,\n it does trigger the OptionSet, but it is triggered through the normal\n path to ensure values like v:option_new are set #38551\n- BG change after startup still triggers autocmds #41146", "style": "conventional", "type": "fix", "scope": "defaults", "breaking": false} {"repo": "neovim", "sha": "2546741d1b36feb7a2e438428992ff26089da184", "date": "2026-08-10T04:21:31-04:00", "subject": "fix(extmarks): undo-redo of a mark explicitly moved during an edit #41252", "body": "Problem:\nA mark moved by nvim_buf_set_extmark() during an edit is misplaced by\nundo and redo. Only splices (\"edits\") are recorded, and replaying them\nreproduces the shifts they caused, never the explicit set: the mark ends\nup wherever the text pushed it.\n\nSolution:\nWhen an open undo block moves an existing mark, record both positions.\nUndo restores the pre-set position, redo re-applies the set.\n\nPartially reverts 18334a4a0cd9 ; ExtmarkSavePos.row/col were unused\nbecause nothing recorded an explicit move, but now `extmark_set()` does.", "style": "conventional", "type": "fix", "scope": "extmarks", "breaking": false} {"repo": "neovim", "sha": "5287a04be4ed1736ae64ca990f10ffa1d430a0c0", "date": "2026-08-10T04:27:49-04:00", "subject": "perf(mbyte): annotate pure functions #41251", "body": "", "style": "conventional", "type": "perf", "scope": "mbyte", "breaking": false} {"repo": "vitest", "sha": "3cf27ed99eeffdf40d286e57a0c6693e7008f1a0", "date": "2026-08-10T10:35:16+02:00", "subject": "docs: add TestMu AI to sponsors", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "c56424a07906864a3e07ca587804fe6b8bb44385", "date": "2026-08-10T10:39:48+02:00", "subject": "docs: add TestMu AI to sponsors (#10904) [backport to v4]", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "eslint", "sha": "585ef37516c0dc29ddb91ce2a2cdcc46fdbbd610", "date": "2026-08-10T11:27:43+02:00", "subject": "docs: update architecture documentation", "body": "* docs: update architecture documentation\n\n* include `js` language\n\n* clarify terminology\n\n* apply review suggestions\n\n* remove mention of directory traversal and file reading", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "f8975657d7451d8e4291406427e6a8ae5e50c8d5", "date": "2026-08-10T17:44:25+08:00", "subject": "chore: update vscode ts settings", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "a833c83a872f00396d017cc83e368bdaf44fcfbd", "date": "2026-08-10T12:08:52+02:00", "subject": "build: minor debug option message fixes/improvements", "body": "- CMAKE_INSTALL.md: sync up option desc with `CMakeLists.txt`.\n- sync `--enable-debug` configure text with help description.\n- say `--disable-debug` is the default in help description.\n\nFollow-up to a15cf1ea07963ab9d92803e23a96ac5541c6ba0f #22512\n\nCloses #22532", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "94d32ce1a63c67bce7b811207095ee6d41d4caed", "date": "2026-08-10T17:25:45+07:00", "subject": "Add ssup_datum_uint32_cmp for comparing OIDs", "body": "Commit 51cd5d6f0 used ssup_datum_unsigned_cmp for the OID comparator,\nwhich compares entire Datums. That gave wrong results, since the upper\nhalf of a Datum containing a 32-bit type is not reliably zero-extended:\nvalues fetched from tuples are sign-extended, while values returned\nby e.g. oidin() are zero-extended. An OID with the high bit set could\ntherefore compare as larger or smaller depending on its source.\n\nFix by adding a sort support comparator that only looks at the low 32\nbits. Also teach radix sort to normalize Datums the same way. This keeps\nOID eligible for radix sort. Add a regression test that sorts OIDs above\n2^31 from both kinds of source, taken from Peter Geoghegan's report.\n\nReported-by: Zsolt Parragi \nReported-by: Peter Geoghegan \nSuggested-by: David Rowley \nReviewed-by: David Rowley \nReviewed-by: Zsolt Parragi \nDiscussion: https://postgr.es/m/CAN4CZFM2afRzSLJJiWKnRjqPwL7Scm2Wd1xXu140diJGuJtzGg@mail.gmail.com\nDiscussion: https://postgr.es/m/CAH2-WznvXyKRah=MQ3ONGHiN8ahJcqpmtyikxCOO96A4usDPpg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "19733fa0b42caffa92463bb8a8098a40d8963d76", "date": "2026-08-10T17:31:29+07:00", "subject": "Rename sort support integer comparators for clarity", "body": "The previous names used \"unsigned\" for Datum-width abbreviated keys\nand \"signed\" for 64-bit signed integers. This may have made more\nsense when Datum width was platform-dependent, but commit 805c64f84\nadded a 32-bit unsigned variant for OIDs, so it makes even less sense\nnow. Rename all to be explicit about signedness and width.\n\nSuggested-by: David Rowley \nDiscussion: https://postgr.es/m/CAApHDvrNMz5ahN1x=uK9yscTogmLupfLBy-_qMW9Z+k=HE5u_A@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "047308b156190ac037339e29a018ed814b770832", "date": "2026-08-10T12:04:37+01:00", "subject": "image: retain ordered placements in the grid", "body": "Replace the per-cell image marker with sparse placement spans attached to\ngrid lines. A placement owns all of its spans and records the input\nprotocol, application image and placement IDs, z-index, and creation order.\n\nThis retains overlapping image layers without storing a list in every grid\ncell. Grid operations move, split, clip, and remove only the affected spans.\n\nUse the input protocol to determine image/text interaction: later text\ndamages SIXEL spans, while Kitty placements remain and are ordered by their\nz-index. Rendering then adapts that one logical scene for each client,\nrather than changing its semantics according to whether the outer terminal\nuses Kitty or SIXEL.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "4f1411d2c82dd98cd725cea436cc0becd095cc2b", "date": "2026-08-10T18:13:27+07:00", "subject": "docs: pin vitepress to 2.0.0-alpha.18 to fix language switcher", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "tmux", "sha": "f2ca8ade2ab117a871d01d967c9e3e4844239c14", "date": "2026-08-10T12:14:02+01:00", "subject": "sixel: scale source raster bounds", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "37c24015846c98abab4529a17e924cb4a476635f", "date": "2026-08-10T18:47:00+07:00", "subject": "fix(browser): prebundle vite module runner with vitest (fix #10836)", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "curl", "sha": "e9b291cf9bd61622f6589d747036d3880eb28b2d", "date": "2026-08-10T13:47:47+02:00", "subject": "header api: add guards", "body": "Add API guards for curl_easy_header and curl_easy_nextheader.\n\nurldata.h: remove prevhead member as not used\n\nCloses #22530", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f2d0bb7f6951cdc1aa1e4a5cad12424f9e36df9b", "date": "2026-08-10T22:18:49+09:00", "subject": "Fix RSA SHA-512 truncated digest signature mappings", "body": "Register the missing RSA signature OID mappings for SHA-512/224 and SHA-512/256.\n\nThis lets certificate request and X.509 verification resolve the digest and key type.\n\nFixes #32252\n\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 12 10:07:33 2026\nMerged-from: https://github.com/openssl/openssl/pull/32269", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "77aac87ddd8062c70c46aa47f8bd76ac91753c72", "date": "2026-08-10T15:19:54+02:00", "subject": "fix(typecheck): report a checker crash on Windows instead of a spawn failure", "body": "", "style": "conventional", "type": "fix", "scope": "typecheck", "breaking": false} {"repo": "vitest", "sha": "d87c96ee4ffdc4078864d72a9aea5bd2d7f30ed3", "date": "2026-08-10T15:20:10+02:00", "subject": "feat!: share the Vite server between inline projects", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": true} {"repo": "tmux", "sha": "c97ceaffe89b9f9a6fba4efbdb71cd144a8ec4c6", "date": "2026-08-10T14:34:16+01:00", "subject": "image: select backend after terminal setup", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "8a31ffc2d4cc5e2f3fb42b6a41b7db802716615d", "date": "2026-08-10T06:38:11-07:00", "subject": "pg_stat_statements: Fix buffer overflow with query normalization", "body": "Since commit 62d712ecfd94, pg_stat_statements has been underestimating\nthe size of the result buffer possible for a normalized query, in cases\nwhere the query includes many squashable lists, causing the normalized\nquery to write past the allocated area.\n\nThe allocated buffer size forgot to account for the comment appended in\na squashable list, \"/*, ... */\". Instead of trying to track down\nprecisely how much space we need, fix by switch to using an expansible\nStringInfo. This not only fixes the bug, but it also makes the code\nsimpler to follow.\n\nAuthor: Álvaro Herrera \nReported-by: Sajeeb Lohani with TrendAI Zero Day Initiative\nReported-by: Yuelin Wang <3020001251@tju.edu.cn>\nDiagnosed-by: Michaël Paquier \nBackpatch-through: 18\nSecurity: CVE-2026-14676\nDiscussion: https://postgr.es/m/19528-7290dd7e6f7dcc22@postgresql.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "08454e8b2defdf8a0b61e550606f78b8f184f2e6", "date": "2026-08-10T06:38:12-07:00", "subject": "Fix multirange type handling in pg_restore_attribute_stats()", "body": "statatt_get_type() unconditionally converted multirange types to their\nunderlying range type. This choice affected all the type information,\nlike atttypid, atttyptype and operators.\n\nThis made the bounds histogram work correctly (range type is required),\nbut it was wrong for all the other stat kinds. MCV values for a\nmultirange column should be parsed as multirange arrays, not range\narrays. It also made the TYPTYPE_MULTIRANGE check for the range stats\nvalidation as dead code, since atttyptype was always TYPTYPE_RANGE\nafter the conversion due to the centralized statatt_get_type().\n\npg_restore_extended_stats() handles the same case correctly: it keeps\nthe original type and explicitly converts to the range type only at the\npoint where range_histogram_bounds is built.\n\nThe fix of this issue is simple: the multirange-to-range conversion\nneeds to be moved from the centralized statatt_get_type() up to where\nattribute stats build their range_histogram_bounds, matching what is\ndone for extended statistics restore.\n\nThe regression tests for multiranges with attribute stats are extended\nto cover this case.\n\nAuthor: OpenAI Security Research Team\nBackpatch-through: 18\nSecurity: CVE-2026-16238", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "8d428c6e677d05122ad2286f0e8d468bb532acb1", "date": "2026-08-10T06:38:12-07:00", "subject": "Add test with multirange type for pg_restore_attribute_stats()", "body": "This commit adds a test for pg_restore_attribute_stats() with the\ninjection of statistics related to a multirange type. This case is\nsupported in statatt_get_type() since its introduction in ce207d2a7901,\nbut there was no test in the main regression test suite to check for the\ncase where attribute stats is restored for a multirange type, as done by\nmultirange_typanalyze().\n\nNote that this has been originally applied as d4504d6f60e0 in v19 and\nnewer branches. This will be used for the test of a follow-up commit.\n\nAuthor: Corey Huinker \nDiscussion: https://postgr.es/m/CADkLM=c3JivzHNXLt-X_JicYknRYwLTiOCHOPiKagm2_vdrFUg@mail.gmail.com\nSecurity: CVE-2026-16238\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "12a6206864a0ba38dc506644bdb3928beea5256a", "date": "2026-08-10T06:38:17-07:00", "subject": "Guard against overlength time zone abbreviations in to_char().", "body": "While typical abbreviations are only a few bytes long, a user-supplied\ntime_zone setting could specify a much longer abbreviation, enough to\noverflow to_char's allocation of 12 bytes per format character. If so,\nthrow an error in the same style as commit 9241c84cb (CVE-2015-0241).\n\nReported-by: Hcamael \nReported-by: Amjad Shahzad \nReported-by: Tan Zhen of AntAISecurityLab \nReported-by: Tomer Fichman \nReported-by: Zheng Yu \nReported-by: Amy Burnett (OpenAI Codex Security)\nReported-by: Rick de Jager \nReported-by: Heewon Song \nReported-by: Sylvie Mayer \nReported-by: Aleksander Alekseev \nReported-by: Hillai Ben Sasson \nAuthor: Tom Lane \nBackpatch-through: 14\nSecurity: CVE-2026-14669", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "4fcccea972bcdce66976da86535e970dea9b999f", "date": "2026-08-10T06:38:18-07:00", "subject": "Move SplitGUCList() to fe_utils", "body": "This is needed so that all versions of pg_upgrade that migrate logical\nreplication slots can parse the new output_plugin_libraries GUC.\n\nBackpatch-through: 17\nSecurity: CVE-2026-6471", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "067a64d40f92e4e012db0367f9ff2d2ae45671ca", "date": "2026-08-10T06:38:18-07:00", "subject": "Reject GSSEncRequest after direct SSL connection", "body": "When a direct SSL connection was established, ProcessStartupPacket()\nstill accepted GSSEncRequest messages. The GSSAPI negotiation would\nthen use raw writes and reads, bypassing the TLS encryption layer.\nAfter the GSS encryption was established, the connection continued to\nuse TLS. This could betray the HBA rules so as the backend does\nprotocol exchanges inconsistent with the connection policies in place,\nwith TLS taking priority over GSS in the backend.\n\nThe SSL negotiation path already guarded against attempts to request\nSSL after a direct SSL request has been processed. The GSS path is now\nguarded the same way when receiving a startup packet.\n\nReported-by: p4p3r \nAuthor: Michael Paquier \nReviewed-by: Jacob Champion \nBackpatch-through: 17\nSecurity: CVE-2026-14681", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "fadbe882d16112fd8c5ccf2b73eb9621dbfe130a", "date": "2026-08-10T06:38:24-07:00", "subject": "Use value of scram_iterations in mock_scram_secret().", "body": "Presently, mock_scram_secret() always uses\nSCRAM_SHA_256_DEFAULT_ITERATIONS, which poses an observable\nresponse discrepancy hazard when scram_iterations is set to\nsomething else. To fix, use the value of the configuration\nparameter instead, and document that unauthenticated users can\ndiscover the existence of roles with passwords created with\ndifferent iteration counts.\n\nReported-by: Radim Marek \nAuthor: Nathan Bossart \nReviewed-by: Michael Paquier \nReviewed-by: Heikki Linnakangas \nReviewed-by: Jacob Champion \nSecurity: CVE-2026-14672\nBackpatch-through: 16", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1e3014f373c77c2bdd52766b5c14c7ea2653b838", "date": "2026-08-10T06:38:34-07:00", "subject": "Harden tsquery code against overflows.", "body": "The only overflow hazards I could find in tsquery construction\nare in QTN2QT(), which builds a flat tsquery datum from the\nQTNode tree representation used by tsquery_or, tsquery_rewrite,\nand allied functions. There are two:\n\n1. It seems theoretically possible for the outputs of cntsize() to\noverflow an int, so I widened them to size_t. There's no hazard\ncertainly in tsquery_or and friends, but tsquery_rewrite could expand\nthe query tree by large multiples (by replacing many identical\nsubtrees with a large replacement tree), so in a 64-bit machine\nwith plenty of available memory it should be possible to build a\nQTNode tree large enough to cause that. If these counters did\noverflow then we'd under-allocate the output tsquery and have a heap\noverwrite problem. size_t is sufficient, since it's counting the size\nof a subset of an in-memory data structure. We also have to fix the\nTSQUERY_TOO_BIG() macro to not get confused if sumlen exceeds\nMaxAllocSize.\n\n2. fillQT() neglects to check that the new \"distance\" value for a\nQI_VAL item fits into the available 20-bit field. It's quite easy\nto reach this, for example by tsquery_or'ing two near-megabyte-sized\ntsquerys. However, the result is only a corrupt tsquery that does\nnot represent the expected query, so perhaps this doesn't rise to\nthe level of a security bug. Nonetheless it should be fixed.\n\nNote: I followed the practice used in other tsquery code of checking\neach distance value as it's assigned, which means that the last\noperand string could extend past the MAXSTRPOS boundary. This is a\nbit different from the pattern used for tsvectors, which insist that\nthe total data length not exceed MAXSTRPOS and thereby avoid making\nper-item checks. Perhaps that should be harmonized sometime, but for\nnow it's okay for the two types to do this differently as long as\neach one is self-consistent.\n\nAuthor: Tom Lane \nReviewed-by: Amit Langote \nBackpatch-through: 14\nSecurity: CVE-2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7c2ba321d35f18f2c5dc7b5c4a212467f7fb2fc5", "date": "2026-08-10T06:38:34-07:00", "subject": "Harden tsvector code against overflows.", "body": "The core of this patch is to prevent array_to_tsvector() from\ngenerating invalid tsvectors. It did not check for overly-long\nlexemes (so that WordEntry.len fields could overflow), nor did it\ncheck that the total \"datalen\" fits within MAXSTRPOS (so that\nWordEntry.pos fields could overflow, and the number of entries\nin the tsvector could be much more than the normal limit).\nWhile the field overflows couldn't do anything much worse than\nproduce a corrupted tsvector value, a sufficiently large number\nof tsvector entries could cause integer overflows in later\nprocessing, such as tsvectorout.\n\nAnother important fix is to prevent tsvectorrecv() from accepting\ninvalid tsvectors. The main problem there is that it did not\nreject empty-string lexemes. Hence, even though it did (mostly)\nenforce the MAXSTRPOS limit, it could still produce a result\nwith an unreasonable number of tsvector entries, if they were\nprimarily empty strings.\n\nAlso, fix tsvectorout's calculation of its required output\nbuffer size: it was multiplying the string lengths by\npg_database_encoding_max_length() for no reason. That contributed\nto the risk of integer overflow there. With valid tsvector input,\nthere's no risk, but there's still no reason to make the output\nbuffer several times bigger than needed.\n\nI also tried to make a couple of related routines more robust,\nand spent some effort on improving the comments in ts_type.h.\nAlso, standardize on a single spelling of the \"string is too long\nfor tsvector\" message, using %zu instead of an assortment of formats.\nThese changes aren't security per se but came out of inspecting the\ncode for problems.\n\nReported-by: Yuhang Wu and Zhenpeng Lin\nReported-by: Zheng Yu \nReported-by: Hcamael \nAuthor: Tom Lane \nReviewed-by: Amit Langote \nBackpatch-through: 14\nSecurity: CVE-2026-14662", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "967acab87f01933859fd73ee206898b4b320e186", "date": "2026-08-10T06:38:35-07:00", "subject": "Obstruct EXTRACT() field name deparse injection.", "body": "The parser accepts any string as an EXTRACT() field name, but\ndeparsing does not quote and escape it accordingly. To fix, quote\nand escape the field name during deparsing as needed. It might be\na good idea to validate the field name during parsing and\ndeparsing, too, but that is left as a future exercise.\n\nReported-by: Ben Morris in collaboration with Claude and Anthropic Research\nAuthor: Nathan Bossart \nReviewed-by: Tom Lane \nReviewed-by: Etsuro Fujita \nSecurity: CVE-2026-15741\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "74c59d062ad7a99fb90e38390e7d5b90e7a69527", "date": "2026-08-10T06:38:35-07:00", "subject": "ecpg: Fix out-of-bound writes due to processing of invalid bytea data", "body": "ECPG assumes that any bytea data it receives from a backend starts with\n'\\x' as its first two bytes, but a check was missed to enforce that. A\nrogue server sending some garbage bytea data would be able to crash a\nclient, resulting in a client-side DoS, in the most common cases.\n\nReported-by: ylwangtju \nBackpatch-through: 14\nSecurity: CVE-2026-16241", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "aff9dac1c198767d997cd8099f9154169168f272", "date": "2026-08-10T06:38:35-07:00", "subject": "Use palloc_array() in pltcl and plperl to avoid overflow", "body": "Some of these could overflow on 32-bit systems with the right input.\nConvert all cases where we called palloc() with multiplication to fix\nthem. Not all of them were bugs, but it's better to be safe than\nsorry.\n\nReported-by: Tulya Project, Team Dhiutsa, Bitecope Technologies Private Ltd\nBackpatch-through: 14\nSecurity: CVE-2026-14677", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a74aa0854d5e7c6c14c7d31f41bc7adc366f3ef7", "date": "2026-08-10T06:38:35-07:00", "subject": "Fix pg_trgm's picksplit function with all-true datums", "body": "The CACHESIGN.sign field is a BITVECP, not a TRGM, so you should not\nuse GETSIGN() on it. You don't get a compiler warning because the\nGETSIGN() macro includes a cast. It resulted in a bogus read beyond\nend of buffer, which would cause bad split decisions or a crash if\nyou're very unlucky.\n\nReported-by: Mehmet D. INCE \nBackpatch-through: 14\nSecurity: CVE-2026-14678", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "70a3b4e1831c06b0cbce601c3361db5140b9daf1", "date": "2026-08-10T06:38:35-07:00", "subject": "Preserve the owner of extended statistics rebuilt by ALTER TABLE.", "body": "When ALTER TABLE ... ALTER COLUMN TYPE (or any subcommand that rebuilds\nthem) drops and re-creates the extended statistics objects depending on\nthe altered column, the re-created objects were owned by the role running\nALTER TABLE rather than by the original owner of the statistics.\n\nRemember each object's owner before dropping it, and restore it on\nre-creation.\n\nCreateStatistics()'s signature changes and CreateStatsStmt gains a field,\nbut no known third-party code calls the former or constructs the latter.\n\nAuthor: Masahiko Sawada \nReviewed-by: Noah Misch \nBackpatch-through: 14\nSecurity: CVE-2026-6469", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "913fe0c316a6d024434e66f084acbb605293265a", "date": "2026-08-10T06:38:35-07:00", "subject": "Return nulls honestly in aggregate \"combine\" functions.", "body": "numeric_combine() and several other state-combining functions for\naggregates cheated for the case of both inputs being NULL: they\nreturned a null pointer without bothering to mark it as a SQL NULL.\nThis was harmless in the expected usage where the result would be\npassed to the same combine function or a related aggregate final\nfunction. But it's bad news from a security standpoint, because\nnow that value can be passed to an internal-accepting function\neven if said function is strict. While a previous patch prevented\nsuch queries from being issued, it seems like good defense-in-depth\nto expend the few additional lines of code needed to do this properly.\nComparable functions such as array_agg_combine() already do so.\n\nReported-by: Amy Burnett (OpenAI Codex Security)\nAuthor: Tom Lane \nBackpatch-through: 14\nSecurity: CVE-2026-14680", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "155dacbc5479a9bfbcb47e88ff68a758c73b6a8c", "date": "2026-08-10T06:38:35-07:00", "subject": "Reject calls from SQL to functions that take or return type internal.", "body": "Allowing that is a security hole, since there are many different\nfunctions with different ideas of what their \"internal\" argument or\nresult is. We already had a defense against the easy case of\n\"'foo'::internal\", but that turns out to be insufficient. Lock down\nboth function and operator syntax. Also disallow attempts to cast to\nor from type internal; those would mostly fail anyway, but we have\ncreated some holes with features such as CoerceViaIO.\n\nReported-by: Amy Burnett (OpenAI Codex Security)\nAuthor: Tom Lane \nReviewed-by: Robert Haas \nBackpatch-through: 14\nSecurity: CVE-2026-14680", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c7f4628382071ef87f5e7fee8939f09db1e5e022", "date": "2026-08-10T06:38:35-07:00", "subject": "Protect some fixed-size arrays that have FUNC_MAX_ARGS elements.", "body": "The maximum number of arguments allowed for an aggregate function\nis FUNC_MAX_ARGS-1 (since the underlying transfn and/or finalfn\nwill be called with one more argument). parse_func.c failed to\nenforce this, allowing construction of calls that would try to\npass FUNC_MAX_ARGS+1 to the underlying functions, resulting in\na memory stomp in the executor. Add correct checking there.\n\nSince it's possible that a bad call has been stored in a view or\nSQL function, also add checks in various aggregate-related and\nwindow-function-related code that there are not more than\nFUNC_MAX_ARGS arguments. These will also protect us against the\npossibility that we're trying to run a stored view that was made\nby a server executable with different FUNC_MAX_ARGS. (Arguably,\nthat scenario does not qualify as a security problem. But let's\njust tighten up all of this while we're here, rather than split\nhairs over whether an overrun is reachable.)\n\nLikewise check in compute_function_hashkey. Here the hazard is\ndirectly from a pg_proc row, but the scenario is the same.\n\nPL/Tcl has a similar issue with a fixed-size string buffer.\nLet's just replace that buffer with a Tcl_DString, removing the\nwhole issue and making the code look more like what's around it.\n\nThere are a lot of other FUNC_MAX_ARGS-sized arrays, but the rest\nhave nearby guards already, some with comments explicitly pointing\nout the hazard of FUNC_MAX_ARGS changing.\n\nI also used palloc_array() in a few related places in funcapi.c.\nThose aren't live hazards AFAICS, but nearby code has been\npalloc_array-ified already, so it seemed inconsistent to not use\nit here.\n\nReported-by: Masahiko Sawada \nAuthor: Tom Lane \nReviewed-by: Masahiko Sawada \nBackpatch-through: 14\nSecurity: CVE-2026-14679", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "57aa21f69127d21e284c5db8eb488b1eccbb6a1f", "date": "2026-08-10T06:38:35-07:00", "subject": "pg_dump: avoid assuming how long pg_proc.protrftypes can be.", "body": "The backend doesn't impose any particular limit on the length of this\narray, and since there could be entries for both input and output\narguments, it's feasible for the length to exceed FUNC_MAX_ARGS\neven without funny business. This could lead to crashes or worse.\n\nMoreover, pg_dump shouldn't rely on hard-coding FUNC_MAX_ARGS in the\nfirst place: it has no business assuming that the backend it's dumping\nfrom was compiled with the same value of FUNC_MAX_ARGS that it is.\nSo the stanza in dumpFunc() that allocates exactly FUNC_MAX_ARGS space\nfor the parsed OID array is fundamentally misguided. And it's broken\nin another way too: if there are exactly FUNC_MAX_ARGS OIDs, then\nparseOidArray won't zero-fill any entries, allowing the subsequent\nloop to run off the end of the array. A crash seems unlikely in\nthis variant, but garbage output is certain.\n\nTo fix, redesign parseOidArray's API so that it does the\narray-mallocing, which simplifies the callers anyway. While we're\nhere, tighten and modernize it a bit; in particular, split it into\nseparate functions for OIDs and integers, as was foreseen long ago.\nThis lets us get rid of the confusing type-punning involved in\nhaving IndxInfo.indkeys be declared as \"Oid *\" when it's really\npotentially-signed ints. Also, most of the callers expect an exact\nnumber of array entries, so make it verify that not just check for\n\"too many\".\n\nI noted while testing that this dumpFunc() stanza isn't even reached\nduring check-world. Add a function with transform to the regression\ntests to rectify that.\n\nReported-by: Masahiko Sawada \nAuthor: Tom Lane \nReviewed-by: Masahiko Sawada \nBackpatch-through: 14\nSecurity: CVE-2026-19385", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "cf44ff5e66353264451ac239b39e9053e8d138e6", "date": "2026-08-10T06:38:35-07:00", "subject": "Replace fixed-size, too-short array with a palloc'd one.", "body": "MatchNamedCall's arggiven array was declared FUNC_MAX_ARGS long,\nbut we may actually use up to pronallargs elements, and that can\nbe more than FUNC_MAX_ARGS if the function has OUT arguments\n(cf. ProcedureCreate). Convert it to a palloc'd array.\n\nReported-by: Zheng Yu \nReported-by: ylwangtju \nAuthor: Tom Lane \nReviewed-by: Michael Paquier \nReviewed-by: Masahiko Sawada \nBackpatch-through: 14\nSecurity: CVE-2026-14679", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "59205c7e9b4a3afbdab4d677c16036cdd1bd080d", "date": "2026-08-10T06:38:35-07:00", "subject": "Be more wary about constant's datatype in scalarineqsel().", "body": "The special case here for estimating conditions involving a ctid\ncolumn failed to check that the RHS constant is of type tid.\nWhile that'd always be true for the built-in operators that\nreference this selectivity estimator, a maliciously constructed\noperator could provide a user-controlled Datum value that would\nget interpreted as an ItemPointer pointer. That at least risks\nSIGSEGV, and perhaps with a bit of sweat it could be used for\nserver memory disclosure.\n\nReported-by: Hcamael \nAuthor: Tom Lane \nReviewed-by: Noah Misch \nBackpatch-through: 14\nSecurity: CVE-2026-14668", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d7fcfead3da81843a800f8bb046d4cb09d1c662d", "date": "2026-08-10T06:38:35-07:00", "subject": "Harden PL/Perl code against \"tied\" Perl arrays and hashes.", "body": "Tied arrays might report different sizes each time they are inspected.\nTo avoid generating a corrupt result array, fix plperl_array_to_datum()\nto read av_len() of each input array only once. If the input does\nappear to get shorter, we'll fill nulls for the now-missing entries,\nwhich seems fine. Conversely, if it gets longer, we'll ignore the new\nentries.\n\nplperl_to_hstore() assumed that Perl's hv_iterinit() returns the\nnumber of entries in the given Perl hash. Usually that's true,\nbut per the Perl docs, \"the return value is currently only meaningful\nfor hashes without tie magic\". That could potentially end in a memory\nstomp. We don't depend on that result value anywhere else, so don't\ndo so here either.\n\nReported-by: Hcamael \nAuthor: Tom Lane \nReviewed-by: Andrew Dunstan \nBackpatch-through: 14\nSecurity: CVE-2026-14670", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "890327639962a0ebbb2cc2b81ba747cc381876c3", "date": "2026-08-10T06:38:35-07:00", "subject": "Fix potential buffer overrun in regexp match/split functions.", "body": "setup_regexp_matches() sizes the buffer used to convert matched\nsubstrings back from pg_wchar form at the smaller of maxlen*eml and\nthe original string's byte length, on the assumption that such a\nconversion cannot produce more bytes than the string it came\nfrom. That assumption holds only for validly encoded input. But\npg_mb2wchar_with_len() silently accepts bytes that are invalid in the\ndatabase encoding, turning each such byte into one pg_wchar, and\nconverting that back can take more bytes than the input did. A string\nmade of such bytes therefore overruns the conversion buffer by up to\nits own length, corrupting the following memory. regexp_match(),\nregexp_matches(), regexp_split_to_table() and regexp_split_to_array()\nare all affected.\n\nFix by dropping the tighter bound and always allocating maxlen*eml + 1\nbytes.\n\nReported-by: Francesco Verardi \nAuthor: Masahiko Sawada \nReviewed-by: Tom Lane \nBackpatch-through: 14\nSecurity: CVE-2026-14664", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "42d333a78233599a25b34b66af940a87446d58c5", "date": "2026-08-10T06:38:36-07:00", "subject": "Fix out-of-bound reads with ascii() for invalid multibyte characters", "body": "This commit addresses two defects in this SQL function, the code\nassuming that:\n- The user-supplied string was long enough to contain a character of the\nlength implied by the first byte. It is possible to provide in input\ndata that was able to disclose a few bytes of server memory, allowing\nout-of-bound reads.\n- Specific bytes had values within the expected range, using a set of\nassertions to validate them. The assertions could be triggered on\ninvalid input. These are replaced by tests and error reports.\n\nReported-by: Hcamael \nAuthor: Michael Paquier \nReviewed-by: Robert Haas \nBackpatch-through: 14\nSecurity: CVE-2026-18024", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5c51ae4556f1e058779f48f95d3ddcf305fb6e8f", "date": "2026-08-10T06:38:36-07:00", "subject": "Teach psql to skip in-line COPY ... FROM STDIN data after a failure.", "body": "If the COPY command fails before sending PGRES_COPY_IN, psql did\nnot realize that it ought to consume any in-line data following\nthe command. Failing to do so leads to trying to execute that\ndata as SQL commands, which in the best case is wrong and in the\nworst case is a SQL-injection hazard.\n\nTo fix:\n\n1. Extend psqlscan.l to recognize COPY ... FROM STDIN. This can\nbe done with a pretty simple extension to the logic that already\nrecognizes nested BEGIN blocks within CREATE FUNCTION et al.\nBut unlike that case, we need to consider and count multiple COPY\ncommands within a single query string (separated by \"\\;\"). The\nfallout from that is that psql_scan_reset must now always be called\nbefore starting a new query string. (The comment for it that claimed\nwe didn't need that because \"the scan state must be INITIAL\" was\nreally obsolete already, since it has long reset more state besides\nstart_state.)\n\n2. Teach handleCopyIn() to read and discard data when passed\nNULL for \"conn\".\n\n3. Add logic to SendQuery() to call handleCopyIn() that way\nif the query string contained COPY ... FROM STDIN command(s)\nthat remain unaccounted-for at the end.\n\nNow that we have this counting logic, we can also detect\nif the backend sends an unexpected PGRES_COPY_IN message.\nThat should never happen, but perhaps a malicious server\ncould try to extract data that way.\n\nA side-effect of doing this is that we have to adjust a number of test\nscripts that thought they needn't write \"\\.\" after a COPY FROM STDIN\nthat they expect to fail. On the whole this is an improvement, since\nthere's now a uniform rule \"write \\. after COPY FROM STDIN, whether\nyou expect it to work or not\". But it is an annoying amount of test\nchurn.\n\nA loose end in this patch is that if it has to skip data, it assumes\nthat that data is text not binary. It seems unduly difficult to\ndetect whether the COPY command requested binary (we could handle the\nold-style COPY BINARY ... syntax, but not the new style with format\noptions). ", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7215c7e9643f21c51a73b0d2eecd0bc3110b683c", "date": "2026-08-10T06:38:36-07:00", "subject": "Save/restore more lexer state when skipping text due to \\if.", "body": "When we implemented \\if ... \\endif in psql, we arranged to\nsave/restore the lexer's parenthesis depth counter across any chunk\nof input that we're ignoring. At the time, that was sufficient,\nbecause no other part of PsqlScanState could need to be restored to\nits prior value. However, commit e717a9a18 and follow-ons added\nmore state fields that ought to be restored to their prior values.\nA problem would only be observed if someone tries to \\if out a\nportion of a CREATE FUNCTION/PROCEDURE command that is relevant to\nBEGIN/END matching, which seems like a pretty unusual usage, so the\nlack of field reports isn't surprising. Nonetheless it's a bug.\n\nTo fix, replace the simple counter field in ConditionalStack\nentries with a pointer to a struct defined by psqlscan_int.h.\n(In the back branches, keep the old field and associated functions\nto minimize the risk of API/ABI breakage, even though it seems\nunlikely that any third-party code is using this. Making the\nnew struct private to psqlscan-related code should prevent API/ABI\nissues for future additions of this type.)\n\nIn itself this is only a minor bug fix, but it's prerequisite\ninfrastructure for the fix for CVE-2026-6464, which will add\nanother such field.\n\nAuthor: Tom Lane \nReviewed-by: Noah Misch \nBackpatch-through: 14\nSecurity: CVE-2026-6464", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f4174aa84a3ce73ac8dd20f0a6b495d00028d3f0", "date": "2026-08-10T06:38:36-07:00", "subject": "Invalidate plan cache after role changes.", "body": "Role membership, role attribute, and database ownership changes may\nimpact the expected behavior of row-level security policies, but\ncurrently the plan cache doesn't take notice. To fix, register\nsyscache callbacks on pg_auth_members, pg_authid, and pg_database\nthat invalidate the role-dependent plans. Changes to other\ndatabases' pg_database rows are ignored.\n\nReported-by: Ilya Staroverov \nReported-by: Shinya Kato \nAuthor: Ilya Staroverov \nAuthor: Shinya Kato \nCo-authored-by: Nathan Bossart \nReviewed-by: Tom Lane \nSecurity: CVE-2026-14666\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b6e45b4f82481b54134640887fc1cdf4791f9b3a", "date": "2026-08-10T06:38:36-07:00", "subject": "Check for USAGE privilege on the composite type in ALTER TABLE OF.", "body": "This omission allowed roles without USAGE on a type to create\ntables that depend on it, which could prevent the owner from\nchanging the type later.\n\nReported-by: Nathan Bossart \nAuthor: Nathan Bossart \nReviewed-by: Robert Haas \nSecurity: CVE-2026-6470\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1a358b8f2a28aaf3a7bfc8cf710de7a39c98436d", "date": "2026-08-10T06:38:36-07:00", "subject": "Check for USAGE privilege on types used by stored expressions.", "body": "This omission allowed roles without USAGE on a type to create\nstored expressions that depend on it, which could prevent the owner\nfrom changing the type later.\n\nThe checks deliberately live in the command paths rather than the\ndependency-recording routines. Those routines also run whenever\nthe server re-derives an existing expression, and re-checking there\nwould break routine maintenance for an owner who has since lost\nUSAGE on a type its objects already reference. (Checking in the\ndependency-recording routines would also require additional\nCommandCounterIncrement() calls to avoid spurious errors.)\n\nThe addition of a parameter to AlterDomainAddConstraint() breaks\nABI compatibility, but we are unaware of any impacted third-party\ncode.\n\nReported-by: Noah Misch \nAuthor: Nathan Bossart \nReviewed-by: Noah Misch \nReviewed-by: Tom Lane \nReviewed-by: Robert Haas \nSecurity: CVE-2026-6470\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "323af0a4ee5e7d2ccebb85d4bd5615a1a6b6fd1f", "date": "2026-08-10T06:38:36-07:00", "subject": "Check for USAGE privilege on the subtype in CREATE TYPE AS RANGE.", "body": "This omission allowed roles without USAGE on a type to create range\ntypes that depend on it, which could prevent the owner from\nchanging the type later.\n\nReported-by: Jingzhou Fu \nAuthor: Nathan Bossart \nReviewed-by: Noah Misch \nReviewed-by: Robert Haas \nSecurity: CVE-2026-6470\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9505175f2a782bb63acd11d8d473a0ccedb719ea", "date": "2026-08-10T06:38:36-07:00", "subject": "Avoid overflow in Levenshtein distance calculations.", "body": "levenshtein() and levenshtein_less_equal() let the caller specify\nthe insertion, deletion, and substitution costs, and\nfuzzystrmatch's corresponding SQL functions accept any 32-bit\ninteger for each. Since the distances are calculated with 32-bit\narithmetic, large costs can cause overflows, thereby producing\nnonsensical results. Certain inputs to levenshtein_less_equal()\ncan even cause out-of-bounds writes. To fix, use 64-bit arithmetic\ninstead, and error whenever the final result won't fit in the\nreturned 32-bit integer.\n\nWe may want to teach these functions to reject negative costs, too,\nbut that didn't seem appropriate for a security fix, and therefore\nit is left as a future exercise.\n\nReported-by: Ben Morris in collaboration with Claude and Anthropic Research\nAuthor: Nathan Bossart \nReviewed-by: Dean Rasheed \nSecurity: CVE-2026-15742\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1f49beef271ef920f129fcb8513c56347803444a", "date": "2026-08-10T06:38:36-07:00", "subject": "Cross-check the type of a portal running EXECUTE or FETCH.", "body": "When an EXECUTE or FETCH statement is executed, there are two portals:\nan outer portal that is created for the EXECUTE or FETCH statement itself,\nand an inner portal for the statement being executed on its behalf.\nBefore this commit, nothing checked that these two portals agreed on\nthe tuple descriptor of the rows being returned. This can be leveraged\nto disclose server memory contents and achieve arbitrary code execution.\n\nTo prevent that, we can make use of an existing safety mechanism,\nadded by Tom Lane in commit 2f48ede080f42b97b594fb14102c82ca1001b80c,\nwhich allows a tuplestore DestReceiver to be informed of the tupleDesc\nrequired by the caller, and which will cause an ERROR to occur if\nthat doesn't match the tupleDesc of what emerges from the executor\n(modulo dropped columns, which aren't an issue in the case at hand).\n\nReported-by: Ben Morris in collaboration with Claude and Anthropic Research\nReported-by: Peter Geoghegan \nReviewed-by: Michael Paquier \nSecurity: CVE-2026-16239", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "bf3842a64f5df489fab31b82aaeb9067ac938fde", "date": "2026-08-10T06:38:36-07:00", "subject": "Add an output_plugin_libraries GUC to bless trusted output plugins", "body": "REPLICATION users were not previously subject to restrictions on output\nplugin paths, so they were able to bypass LOAD-time protections during\nlogical decoding. Unfortunately, adding the standard LOAD restrictions\nnow would retroactively require all third-party output plugins to be\ninstalled under the $libdir/plugins directory. This would prevent the\nuse of dynamic_library_path, introduce a wire incompatibility for\nclients, and require all plugin authors to check that their libraries\nare safe for use by any unprivileged user; we want to avoid that.\n\nInstead, introduce an output_plugin_libraries GUC so that DBAs can\nspecify the output plugins that are trusted for use in logical decoding.\nFor simplicity, superusers are subject to the restriction as well\n(though they're free to modify the GUC at will during a session, so no\npower is actually lost).\n\nThe default setting is 'pgoutput, test_decoding'. If other third-party\nplugins are in use, DBAs will need to modify this parameter after they\nupdate. Some pointers have been added to the documentation to assist\nwith this.\n\nAuthor: Jacob Champion \nReported-by: Vladimir Tokarev \nReported-by: Yu Kunpeng \nReviewed-by: Álvaro Herrera \nReviewed-by: Noah Misch \nReviewed-by: Robert Haas \nReviewed-by: Tom Lane \nBackpatch-through: 14\nSecurity: CVE-2026-6471", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "39d792040c7cc02b8a805b96b718e723a15f1641", "date": "2026-08-10T06:38:36-07:00", "subject": "Empty search_path in amcheck.", "body": "A grantee of amcheck function EXECUTE privilege could execute arbitrary\nfunctions as the owners of expression indexes that depend on the search\npath. An expression like (lower(col)) was not vulnerable, because\nlower() is resolved at CREATE INDEX time. However, an expression\ncalling an sql-language or plpgsql-language function often was\nvulnerable, even if it used search_path only to find objects in\npg_catalog. The amcheck documentation has been warning about data\ndisclosure after such a GRANT, not about function execution.\n\nThis might cause new amcheck errors when index expressions rely on a\nbroader search_path. Such indexes have seen errors during auto-analyze\nsince CVE-2018-1058 commit 582edc369cdbd348d68441fc50fa26a84afd0c1a, and\nv17 amcheck always worked this way. Hence, the risk is low.\n\nLeave a comment on the one other sandbox entrance that doesn't empty\nsearch_path. In its case, the choice was valid.\n\nBack-patch to v14 (all supported versions), but v17 was safe already.\nCommit 2af07e2f749a9208ca1ed84fa1d8fe0e75833288 (v17) unintentionally\nblocked the attack, and commit d70b17636ddf1ea2c71d1c7bc477372b36ccb66b\n(v18) unintentionally removed that protection. Hence, this adds to v17\njust a test and a comment. While emptying search_path became more\nwidespread in commit 2af07e2f749a9208ca1ed84fa1d8fe0e75833288 (v17),\nnone of its other changes blocked an attack available in v16, even when\nconsidering GRANT. For example, brin_summarize_range() has had an owner\ncheck that GRANT does not override.\n\nReported-by: 王跃林 \nReported-by: Jacob Brazeal \nBackpatch-through: 14\nSecurity: CVE-2026-14673", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2006fca401e07fc116a5d1777fc0f22bf49c9fae", "date": "2026-08-10T06:38:37-07:00", "subject": "psql: Don't do backquote expansion in \\unrestrict.", "body": "This oversight in commit 71ea0d6795 allows a malicious server to\ninject shell commands into plain-text dump output that are run at\nrestore time on the machine running psql. To fix, interpret all\ntext after \\unrestrict until the end of the line as its argument.\n\nReported-by: Lucas Velgus \nReported-by: Filip Janus \nReported-by: Daniel Bakker \nAuthor: Nathan Bossart \nReviewed-by: Robert Haas \nReviewed-by: Noah Misch \nSecurity: CVE-2026-18408\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "e2c48c81fcf858266625e835becff3f17478beea", "date": "2026-08-10T06:38:37-07:00", "subject": "pgcrypto: Add option to revert to prior decryption behavior", "body": "The previous commit raises an ERROR during PGP operations if OpenSSL\ndoes not support the cipher in use. However, any existing messages\ncreated with faulty encryption will no longer be accessible via\npgp_[sym|pub]_decrypt().\n\nTo help users out of this situation, add a new ignore-cipher-failure\noption which reverts to the broken behavior during decryption only. A\nfaulty encryption wrapper, created by an OpenSSL configuration that does\nnot support the cipher, can then be stripped back off by that same\nOpenSSL in order to safely reencrypt it. (Note that when OpenSSL does\nsupport the cipher, corrupted messages will not be decrypted regardless\nof the ignore-cipher-failure setting; this is unchanged.)\n\nThe new tests add a corrupted Blowfish message for both public- and\nsymmetric-key decryption, resulting in the following test matrix:\n\n- Blowfish supported, default behavior: fails to decrypt\n- Blowfish supported, ignore-cipher-failure: fails to decrypt\n- Blowfish unsupported, default behavior: fails to load cipher\n- Blowfish unsupported, ignore-cipher-failure: strips faulty encryption\n\nThe previous commit's change to the pubkey tests is expanded similarly:\ncorrectly encrypted messages cannot be decrypted by an OpenSSL that does\nnot support the cipher, regardless of the option's setting, though the\nfailure mode will change.\n\nSuggested-by: Noah Misch \nReviewed-by: Daniel Gustafsson \nReviewed-by: Noah Misch \nSecurity: CVE-2026-14663\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7e29eb6f4a154124602658531df2555e2693e4dc", "date": "2026-08-10T06:38:37-07:00", "subject": "Fix errorhandling for PGP encryption", "body": "PGP encryption was using px_cipher_encrypt without checking if any\nerror was returned. When OpenSSL is running in FIPS mode, or when\nthe legacy provider hasn't been loaded, not all ciphers which are\nsupported by the PGP code are available and fail the init step in\npx_cipher_encrypt. Since the PGP encryption failed to notice this\nit XORed the non-encrypted block with the plaintext, effectively\ndisabling the encryption.\n\nThis was found due to a report of PGP encryption not respecting\nthe pgcrypto.builtin_crypto_enabled flag and allowing Blowfish\nand DES. This however turned out to be a false positive, since\nthe PGP code only use ciphers from OpenSSL and not the built in\nciphers.\n\nBug: #19457\nReported-by: Shishir Sharma \nReviewed-by: Jacob Champion \nDiscussion: https://postgr.es/m/19457-4bab15c17aea36c7@postgresql.org\nSecurity: CVE-2026-14663\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "c2426fcf083ff8d160bc8f81eadabc7a40f96c34", "date": "2026-08-10T15:08:26+01:00", "subject": "image: preserve fractional cell extents", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "7bd1ba9cad3c900289768ce2d099f2de145f64e2", "date": "2026-08-10T15:12:47+01:00", "subject": "sixel: preserve fractional raster extents", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "e805bafbe781b96bdf37b684c5b0618863f177c2", "date": "2026-08-10T17:13:34+03:00", "subject": "fix(docs-infra): contain scroll within version picker dropdown", "body": "Scrolling to the top or bottom of the version picker's dropdown list\nlet the scroll event chain into the page behind it, causing the whole\napp to scroll. The search dialog and search history dropdowns already\nguard against this with overscroll-behavior: contain; apply the same\nfix to the version picker.\n\n(cherry picked from commit 768c8c29dc3388017bcef37ce5eb01c15e13533f)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "768c8c29dc3388017bcef37ce5eb01c15e13533f", "date": "2026-08-10T17:13:34+03:00", "subject": "fix(docs-infra): contain scroll within version picker dropdown", "body": "Scrolling to the top or bottom of the version picker's dropdown list\nlet the scroll event chain into the page behind it, causing the whole\napp to scroll. The search dialog and search history dropdowns already\nguard against this with overscroll-behavior: contain; apply the same\nfix to the version picker.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "tmux", "sha": "4c3983e286a4484eb13a50c3948492f9ae0958ba", "date": "2026-08-10T15:46:47+01:00", "subject": "image: add manual visual checks", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d14b7361938d9695f10b0581f5875ca701146961", "date": "2026-08-10T15:52:00+01:00", "subject": "DTLS 1.3 Wait rather than spin in the blocking accept path", "body": "SSL_accept_connection() looped calling ossl_dtls_tick() with nothing in\nbetween, so it only avoided spinning if the application supplied a\nblocking network BIO and each tick slept inside BIO_recvmmsg().\n\nA listener cannot require that. It demultiplexes one socket to many\nconnections and ossl_dgram_demux_pump() holds the demux lock across the\nreceive, so a thread blocked in accept stalls reads on every established\nconnection until a datagram for any peer arrives. QUIC takes the\nopposite approach, keeping the socket non-blocking and waiting for\nreadiness instead.\n\nDo the same: set the listener's network BIO non-blocking and wait\nbetween ticks via ossl_dtls_block_until_ready(), which reuses\nSSL_poll()'s blocking machinery. It reads nothing out, so the caller\nre-tests its own condition and waits again if the wakeup was not for it.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Jakub Zelenka \nMerge-date: Mon Aug 17 08:29:51 2026\nMerged-from: https://github.com/openssl/openssl/pull/32324", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a32011d667cd7a35b3931034706716f304faf947", "date": "2026-08-10T16:02:37+01:00", "subject": "DTLS 1.3 Name the accept flag in the no-BIO test comments", "body": "Both tests describe themselves in terms of \"blocking mode\" and\n\"non-blocking behaviour\", when what they actually vary is the\nSSL_ACCEPT_CONNECTION_NO_BLOCK flag. That reads as though they concern a\nmode on the object, which invites confusion with\nSSL_set_blocking_mode(). Name the flag, and say what each case is really\nasserting.\n\nComment only, no functional change.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Jakub Zelenka \nMerge-date: Mon Aug 17 08:29:53 2026\nMerged-from: https://github.com/openssl/openssl/pull/32324", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "4fc6e319bf63f5fbcbbe47f408cf0d8efe1c771f", "date": "2026-08-10T16:13:57Z", "subject": "build: update dependency @types/jsdom to v30", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "18ec528b8c3804f4c101e3a8d091ada7c5a40bb4", "date": "2026-08-10T12:35:49-04:00", "subject": "Update .abi-compliance-history for security fixes.", "body": "Security: CVE-2026-6470, CVE-2026-6471, CVE-2026-6469", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "0df9d5eb6516106188b879536f7f813a9ca8811c", "date": "2026-08-10T10:07:03-07:00", "subject": "fix(core): ensure i18n_util hasOwnProperty checks are safe for property renaming", "body": "`getTIcu` and `setTIcu` performed `hasOwnProperty` checks with string\nliterals (`'currentCaseLViewIndex'` and `'tView'`). Under Closure Compiler\nproperty renaming optimizations, these properties are minified but the\nstring literals are not, causing property lookups and type discriminations\nto fail.\n\nThis commit defines closure-safe property constants using\n`getClosureSafeProperty` so the property names are renamed consistently.\n\n(cherry picked from commit 8925a4ee491aebaf6a1a74880c73dfb20e0a4ba1)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "8925a4ee491aebaf6a1a74880c73dfb20e0a4ba1", "date": "2026-08-10T10:07:03-07:00", "subject": "fix(core): ensure i18n_util hasOwnProperty checks are safe for property renaming", "body": "`getTIcu` and `setTIcu` performed `hasOwnProperty` checks with string\nliterals (`'currentCaseLViewIndex'` and `'tView'`). Under Closure Compiler\nproperty renaming optimizations, these properties are minified but the\nstring literals are not, causing property lookups and type discriminations\nto fail.\n\nThis commit defines closure-safe property constants using\n`getClosureSafeProperty` so the property names are renamed consistently.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "curl", "sha": "1cf6be5a0b4a1db8c501dcda8ac98e09970fd372", "date": "2026-08-10T19:09:17+02:00", "subject": "GHA/linux: replace raw C option with `CMAKE_BUILD_TYPE=Debug` in tsan job", "body": "Ref: #22533\nFollow-up to 2a46df31fdb91851895bc46d81f0065e6cafc80b #18274\n\nCloses #22534", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "43680a50cf289778cba1f20f752ba6a9cb81fd66", "date": "2026-08-10T19:26:15+02:00", "subject": "appveyor: sync Debug options in two jobs", "body": "Also drop `!DEBUGBUILD` from job names, it's redundant now.\n\nFollow-up to 5050196dcc74b8369bcfad35d3b0926d0ff96180 #22513\n\nCloses #22535", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2106c2525450f163640256c9099ae3fa2ce0b57f", "date": "2026-08-10T13:44:21-04:00", "subject": "Reject DSA parameters where N exceeds 512", "body": "Update the FFC parameter validation to explicitly reject DSA configurations\nwhere the N parameter is greater than 512. Previously, the logic accepted any\nN >= 256 if L >= 3072. This change enforces the maximum supported limit for N\nand raises an error to prevent the use of invalid or unsupported parameter\ncombinations.\n\nThis is needed because generate_q_fips186_4() allocates a `md` buffer that is\nat most `EVP_MAX_MD_SIZE` long (64 bytes). If N is greater than 512 then the\nqsize parameter passed in input to generate_q_fips186_4() will be greater\nthan 64 causing internal operations on the buffer to overflow.\n\nThis was found by Red Hat with AISLE, but deemed not a security issue\nbecause there is no reasonable way to cause an attacket to set the qbits\nvalues directly, if this ever happen it is just misuse of an API by an\napplication.\n\nSigned-off-by: Simo Sorce \n\nReviewed-by: Richard Levitte \nReviewed-by: Tomas Mraz \nReviewed-by: Dmitry Belyavskiy \nMergeDate: Tue Aug 18 12:41:55 2026\n(Merged from https://github.com/openssl/openssl/pull/32271)\n\n(cherry picked from commit 1bf213943b0f422ced4b0c2db51f626f488f89c7)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1bf213943b0f422ced4b0c2db51f626f488f89c7", "date": "2026-08-10T13:44:21-04:00", "subject": "Reject DSA parameters where N exceeds 512", "body": "Update the FFC parameter validation to explicitly reject DSA configurations\nwhere the N parameter is greater than 512. Previously, the logic accepted any\nN >= 256 if L >= 3072. This change enforces the maximum supported limit for N\nand raises an error to prevent the use of invalid or unsupported parameter\ncombinations.\n\nThis is needed because generate_q_fips186_4() allocates a `md` buffer that is\nat most `EVP_MAX_MD_SIZE` long (64 bytes). If N is greater than 512 then the\nqsize parameter passed in input to generate_q_fips186_4() will be greater\nthan 64 causing internal operations on the buffer to overflow.\n\nThis was found by Red Hat with AISLE, but deemed not a security issue\nbecause there is no reasonable way to cause an attacket to set the qbits\nvalues directly, if this ever happen it is just misuse of an API by an\napplication.\n\nSigned-off-by: Simo Sorce \n\nReviewed-by: Richard Levitte \nReviewed-by: Tomas Mraz \nReviewed-by: Dmitry Belyavskiy \nMergeDate: Tue Aug 18 12:41:55 2026\n(Merged from https://github.com/openssl/openssl/pull/32271)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a62c45948a9da65a9c46dd79ac5f076f26accc42", "date": "2026-08-10T12:25:30-06:00", "subject": "Fix provider compat test.", "body": "The test added to catch invalid input will fail on older providers\nthat will ignore the \"1x\" in the test and treat it as \"1\".\n\nGate this test on 4.1.0\n\nReviewed-by: Neil Horman \nReviewed-by: Milan Broz \nMergeDate: Tue Aug 11 05:03:01 2026\n(Merged from https://github.com/openssl/openssl/pull/32272)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "f466d6289ebeed6fb0d234527761279dd54fd655", "date": "2026-08-10T15:38:55-03:00", "subject": "compat: use socklen_t for getsockopt length", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "bec8012214afddccf72dfe680a755e394eae6312", "date": "2026-08-10T12:42:40-07:00", "subject": "libcurl.pc: Add Copyright tag to the pkgconf file", "body": "Follow-up to 3f1c033afe008123680306663cc01279e831ac2d #22519", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "6c7993d4d9f8722a869207f86a31f20fa9e8780d", "date": "2026-08-10T22:49:53+02:00", "subject": "build: strip trailing spaces from `libcurl.pc`", "body": "Fixing:\n```\n$ pccritic libcurl.pc\n[...]\n [info ] file has trailing whitespace on one or more lines (style/PC062)\n```\n\nCloses #22536", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "curl", "sha": "3ca86602ef9a2ce8a2e22813736a96ba544471cf", "date": "2026-08-10T23:24:18+02:00", "subject": "spnego_sspi: pass channel bindings on initial context", "body": "HTTP Negotiate can create and send a Kerberos token before receiving a\nserver challenge. The SSPI backend only supplied Schannel endpoint\nbindings when a challenge token was present, causing IIS with EPA\nenabled to reject preemptive Kerberos authentication with\nSTATUS_BAD_BINDINGS.\n\nBuild the SSPI input descriptor independently of the challenge token so\nSchannel endpoint bindings are included on both initial and subsequent\nInitializeSecurityContext calls.\n\nAssisted-by: martin-fzi@users.noreply.github.com\n\nFixes https://github.com/curl/curl/issues/22466\nCloses https://github.com/curl/curl/pull/22537", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "65e2aed04ffd3a312d5c5a7d789f42eb03d50505", "date": "2026-08-11T00:38:09+03:00", "subject": "fix(docs-infra): contain scroll within adev dropdown menus", "body": "Scrolling to the top or bottom of a dropdown let the scroll event chain into\nthe page behind it, causing the whole app to scroll. #70137 fixed this for the\nversion picker, but the same containment was missing on every other menu.\n\nThe API reference filter, the tutorial step list and the update guide's version\ndropdowns already had a scroll container, so they only needed\noverscroll-behavior.\n\nThe social and theme mini-menus had neither a height cap nor overflow, so\noverscroll-behavior alone would have been inert on them. They now share the\nversion picker's max-height/overflow-y/overscroll-behavior, hoisted onto\n.adev-mini-menu, which also keeps their items reachable when the viewport is\ntoo short to fit the whole menu.\n\n(cherry picked from commit dd2971382887feaf123204dc484d76f920bf0a0b)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "dd2971382887feaf123204dc484d76f920bf0a0b", "date": "2026-08-11T00:38:09+03:00", "subject": "fix(docs-infra): contain scroll within adev dropdown menus", "body": "Scrolling to the top or bottom of a dropdown let the scroll event chain into\nthe page behind it, causing the whole app to scroll. #70137 fixed this for the\nversion picker, but the same containment was missing on every other menu.\n\nThe API reference filter, the tutorial step list and the update guide's version\ndropdowns already had a scroll container, so they only needed\noverscroll-behavior.\n\nThe social and theme mini-menus had neither a height cap nor overflow, so\noverscroll-behavior alone would have been inert on them. They now share the\nversion picker's max-height/overflow-y/overscroll-behavior, hoisted onto\n.adev-mini-menu, which also keeps their items reachable when the viewport is\ntoo short to fit the whole menu.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "openssl", "sha": "e11738aa0f7e2ac4ed7c36e15ac69c2c3b8c5159", "date": "2026-08-11T00:00:14+02:00", "subject": "ci: remove enable-ssl3 from prov-compat-label tests", "body": "This is a partial backport of patch\n60c15b2aff (\"Remove support for SSLv3\") to fix CI run.\n\nReviewed-by: Neil Horman \nReviewed-by: Nikola Pajkovsky \nMergeDate: Tue Aug 11 05:21:41 2026\n(Merged from https://github.com/openssl/openssl/pull/32278)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "fb519c53692d4d1bdd7207acaefc95bc73816c10", "date": "2026-08-10T16:42:43-06:00", "subject": "Build the software once before installing it", "body": "install_dev, install_modules and install_runtime each reach the build\nthrough a recursive make of their own, and the sets they build overlap:\napps/libapps.a is one of the libraries and is also needed to link the\ninstallable programs. Under \"make -j\" the three run concurrently, so the\narchive gets built twice at the same time and one job removes it while\nthe other is indexing it, failing the install with a ranlib error.\n\nBuild the installable software in a single make before the install\ntargets run, so they find everything current.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:02:57 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "574749016d6286c7a075041ef70edef7ee50c4be", "date": "2026-08-11T01:16:32+02:00", "subject": "tests: verify AWS SigV4 redirect auth scope", "body": "Closes #22538", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "49be7ee7839885469abe51428088aab900e76d32", "date": "2026-08-10T17:47:33-06:00", "subject": "Stop reloading Pod::Man for every manual page", "body": "pod2man is a perl program that compiles Pod::Man before it reads a\nsingle pod, and there are over nine hundred pages, so formatting the\nmanual spent most of its effort compiling the same module again and\nagain -- 24 seconds of user time and 6 of system time to produce\nabout two and a half seconds of formatting.\n\nFormat the manual in one invocation that loads the module once and\ndivides the pages between a few forked workers. No target depends on\nan individual page, since the install recipes take them as arguments,\nso a stamp is enough to order the work. The per-page rules stay,\ncalling the same script with a single pod, so one page can still be\nbuilt by name, and that path is quicker than pod2man was too.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:10 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ad3d4e4656586471d70726c635b946bbe89944c0", "date": "2026-08-10T17:58:23-06:00", "subject": "Stop reloading Pod::Html for every manual page", "body": "Pod::Html is compiled before the first pod is read, so converting the\nmanual to HTML started nine hundred perl processes to do about a\nsecond of work.\n\nConvert the whole set in one invocation that loads the module once and\ndivides the pages between a few forked workers. Nothing depends on an\nindividual page, since the install recipe takes them as arguments, so\na stamp is enough to order the work, and the per-page rules stay for\nbuilding one page by name. The single page option the VMS and Windows\nbuild files use is unchanged, and both paths convert through the same\ncode.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:11 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "ecc09cd9dfcdd00379e0324f88195d206b951b98", "date": "2026-08-11T08:12:31+08:00", "subject": "fix(runtime-vapor): isolate cached component props and dynamic slots", "body": "Closes #15228\n\nVapor component props can retain lazy subscriptions to live parent sources. When the same parent value controls a v-if branch and is passed as a prop, a cached component can observe a transient value before the branch is deactivated. This can run render effects and flush: 'sync' watchers with invalid props, unlike VDOM where a cached vnode retains its last patched inputs.\n\nCommit dynamic raw prop sources through an effect owned by the KeepAlive branch scope. Cached components read the committed shallow values instead of subscribing directly to the parent sources. When the branch is removed, its scope is paused before queued commit jobs run, so the cached component continues to observe its last committed props. Activation resumes the scope and synchronizes the latest values.\n\nKeep component scopes and their local effects active, preserve raw source ordering and key spelling, reconcile dynamic key additions and removals, and skip the commit bridge for v-once components that already snapshot their raw props.\n\nApply the same commit boundary to dynamic slot descriptors so cached components using useSlots() retain their last committed slot table while deactivated. Static slot entries and slot function closure semantics remain unchanged.", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "2666662920c49e1420c2b7cf585df11d881ea733", "date": "2026-08-11T05:15:59+05:00", "subject": "fix(compiler-vapor): wrap custom directive value in parens", "body": "close #15253", "style": "conventional", "type": "fix", "scope": "compiler-vapor", "breaking": false} {"repo": "vue-core", "sha": "a9e6e806dc03dae0d644c608557f5accd7b13547", "date": "2026-08-11T05:18:10+05:00", "subject": "fix(compiler-vapor): keep the space after a quoted attribute value", "body": "close #15257", "style": "conventional", "type": "fix", "scope": "compiler-vapor", "breaking": false} {"repo": "vue-core", "sha": "b140916b5da8fe2545418651999356778b9d65a8", "date": "2026-08-11T08:41:03+08:00", "subject": "fix(runtime-vapor): preserve vdom props normalization in interop", "body": "Closes #15254", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "ee0709d501071068bea458170a0c356b9e85bf6f", "date": "2026-08-11T08:46:52+08:00", "subject": "demo(FloatButton): demo update", "body": "* demo(FloatButton): demo update\n\n* demo(FloatButton): demo update\n\n---------", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c9aa42db82c3cf436e1e80481e7fb7eb345fb051", "date": "2026-08-10T18:47:02-06:00", "subject": "Make most generated files read-only, and write them atomically", "body": "Generated files were left writable, so a hand edit looked like it had\nworked and was silently lost at the next build. Only the pods and the\ngenerated scripts were protected, by a chmod either side of the\nrecipe, which cost a shell and two processes for each file.\n\nRoute the generators that write to standard output through a helper\nthat collects the output, sets the mode and renames the result into\nplace. Renaming needs no permission on the file it replaces, so there\nis no unlocking step, and a generator that fails part way now leaves\nthe previous file rather than a truncated one with a fresh timestamp\nthat make would believe.\n\nThe perlasm modules name their own output rather than writing to\nstandard output, so they are unchanged and their output stays\nwritable.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:03:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32243", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d29d469becec1f2ac082d82863ae1c29fca9dd97", "date": "2026-08-11T13:30:07+12:00", "subject": "Show estimated number of groups for Incremental Sort in EXPLAIN", "body": "Incremental Sort's costs heavily depend on the estimated number of input\ngroups with equal presorted key values. Overestimations can cause the\nplanner to choose Incremental Sort over Sort when Sort would have\nbeen a better choice.\n\nHere, we add the planner's estimate to EXPLAIN to allow easier\nunderstanding of why Incremental Sort has been chosen.\n\nAuthor: Ilia Evdokimov \nReviewed-by: Enrique Sánchez \nReviewed-by: David Rowley \nReviewed-by: solai v \nDiscussion: https://postgr.es/m/10682fef-3748-43f5-a932-7adcdd9bd2b8%40tantorlabs.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "0cba7fe9f0b7eef456bcf48ca22d8e14b01ccb26", "date": "2026-08-11T10:16:30+08:00", "subject": "docs: fix highlight lines", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "a0cfcf72f8ef8bf0f2f11d553333b9bb31f1d316", "date": "2026-08-11T10:45:36+08:00", "subject": "chore: fix ts errors in build test cases", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vite", "sha": "8fb76752836f61224d3095b502fa237b478a06b2", "date": "2026-08-11T10:50:06+08:00", "subject": "fix(optimizer): close custom extension analysis bundles", "body": "", "style": "conventional", "type": "fix", "scope": "optimizer", "breaking": false} {"repo": "ant-design", "sha": "473d255738c8fe2b81e5b64c328c42b17603eb51", "date": "2026-08-11T11:31:20+08:00", "subject": "refactor(BorderBeam): optimize border style handling", "body": "* refactor(BorderBeam): optimize border style handling\n\n* update\n\n* update\n\n* update\n\n* update\n\n* update FAQ\n\n* update FAQ\n\n* update snap\n\n* update\n\n* add test case\n\n* update test case\n\n---------", "style": "conventional", "type": "refactor", "scope": "BorderBeam", "breaking": false} {"repo": "vue-core", "sha": "fc659725347022e7a105e5b86fd6352e13287d85", "date": "2026-08-11T11:36:48+08:00", "subject": "fix(runtime-vapor): align app unmount lifecycle with vdom", "body": "close #15234", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "eslint", "sha": "55db4791120ae591d88089c43127b7b0e16866d4", "date": "2026-08-11T07:35:08+03:00", "subject": "feat: add checkConditionalExpressions to `no-unmodified-loop-condition`", "body": "* feat: add checkConditionalExpressions to `no-unmodified-loop-condition`\n\n* apply suggestions", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "curl", "sha": "419f6404d46d8f5e404988149a0881d015113bb9", "date": "2026-08-11T06:43:06+02:00", "subject": "schannel: reuse the send buffer", "body": "- Reuse a buffer for schannel_send instead of allocating and freeing a\n buffer for every send.\n\nCloses https://github.com/curl/curl/pull/22540", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "684b57972e1ddf25e076fb36189c60bbcacee635", "date": "2026-08-11T13:48:55+09:00", "subject": "fix: prefer-template invalid autofix creates a tagged template call", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "eslint", "sha": "8e6962219a605c5f5add10953aa31027da839194", "date": "2026-08-11T14:01:03+09:00", "subject": "fix: Prevent no-var autofix when var is shadowed by catch parameter", "body": "* fix: Prevent no-var autofix when var is shadowed by catch parameter\n\nA `var` declaration may reuse the name of a simple catch clause parameter,\nbut the equivalent `let` declaration may not. Directly inside the catch\nblock, `let` is a syntax error. In a nested block, `let` is valid but\nassigns a separate binding and leaves the catch parameter untouched.\n\nThe catch parameter and the `var` are distinct variables in distinct\nscopes, so neither has more than one definition and `isRedeclared` does\nnot detect the conflict.\n\nFixes #21194\n\n* fix: Simplify catch parameter check with getVariableByName\n\nResolve the declared name from the scope of the declaration instead of\nwalking the scope chain looking for a catch scope. A simple catch parameter\nis the only binding a `var` is allowed to shadow, so a name that does not\nresolve to the variable being declared is shadowed by one.", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "f5378b88a974e565f767f2a041972aa942a69c5d", "date": "2026-08-11T11:42:14+05:30", "subject": "content_encoding: exact-match the identity transfer-coding token", "body": "Verified by test 2119\n\nCloses #22541", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "d574d1d59a503b2200f16af84edb79843010c694", "date": "2026-08-11T14:20:16+08:00", "subject": "fix(compiler-vapor): preserve modifiers on merged event handlers", "body": "close #15264", "style": "conventional", "type": "fix", "scope": "compiler-vapor", "breaking": false} {"repo": "vitest", "sha": "90d696d44052d212ecb919fd536a6ab2ff5a3c5f", "date": "2026-08-11T15:20:37+09:00", "subject": "fix(browser): restore dependency reload warning", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "vitest", "sha": "6ac8cf2d3e72bf7bc3f6314677d3b85b3991acb1", "date": "2026-08-11T11:21:36+05:00", "subject": "docs(browser): clarify special keys supported by userEvent.keyboard (closes #7118)", "body": "", "style": "conventional", "type": "docs", "scope": "browser", "breaking": false} {"repo": "ant-design", "sha": "8b446f6abb3c548ceaadeaef8530b14073334931", "date": "2026-08-11T15:16:36+08:00", "subject": "test(BorderBeam): avoid testing internal hook directly", "body": "", "style": "conventional", "type": "test", "scope": "BorderBeam", "breaking": false} {"repo": "ant-design", "sha": "2d319ddd937f6a8d9dcbd01e55937073c7531d7f", "date": "2026-08-11T15:17:02+08:00", "subject": "site: enable FloatButton BackTop scroll progress", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1d1d7b0e9c99a6c7b2bb874b5c7f6806e5940f04", "date": "2026-08-11T09:21:34+02:00", "subject": "Use frontend logging API in fe_utils/string_utils.c", "body": "This simplifies error message printing a bit.\n\nAuthor: Chao Li \nDiscussion: https://www.postgresql.org/message-id/flat/c855d318-4545-4dc7-b466-2e24fbf64725%40eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c83a133f2bb531fe56284e35809c14eed1d8bd6e", "date": "2026-08-11T09:21:34+02:00", "subject": "Make printTableAddCell/printTableAddHeader string argument const", "body": "These functions would sometimes overwrite the string argument they are\npassed, namely via mbvalidate(), which removes invalid UTF-8\ncharacters (or potentially analogously in other encodings, but that is\nnot implemented). However, many callers are not expecting that. In\nmany callers, the input value comes directly from libpq structures,\nsuch as from PQgetvalue() or PQsslAttribute(). The latter actually\nhas a const char * return type, and that was cast away. But even the\nformer is not expecting its return value to be modified.\n\nFix that by making these arguments const. Internally, we add a\nseparate function that does only the checking part of mbvalidate().\nOnly if the validation returns a negative result, we make a copy and\nrun mbvalidate() on the copy. printTableAddCell() already had\ninternal infrastructure for keeping track of what values needed to be\nfreed. We add the same for printTableAddHeader().\n\nIn passing, also simplify the code a bit. There were essentially\nduplicate mechanisms for keeping track of the most recently added\ncell (fields .cell and .cellsadded). Make that consistent by using an\ninteger counter for everything. That makes the code arguably easier\nto read than with the \"current pointer\" approaches.\n\nReviewed-by: Chao Li \nDiscussion: https://www.postgresql.org/message-id/flat/c855d318-4545-4dc7-b466-2e24fbf64725%40eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "59aec847fc187cfe9000fcb071c32ac38bf71948", "date": "2026-08-11T09:21:34+02:00", "subject": "Use frontend logging API in fe_utils/print.c", "body": "This simplifies error message printing a bit. Also adjust the error\nmessages a bit for style, while we're here.\n\nReviewed-by: Chao Li \nDiscussion: https://www.postgresql.org/message-id/flat/c855d318-4545-4dc7-b466-2e24fbf64725%40eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0a92b9b76295cac6ec2ba445523d9422592f0cef", "date": "2026-08-11T09:21:34+02:00", "subject": "Remove useless ENABLE_NLS conditionals", "body": "This already works even without those conditionals, since _() expands\nto a no-op if NLS is disabled.\n\nReviewed-by: Chao Li \nDiscussion: https://www.postgresql.org/message-id/flat/c855d318-4545-4dc7-b466-2e24fbf64725%40eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "370f3366c4c214abb0c34b2e2bd7fd42802112bf", "date": "2026-08-11T09:21:34+02:00", "subject": "Remove useless/confusing const qualifiers", "body": "The const qualifiers removed here did not qualify what is being\npointed to, which is the normal use in function prototypes. Instead,\nthey qualify the variable itself, so that it cannot be changed inside\nthe function. That is itself not wrong, but it's not a normal style,\nand it seems distracting here.\n\nReviewed-by: Chao Li \nDiscussion: https://www.postgresql.org/message-id/flat/c855d318-4545-4dc7-b466-2e24fbf64725%40eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "a2b40db9a83b36ed9da3a16403cf8f040262d73f", "date": "2026-08-11T15:28:17+08:00", "subject": "fix(server-renderer): reject CR in attribute names", "body": "", "style": "conventional", "type": "fix", "scope": "server-renderer", "breaking": false} {"repo": "vitest", "sha": "6093d764a18c21ea171b311d13f004079af092f0", "date": "2026-08-11T09:43:48+02:00", "subject": "fix(spy)!: preserve class mock prototype methods on instances", "body": "", "style": "conventional", "type": "fix", "scope": "spy", "breaking": true} {"repo": "vitest", "sha": "a72f46eac0a2180fbb9d327031ed2153873c4cf3", "date": "2026-08-11T16:44:31+09:00", "subject": "fix(browser): configure internal optimizer dependencies", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "vite", "sha": "2c485d6fd41c15909a5d37407cd2cf001bd8ea59", "date": "2026-08-11T16:45:44+09:00", "subject": "docs: fix grammar in Vite 5.1 blog post", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "9c75eb1ff075455e91d43729827213a159216c4f", "date": "2026-08-11T15:46:32+08:00", "subject": "docs: correct the type of `server.middlewareMode`", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "0441a1da8818cd685504198210c2271500a4e2fd", "date": "2026-08-11T16:08:11+08:00", "subject": "refactor(runtime-vapor): formalize fragment and interop boundaries", "body": "Use explicit fragment role flags and separate the internal bridge interfaces\nwithout changing the unified interop plugin.", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "neovim", "sha": "4a3197f2ad7ef54047264c08f77c8b4f26cac4ed", "date": "2026-08-11T01:08:33-07:00", "subject": "refactor(defaults): edit global cwd on \"1-\" #41254", "body": "Problem:\n`1-` does nothing from a directory buffer, because we are already in the\nbuffer-local CWD. It's also unintuitive that this mapping behaves\ndifferently based on the resolved CWD.\n\nSolution:\nHave `1-` open the global CWD.", "style": "conventional", "type": "refactor", "scope": "defaults", "breaking": false} {"repo": "curl", "sha": "943ac04fc4d7d554d9064c7cd03fbe1830ff25d4", "date": "2026-08-11T10:31:55+02:00", "subject": "CI: dump `libcurl.pc` (and `curl-config`) to log, verify `.pc` with `pccritic` tool", "body": "`pccritic` is available on the latest macOS runner revision (not every\njob runs get it yet), Cygwin, MSYS2/mingw-w64 (one job, others will\nfollow on the next msys2/setup-msys2 Action bump). It will be\nautomatically used on BSDs, once they start offering it. Missing from\nLinux jobs.\n\nAlso:\n- GHA/windows: convert `libcurl.pc` to Unix newlines to bump score\n 96 to 100 in mingw-w64 jobs.\n CMake-specific fix pending in: #22556\n- GHA: verify with minimum `pccritic` score 94 to 100 (depending on CI\n job).\n- build: fill `Maintainer:` `libcurl.pc` property when run curl CI.\n\nIssues remaining:\n\nmacOS:\n```\n[info ] Cflags adds an include path already on the default search path (pkgconf strips it): -I/usr/local/include (cflags/PC031)\n[info ] Libs adds a library path already on the default search path (pkgconf strips it): -L/usr/local/lib (libs/PC040)\n[minor ] Libs.private contains an unexpected flag: /Applications/Xcode_26.6.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX26.5.sdk/System/Library/Frameworks/GSS.framework (libs/PC041)\n```\n\nWindows:\n```\n[minor ] file uses CRLF line endings (style/PC061)\n```\n\nRef: #22556\nFollow-up to b9254da6175360102176303ecf9dbb5519b0ffcf #22548\nFollow-up to 852f5d34da772de65d057897278aa1e75dd060d1 #22553\nFollow-up to 4d9ba9fef48e43903815e0c0e96f917eaf9f185c #22545\nFollow-up to 6c7993d4d9f8722a869207f86a31f20fa9e8780d #22536\nFollow-up to 3f1c033afe008123680306663cc01279e831ac2d #22519\n\nCloses #22543", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "curl", "sha": "45a84dfe0658ccab05ff1a9eee3afc3206fe0374", "date": "2026-08-11T10:38:48+02:00", "subject": "scorecard: report 'rss-max'", "body": "In addition report the maximum of the sampple RSS values. Increase the\nsampling frequency for accuracy.\n\nCloses #22542", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "4d9ba9fef48e43903815e0c0e96f917eaf9f185c", "date": "2026-08-11T11:03:46+02:00", "subject": "libcurl.pc: add `License` tag", "body": "Also to improve `pccritic` score.\n\nIdea-by: Dan Fandrich\n\nFollow-up to 3f1c033afe008123680306663cc01279e831ac2d #22519\nCherry-picked from #22543\n\nCloses #22545", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "2c22d3069aef507d6a6876a6d20616fe5e50c6a3", "date": "2026-08-11T11:19:28+02:00", "subject": "build: strip duplicate spaces after `Libs.private:` in `libcurl.pc`", "body": "When `LIBCURL_PC_LDFLAGS_PRIVATE` is empty and `LIBCURL_PC_LIBS_PRIVATE`\nis not.\n\nTo keep it tidy.\n\nCloses #22544", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "82d6f78a6112edd84878f205dcca351b8da62a43", "date": "2026-08-11T18:25:05+08:00", "subject": "feat(Tree): align `selectable={false}` with disabled style in non-checkable mode", "body": "Bump `@rc-component/tree` to `~1.5.0` and `@rc-component/tree-select` to\n`~1.17.0` to pick up react-component/tree#1066, which adds a\n`treenode-unselectable` class when a non-checkable Tree node has\n`selectable={false}`.\n\nIn non-checkable mode such a node can neither be selected nor checked, so\nalign its visual with `disabled` (disabled text color + not-allowed cursor,\nno hover highlight). Checkable mode keeps the existing behavior — a\n`selectable={false}` node stays checkable and no longer shows the\ndisabled look.", "style": "conventional", "type": "feat", "scope": "Tree", "breaking": false} {"repo": "openssl", "sha": "bfa50e13f857dc85428a693fbad3b60c54c1511b", "date": "2026-08-11T12:26:47+02:00", "subject": "statem: test tls_construct_client_key_exchange", "body": "Call tls_construct_client_key_exchange() directly for each key exchange\nthe dispatcher can select, priming the connection with the state the\nreal state machine would have set up: a minimal session, the negotiated\ncipher and the server key material. The kRSA test decrypts the emitted\npremaster secret and checks it against the saved one, kECDHE/kDHE check\nthe encoded public key and the derived secret, and kPSK checks the\nemitted identity plus the PSK stashed for the secret derivation. Add\ndeterministic tests for the error branches: a missing server\ncertificate or ephemeral key, the PSK callback failures, a WPACKET\noverflow and a cipher matching no key exchange. Out-of-memory branches\nare exercised with the mfail tests.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Ryan Hooper \nReviewed-by: Tomas Mraz \nMerge-date: Mon Sep 7 13:26:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/32299", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "b4d66fee14d970f45b8a6f3d7d6aee73ca9b88ab", "date": "2026-08-11T18:27:53+08:00", "subject": "test(define): close rolldown bundler after generate", "body": "", "style": "conventional", "type": "test", "scope": "define", "breaking": false} {"repo": "vitest", "sha": "94a9b1dbf623b0cf862e18d049627dfa0fac0605", "date": "2026-08-11T19:26:34+08:00", "subject": "docs: normalize ordered list number", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b2f38705af77a209a72478e36ca66731779cb868", "date": "2026-08-11T20:32:51+09:00", "subject": "ci: fix race condition in parallel Perl 5.10 build", "body": "The Perl 5.10 makefile can run multiple configpm recipes concurrently\nduring a parallel build. They update Config.pm and Config_heavy.pl in\nplace, so one process may read a partially written Config.pm and fail.\n\nBuild this old Perl release serially. Other workflow builds continue\nto use all available CPUs.\n\nReviewed-by: Richard Levitte \nReviewed-by: Eugene Syromiatnikov \nMergeDate: Thu Aug 13 11:37:13 2026\n(Merged from https://github.com/openssl/openssl/pull/32302)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vitest", "sha": "99e596d33c76eb2bfa3dc0d8ce1531f0fcd4160a", "date": "2026-08-11T19:35:26+08:00", "subject": "fix(expect): handle nullish toHaveProperty values (fix #10735)", "body": "", "style": "conventional", "type": "fix", "scope": "expect", "breaking": false} {"repo": "vitest", "sha": "658affeb33889cf2bd32dc2d4f289e12cb902c01", "date": "2026-08-11T12:54:22+01:00", "subject": "fix(reporters): print parent describe suites when the passed-test list is hidden (fix #10606)", "body": "", "style": "conventional", "type": "fix", "scope": "reporters", "breaking": false} {"repo": "openssl", "sha": "0ee595cce78c6412abef94d3c761dae53505af85", "date": "2026-08-11T07:59:18-04:00", "subject": "Test EVP_PKEY_get_size for zero return", "body": "The test_rsasve_degenerate_ciphertext calls EVP_PKEY_get_size and uses\nthat returned value as an array index (offset by -1), but fails to check\nif the returned size is greater than zero (which can be returned in an\nerror case), leading to an array index underflow.\n\nEnsure that the returned value is greater than zero\n\nFixes https://scan5.scan.coverity.com/#/project-view/60762/10222?selectedIssue=1699973\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Paul Dale \nMergeDate: Thu Aug 13 11:54:49 2026\n(Merged from https://github.com/openssl/openssl/pull/32309)\n\n(cherry picked from commit 05103ee62c17dfe7de61af2ec91130527643678a)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "05103ee62c17dfe7de61af2ec91130527643678a", "date": "2026-08-11T07:59:18-04:00", "subject": "Test EVP_PKEY_get_size for zero return", "body": "The test_rsasve_degenerate_ciphertext calls EVP_PKEY_get_size and uses\nthat returned value as an array index (offset by -1), but fails to check\nif the returned size is greater than zero (which can be returned in an\nerror case), leading to an array index underflow.\n\nEnsure that the returned value is greater than zero\n\nFixes https://scan5.scan.coverity.com/#/project-view/60762/10222?selectedIssue=1699973\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Paul Dale \nMergeDate: Thu Aug 13 11:54:49 2026\n(Merged from https://github.com/openssl/openssl/pull/32309)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "8a1287ea67975602e2d221ac70b1c1a8ce2ed728", "date": "2026-08-11T20:00:28+08:00", "subject": "fix(locale): refine Traditional Chinese terminology", "body": "", "style": "conventional", "type": "fix", "scope": "locale", "breaking": false} {"repo": "vitest", "sha": "3b5bbd8b42ac51e11dcf7adb49eb105a6dc05d67", "date": "2026-08-11T15:11:54+03:00", "subject": "feat(browser)!: save failure screenshots in `attachmentsDir`", "body": "", "style": "conventional", "type": "feat", "scope": "browser", "breaking": true} {"repo": "vitest", "sha": "2e5df7543f730c53ecf488f1614c7687b0f0dff4", "date": "2026-08-11T14:12:12+02:00", "subject": "fix(vitest): remove broken \"./src/*\" export from package.json", "body": "", "style": "conventional", "type": "fix", "scope": "vitest", "breaking": false} {"repo": "openssl", "sha": "e38098c20a0cb13daedc2ed0ec8c417814a60d1c", "date": "2026-08-11T21:46:40+09:00", "subject": "cms: clear temporary key buffers on failure", "body": "Use OPENSSL_clear_free() with the original allocation length on the\nKARI, KEKRI, and PWRI failure paths. This is defense-in-depth\nhardening at the CMS/provider boundary.\n\nReviewed-by: Tomas Mraz \nReviewed-by: Dmitry Belyavskiy \nMerge-date: Fri Aug 14 06:38:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/32303\n(cherry picked from commit 57b84ed1cf9b6c6fe8230be85869351ce2e8e839)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "57b84ed1cf9b6c6fe8230be85869351ce2e8e839", "date": "2026-08-11T21:46:40+09:00", "subject": "cms: clear temporary key buffers on failure", "body": "Use OPENSSL_clear_free() with the original allocation length on the\nKARI, KEKRI, and PWRI failure paths. This is defense-in-depth\nhardening at the CMS/provider boundary.\n\nReviewed-by: Tomas Mraz \nReviewed-by: Dmitry Belyavskiy \nMerge-date: Fri Aug 14 06:38:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/32303\n(cherry picked from commit de80bb12501c32bd92cfd02b56da778925da992c)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "de80bb12501c32bd92cfd02b56da778925da992c", "date": "2026-08-11T21:46:40+09:00", "subject": "cms: clear temporary key buffers on failure", "body": "Use OPENSSL_clear_free() with the original allocation length on the\nKARI, KEKRI, and PWRI failure paths. This is defense-in-depth\nhardening at the CMS/provider boundary.\n\nReviewed-by: Tomas Mraz \nReviewed-by: Dmitry Belyavskiy \nMerge-date: Fri Aug 14 06:38:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/32303", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "6a99eac1bf83043622d1bff6766cf52eb78c07b2", "date": "2026-08-11T14:54:36+02:00", "subject": "chore(vitest): remove dead code that the \"./src/*\" export was hiding from knip", "body": "", "style": "conventional", "type": "chore", "scope": "vitest", "breaking": false} {"repo": "angular", "sha": "d9620e0f1b85ff5896e4d725c52c5ad6c30e1af9", "date": "2026-08-11T18:25:30+05:30", "subject": "fix(animations): detect object trigger values with Object.hasOwn", "body": "StateValue and AnimationTransitionNamespace.trigger detect the {value,\nparams} object form of a trigger binding by calling hasOwnProperty on the\nbound value. When that value is an object from untrusted data (for example\na parsed JSON payload) carrying an own hasOwnProperty key, the shadowed\nproperty is called as a method and throws, breaking the animation flush.\nUse Object.hasOwn for the check so a shadowing key no longer matters.\n\n(cherry picked from commit c73a001fbf766b350fab4e64244cfaee23fdc8bc)", "style": "conventional", "type": "fix", "scope": "animations", "breaking": false} {"repo": "angular", "sha": "c73a001fbf766b350fab4e64244cfaee23fdc8bc", "date": "2026-08-11T18:25:30+05:30", "subject": "fix(animations): detect object trigger values with Object.hasOwn", "body": "StateValue and AnimationTransitionNamespace.trigger detect the {value,\nparams} object form of a trigger binding by calling hasOwnProperty on the\nbound value. When that value is an object from untrusted data (for example\na parsed JSON payload) carrying an own hasOwnProperty key, the shadowed\nproperty is called as a method and throws, breaking the animation flush.\nUse Object.hasOwn for the check so a shadowing key no longer matters.", "style": "conventional", "type": "fix", "scope": "animations", "breaking": false} {"repo": "vitest", "sha": "81001a49eed24be231b7beb0339773037f0945fb", "date": "2026-08-11T22:06:45+09:00", "subject": "ci: trigger pkg.pr.new on backport branch and workflow dispatch", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vitest", "sha": "7af4b6ed3bb4aa61e008af8d747abd2c1936b389", "date": "2026-08-11T22:07:16+09:00", "subject": "test: remove unused empty fixture configs", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vitest", "sha": "a7fa111fef94bdc80ca9614c4e20b56f3393c920", "date": "2026-08-11T15:43:29+02:00", "subject": "chore: release v5.0.0-rc.1", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5b8501c9c7058de27524b8cf44ee687f75b1ef6a", "date": "2026-08-11T10:17:22-04:00", "subject": "DTLS 1.3 CI nightly NO DTLS 1.2 Failures Fix", "body": "dtlsv1listentest uses DTLSv1_listen(), a DTLS 1.2-only API that crashes\nwhen DTLS 1.2 is disabled. Guard the test in build.info under\n!$disabled{dtls1_2} and add a matching recipe skip so the test is neither\ncompiled nor run when DTLS 1.2 is unavailable.\n\nThe dtls_ccs_reorder test file is specific to DTLS 1.0/1.2 CCS reordering,\nwhich does not apply to DTLS 1.3. The recipe only skipped when all DTLS was\ndisabled, so when both DTLS 1.0 and 1.2 were disabled while DTLS 1.3\nremained, the test still ran. Add a recipe skip for when both DTLS 1.0 and\nDTLS 1.2 are disabled, matching the build.info condition.\n\nssl_ctx_test had two test table entries that set DTLS1_2_VERSION as a max\nprotocol version and expected success. When DTLS 1.2 is disabled that\nversion is not valid for any protocol, so SSL_CTX_set_max_proto_version()\nreturns 0. Guard both entries with #ifndef OPENSSL_NO_DTLS1_2.\n\ndtlstest contained a dead #ifdef OPENSSL_NO_DTLS1_2 block inside\ntest_listen(), which is itself wrapped in #ifndef OPENSSL_NO_DTLS1_2.\nThe inner ifdef could never be true; remove it.\n\nAssisted-by: Claude:claude-sonnet-4-6\nReviewed-by: Tomas Mraz \nReviewed-by: Matt Caswell \nMerge-date: Thu Aug 20 09:55:38 2026\nMerged-from: https://github.com/openssl/openssl/pull/32340", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "1fa9837ec26533512fdcad8baebf249771bd340a", "date": "2026-08-11T16:26:08+02:00", "subject": "chore: reference local knip schema instead of unpkg", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "b9254da6175360102176303ecf9dbb5519b0ffcf", "date": "2026-08-11T16:32:07+02:00", "subject": "build: clear `Require.private` for static-only builds in `libcurl.pc`", "body": "Before this patch static-only libcurl builds stored the list of module\ndependencies in both `Require` and `Require.private`. The new `pccritic`\ntool flagged this as an issue. Having (or not) a duplicate in\n`Require.private` does not change the output created by `pkgconf` and\n`pkg-config`, as tested with both with and without `--static` option.\nThus, fix by deleting the private copy.\n\nFixing:\n```\n$ pccritic libcurl.pc\nbld/libcurl.pc\n score: 63/100 grade: D (0 critical, 0 major, 9 minor, 1 info)\n [minor ] 'libssh2' appears in both Requires and Requires.private (requires/PC051)\n [minor ] 'libidn2' appears in both Requires and Requires.private (requires/PC051)\n [minor ] 'openssl' appears in both Requires and Requires.private (requires/PC051)\n[...]\n```\nRef: https://github.com/curl/curl/actions/runs/31495008715/job/93790403026#step:10:210 (macOS, CM clang OpenSSL torture)\nRef: https://github.com/curl/curl/actions/runs/31495008825/job/93790433646#step:9:216 (mingw, AM ucrt-x86_64 wolfssl c-ares U)\n\nFollow-up to f057de5a1a950a90d1920021db152a4b695f1a8a #13911\nFollow-up to 98e5904165859679cd78825bcccb52306ee3bb66 #5373\nCherry-picked from #22543\n\nCloses #22548", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1c26ae1e1dd1159e785c75ea3ed6f1c72557cc56", "date": "2026-08-11T15:41:03+01:00", "subject": "DTLS 1.3 Support SSL_set_blocking_mode() on a DTLS listener", "body": "A DTLS listener cannot take its blocking behaviour from its network BIO\nthe way an ordinary DTLS object does. It demultiplexes one socket to\nmany connections, so a read for one of them must not be allowed to\nblock and stall the others, and the connections it hands out have no BIO\nof their own to configure - they read from a queue the listener fills.\nBlocking was therefore unavailable to them, SSL_set_blocking_mode()\nbeing QUIC only.\n\nExtend it to a DTLS listener and to the connections created from one,\nwith the same semantics QUIC has. Blocking is the default, a connection\nfollows its listener unless given a setting of its own, and asking for\nblocking fails where it cannot be delivered, which for a listener means\na network BIO with no poll descriptor to wait on.\nSSL_accept_connection() waits only where the caller did not pass\nSSL_ACCEPT_CONNECTION_NO_BLOCK and the listener is in blocking mode,\nmatching what QUIC does with the same flag.\n\nOnly the accept path acts on the mode so far; reads and writes follow.\n\nNote that this does not extend to a DTLS object which did not come from\na listener. Such an object has its own BIO and already offers both modes\nthrough it, so unlike QUIC - where the socket is always nonblocking and\nnothing works without emulation - there is nothing missing to provide.\n\nWhile documenting the above, correct the existing description of which\nQUIC objects these functions apply to. It said connection objects only,\nbut they have always been accepted for QUIC stream and listener objects\ntoo.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Jakub Zelenka \nMerge-date: Mon Aug 17 08:29:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32324", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "35843dd635bab1627902b08a605e2a6334cc2fc7", "date": "2026-08-11T17:09:26+02:00", "subject": "progress: cleanup, less memory", "body": "Keep only a single `start` time and calculate the rest in delta\nmicroseconds since then. Separate timestamps and durations into separate\nsub-structures for clarity.\n\nremove progress.t_acceptdata\nIt was recorded in FTP accept but never used anywhere.\n\nCloses #22547", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "98bc21182155ce8abbfaefc6867db5456f8b30da", "date": "2026-08-11T10:20:18-05:00", "subject": "test(platform-server): clean up viewport hydration observer", "body": "Trigger routed viewport hydration so its cached observer is removed before later randomized specs run.\n\n(cherry picked from commit 870a525bd3c9eaf98d79e426d8ddf2625afe474d)", "style": "conventional", "type": "test", "scope": "platform-server", "breaking": false} {"repo": "angular", "sha": "870a525bd3c9eaf98d79e426d8ddf2625afe474d", "date": "2026-08-11T10:20:18-05:00", "subject": "test(platform-server): clean up viewport hydration observer", "body": "Trigger routed viewport hydration so its cached observer is removed before later randomized specs run.", "style": "conventional", "type": "test", "scope": "platform-server", "breaking": false} {"repo": "ant-design", "sha": "22afd656ed35aa64b092cab32c0ac9b343c355e6", "date": "2026-08-12T00:35:33+08:00", "subject": "fix: use Taiwan week terminology in DatePicker", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "angular", "sha": "ab38005c6e4dcb33348cf90b4ace69f0536583a7", "date": "2026-08-11T16:40:19Z", "subject": "build: update dependency node to v24", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "d22eb386a5651c43c23746b1b3fe28f27e2f05ea", "date": "2026-08-11T11:54:09-05:00", "subject": "refactor(core): distinguish animations that share a name", "body": "animate.leave waits for the longest-running animation before removing an element. When multiple animations use the same keyframe name, their animationend events have the same animationName, so a shorter animation can be mistaken for the longest one and remove the element too early.\n\nTrack the longest animation duration returned by getAnimations() and compare it with the duration from event.animation when available. Keep the existing name/property checks as a fallback for older browsers and computed-style detection.\n\nAllow a 1ms tolerance for rounding differences between CSSOM and Web Animations and add a regression test using values equivalent to fractional calc() durations.\n\n(cherry picked from commit a26fbfa641b965ad581a3919c06d2c51d8b16125)", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "a26fbfa641b965ad581a3919c06d2c51d8b16125", "date": "2026-08-11T11:54:09-05:00", "subject": "refactor(core): distinguish animations that share a name", "body": "animate.leave waits for the longest-running animation before removing an element. When multiple animations use the same keyframe name, their animationend events have the same animationName, so a shorter animation can be mistaken for the longest one and remove the element too early.\n\nTrack the longest animation duration returned by getAnimations() and compare it with the duration from event.animation when available. Keep the existing name/property checks as a fallback for older browsers and computed-style detection.\n\nAllow a 1ms tolerance for rounding differences between CSSOM and Web Animations and add a regression test using values equivalent to fractional calc() durations.", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "66c5ea75449ee52ca4128adcc365272bb384e996", "date": "2026-08-11T20:00:02+03:00", "subject": "refactor(devtools): improve profiler ux", "body": "- Add info tooltips that describe the visualizations and other\nparts of the UI\n- Improve the frame selector by adding axes labels and a\nhorizontal line for frames exceeding 60 fps\n- Use a precise frame rate calculations instead of approximations\n- Introduce improvements to the details panel\n- Improve the bar chart visualization\n- And more", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "openssl", "sha": "71a0acd3a5c200a0590499bcd660f437e9cf5efb", "date": "2026-08-11T11:28:43-06:00", "subject": "Don't do pointer arithmetic on a NULL pointer in OPENSSL_uni2asc", "body": "Reviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:52 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "49b7ee56918c75d4f19d11aa8fb436129d484322", "date": "2026-08-11T11:29:12-06:00", "subject": "Don't hit the legacy strlen case in ASN1_PRINTABLE_type", "body": "Reviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:50 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ea15c2442d3028fb0c6311e577d4b5f6726a4408", "date": "2026-08-11T14:14:58-04:00", "subject": "DTLS 1.3 Fix DTLS_get_data_mtu() and extend MTU test to cover DTLS 1.3", "body": "DTLS_get_data_mtu() included the explicit IV overhead from\nssl_cipher_get_overhead() for AES-GCM and CCM ciphers, but DTLS 1.3 uses\nimplicit IVs so no IV bytes appear on the wire. Fix by subtracting\nEVP_GCM_TLS_EXPLICIT_IV_LEN or EVP_CCM_TLS_EXPLICIT_IV_LEN.\n\nRestructure the MTU test to pin DTLS 1.2 in the PSK cipher loop (previously\nDTLS 1.3 was silently negotiated instead) and add a DTLS 1.3 loop using\ncertificate-based auth. Guard TLS_CHACHA20_POLY1305_SHA256 with both\nOPENSSL_NO_CHACHA and OPENSSL_NO_POLY1305.\n\nAssisted-by: Claude:claude-sonnet-4-6\nReviewed-by: Tomas Mraz \nReviewed-by: Matt Caswell \nMerge-date: Thu Aug 20 09:55:39 2026\nMerged-from: https://github.com/openssl/openssl/pull/32340", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5f003855e7f05bd03c5531a58838540f62b8b19e", "date": "2026-08-11T21:18:45+03:00", "subject": "Fix pg_strupper/lower/title/fold() functions to work with C locale", "body": "Starting with commit 011384ba45, calling to_date() or to_timestamp()\nwith \"TMMonth\" or other TM-prefixed format keyword, with the C locale,\nwould crash.\n\nIn REL_19_STABLE and above, pg_strupper(), pg_strlower(),\npg_strtitle(), and pg_strfold() functions have a special case for the\nC locale, but that was missing in REL_18_STABLE. On REL_18_STABLE,\nthe functions call the libc function even in C locale, even though the\nnative locale object is NULL. On Linux, the underlying libc functions\nwill crash when called with NULL locale. (On macOS, they reportedly\ndo not, but even then it's not clear if they will do what you'd\nexpect.)\n\nThis went unnoticed because until commit 011384ba45, we never called\nthese functions in C locale, all the callers had a special codepath\nfor C locale. We could add a special path in the new callers too, but\nit's an accident waiting to happen, so let's backport the C\nlocale-specific handling from REL_19_STABLE to REL_18_STABLE. Older\nversions did not have these functions at all, hence no problem.\n\nThis applies to REL_18_STABLE only.\n\nReported-by: Masashi Kamura \nReviewed-by: Álvaro Herrera \nDiscussion: https://www.postgresql.org/message-id/OS9PR01MB1317436E07D06281AD1A0452F94DD2%40OS9PR01MB13174.jpnprd01.prod.outlook.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dad6ff265ab7923e7d59341507ef9db627a97008", "date": "2026-08-11T12:26:43-06:00", "subject": "Use an int for the return value of i2a_ASN1_OBJECT", "body": "And as a result in this loop. I changed this to a size_t\nwhich was needed where the variable was used later on and\nmissed this case at the start.\n\nWhile harmless, coverity notices. so let's do it right\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMerge-date: Fri Aug 14 06:29:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/32310", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "841d0cb0f9b8b1b25e7b299ac0eb134074b45a08", "date": "2026-08-11T14:37:01-04:00", "subject": "Release notes for 18.6.", "body": "Farewell, 18.5; we hardly knew ye.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "bc86dc738e0e4c56a0635efd3d18b367b7e41f9b", "date": "2026-08-11T21:53:59+02:00", "subject": "libpq: remove unused field ssl_handshake_started from PGconn", "body": "The ssl_handshake_started field was added by d39a49c1e45 to allow\nconnection_failed() skip retrying with a different SSL negotiation\nmethod, once the server had reported an error after the TLS handshake\nhad already started. That logic was dropped by fb5718f35ff, which\nremoved the ability to fall back from direct to negotiated SSL, and\nwith it the only consumer of the data in this field. Fix by removing\nthe field.\n\nAuthor: Aleksander Alekseev \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CAJ7c6TPkHjVFERp1Fd-idMe=iRJnDX2Nzy_ZOkV=y4OLLgf89Q@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "e9660b1801a0280fa5ba41e7c330cb6d597199ec", "date": "2026-08-11T13:28:57-07:00", "subject": "fix(compiler-cli): correctly resolve symbol for SafePropertyRead in chained optional navigation", "body": "When resolving template symbols for SafePropertyRead in TCBs emitted with optional chaining (strictSafeNavigationTypes: true), SymbolBuilder falls back to finding a TS node matching the AST expression's nameSpan. It then traverses up through parent nodes to find the enclosing expression.\n\nPreviously, the traversal loop checked isAccessExpression(node.parent) without verifying whether node was the accessed member name or the expression receiver. When multiple optional navigation expressions are chained (e.g. route?.data?.['icon']), the parent of ((route)?.data) is an access expression where ((route)?.data) is the receiver. Because isAccessExpression was true, the loop continued ascending into the outer access expression, causing symbol resolution for data to erroneously return the symbol and TCB location of icon.\n\nThis commit refines the parent traversal condition so that it only climbs into a parent PropertyAccessExpression if node is the accessed name (node.parent.name === node), preventing escape into outer receiver expressions.\n\n(cherry picked from commit e8aa222e7d98d2e54c5604107a929a11cb1ef62d)", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "e8aa222e7d98d2e54c5604107a929a11cb1ef62d", "date": "2026-08-11T13:28:57-07:00", "subject": "fix(compiler-cli): correctly resolve symbol for SafePropertyRead in chained optional navigation", "body": "When resolving template symbols for SafePropertyRead in TCBs emitted with optional chaining (strictSafeNavigationTypes: true), SymbolBuilder falls back to finding a TS node matching the AST expression's nameSpan. It then traverses up through parent nodes to find the enclosing expression.\n\nPreviously, the traversal loop checked isAccessExpression(node.parent) without verifying whether node was the accessed member name or the expression receiver. When multiple optional navigation expressions are chained (e.g. route?.data?.['icon']), the parent of ((route)?.data) is an access expression where ((route)?.data) is the receiver. Because isAccessExpression was true, the loop continued ascending into the outer access expression, causing symbol resolution for data to erroneously return the symbol and TCB location of icon.\n\nThis commit refines the parent traversal condition so that it only climbs into a parent PropertyAccessExpression if node is the accessed name (node.parent.name === node), preventing escape into outer receiver expressions.", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "openssl", "sha": "c059db1feb2e070971836338159f289d9b71880e", "date": "2026-08-11T16:40:14-04:00", "subject": "DTLS 1.3 Fix no-dtls1_3 build configuration CI failures", "body": "Guard the DTLS1_3_VERSION entry in dtls_version_table with\nOPENSSL_NO_DTLS1_3 (NULL fallback), matching the pattern used by\ntls_version_table for TLS1_3. Add a min_version clamp for DTLS1_3 in\nssl_check_allowed_versions() to fix SSL_CTX_set_min_proto_version()\nincorrectly returning failure under no-dtls1_3, caused by DTLS inverted\nwire-number ordering making the range check fire after max was clamped to\nDTLS1_2. Guard _dtls13 test functions and their ADD_TEST calls in\ndtlstest.c. Guard DTLS 1.3-only helpers and tests in\ndtlsssllistenertest.c, and move SSL_poll tests outside the DTLS 1.3\nguard so they run for any DTLS version. Add no-dtls1_3 to the daily CI\nchecker matrix.\n\nAssisted-by: Claude:claude-sonnet-4-6\nReviewed-by: Tomas Mraz \nReviewed-by: Matt Caswell \nMerge-date: Thu Aug 20 09:55:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/32340", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b451e44c39bbb65a0460ce1b960810bd425df2a3", "date": "2026-08-11T23:22:47+02:00", "subject": "test: let the kernel pick the DTLS client port in TLSProxy", "body": "For DTLS, TLSProxy picked a random port for s_client to -bind to,\nprobing it for availability first. This was racy in two ways: the\nprobe socket was TCP while s_client binds UDP (the two port spaces\nare independent, so a UDP socket held by a parallel test job was\ninvisible to the probe), and even a UDP probe would leave a window\nbetween closing the probe socket and s_client binding the port.\nWith many tests now running under the DTLS proxy in parallel CI\njobs this caused intermittent BIO_bind EADDRINUSE failures.\n\nInstead, drop -bind and the port probing entirely. The kernel\nassigns s_client's source port atomically on first send, and the\nproxy, whose UDP socket is now left unconnected, learns the\nclient's address from the datagrams it receives, in the same way\nany regular UDP server does.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Tomas Mraz \nReviewed-by: Frederik Wedel-Heinen \nMerge-date: Thu Aug 13 13:42:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/32312", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2e29cb3b815df71f02158dbe538ca7906291aa5c", "date": "2026-08-11T18:17:48-04:00", "subject": "Tweak allocation rule for tzload()'s \"union local_storage\" variable.", "body": "By default, allocate this via malloc, as we've been doing since commit\n62c8421e8. Commit aeb07c55f adopted upstream tzdb's default of\nallocating it on the stack, but that still doesn't seem like a good\nidea for the reasons given in 62c8421e8 (and now memorialized in a\ncomment, in hopes that we don't make the same mistake again).\n\nHowever, under USE_VALGRIND, put it on the stack as upstream does.\nThis accidentally prevents a crash when Python 3.14 is used under\nValgrind. The reasons for that are obscure, and it's most likely\nnot our bug, and even if we figured it out it'd be nice to have a fix\nfor buildfarm member skink now rather than after persuading the guilty\nparty to fix it.\n\nIn the normal non-USE_VALGRIND case, this has no effect on the logic\nin released branches, and it reverts master to match them.\n\nReported-by: Alexander Lakhin \nAuthor: Tom Lane \nDiscussion: https://postgr.es/m/f071e691-5930-4738-9dbd-43ed2da367fe@gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "faaef82348e2e98e03c85f22658ed2d41a023257", "date": "2026-08-12T00:27:22+02:00", "subject": "Add cms_verify fuzzer", "body": "Fuzz CMS_verify() with CMS_NO_SIGNER_CERT_VERIFY\nto exercise signature verification without requiring\ntrusted certificate store.\n\nMotivated by : CVE-2026-45447 which affects PKCS7_verify(),\nand add coverage for similar CMS_verify().\n\nAssisted-by: ChatGPT:gpt-5.6\nReviewed-by: Jakub Zelenka \nReviewed-by: Milan Broz \nMerge-date: Tue Sep 15 15:00:01 2026\nMerged-from: https://github.com/openssl/openssl/pull/32342", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "299607d0fb7a685aa6d15d0f64d58e8fd294fb9e", "date": "2026-08-11T15:32:19-07:00", "subject": "docs: clarify animate.leave child animation behavior", "body": "This updates the docs to be explicit about `animate.leave` and nested element removal order. It clearly specifies that `animate.leave` will only fire nested animations within the same component template.\n\ncloses: #70131\n(cherry picked from commit 3d32db6a493e24c09da9bff6f01c7a73ae804f44)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "3d32db6a493e24c09da9bff6f01c7a73ae804f44", "date": "2026-08-11T15:32:19-07:00", "subject": "docs: clarify animate.leave child animation behavior", "body": "This updates the docs to be explicit about `animate.leave` and nested element removal order. It clearly specifies that `animate.leave` will only fire nested animations within the same component template.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5838806ce696dd0759c5e4b87754e802612135e0", "date": "2026-08-12T01:08:21+02:00", "subject": "Change wal_compression=on to the first of zstd, lz4, pglz", "body": "Previously, wal_compression=on was an alias for pglz, with the\nassumption that users could make an informed choice to pick a better\noption. But in practice, users rarely got to that second step.\n\nMany users don't want to be choosing algorithms - they just want WAL\ncompression, and expect that to work well. The configuration parameter\nis set by administrators, who do not control the workload, and so are\nnot in a position to evaluate the options anyway. Some users may not\neven realize there are other options, as previously \"on\" was the only\nchoice available.\n\nThis change maps \"on\" to non-pglz options, supported by the build. Both\nlz4 and zstd are faster, with a comparable (or better) compression\nratio. We prefer zstd over lz4 - per our testing the better compression\nratio pays for the lower (de)compression speed.\n\nLike for TOAST compression, the value depends on algorithms supported by\nthe PostgreSQL build, with lz4 and zstd being optional. But most builds\nwill have at least one of these external libraries. If neither zstd or\nlz4 is supported, we fallback to pglz.\n\nThis only affects what \"on\" means. Users can still make the informed\nchoice and explicitly select a compression algorithm if it works better\nfor their system. The default value for \"wal_compression\" remains \"off.\"\n\nInitial proposal and patch by wenhui qiu, reviews and patch adjustments\nby Christoph Berg. A number of other people participated in the\ndiscussion. Benchmarks by me.\n\nBackpatch to 19.\n\nAuthor: wenhui qiu \nReviewed-by: Christoph Berg \nReviewed-by: Michael Paquier \nDiscussion: https://postgr.es/m/CAGjGUAL1b=Mwd1SCvLbo+fivEr9KDpFcu4jmqKCZXwT=6CiiGQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "db1b675c0472dcea7372b6b523fc2ffb2ec2d1f9", "date": "2026-08-12T01:49:32+02:00", "subject": "Add test for ASN1_TYPE memory leak in asn1_multi()", "body": "Add an MFAIL regression test for X509V3_EXT_add_nconf().\n\nTargeting bug : (https://github.com/openssl/openssl/pull/32206)\nIn asn1_gen.c asn1_multi() function variable allocated as\nASN1_TYPE *typ = generate_v3() is not freed if\nif (!sk_ASN1_TYPE_push(sk, typ)) branch fails.\n\nThe issue was found by the x509v3 fuzzer MFAIL test.\n\nReviewed-by: Milan Broz \nReviewed-by: Jakub Zelenka \nMerge-date: Mon Sep 7 14:08:11 2026\nMerged-from: https://github.com/openssl/openssl/pull/32330", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "66cdfbae498bb601c3c4a3a03722207b76874754", "date": "2026-08-12T07:50:49+08:00", "subject": "fix(runtime-vapor): avoid redundant transition block resolution", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "75a5f746bee108d111231f09481c81acc9f07ea9", "date": "2026-08-12T08:02:51+08:00", "subject": "chore(runtime-vapor): tree-shake teleport target anchor hydration", "body": "", "style": "conventional", "type": "chore", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "afd63e0ad23ed826aa953977a5306a3703503afa", "date": "2026-08-12T09:09:58+09:00", "subject": "Improve GrantLockLocal() handling on out-of-memory errors", "body": "On the first LockAcquire for a lock tag, LockAcquireExtended() creates\na LOCALLOCK entry in the hash table LockMethodLocalHash, with lockOwners\nset to NULL. lockOwners is then allocated in TopMemoryContext.\n\nIf the allocation fails, the hash entry would persist with lockOwners\nstill NULL and maxLockOwners already set to 8. A follow-up\nLockAcquire() on the same entry would crash on a NULL pointer\ndereference.\n\nThis commit adds an extra layer of safety for such a case, by\nre-attempting MemoryContextAlloc() if a LOCALLOCK entry exists but its\nlockOwners is still NULL due to a previous in-flight allocation failure.\n\nThis problem is unlikely going to show up in practice, so no backpatch\nis done.\n\nAuthor: Bryan Green \nAuthor: Mark Dilger \nDiscussion: https://postgr.es/m/c9a5eeb1-4d31-46e1-bb9e-33448a23a5d3@gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "a4ae8849aeedb9a808ceb4ead1f8eb7c769e03ba", "date": "2026-08-12T08:21:52+08:00", "subject": "docs: fix typos and cross-language link in docs", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "feb15361b34a1bbbc586ee3dc0ad970f6434aad3", "date": "2026-08-12T09:45:03+09:00", "subject": "Fix unlink() error report in KeepFileRestoredFromArchive()", "body": "KeepFileRestoredFromArchive() has been using an incorrect WAL segment\nname in the error report for an unlink() failure.\n\nEven without 1200dfd60c36, this was of consequence only on WIN32, where\nwe would miss to report a segment name suffixed by \".deleted%u\".\n\nOversight in 60df192aea0e, for what looks like a copy-pasto to me.\n\nReported-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAHGQGwHSbJnUmWCuF-QNUsGYF=iSSuzLRx1z_dLfYFwfbTu4Cw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "b08fe48a80ecd2e40440f20d1ab014ac00851cc1", "date": "2026-08-12T08:58:56+08:00", "subject": "docs(BorderBeam): document size limits", "body": "", "style": "conventional", "type": "docs", "scope": "BorderBeam", "breaking": false} {"repo": "openssl", "sha": "6864581ff313019daa18dc8b5ed400329c28caf6", "date": "2026-08-12T10:19:04+09:00", "subject": "doc: correct pkcs12 passin and passout descriptions", "body": "The pkcs12 documentation incorrectly says that -passin is used to\nencrypt output private keys. The implementation uses -passin for input\nand -passout for output, with direction-specific roles depending on\n-export.\n\nDescribe both options explicitly and link -passout to the shared\npassphrase-options documentation.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Richard Levitte \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 1 14:10:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32315\n(cherry picked from commit 2c5f67b894e4f0703ab23262800897b81586651a)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2c5f67b894e4f0703ab23262800897b81586651a", "date": "2026-08-12T10:19:04+09:00", "subject": "doc: correct pkcs12 passin and passout descriptions", "body": "The pkcs12 documentation incorrectly says that -passin is used to\nencrypt output private keys. The implementation uses -passin for input\nand -passout for output, with direction-specific roles depending on\n-export.\n\nDescribe both options explicitly and link -passout to the shared\npassphrase-options documentation.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Richard Levitte \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 1 14:10:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32315", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "0a65e60af86cadedab2531eb8c2b039dab179941", "date": "2026-08-12T09:28:18+08:00", "subject": "fix(compiler-vapor): propagate component root through Transition", "body": "Close #15274\n\nTreat a root Transition as transparent when determining its single-root\nchild so fallthrough class and style bindings use root merging.", "style": "conventional", "type": "fix", "scope": "compiler-vapor", "breaking": false} {"repo": "ant-design", "sha": "480908306434536cb8d0daf49c3a6d43a3a855c9", "date": "2026-08-12T10:53:48+08:00", "subject": "docs: add guide for AI-assisted open source contributions", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "72a6dad1c9119c5ddebdb16d218213f4201f52a2", "date": "2026-08-12T12:06:05+09:00", "subject": "Move index statistics into their own stats kind", "body": "This commit splits the relation and index stats data, where the index\ndata is now moved to a new kind called PGSTAT_KIND_INDEX. A benefit of\nthis refactoring is a reduction of the shared memory required for each\nindex, the new PgStat_StatIdxEntry including only the fields required\nfor indexes. This leads to less data flushed to disk on shutdown.\nAnother benefit is that the addition of new fields for relation-level\nstats is easier to think about. Note that indexes have no need for most\nof the xact-level fields that relations care about, clarifying the role\nof each counter for each relkind.\n\nSystem views are updated to use the index-level functions. Note this\nsplit means a possible break for tools that had the idea to call the\nSQL functions for relation stats (not documented), as now these would\nreturn 0 when queried for an index. These tools can use the new index\nfunctions to achieve the same result.\n\nAs of this change, the pending statistics are still controlled by\nPgStat_TableStatus for both relation and index kinds. This is left as a\nfollow-up piece, this one being good enough on its own. This commit is\nfocused on the changes required for the SQL interface, shared memory\nmanagement and the on-disk format.\n\nBump catalog version.\nBump PGSTAT_FILE_FORMAT_ID.\n\nAuthor: Bertrand Drouvot \nAuthor: Michael Paquier \nDiscussion: https://postgr.es/m/f572abe7-a1bb-e13b-48c7-2ca150546822@gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "9e4c9c5d73cfb1bd3bccc236bd6018b5c7dd3bd8", "date": "2026-08-12T11:10:22+08:00", "subject": "fix(Transfer): respect locale.remove in one-way mode", "body": "", "style": "conventional", "type": "fix", "scope": "Transfer", "breaking": false} {"repo": "ant-design", "sha": "0b0edf011ab35f3d7a9ae4d1b525e76d91edac6a", "date": "2026-08-12T11:10:42+08:00", "subject": "fix: localize Modal and Notification close button names", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "6455d91fabd6e58f3e89014105640e8ceb9ff502", "date": "2026-08-12T11:12:50+08:00", "subject": "fix: use localized names for Upload actions", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9659b8706b67184f0c26563481362a9b0b58cb1f", "date": "2026-08-12T14:36:38+10:00", "subject": "Revised RSASVE degenerate ciphertext check.", "body": "The additional ciphertext check is now applied in rsasve_recover() where it\nbelongs, and not in the underlying RSA primitives, where it remains conditional\ndefined(FIPS_MODULE).\n\nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMergeDate: Mon Aug 17 14:02:44 2026\n(Merged from https://github.com/openssl/openssl/pull/32314)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f49a4e20518c06227d857166cc7a37fe3a8eefef", "date": "2026-08-12T14:36:38+10:00", "subject": "Revised RSASVE degenerate ciphertext check.", "body": "The additional ciphertext check is now applied in rsasve_recover() where it\nbelongs, and not in the underlying RSA primitives, where it remains conditional\ndefined(FIPS_MODULE).\n\nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMergeDate: Mon Aug 17 14:01:18 2026\n(Merged from https://github.com/openssl/openssl/pull/32314)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9e07a21a2687e7288bccdab0c281a4c485d55f62", "date": "2026-08-12T14:36:38+10:00", "subject": "Revised RSASVE degenerate ciphertext check.", "body": "The additional ciphertext check is now applied in rsasve_recover() where it\nbelongs, and not in the underlying RSA primitives, where it remains conditional\ndefined(FIPS_MODULE).\n\nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMergeDate: Mon Aug 17 14:00:11 2026\n(Merged from https://github.com/openssl/openssl/pull/32314)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "476166d0b6f0656141bccda98c9d29e74746ad67", "date": "2026-08-12T14:36:38+10:00", "subject": "Revised RSASVE degenerate ciphertext check.", "body": "The additional ciphertext check is now applied in rsasve_recover() where it\nbelongs, and not in the underlying RSA primitives, where it remains conditional\ndefined(FIPS_MODULE).\n\nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMergeDate: Mon Aug 17 13:58:56 2026\n(Merged from https://github.com/openssl/openssl/pull/32314)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3b6a8a1fd014107e6d27a1dfe70f9018531aeba8", "date": "2026-08-12T14:36:38+10:00", "subject": "Revised RSASVE degenerate ciphertext check.", "body": "The additional ciphertext check is now applied in rsasve_recover() where it\nbelongs, and not in the underlying RSA primitives, where it remains conditional\ndefined(FIPS_MODULE).\n\nReviewed-by: Milan Broz \nReviewed-by: Nikola Pajkovsky \nMergeDate: Mon Aug 17 13:56:53 2026\n(Merged from https://github.com/openssl/openssl/pull/32314)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "dc1e7a8416937edefe04cf836ee202a6fc03bedd", "date": "2026-08-12T00:43:34-04:00", "subject": "chore: update ecosystem plugins", "body": "* chore: update ecosystem plugins\n\n* unpin `core-js-compat`\n\n---------", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vitest", "sha": "34e4c008e63b6c33634cc2d6f89c521802f59286", "date": "2026-08-12T14:01:42+09:00", "subject": "test: strip rc versions from reporter output snapshot", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "961eadb1871caaf4e80240555d1bdb743420ddc2", "date": "2026-08-12T13:32:38+08:00", "subject": "fix(runtime-vapor): respect inheritAttrs ownership for component roots", "body": "close #15277", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "33c77cfad3e4d5749a5827f7c8c6394e7cab0b30", "date": "2026-08-12T14:42:57+09:00", "subject": "Avoid unnecessary matview populated-state updates", "body": "REFRESH MATERIALIZED VIEW currently calls SetMatViewPopulatedState()\neven when the materialized view's relispopulated value already matches\nthe state requested by the command. That performs a catalog update of\npg_class without changing the visible state of the materialized view.\n\nThe extra update creates a dead pg_class tuple and causes the usual\ncatalog-update side effects, including relcache invalidation and WAL,\nfor no semantic benefit. The effect is especially visible for\nREFRESH MATERIALIZED VIEW CONCURRENTLY, where the refresh updates the\nmaterialized view contents with DML and otherwise need not change\npg_class just to keep relispopulated set to true.\n\nCheck the current populated state before calling SetMatViewPopulatedState(),\nand update pg_class only when the value would actually change. This\npreserves the required state transitions for WITH DATA and WITH NO DATA\nwhile avoiding catalog churn for repeated refreshes that leave the\npopulated state unchanged.\n\nAuthor: ChangAo Chen \nReviewed-by: Zizhuan Liu <44973863@qq.com>\nReviewed-by: Adam Brusselback \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/tencent_DF2D8533B5412908326F7806075B7637A108@qq.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "cc1ea710dc88186aab4b3c129883d710781fcfb6", "date": "2026-08-12T11:28:30+05:30", "subject": "Remove unnecessary list_free() calls in OpenTableList().", "body": "OpenTableList() is invoked at most once per DDL statement, from\nCreatePublication() and AlterPublicationTables(), and everything it\nbuilds (relids, relids_with_rf, and relids_with_collist) is allocated\nin the command's own memory context. That context is reclaimed as\nsoon as the command finishes, so explicitly freeing relids and\nrelids_with_rf here served no purpose.\n\nReported-by: Peter Smith \nSuggested-by: Tom Lane \nAuthor: Peter Smith \nReviewed-by: shveta malik \nReviewed-by: shihao zhong \nReviewed-by: surya poondla \nDiscussion: https://postgr.es/m/CAHut+Pt3bgyt==d6GCj=p9qu4sVSH_KQWLN7cMaH0RdJCO-K_Q@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "9509fd20ac0865c1d5d036b49f347e631ad72294", "date": "2026-08-12T15:19:28+08:00", "subject": "demo: add drag-sorting demo", "body": "* demo: add drag-sorting demo\n\n* Refine Listy drag sorting handle UI in demo\n\n---------", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d47336e5f6f1f20b4522d6073c30c6dbd5b07fb6", "date": "2026-08-12T16:23:21+09:00", "subject": "Fix poll_query_until() timeout handling with an undefined query in Cluster.pm", "body": "On timeout, poll_query_until() would fail to print its report when\n\"$query\" was undefined. Instead of the expected report, this prevented\nthe contents of stdout and stderr from being shown.\n\nNote that some of the in-core tests use an undefined query for\nconnection-only polls.\n\nAuthor: Bryan Green \nReviewed-by: Jonathan Gonzalez V. \nDiscussion: https://postgr.es/m/0f515c6d-6032-4c5f-80ac-5c78faae9522@gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7cda7b50d562a03f1c2e0c1d2f2e970acbd769e8", "date": "2026-08-12T09:41:44+02:00", "subject": "Use per-query timeout in subscription test", "body": "A buildfarm failure in subscription/038_walsnd_shutdown_timeout was\ndiagnosed to the background psql session timing out due to a backup\ntaking a long time. The session was using a single timeout for all\nqueries, so switch to resetting the timer for each query to make it\nsurvive slow backups due to constrained buildfarm machines.\n\nBackpatch to v19 where the test was introduced.\n\nAuthor: Hayato Kuroda \nDiscussion: https://postgr.es/m/OS9PR01MB12149AA18EEA475D2AA22FD3BF5D12@OS9PR01MB12149.jpnprd01.prod.outlook.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "6bc52681ed46ef892b93106dfde06646737dabfc", "date": "2026-08-12T15:59:23+08:00", "subject": "fix(Listy): improve virtual scrollbar track hover feedback", "body": "", "style": "conventional", "type": "fix", "scope": "Listy", "breaking": false} {"repo": "postgres", "sha": "d6bf0ab1704c62f94875d25dd71c4baa1feac5be", "date": "2026-08-12T17:03:46+09:00", "subject": "Add common LSN parser for user-supplied locations", "body": "pg_waldump (--start/--end), pg_recvlogical (--startpos/--endpos), and\npg_receivewal (--endpos) parsed user-supplied WAL locations with\nsscanf(\"%X/%08X\"). That accepts several forms rejected by pg_lsn input\nand can proceed with a different location than the user specified:\noverlong components, trailing characters, leading whitespace, signs, and\n0x prefixes.\n\nAdd pg_parse_lsn() to src/common and use it for those command-line\noptions. The helper accepts the pg_lsn text syntax only: one to eight\nhexadecimal digits, a slash, one to eight hexadecimal digits, and no\ntrailing characters. Each frontend tool keeps its existing error\nmessage.\n\nUse the same helper from pg_lsn_in_safe(), leaving only backend-specific\nsoft error reporting there. This keeps frontend command-line validation\nand backend pg_lsn input tied to one parser.\n\nBug: #19598\nReported-by: Michael Malis \nAuthor: Zexin Li \nReviewed-by: Ayush Tiwari \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/19598-aa67c8f4331611b4@postgresql.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "032d3189217a74b366f9702b16c55a0bbc75927b", "date": "2026-08-12T17:12:11+09:00", "subject": "fix(browser): trigger playwright/chromium gc on lower disk availability", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "ant-design", "sha": "315d5c01174324da83b20c95592db7e3ade45d68", "date": "2026-08-12T16:24:08+08:00", "subject": "docs: avoid duplicate customize theme toc entries", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "cd238bb7addf5b1f174df47ccc61651a545c4876", "date": "2026-08-12T10:36:20+02:00", "subject": "tool_operate: only check for schannel if on windows", "body": "It can never be true elsewhere.\n\nReported-by: Max Dymond\nCloses #22552", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "355b8ad8d19b818f1493a8ae092bb5b14587b5a5", "date": "2026-08-12T13:42:37+05:00", "subject": "fix(Select): keep customize input aligned on clear hover", "body": "* fix(Select): keep customize input aligned on clear hover\n\n* test(AutoComplete): cover both customize inputs on clear hover\n\n---------", "style": "conventional", "type": "fix", "scope": "Select", "breaking": false} {"repo": "vite", "sha": "dc780909227e1b57c3b5f7ef9014acc026228dc5", "date": "2026-08-12T16:48:38+08:00", "subject": "feat(bundled-dev): serve static files with fs checks after the bundle lookup", "body": "", "style": "conventional", "type": "feat", "scope": "bundled-dev", "breaking": false} {"repo": "curl", "sha": "852f5d34da772de65d057897278aa1e75dd060d1", "date": "2026-08-12T10:52:42+02:00", "subject": "build: strip empty properties from `libcurl.pc`", "body": "To keep it tidy and to support properties in the template that only\nappear in the generated `libcurl.pc` when filled.\n\nE.g. `Required.private` may remain empty after #22548.\n\nAlso:\n- GHA/windows: dump `libcurl.pc` to log in every job\n- tidy up a pre-existing regex in the updated command.\n 2c22d3069aef507d6a6876a6d20616fe5e50c6a3 #22544\n\nRef: b9254da6175360102176303ecf9dbb5519b0ffcf #22548\nCherry-picked from #22543\n\nCloses #22553", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "curl", "sha": "acd9411b01d2fff0bb6aa7c82c1d44c280751d1e", "date": "2026-08-12T11:02:20+02:00", "subject": "configure: replace stray raw `grep` with `$GREP`", "body": "Follow-up to 7cc2e8b349df28d55d5f40bfae323485df9f0cf2\nFollow-up to bfaab6ed6bb3e8865cd8dd60375fbdc39203a862\nCherry-picked from #22543\n\nCloses #22554", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6d6078160e43b6e79edf18eb68d6f9561a974ced", "date": "2026-08-12T10:17:50+01:00", "subject": "DTLS 1.3 Keep listener driven connections out of blocking mode", "body": "A DTLS listener drives the handshakes of connections which have not been\naccepted yet from inside its own tick: dtls_listener_drive_pending()\ncalls SSL_accept() on each of them. Those connections inherit the\nlistener's blocking mode, so once a listener connection is able to block\nwaiting for a datagram, that SSL_accept() can block too - and it must\nnot, because the listener is the only thing which will ever deliver the\ndatagram it is waiting for. Nothing else runs while the tick is inside\nthat call, so the wait cannot end.\n\nAdd force_nonblocking to DTLS1_STATE, set around the SSL_accept() that\ndtls_listener_drive_pending() performs, and consulted by\nossl_dtls_desires_blocking(). Ownership of a connection passes to the\napplication when SSL_accept_connection() returns it, and by then the\nflag is clear, so an application which asked for blocking still gets it.\n\nThe flag has to survive SSL_clear(), which memsets d1. SSL_clear() can\nrun inside the very SSL_accept() the listener is driving:\ndrive_single_connection() sets TLS1_FLAGS_STATELESS to suppress it, but\nonly where cookie validation is enabled. Losing the flag there is not\nmerely a missed optimisation - the connection blocks inside the tick,\npoll_translate_ssl_dtls_conn() ticks the listener again from another\nthread, drive_pending re-enters and collects the same connection twice,\nand registering it a second time trips the old == NULL assertion in the\nconnection lookup. In a release build that appears later as a\nuse-after-free rather than as anything to do with blocking mode.\n\nNothing blocks in a read yet, so this commit changes no behaviour on its\nown. It lands first so that the commit which adds blocking reads cannot\ndeadlock at any point in the series.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Jakub Zelenka \nMerge-date: Mon Aug 17 08:29:56 2026\nMerged-from: https://github.com/openssl/openssl/pull/32324", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e4d3898697cf553f45989db145d8a57842082fc6", "date": "2026-08-12T10:18:58+01:00", "subject": "DTLS 1.3 Support blocking reads on listener connections", "body": "A connection created by a DTLS listener has no BIO of its own. The\nlistener owns the socket and demultiplexes datagrams into a per\nconnection queue, so there is nothing for a read to block in, and every\nread on an empty queue reported WANT_READ however the connection was\nconfigured. SSL_set_blocking_mode() therefore had no effect on reads.\n\nWait in the record layer instead, where a blocking BIO would have. When\nrlayer_dtls_get_urxe_packet() finds the queue empty and the connection is\nin blocking mode, ossl_dtls_conn_wait_for_datagram() waits for one to\narrive. The state machine above is untouched, so SSL_read(), SSL_peek(),\nSSL_do_handshake() and SSL_accept() all become blocking together, which\nis what an application moving from a plain DTLS object expects.\n\nThe wait loops, because the socket is shared: a wakeup may be for a\ndatagram belonging to a different connection, in which case our queue is\nstill empty and there is nothing to return. It also handles events after\neach wakeup, since while a thread is in here nothing else services the\nconnection, and the retransmission timer has to keep running - the poll\ntranslation folds the event timeout into the deadline so the wait ends in\ntime for it.\n\nExisting tests drive both ends of a connection from a single thread,\nwhich cannot work if the server side blocks, so create_dtls_listener() in\ndtlsssllistenertest.c now turns blocking mode off for its callers, with\ncreate_dtls_listener_unconfigured() for the one test which has to observe\nthe default. test_dtls_multithread opts out for the same reason.\n\nThe new test_dtls_blocking_read has the client stay silent for a while\nafter its handshake, so the server thread is already parked in\nSSL_read_ex() when the write finally comes. Its second iteration\nhandshakes in non-blocking mode and switches to blocking only for the\nread, which pins the read path on its own: in the first iteration the\nhandshake is what fails without this change, and the read is never\nreached.\n\nAssisted-by: Cl", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "38a095dc3a8cff7060531a18c0fa5b7d42a1d66b", "date": "2026-08-12T17:27:24+08:00", "subject": "fix(vapor): preserve dynamic v-for slot state", "body": "close #15276\n\nMake createForSlots retain slot records across source updates. Reconcile keyed slots by their template key and unkeyed slots by their resolved slot name, updating loop alias refs without recreating slot functions.\n\nPass a stable record identity to slot fragments so unchanged slots retain their mounted content and local state. Reuse the resolved records array when slot resolution is unchanged to avoid unnecessary outlet updates.\n\nUpdate compiler-vapor to emit separate source, slot factory, name, and key callbacks, and add regression coverage for keyed and unkeyed dynamic slots.", "style": "conventional", "type": "fix", "scope": "vapor", "breaking": false} {"repo": "curl", "sha": "d064bc2adbf2aa0762c89ea4166cecd55693bb62", "date": "2026-08-12T11:44:04+02:00", "subject": "thrdqueue: drop name strdups from Curl_thrdq_create", "body": "The only usage of the thread queue and pool is for \"DNS\" which is a\nstatic string. No point in strdup'ing it (twice). Future users are\n*likely* to also used fixed strings, otherwise we revert this change\nwhen dynamic names are introduced.\n\nReported-by: Max Dymond\nCloses #22555", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "fd56954c9fe69bdc899d7cad93ec0b22b663cb30", "date": "2026-08-12T11:53:10+02:00", "subject": "Fix stale comments/signature after commit 28d534e2ae0a", "body": "Commit db89a47115f0 added an 'options' argument to table_tuple_update()\nand table_tuple_delete(), correctly documenting it as recognizing no\nvalues. It also (correctly) added pg_attribute_unused() to the\ncorresponding heap_update() argument. Commit 28d534e2ae0a then added\nTABLE_UPDATE_NO_LOGICAL and TABLE_DELETE_NO_LOGICAL, but failed to\n(correctly) update the comment and heap_update()'s signature.\n\nIn table_tuple_update(), move the correct explanation of 'options' to\nthe right place, remove the bogus one.\n\nIn table_tuple_delete(), add the missing TABLE_DELETE_NO_LOGICAL doc.\n\nIn heap_update(), remove the pg_attribute_unused() marker.\n\nAuthor: Nikhil Sontakke \nBackpatch-through: 19\nDiscussion: https://postgr.es/m/CA+UBoq21SzkjThYMSwSpVU-jwzpuEsf22Hq1FPbcKcDc43C29g@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "eae88a7473475511220767b7f2ed792da82a79d6", "date": "2026-08-12T12:11:37+02:00", "subject": "urlapi: avoid dedotdotify() if possible", "body": "The dedotdotify() function that removes ./ and ../ sequences from paths\njuggles memory and is slow. Now needs_dedotdot() is called first to\ndetermine if the removal process is necessary and otherwise avoids doing\nit. Avoids unnecessary memory operations.\n\nAdjusted unit test 1395 accordingly because now a lot of input strings\nreturn NULL for \"no change necessary\".\n\nSuggested-by: Max Dymond\n\nCloses #22557", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a02eac27f0f254083783d7e6d52b7bb96f5020d3", "date": "2026-08-12T12:11:37+02:00", "subject": "urlapi: avoid dedotdotify() if possible", "body": "The dedotdotify() function that removes ./ and ../ sequences from paths\njuggles memory and is slow. Now needs_dedotdot() is called first to\ndetermine if the removal process is necessary and otherwise avoids doing\nit. Avoids unnecessary memory operations.\n\nAdjusted unit test 1395 accordingly because now a lot of input strings\nreturn NULL for \"no change necessary\".", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "b298df6092b383cac13bf9b5983d319e901be075", "date": "2026-08-12T11:17:17+01:00", "subject": "fix(vitest): render non-Error causes from env setup so the actual diagnostic is not dropped", "body": "", "style": "conventional", "type": "fix", "scope": "vitest", "breaking": false} {"repo": "openssl", "sha": "f2b3c3cbff9df70bdf1baccc937a0362fe6881b0", "date": "2026-08-12T11:19:05+01:00", "subject": "DTLS 1.3 Support blocking writes on listener connections", "body": "A listener connection writes through the listener's socket, which is\nshared with every other connection and is always non-blocking. When a\nsend cannot be completed the DTLS record layer discards the datagram and\nreports that the write should be retried - a fair default for an\nunreliable transport, but not what an application which asked for\nblocking writes expects, since it gets neither the data sent nor a call\nwhich waited.\n\nAdd an rlayer callback, OSSL_FUNC_RLAYER_BLOCK_FOR_WRITE, which the\nrecord layer calls instead of reporting a retry. For a listener\nconnection in blocking mode it waits for the socket to become writable\nand returns 1, and tls_retry_write_records() goes round its loop again to\nrepeat the same send: nothing has been consumed at that point, so the\nretry is a genuine second attempt rather than a resend.\n\nOnly the BIO_sendmmsg() branch is hooked. That is the one listener\nconnections take, because their peer address is set, and the BIO_write()\nbranch below it belongs to connections which have a BIO of their own to\nblock in. A NULL callback leaves both paths exactly as they were, so TLS\nand standalone DTLS are untouched.\n\nThe wait itself is one wait per call, not a loop: the caller retries the\nsend and comes back here if it still cannot proceed, so a wakeup which\nturns out not to have left room in the socket buffer costs another\nattempt rather than a lost datagram.\n\nA loopback socket's send buffer does not fill, so the test supplies the\ntransient failure with a filter BIO in front of the listener's write BIO\nwhich rejects one send with a non-fatal error. Its ctrl forwards\neverything to the socket underneath, the poll descriptors included, since\nthose are what the wait itself needs. The client is then read to confirm\nthe datagram really was sent rather than merely reported as sent.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Jakub Zelenka \nMerge-date: Mon Aug", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "667c13954daff2d1bec8a866702da8934b9ce539", "date": "2026-08-12T19:20:45+09:00", "subject": "fix(snapshot): support no-unsafe-eval CSP by evaluating snapshot files on server", "body": "", "style": "conventional", "type": "fix", "scope": "snapshot", "breaking": false} {"repo": "curl", "sha": "26865e74374a891ea74240a556b0607cc9a44a82", "date": "2026-08-12T12:32:44+02:00", "subject": "drop unused variable, clarify comment", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a58528d130773485e62bb647f6cc8ffc3b440b51", "date": "2026-08-12T12:48:33+02:00", "subject": "Fix error handling in asn1_multi()", "body": "Free the partially constructed ASN1_TYPE and return NULL when\nASN1_STRING_type_new() fails. Previously asn1_multi() returned an\nincomplete object.\n\nFound by x509v3 mfail regression test.\n\nReviewed-by: Milan Broz \nReviewed-by: Jakub Zelenka \nMerge-date: Mon Sep 7 14:08:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32330", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "2c71c89942edb88fb9c34b59c9517cf1c98e2021", "date": "2026-08-12T19:09:56+08:00", "subject": "docs: refine AI-assisted maintenance guide", "body": "* docs: refine AI-assisted maintenance guide\n\n* docs: fix Markdown parsing in AI maintenance guide", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5a8c9d45a0675c8a705f1a99934834a687a1fb8b", "date": "2026-08-12T12:20:29+01:00", "subject": "Skip the common test options in dtls_multithread_test", "body": "setup_tests() read its two positional arguments without first calling\ntest_skip_common_options(), so the option cursor was never advanced from\nthe test's point of view and test_get_argument(0) returned the option\nitself. Passing -test or -iter therefore left the certificate path set to\nthe literal string \"-test\", and the run failed on that:\n\n error:80000002:system library:file_ctrl:No such file or directory:\n ...:calling fopen(-test, r)\n\nThe option was not being ignored: the framework parses it in\nprocess_shared_options() and the selected sub-test really does run. Only\nthe positional arguments were shifted, which is what made the failure\nlook unrelated to the option.\n\nAlso declare the usage string, so -help describes the two arguments the\ntest needs rather than showing none.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Jakub Zelenka \nMerge-date: Mon Aug 17 08:36:26 2026\nMerged-from: https://github.com/openssl/openssl/pull/32331", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "fbfb2499f03567611d2dc649bfadafe2b3341ef7", "date": "2026-08-12T14:04:48+02:00", "subject": "keylog: add a random size argument to Curl_tls_keylog_write()", "body": "To allow the function to verify that the buffer is large enough. Avoids\npossible future internal mishaps.\n\nCloses #22560", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "1b79a5ff6145e161e70c6690f0f885c728eae02c", "date": "2026-08-12T08:32:21-04:00", "subject": "build(vim-patch): n/a popup_ api #41270", "body": "- enums\n- macros\n- struct members\n- popup_.*()\n\n8.1.1523\n\n* build(vim-patch): 8.1.0863 is almost auto n/a\n\nCan't ignore os_unix.c hunk for \"failed:\" .", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "curl", "sha": "743096a4964bf921f7b39cc7b739fd4fab6ad39c", "date": "2026-08-12T15:49:39+02:00", "subject": "uint-bset: add slot0 member", "body": "Add a fixed slot0 member for uint32_bset to have no allocations for\nsizes < 64. This is a common use case for curl_easy_perform().\n\nCloses #22561", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "eca5ded9f6c6281f46abb9cae073c5132e082df5", "date": "2026-08-12T16:23:58+02:00", "subject": "mailmap: Ralf Mueller", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "aaf67e4825f32d764296fabfa3a86c4da90cec16", "date": "2026-08-12T16:48:21+02:00", "subject": "Avoid recalculating pgprocno in ProcArrayAdd()", "body": "The value is already stored in a local variable a few lines above.\nThis is just a cosmetic change.\n\nAuthor: Shinya Kato \nReviewed-by: Chao Li \nReviewed-by: solai v \nReviewed-by: Pierre Forstmann \nDiscussion: https://postgr.es/m/CAOzEurTJO2aqd9ewQmo9vjCrFUAD8Lk+raujCfK5oWoLGUP_3w@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "507ee5d79215ee8026db35b03925087c48e03318", "date": "2026-08-12T17:30:40+02:00", "subject": "progress: fix progress meter update to run every second again", "body": "Closes #22562", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "axios", "sha": "d19040bda7a8be2f82c3c6e1a5bc03917daee39a", "date": "2026-08-12T18:27:45+02:00", "subject": "fix: harden runtime option handling", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vitest", "sha": "4363cdb49999addab6f2b9d6402cfea2315c13e3", "date": "2026-08-12T12:33:02-04:00", "subject": "feat(reporter): add custom GitHub Actions summary title", "body": "", "style": "conventional", "type": "feat", "scope": "reporter", "breaking": false} {"repo": "vue-core", "sha": "7db14544ea622297802635d7e953633f587a1c19", "date": "2026-08-12T21:41:07+05:00", "subject": "fix(runtime-vapor): release unmounted v-for component items from the parent scope", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "7774249d0b66fff0b60823783db24090981020ac", "date": "2026-08-12T12:48:53-04:00", "subject": "Fix float8_regr_accum() for Inf/NaN with constant other input.", "body": "Commit 649828769 improved this code to keep Sxx, Syy, and Sxy\nexactly zero so long as we see only a single value of the input(s).\nHowever, if any values of the other input are Inf or NaN, we'd\nbetter set Sxy to NaN instead. Otherwise we risk reporting zero\nvariance when the result is really undefined. The old coding handled\nthis implicitly, but in the short-circuit path we have to take care\nof it explicitly.\n\nBug: #19615\nReported-by: Junwen An \nAuthor: Andrey Rachitskiy \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/19615-c7e390593416f6b6@postgresql.org\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "864ef0fb1c91e83dff143d6de3f6f85390b6415b", "date": "2026-08-12T19:15:48+02:00", "subject": "Reject REPLICA IDENTITY USING INDEX on column with invalid NOT NULL", "body": "ALTER TABLE ... REPLICA IDENTITY USING INDEX verified key columns by\nreading pg_attribute.attnotnull, but commit a379061a22a8 made\nattnotnull true also for unvalidated (NOT VALID) not-null constraints,\nwhich do not prove the column null-free. An index over such a column\ncould thus be marked as replica identity even though the column might\ncontain NULLs, causing apply-side divergence for UPDATE/DELETE on the\nnullable rows.\n\nFix by additionally requiring convalidated for the underlying\nconstraint, mirroring the fix d9ffc27291f applied to ATExecAddIdentity\nfor the analogous identity-column case.\n\nAuthor: Ante Krešić \nReviewed-by: Aleksander Alekseev \nReviewed-by: solai v \nBackpatch-through: 18\nDiscussion: https://postgr.es/m/CABXQ4dJUibZzN91qvWmsfA7MUDn9YRCNyu3CcukdyokbH1=41Q@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "701178f81421442a622da7df23029feb8d138374", "date": "2026-08-12T20:58:04+02:00", "subject": "test: use portable function-name macro in QUIC test helper", "body": "Use OPENSSL_FUNC instead of __func__ so the helper compiles in strict\nANSI/C90 builds.\n\nFixes: 0461a5636dec (\"test/quicapitest.c: add test for pending connections limit enforcement\")\nSigned-off-by: Nikola Pajkovsky \nReviewed-by: Viktor Dukhovni \nReviewed-by: Saša Nedvědický \nReviewed-by: Milan Broz \nMerge-date: Thu Aug 13 07:18:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32337\n(cherry picked from commit 0b9dacf5ef252775a9fab0e876a05415b23651ab)\n(cherry picked from commit 7df38b8e2a11ce14e2c2b8257c290ae7647fd5bd)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7df38b8e2a11ce14e2c2b8257c290ae7647fd5bd", "date": "2026-08-12T20:58:04+02:00", "subject": "test: use portable function-name macro in QUIC test helper", "body": "Use OPENSSL_FUNC instead of __func__ so the helper compiles in strict\nANSI/C90 builds.\n\nFixes: 0461a5636dec (\"test/quicapitest.c: add test for pending connections limit enforcement\")\nSigned-off-by: Nikola Pajkovsky \nReviewed-by: Viktor Dukhovni \nReviewed-by: Saša Nedvědický \nReviewed-by: Milan Broz \nMerge-date: Thu Aug 13 07:18:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32337\n(cherry picked from commit 0b9dacf5ef252775a9fab0e876a05415b23651ab)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0b9dacf5ef252775a9fab0e876a05415b23651ab", "date": "2026-08-12T20:58:04+02:00", "subject": "test: use portable function-name macro in QUIC test helper", "body": "Use OPENSSL_FUNC instead of __func__ so the helper compiles in strict\nANSI/C90 builds.\n\nFixes: 0461a5636dec (\"test/quicapitest.c: add test for pending connections limit enforcement\")\nSigned-off-by: Nikola Pajkovsky \nReviewed-by: Viktor Dukhovni \nReviewed-by: Saša Nedvědický \nReviewed-by: Milan Broz \nMerge-date: Thu Aug 13 07:18:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32337", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3703e9d7911acba31b97aaeaa5b319acce428dba", "date": "2026-08-12T15:12:37-04:00", "subject": "DTLS 1.3 Fix TLSProxy DTLS retransmission in valgrind CI", "body": "Valgrind slows execution down, causing DTLS retransmission timers to fire\nduring otherwise normal handshakes. The client retransmits its ClientHello\nbefore the server responds, and TLSProxy recorded the duplicate into\nmessage_list unconditionally. This shifted every subsequent message index\nby one, breaking three DTLS 1.3 subtests:\n\n70-test_tls13alerts.t: alert_filter corrupted message_list[1] expecting\nthe ServerHello, but found the retransmitted ClientHello instead. The\nreal ServerHello passed through uncorrupted, the handshake succeeded,\nand Message->fail() returned false.\n\n70-test_tls13messages.t / 70-test_tls13kexmodes.t: checkhandshake.pm\nwalks message_list sequentially comparing each .mt to an expected\nsequence. The extra ClientHello at position 0 shifted all five\nsubsequent checks.\n\nFix by deduplicating message_list in DTLS mode using (sender, msgseq) as\nthe key. Retransmitted DTLS handshake messages always carry the same\nmessage sequence number as the original, so duplicates are naturally\ndiscarded. HelloRetryRequest is unaffected becasue the retried ClientHello\ngets an incremented msgseq. The seen_msgseq hash is reset by clearClient\nso there is no state leakage between subtests.\n\nAssisted-by: Claude:claude-sonnet-4-6\nReviewed-by: Tomas Mraz \nReviewed-by: Matt Caswell \nMerge-date: Thu Aug 20 09:55:42 2026\nMerged-from: https://github.com/openssl/openssl/pull/32340", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9a4cf2d265ce53f6cf4a8b5845c282fb3c79ac15", "date": "2026-08-12T16:14:10-04:00", "subject": "Fix build break on evp_mac_fetch_from_prov", "body": "Recent backport applied cleanly to 3.6, but 3.6 includes an extra call\nto evp_mac_fetch_from_prov, which uses a no longer in existance\nparameter.\n\nFix that up.\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Paul Dale \nReviewed-by: Viktor Dukhovni \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 13 08:57:03 2026\nMerged-from: https://github.com/openssl/openssl/pull/32341", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "3089e7eec8b694ff36a3aa700c4a319d656806e7", "date": "2026-08-12T23:36:38+02:00", "subject": "multi: hold timeout values in 'int' instead of 'long'", "body": "Since this code needs to work (identically) on systems with 32 bit longs\nanyway, this might just as well always use 32 bit to save space and make\nthe code work the same everywhere.\n\nCloses #22564", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "6a6edec2c6ee7c401b6262c385483e0a7f237dd4", "date": "2026-08-13T00:03:27+02:00", "subject": "multihandle: move two struct fields", "body": "To avoid holes on 64-bit architectures.\n\nCloses #22565", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "bf1f63476dc82b12a330798047031cf2e63f142c", "date": "2026-08-12T22:11:13Z", "subject": "build: update cross-repo angular dependencies to v22.2.0-next.1", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "42bc497edf555dbad92d6868acf85d7b1efe79ac", "date": "2026-08-12T22:13:37Z", "subject": "build: update cross-repo angular dependencies to v22.1.2", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "curl", "sha": "dea00a6d7386b91dfcf9b87eb5b2d5c24186aea9", "date": "2026-08-13T01:01:07+02:00", "subject": "cmake: use built-in variable and target property dump functions with CMake 4.5+", "body": "Their output is nearly identical to curl's implementations. Minor\ndifferences: it goes to stdout (curl: stderr), marks cache variables,\nbut doesn't mark 'advanced' ones.\n\nAlso:\n- indent curl's output to match CMake's.\n- replace single quotes with double to match CMake output.\n- alpha-sort target properties to match CMake.\n- alpha-sort config types to match CMake.\n- tested OK with cmake-4.4.20260811-g567ed0e and 4.4.20260812-g760e4fb.\n\nRefs:\nhttps://gitlab.kitware.com/cmake/cmake/-/commit/f723293c7f625b14bd48627042fc04e67b76c513\nhttps://gitlab.kitware.com/cmake/cmake/-/merge_requests/11834\nhttps://gitlab.kitware.com/cmake/cmake/-/work_items/27513\nhttps://cmake.org/cmake/help/git-master/command/cmake_language.html#command:cmake_language\n\nFollow-up to 855acb3bb07e0dd06b5722218eb5fded333f7ce0 #17701\nFollow-up to f58342ae21fe4a04edddd98f255d8b0c3d66ae3a #15562\nFollow-up to cd51bb503ae1779a45c90898b89da6e19c616521 #14439\n\nCloses #22566", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "e53bb2bca7ed4bd9f12f2c8749f03d281f561ee6", "date": "2026-08-12T16:05:05-07:00", "subject": "release: bump VSCode extension version to 22.1.0", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "dc8c0717a05451694b6d89d668fedb1b3f15c767", "date": "2026-08-12T16:07:05-07:00", "subject": "refactor(router): Create transactional router resource", "body": "This commit adds an implementation of a router resource (not currently exposed for public use) which\ndefines the behavior of a resource dependent on the Router navigation lifecycle.\n\n(cherry picked from commit c111f3fb0c2ee03ea9d96ea31306faddd519e575)", "style": "conventional", "type": "refactor", "scope": "router", "breaking": false} {"repo": "angular", "sha": "c111f3fb0c2ee03ea9d96ea31306faddd519e575", "date": "2026-08-12T16:07:05-07:00", "subject": "refactor(router): Create transactional router resource", "body": "This commit adds an implementation of a router resource (not currently exposed for public use) which\ndefines the behavior of a resource dependent on the Router navigation lifecycle.", "style": "conventional", "type": "refactor", "scope": "router", "breaking": false} {"repo": "angular", "sha": "910f391674bf9858885cb779a35f1b650aabc1d3", "date": "2026-08-12T16:24:14-07:00", "subject": "build(vscode-extension): dynamically resolve upstream remote and use token for https pushes", "body": "Previously, the release script hardcoded the upstream repository URL as an unauthenticated HTTPS URL (https://github.com/angular/angular.git). Although the script verified that a GITHUB_TOKEN environment variable was present, it only used that token for REST API calls (such as creating the GitHub release) and did not provide it to Git commands. As a result, users who authenticate to GitHub via SSH (and do not have an HTTPS Git credential helper configured) were prompted interactively for GitHub login credentials when pushing the release tag.\n\nThis change dynamically resolves the upstream remote name from the user's configured remotes by checking for any remote pointing to angular/angular. When pushing over SSH, Git uses the user's existing SSH credentials. When pushing over HTTPS (or falling back), the script injects GITHUB_TOKEN into the push URL to prevent interactive authentication prompts.\n\n(cherry picked from commit 6f848db435ed7b675f0d8955f93c40da69904a20)", "style": "conventional", "type": "build", "scope": "vscode-extension", "breaking": false} {"repo": "angular", "sha": "6f848db435ed7b675f0d8955f93c40da69904a20", "date": "2026-08-12T16:24:14-07:00", "subject": "build(vscode-extension): dynamically resolve upstream remote and use token for https pushes", "body": "Previously, the release script hardcoded the upstream repository URL as an unauthenticated HTTPS URL (https://github.com/angular/angular.git). Although the script verified that a GITHUB_TOKEN environment variable was present, it only used that token for REST API calls (such as creating the GitHub release) and did not provide it to Git commands. As a result, users who authenticate to GitHub via SSH (and do not have an HTTPS Git credential helper configured) were prompted interactively for GitHub login credentials when pushing the release tag.\n\nThis change dynamically resolves the upstream remote name from the user's configured remotes by checking for any remote pointing to angular/angular. When pushing over SSH, Git uses the user's existing SSH credentials. When pushing over HTTPS (or falling back), the script injects GITHUB_TOKEN into the push URL to prevent interactive authentication prompts.", "style": "conventional", "type": "build", "scope": "vscode-extension", "breaking": false} {"repo": "openssl", "sha": "8274744b92217ea857baac7b6bac2de4e95bc92d", "date": "2026-08-12T17:25:46-06:00", "subject": "Add a test that generated files are gitignored and untracked", "body": "Every GENERATE target must be absent from the index and covered by a\ncommitted .gitignore. The deliberately committed generated files, such\nas util/libcrypto.num and doc/build.info, come from \"make update\" rather\nthan GENERATE, and are not checked. Skips without git or a checkout.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMergeDate: Tue Aug 18 15:27:39 2026\n(Merged from https://github.com/openssl/openssl/pull/32343)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "5e7efa0087806738f2737fb2b0982569a8444dee", "date": "2026-08-13T10:49:36+08:00", "subject": "chore: upgrade markdown-it-image-size", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vite", "sha": "74a25c324e4f1cd4a73ec594594f6944911fd6ed", "date": "2026-08-13T11:05:41+08:00", "subject": "docs(server-options): update sourcemapIgnoreList Rolldown references", "body": "", "style": "conventional", "type": "docs", "scope": "server-options", "breaking": false} {"repo": "angular", "sha": "8adaa09f959c2e758de5811e7b9f6abd99be8b7d", "date": "2026-08-12T23:21:58-05:00", "subject": "fix(vscode-extension): prevent command URI injection in TSDK approval", "body": "VS Code parses notification text for links and permits `command:` URIs. Interpolating the workspace-controlled TSDK path can therefore let a malicious path add a command link to the approval prompt.\n\nRemove the path from the notification instead of attempting to sanitize or escape it. This keeps the prompt static and matches VS Code's TypeScript extension workspace-version approval flow.\n\nFixes #70176\n\n(cherry picked from commit ba13a3ce22fdf875b14c1cbbe200dcfa23717bf4)", "style": "conventional", "type": "fix", "scope": "vscode-extension", "breaking": false} {"repo": "angular", "sha": "ba13a3ce22fdf875b14c1cbbe200dcfa23717bf4", "date": "2026-08-12T23:21:58-05:00", "subject": "fix(vscode-extension): prevent command URI injection in TSDK approval", "body": "VS Code parses notification text for links and permits `command:` URIs. Interpolating the workspace-controlled TSDK path can therefore let a malicious path add a command link to the approval prompt.\n\nRemove the path from the notification instead of attempting to sanitize or escape it. This keeps the prompt static and matches VS Code's TypeScript extension workspace-version approval flow.\n\nFixes #70176", "style": "conventional", "type": "fix", "scope": "vscode-extension", "breaking": false} {"repo": "postgres", "sha": "ed23f7ce7f8f633175ae785d18577fbc440139ae", "date": "2026-08-13T13:26:47+09:00", "subject": "doc: Clarify pgbench reporting with --continue-on-error", "body": "Clarify that transactions failed under --continue-on-error are reported\nseparately and are not counted as transactions actually processed, while\ntime spent on them can lower reported TPS. Also adjust the latency\ndescription to avoid implying that failed transactions are generally included\nin latency measurements.\n\nBackpatch to v19, where pgbench's --continue-on-error option was introduced.\n\nAuthor: Chao Li \nReviewed-by: Yugo Nagata \nReviewed-by: Xuneng Zhou \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/597C9755-4386-488D-A289-16D50AD81FBF@gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "d615bcdb23d96c1ca5ce1ee45e21d8d87381106f", "date": "2026-08-13T12:27:32+08:00", "subject": "build: use JSON import attributes instead of readFIleSync in rolldown configs", "body": "", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "dc6c9c540b5a4bf5aa2aedc8959b0d015004d45c", "date": "2026-08-13T10:10:46+05:30", "subject": "Doc: Clarify ALTER PUBLICATION's REFRESH PUBLICATION wording.", "body": "The note used a confusing \"tables/except tables/schemas\" slash-list,\nomitted DROP from the operations requiring a subscriber refresh, and\ncalled out \"unset ALL SEQUENCES\" without its symmetric \"unset ALL TABLES\"\ncase. Reword for clarity and consistency.\n\nAuthor: Peter Smith \nReviewed-by: Amit Kapila \nBackpatch-through: 19\nDiscussion: https://postgr.es/m/CAHut+PuGKrWhVMvuFyf-wSdF-Tai_jPTz2UCSCqsYM-82ATY5w@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "38cd0ba1b8ce3ee86041628d340300adac65c0ab", "date": "2026-08-13T00:45:53-04:00", "subject": "build(vim-patch): n/a autocmd from 8.1.1218", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "postgres", "sha": "9e725e77a5d8356a2237a15025b57d75b08e87e4", "date": "2026-08-13T14:22:39+09:00", "subject": "vacuumdb: Use ANALYZE ONLY for partitioned tables", "body": "When vacuumdb --analyze-only or --analyze-in-stages enumerates\ntables itself, it can select both a partitioned table and its\npartitions as separate work items. Previously, it generated plain\nANALYZE for the partitioned table, which recursively analyzed its\npartitions as well.\n\nThis caused duplicate work when the partitions were already selected by\nvacuumdb. With --analyze-in-stages, the redundant work could be\nrepeated at every stage.\n\nFix this by generating ANALYZE ONLY for automatically selected\npartitioned tables when connected to servers that support it. This\nupdates inherited statistics for the partitioned table while leaving\nper-partition statistics to the separately selected partition entries.\nFor older servers that do not support ANALYZE ONLY, vacuumdb falls back\nto the previous plain ANALYZE command.\n\nThis also applies when partitioned tables are selected by schema\nfilters. In that case, partitions matching the filters are analyzed as\nseparate targets, while per-partition statistics for partitions outside\nthe filters are not updated via the partitioned parent.\n\nThis commit does not change the behavior for partitioned tables\nspecified explicitly with --table, i.e., in that case, vacuumdb\ncontinues to generate plain ANALYZE, recursively analyzing the specified\npartitioned table and its partitions.\n\nSuggested-by: Justin Pryzby \nAuthor: Fujii Masao \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/aKZ7lMYGnnIo35c0@pryzbyj2023\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4f08f8ed1fb746f42b8510991fbab46fa0093fa6", "date": "2026-08-13T15:23:33+10:00", "subject": "Deal with riscv64 qemu-user stderr noise", "body": "When the RISCV vector extensions version is not explicitly\nspecified, qemu issues warnings to stderr when running each\nprogram. This broke some CI jobs.\n\nPin the version to the (current) \"v1.0\" default.\n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Saša Nedvědický \nReviewed-by: Igor Ustinov \nMergeDate: Thu Aug 13 06:32:59 2026\n(Merged from https://github.com/openssl/openssl/pull/32345)\n\n(cherry picked from commit 31e93626b5f24b84947a954929106b4cfd5a03e7)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "31e93626b5f24b84947a954929106b4cfd5a03e7", "date": "2026-08-13T15:23:33+10:00", "subject": "Deal with riscv64 qemu-user stderr noise", "body": "When the RISCV vector extensions version is not explicitly\nspecified, qemu issues warnings to stderr when running each\nprogram. This broke some CI jobs.\n\nPin the version to the (current) \"v1.0\" default.\n\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Saša Nedvědický \nReviewed-by: Igor Ustinov \nMergeDate: Thu Aug 13 06:32:59 2026\n(Merged from https://github.com/openssl/openssl/pull/32345)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "73f28b2300bef736dcf64a2d40d5d77ef1575880", "date": "2026-08-13T01:25:35-04:00", "subject": "fix(core): scan pre-registered dynamic imports of lazy modules", "body": "A lazily loaded `DynamicModule` whose dynamic metadata carries `imports`\nnever had those imports scanned, so their providers could not resolve.\n\n`NestContainer#setModule` pre-registers a dynamic module's `imports` via\n`addDynamicMetadata` -> `addDynamicModules` -> `addModule` before the\nscanner recurses into them. Those modules therefore report\n`inserted: false`, and the lazy skip condition treated them as already\nloaded, leaving them registered but never scanned or instantiated.\n\nSkip only modules that were actually instantiated, and bind global\nproviders on every lazy pass: a rescanned module reports\n`inserted: false`, so `lazy && moduleInserted` skipped binding `@Global`\nproviders into it.\n\nCloses #17462", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "vue-core", "sha": "f1a7af549d99753c1ceab0b53ea7d639a55a70e5", "date": "2026-08-13T13:42:20+08:00", "subject": "fix(runtime-vapor): clean up removed v-for components", "body": "Capture cleanup callbacks registered while rendering component v-for items\nand transfer them from the owner scope to the corresponding ForBlock.\n\nRun the captured callbacks after structural removal so removed component\ninstances and their DOM can be released immediately. This preserves the\ncomponent v-for optimization by avoiding an additional EffectScope for every\nitem.", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "openssl", "sha": "39a7a0f54975988a8452f9b83dbeb7224e72d914", "date": "2026-08-13T15:00:13+09:00", "subject": "test: fix FIPS identity DRBG regression coverage", "body": "Use the configuration filename that was actually added. Set the random\nconfiguration property to fips=yes so the existing RSA checks use a DRBG\nfrom the renamed FIPS provider, rather than one from the default\nprovider.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Norbert Pocs \nMerge-date: Fri Aug 21 11:51:15 2026\nMerged-from: https://github.com/openssl/openssl/pull/32357", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "431a34b15ef81a0d32a69d87e4b7a7065b517c60", "date": "2026-08-13T15:00:13+09:00", "subject": "test: fix FIPS identity DRBG regression coverage", "body": "Use the configuration filename that was actually added. Set the random\nconfiguration property to fips=yes so the existing RSA checks use a DRBG\nfrom the renamed FIPS provider, rather than one from the default\nprovider.\n\nAssisted-by: Codex:gpt-5.6-sol\n\nReviewed-by: Simo Sorce \nReviewed-by: Andrew Dinh \nMergeDate: Tue Aug 18 16:22:41 2026\n(Merged from https://github.com/openssl/openssl/pull/32355)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "03fcd04c7c7519d552ca1e0d5d2619215d654b59", "date": "2026-08-13T15:00:13+09:00", "subject": "test: fix FIPS identity DRBG regression coverage", "body": "Use the configuration filename that was actually added. Set the random\nconfiguration property to fips=yes so the existing RSA checks use a DRBG\nfrom the renamed FIPS provider, rather than one from the default\nprovider.\n\nAssisted-by: Codex:gpt-5.6-sol\n\nReviewed-by: Richard Levitte \nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Neil Horman \nMergeDate: Fri Aug 14 12:41:55 2026\n(Merged from https://github.com/openssl/openssl/pull/32346)\n\n(cherry picked from commit fd2e2860a415c6febbb180ab4133dfa0e9d8d6a6)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "fd2e2860a415c6febbb180ab4133dfa0e9d8d6a6", "date": "2026-08-13T15:00:13+09:00", "subject": "test: fix FIPS identity DRBG regression coverage", "body": "Use the configuration filename that was actually added. Set the random\nconfiguration property to fips=yes so the existing RSA checks use a DRBG\nfrom the renamed FIPS provider, rather than one from the default\nprovider.\n\nAssisted-by: Codex:gpt-5.6-sol\n\nReviewed-by: Richard Levitte \nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Neil Horman \nMergeDate: Fri Aug 14 12:41:55 2026\n(Merged from https://github.com/openssl/openssl/pull/32346)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "angular", "sha": "1aeec887b4240dd40c544dc2b0d19f0908571446", "date": "2026-08-13T06:58:21Z", "subject": "build: update dependency tar.bzl to v0.10.8", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "aac355dcb0e8fe0d694699ea18ab59031f253cbe", "date": "2026-08-13T15:26:45+08:00", "subject": "fix(runtime-vapor): remove v-for components before scope cleanup", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "22b9b41bc1457553dddf78ad2ab7992526e5ba41", "date": "2026-08-13T15:31:09+08:00", "subject": "fix(runtime-vapor): scope prop source caches to consumers", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vitest", "sha": "e75c8d159cde1f5946dc3422799cdeb42277b648", "date": "2026-08-13T16:47:11+09:00", "subject": "chore: add bundled workspace dep build gotcha in AGENTS.md", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vitest", "sha": "68352a3960f80c5d51ae5b77381266119c9be57a", "date": "2026-08-13T16:49:13+09:00", "subject": "test(browser): update browser mode snapshot error assertion", "body": "", "style": "conventional", "type": "test", "scope": "browser", "breaking": false} {"repo": "vue-core", "sha": "e4d972543470f260dbb73d31c868f32f7fb4ff69", "date": "2026-08-13T15:54:14+08:00", "subject": "test(runtime-vapor): cover component v-for effect cleanup", "body": "", "style": "conventional", "type": "test", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "75a8792064fbbd2cb360477987ce06cfc7223b88", "date": "2026-08-13T16:04:52+08:00", "subject": "fix(runtime-vapor): normalize declared style props (fix #15285)", "body": "Closes #15285", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "e8a09b3f42b83afa2add18c2e533f8163b44f934", "date": "2026-08-13T16:26:36+08:00", "subject": "fix(runtime-vapor): scope fallback prop caches to active consumers", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "9d034539ece1e04f6fb7cb9b7ea02111312f7456", "date": "2026-08-13T10:39:28+02:00", "subject": "tool_operate: limit `is_using_schannel()` call to Windows", "body": "Follow-up to cd238bb7addf5b1f174df47ccc61651a545c4876 #22552\n\nCloses #22568", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "2720362818cdeb2a940171e4ab6f21cf78c6a302", "date": "2026-08-13T04:03:59-05:00", "subject": "feat(router): add containsTree as public API", "body": "Export containsTree from @angular/router to enable direct UrlTree subset matching.", "style": "conventional", "type": "feat", "scope": "router", "breaking": false} {"repo": "openssl", "sha": "79fd1a80fe0339c683b5cbce090812d49c5c69d0", "date": "2026-08-13T11:17:51+02:00", "subject": "doc: drop OPENSSL_armcap(3) references in EVP_CIPHER-SM4", "body": "The OPENSSL_armcap(3) man page does not exist on the 3.4 branch,\nwhich makes 'make doc-nits' (check_docs CI job) fail with:\n\n doc/man7/EVP_CIPHER-SM4.pod:1: reference to non-existing OPENSSL_armcap(3)\n\nRemove the references, keeping the existing OPENSSL_ia32cap(3) and\nOPENSSL_riscvcap(3) ones.\n\nFixes: d0aa08287cae \"Document SM4 table-lookup side-channel risk\"\nAssisted-by: Pi:moonshotai/kimi-k3\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 13 13:24:24 2026\nMerged-from: https://github.com/openssl/openssl/pull/32350", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c37abfdd65030bb11b708e43504b47fc1c96ab31", "date": "2026-08-13T11:18:56+02:00", "subject": "test: use portable function-name macro in QUIC test helper", "body": "Signed-off-by: Nikola Pajkovsky \n\nReviewed-by: Saša Nedvědický \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Viktor Dukhovni \nMergeDate: Thu Aug 13 12:15:34 2026\n(Merged from https://github.com/openssl/openssl/pull/32349)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "curl", "sha": "ca26e728a62b8f22032f29445c8e4e865cb9e3ad", "date": "2026-08-13T11:34:42+02:00", "subject": "conncache: don't assume curl_off_t increment wrap-around", "body": "In the totally unlikely event that next_easy_id would ever wrap, avoid\nundefined behavior. It is a signed type.\n\nAlso: update the related comment in urldata.h that mentioned LONG_MAX\nbut is nowadays CURL_OFF_T_MAX\n\nCloses #22569", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "5510e44fcc3a401214cd5f20c415d8af193a7b11", "date": "2026-08-13T11:51:26+02:00", "subject": "tool_operate: refactor schannel checks and move win32 cert checks to function", "body": "Follow-up to 9d034539ece1e\n\nCloses #22570", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "30eb1faeada87e77562fc88a8ab1acc07610a23d", "date": "2026-08-13T11:51:26+02:00", "subject": "tool_operate: reduce ifdefs for the is_using_schannel check", "body": "Follow-up to 9d034539ece1e", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "2f0f5c29931548a39e0d8138e0bdafb135f628d4", "date": "2026-08-13T13:09:13+02:00", "subject": "tool_operate: move the win32 cert checks into its own function", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "e8a8ccc3ca192178fd3a179a06a91368008899dc", "date": "2026-08-13T13:18:12+02:00", "subject": "pgcrypto: remove unused field mdc_checked from PGP_Context", "body": "mdc_finish used to set this field once it had successfully verified\nthe MDC packet, with its only reader in the same function. This was\nremoved by a59ee881978 when mdc_finish stopped being called through\nthe pullf_read API, the field has been unused since.\n\nNote that MDC verification does not rely on it: process_data_packets\ntracks whether an MDC packet was seen in its local got_mdc variable\nand rejects the message if a required MDC is missing, while the\nuse_mdcbuf_filter path checks the hash in mdcbuf_finish().\n\nAuthor: Aleksander Alekseev \nDiscussion: https://postgr.es/m/CAJ7c6TNmHnnCkn3uZwRUE7WWYftW--48JFoDVDqoZqMrLYxUZw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "7a061e9a39cd8c55dcd6b0556f6cdff460290215", "date": "2026-08-13T13:19:29+02:00", "subject": "repair build", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "4e26486343180148be03681c96b25814f4eb490f", "date": "2026-08-13T13:21:40+02:00", "subject": "more fix", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "7103a93b05bc69ea98ed9d05d02fa9eeba533f2f", "date": "2026-08-13T13:23:53+02:00", "subject": "connections: use admin handles only for maintenance", "body": "Maintenance tasks like connection upkeep, liveliness checks and cache\neviction must only run on an admin handle, not the application transfer\nfrom the calling context.\n\nThen application transfers will get only attached to the connections\nthey actually work on. This makes `data->state.recent_conn_id`\nmeaningful for checks if a subsequent address is finding the previous\nconnection again.\n\nOne thing led to another:\n- add Curl_get_admin(data), to obtain an admin handle for\n an application handle, inheriting some properties for\n connection operation.\n- rename `Curl_easy *data` parameter to `Curl_easy *admin`\n where only admin handles should be passed. Add DEBUGASSERTs\n in the called function.\n- rename `Curl_conn_terminate()` to `Curl_conn_close()`\n because this makes connections enter the shutdown close\n where possible.\n- rename `Curl_cshutdn_terminate()` to `Curl_conn_terminate()`\n since this definitely kills the connection and does not\n involve a `cshutdn` instance.\n- add Curl_share_lock_share(), Curl_share_unlock_share()\n so it possible to operate on a share without the passed\n easy handle having the share set.\n This catches a case where a shared needed to be locked\n but was not before. Adjust test1554 results.\n\nFixes #22567\nReported-by: cybertron10 on github\nCloses #22572", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "bf4a17c0abf787029a0ac9bb8189a9b38ca8aea7", "date": "2026-08-13T13:39:06+02:00", "subject": "fixup comment", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "886da9780f2d6607f5e78170f58cdf2062c99219", "date": "2026-08-13T13:57:01+02:00", "subject": "docs: update naming-conventions skill", "body": "(cherry picked from commit 560d3f04122889957bdca86787bc165a051bcfaa)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "560d3f04122889957bdca86787bc165a051bcfaa", "date": "2026-08-13T13:57:01+02:00", "subject": "docs: update naming-conventions skill", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "82b289c9527890a0766b55be2214b3749283307a", "date": "2026-08-13T14:02:26+02:00", "subject": "test_20_11: more leniency", "body": "Loosen the RSS increase check slightly to make the test less flaky.\nAlso use 'rss-max' stat.\n\nCloses #22574", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a2b64736bb657163dfb085f279b4e1272b647448", "date": "2026-08-13T15:03:15+03:00", "subject": "Fix missing hash_seq_term()", "body": "This was harmless because the function (ThereAreNoReadyPortals()) has\nonly one caller, and that caller throws an error if the function\nreturns false and transaction abort cleans up the in-progress hash\nsearches. But if someone called the function in some other way,\nthey'd get a warning about a leaked hash_seq_search scan.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "776348900b40eb19766ebc88df034d368543707c", "date": "2026-08-13T15:03:22+03:00", "subject": "Remove unnecessary #include ", "body": "Commit cba5b994c9 removed the call to poll() that needed it.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2ae459b08523ab110d1e2d60b46eb6bcd253bbdc", "date": "2026-08-13T15:04:40+03:00", "subject": "Fix misc spelling mistakes in comments, docs and tests", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "0a708e95bf650339c434be65b0a0f2b246370dc9", "date": "2026-08-13T17:35:02+05:30", "subject": "docs: note SafeResourceUrl audio src change", "body": "(cherry picked from commit 647bf8e1436d5479b475d6364c427662586e4c88)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "647bf8e1436d5479b475d6364c427662586e4c88", "date": "2026-08-13T17:35:02+05:30", "subject": "docs: note SafeResourceUrl audio src change", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "76096c769153f183d7a8554bc392b41ebb5493bd", "date": "2026-08-13T17:06:57+05:00", "subject": "fix(pool): worker init to handle queued synchronous messages", "body": "", "style": "conventional", "type": "fix", "scope": "pool", "breaking": false} {"repo": "curl", "sha": "3298dd1f94cfc69b260c0241c9900a073f31c81d", "date": "2026-08-13T14:51:36+02:00", "subject": "call it win32_using_schannel", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "b78adb98bac2d589551a5b09bb0a7e39887321f4", "date": "2026-08-13T15:52:30+03:00", "subject": "fix(wasm): remove custom log path and disable autoread", "body": "", "style": "conventional", "type": "fix", "scope": "wasm", "breaking": false} {"repo": "curl", "sha": "dbdd0e0a3ce7d3388475604469a7b0dfc2526f37", "date": "2026-08-13T14:52:33+02:00", "subject": "combine ifdef sections", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "faf8345eef8c9a59f254d3e4bafab1a9c125ee21", "date": "2026-08-13T15:17:16+02:00", "subject": "fix(diagnostic): don't accumulate BufRead autocmds for unloaded buffers #40869", "body": "Problem:\nvim.diagnostic.set() defers extmark position computation for an\nunloaded buffer via a once=true BufRead autocmd, registering a new one\non every call without replacing the previous one. Each pending autocmd\nalso retains that call's diagnostics.\n\nSolution:\nInstead of registering an autocmd per set() call, register a single\nstatic BufRead autocmd that computes positions from the diagnostic\ncache for any buffer with cached diagnostics when it is read. This\nremoves the per-call registration entirely (nothing left to\naccumulate) and means diagnostics cleared while the buffer was\nunloaded no longer produce stale extmarks.", "style": "conventional", "type": "fix", "scope": "diagnostic", "breaking": false} {"repo": "angular", "sha": "5a516cdc373158c3dc23cc03a627e4507872ef71", "date": "2026-08-13T13:55:46Z", "subject": "build: update all github actions", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c856bc543403e023426b4a381040ae0bb2c7dd05", "date": "2026-08-13T22:13:13+07:00", "subject": "Extend qlog coverage to radix test", "body": "So that frame-type check covers both frameworks\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Norbert Pocs \nMerge-date: Wed Sep 2 06:32:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32335", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "07605ba793b9c2e23095f1726ad19ffd0d36676d", "date": "2026-08-13T18:18:35+03:00", "subject": "feat(devtools): implement CD analyzer", "body": "Implement component highlighting on change detection cycles along with embedding\nthe component-specific data to the directive/component tree explorer.\n\nCloses #59057\n\nPR Close #70254", "style": "conventional", "type": "feat", "scope": "devtools", "breaking": false} {"repo": "postgres", "sha": "ef9063a8870fb5fb1666da3399ff666201eccf70", "date": "2026-08-13T11:20:27-04:00", "subject": "Consistently enforce tsvector/tsquery maximum lengths.", "body": "Some places rejected individual tokens longer than MAXSTRLEN, while\nothers rejected ones longer than MAXSTRLEN-1. The data structure is\nperfectly capable of handling MAXSTRLEN, so there's nothing wrong\nwith using the looser bound. Moreover, as things stand there is a\ndump/reload hazard: some code paths permit construction of a tsvector\nor tsquery that would later be rejected by tsvectorin or tsqueryin.\nSo standardize on using MAXSTRLEN.\n\nIdentical remarks apply to MAXSTRPOS (the total data length),\nso fix that too.\n\nBack-patch, in hopes of avoiding cases where a value acceptable to\none supported release is not acceptable to another.\n\nAuthor: Tom Lane \nReviewed-by: Zsolt Parragi \nDiscussion: https://postgr.es/m/CAN4CZFNYQo4zfbRR435uD0vSfuy5y7dnFOXDfKr9zYoL1JnAxA@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "68b5e6013878b1842e70c2af2f23d67427216910", "date": "2026-08-13T17:27:57+02:00", "subject": "dict: avoid busy-loop in sendf() when the socket is not writable", "body": "Curl_xfer_send() reports CURLE_AGAIN as a successful zero-byte send, so\nthe retry loop spun at 100% CPU and ignored the timeout when the peer\nstopped draining. Wait for writability and check the remaining time\nbetween retries.\n\nCloses #22576", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "48a0fd6e8a8d14bdc1d901ee5615f4b0ab698fe8", "date": "2026-08-13T15:44:06Z", "subject": "feat(compiler): allow template to access private props", "body": "To allow this we'll catch the errors during typechecking and discard it.\n\nAs context, when setting `isolatedDeclarations: true` this requires developers to explicitly type every property but the `private` ones. By allowing private properties to be used in templates we discard the actually for explicit typing for template only properties.", "style": "conventional", "type": "feat", "scope": "compiler", "breaking": false} {"repo": "openssl", "sha": "5f61ae81549a6872869469bcb2784cdc08be9271", "date": "2026-08-13T17:58:42+02:00", "subject": "Fix memory leak in PKCS12_parse", "body": "If appending an additional certificate to the CA stack fails\nPKCS12_parse() leaves certificates appended earlier during the call\nin the output stack. If PKCS12_parse() allocated the stack\nthen the whole stack might be leaked if caller assumes failure.\n\nFix:\nRecord the original stack and its size then on failure\nremove certificates appended during the failed call.\nFree the stack if it was allocated by PKCS12_parse().\n\nFound by : store fuzzer (MFAIL-test)\n\nAssisted-by: ChatGPT:gpt-5.6\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Mounir Idrassi \nMerge-date: Thu Aug 27 16:49:42 2026\nMerged-from: https://github.com/openssl/openssl/pull/32407", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "640460d6065cdcd8fe407c7b35883f81f5045473", "date": "2026-08-13T19:02:27+03:00", "subject": "fix(platform-browser): throw a descriptive error when insertBefore reference node is missing", "body": "Angular's internal LView/TNode bookkeeping can get out of sync with the\nreal DOM: manual DOM manipulation, a browser extension, or an edge case in\nAngular's own view-insertion/reordering code can all leave Angular believing\na node is still attached at a given position when it isn't. The next time\nAngular's renderer calls `insertBefore` relative to that stale reference\nnode, the native DOM API throws an opaque `NotFoundError` with no indication\nof which node or component was involved, making these errors effectively\nundebuggable in production:\n\n NotFoundError: Failed to execute 'insertBefore' on 'Node': The node\n before which the new node is to be inserted is not a child of this node.\n at Node.insertBefore (native)\n at DefaultDomRenderer2.insertBefore (packages/platform-browser/src/dom/dom_renderer.ts)\n at nativeInsertBefore (packages/core/src/render3/dom_node_manipulation.ts)\n at nativeAppendOrInsertBefore (packages/core/src/render3/dom_node_manipulation.ts)\n ... (called while Angular inserts or moves a view during change detection)\n\nCheck the reference node's actual parent against the expected parent before\ncalling the native `insertBefore`, and throw a descriptive `RuntimeError`\n(NG05106) instead, following the same pattern already used for hydration\nnode mismatches.\n\n(cherry picked from commit 1cb3d606bf5bb01ec9c9e9289703a79f64c76e89)", "style": "conventional", "type": "fix", "scope": "platform-browser", "breaking": false} {"repo": "angular", "sha": "1cb3d606bf5bb01ec9c9e9289703a79f64c76e89", "date": "2026-08-13T19:02:27+03:00", "subject": "fix(platform-browser): throw a descriptive error when insertBefore reference node is missing", "body": "Angular's internal LView/TNode bookkeeping can get out of sync with the\nreal DOM: manual DOM manipulation, a browser extension, or an edge case in\nAngular's own view-insertion/reordering code can all leave Angular believing\na node is still attached at a given position when it isn't. The next time\nAngular's renderer calls `insertBefore` relative to that stale reference\nnode, the native DOM API throws an opaque `NotFoundError` with no indication\nof which node or component was involved, making these errors effectively\nundebuggable in production:\n\n NotFoundError: Failed to execute 'insertBefore' on 'Node': The node\n before which the new node is to be inserted is not a child of this node.\n at Node.insertBefore (native)\n at DefaultDomRenderer2.insertBefore (packages/platform-browser/src/dom/dom_renderer.ts)\n at nativeInsertBefore (packages/core/src/render3/dom_node_manipulation.ts)\n at nativeAppendOrInsertBefore (packages/core/src/render3/dom_node_manipulation.ts)\n ... (called while Angular inserts or moves a view during change detection)\n\nCheck the reference node's actual parent against the expected parent before\ncalling the native `insertBefore`, and throw a descriptive `RuntimeError`\n(NG05106) instead, following the same pattern already used for hydration\nnode mismatches.", "style": "conventional", "type": "fix", "scope": "platform-browser", "breaking": false} {"repo": "angular", "sha": "a9f8fb2ef9ace8d6188d91abe375795c18771ba0", "date": "2026-08-13T16:43:17Z", "subject": "build: update cross-repo angular dependencies to v22.1.4", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "95d131138c5df177ebfb839584f1308adca420e3", "date": "2026-08-13T10:22:30-07:00", "subject": "refactor(router): rename activated route injector feature to resource feature", "body": "Activated Route Injector feature was developed for resources. This moves the internal\nnaming (files and vars) to reflect this now that we have landed the core bits of plumbing\nand are ready to implement router resources.\n\nPR Close #70211", "style": "conventional", "type": "refactor", "scope": "router", "breaking": false} {"repo": "vite", "sha": "a20a35ec0685e374519864d0f41dd5f6e9ba0271", "date": "2026-08-13T23:01:22+05:30", "subject": "docs(build): note cssTarget precedence", "body": "", "style": "conventional", "type": "docs", "scope": "build", "breaking": false} {"repo": "vite", "sha": "dcf88bd2ad2b1a8845f9029587cc8c825e382d42", "date": "2026-08-13T22:42:42+05:00", "subject": "fix(define): fix match escaped dots to support $-prefixed define keys", "body": "", "style": "conventional", "type": "fix", "scope": "define", "breaking": false} {"repo": "axios", "sha": "2d2a21af8a433089474a2149781799c93acbcf3c", "date": "2026-08-13T23:25:34+05:30", "subject": "fix(interceptors): tolerate nullish handlers in syncHandlerEntries", "body": "", "style": "conventional", "type": "fix", "scope": "interceptors", "breaking": false} {"repo": "axios", "sha": "d8a919fd81403d59058c0e9dbefc540407dee83f", "date": "2026-08-13T20:05:22+02:00", "subject": "fix(xhr): flush final progress during the live loadend dispatch", "body": "", "style": "conventional", "type": "fix", "scope": "xhr", "breaking": false} {"repo": "axios", "sha": "e6824eec5fcf9da467a9792724396badc490c469", "date": "2026-08-13T23:44:45+05:30", "subject": "fix: core methodList, HTTP adapter errors, and add tests", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4e2810c8d5849ea44bb1546822d11438b0d97b2d", "date": "2026-08-13T20:58:07+02:00", "subject": "test_pending_limit(): Use OPENSSL_FUNC instead of __func__", "body": "This is needed on 3.5 as that supports -ansi builds.\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Richard Levitte \nReviewed-by: Milan Broz \nMerge-date: Fri Aug 21 10:49:43 2026\nMerged-from: https://github.com/openssl/openssl/pull/32356", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "19720bf9a30f372f9c9a169fb3d983905efa9654", "date": "2026-08-13T20:58:07+02:00", "subject": "test_pending_limit(): Use OPENSSL_FUNC instead of __func__", "body": "This is needed on 3.5 as that supports -ansi builds.\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Richard Levitte \nReviewed-by: Milan Broz \nMerge-date: Fri Aug 21 10:49:31 2026\nMerged-from: https://github.com/openssl/openssl/pull/32356", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "85350202bfef7c85a6c4971fa033708e0d1471c8", "date": "2026-08-13T20:58:07+02:00", "subject": "test_pending_limit(): Use OPENSSL_FUNC instead of __func__", "body": "This is needed on 3.5 as that supports -ansi builds.\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Richard Levitte \nReviewed-by: Milan Broz \nMerge-date: Fri Aug 21 10:49:15 2026\nMerged-from: https://github.com/openssl/openssl/pull/32356", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c7c42afab91c46e6baaa0e2801626e13acd6c30a", "date": "2026-08-13T21:14:02+02:00", "subject": "file: support directory listing on Windows", "body": "Closes #22577", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4972eeeb6aad64a6f362fd7b0b1e73e8e16e0ac8", "date": "2026-08-13T22:10:22+02:00", "subject": "Fix NULL dereference in try_key_value()", "body": "In store_result.c:try_key_value() return value of\nOSSL_DECODER_CTX_new_for_pkey() is not checked so on\nfailed memory allocation OSSL_DECODER_CTX_set_passphrase_cb()\ndereferences NULL without any check.\n\nFound by : store fuzzer (MFAIL-test)\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Neil Horman \nMergeDate: Tue Sep 15 14:20:04 2026\n(Merged from https://github.com/openssl/openssl/pull/32413)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d7db169fa13a7e4db9013816c7a2c498bd304fba", "date": "2026-08-13T15:28:34-05:00", "subject": "Reject CLUSTER (ANALYZE).", "body": "The introduction of the REPACK command also added support for\nCLUSTER (ANALYZE). Presumably this was unintentional, as the\nCLUSTER docs make no mention of it. CLUSTER (ANALYZE) on an\nordinary table succeeds and does indeed analyze the table, but a\ndatabase-wide CLUSTER (ANALYZE) or one on a partitioned table\nfails with an error like:\n\n ERROR: cannot execute REPACK (ANALYZE) on multiple tables\n\nRather than improving support for CLUSTER (ANALYZE) and thereby\nencouraging folks to use CLUSTER instead of REPACK, let's just\nreject it. This commit does so by teaching ExecRepack() to ERROR\nfor the ANALYZE option in anything except REPACK commands. Note\nthat VACUUM (FULL, ANALYZE) does not go through ExecRepack() and\ntherefore is unaffected by this change.\n\nOversight in commit ac58465e06.\n\nReported-by: Zsolt Parragi \nAuthor: Zsolt Parragi \nDiscussion: https://postgr.es/m/CAN4CZFMVcgv1b2G-i%2Bkhs2MDnzWSr7O_53j_x%2BuhAxUV5rwWqw%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f666172c4651d5a0a5f6425f899b3ef1a1b2eaac", "date": "2026-08-13T23:17:46+02:00", "subject": "Fix memory leak in ossl_decoder_instance_new_forprov()", "body": "On failed memory allocation in decoder_lib.c and ossl_decoder_instance_new_forprov()\nallocation of return value OSSL_DECODER_INSTANCE* is not checked\ncausing decoderctx to leak.\n\nFound by : store fuzzer (MFAIL-test)\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Neil Horman \nMergeDate: Tue Sep 15 14:20:03 2026\n(Merged from https://github.com/openssl/openssl/pull/32413)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "c04931c88b1c7959cb95abd28cf5b3614d97e168", "date": "2026-08-14T00:20:55+03:00", "subject": "fix(core): throw a descriptive error instead of crashing when a hydration node is missing", "body": "locateOrCreateElementNodeImpl looks up the DOM node for an element\nduring hydration and immediately checks its nodeType. If the\nclient-rendered DOM has fewer nodes than the server-rendered HTML,\nthe lookup returns null, and in production that null flows straight\ninto the nodeType check and crashes with a raw, uncoded\n\"Cannot read properties of null (reading 'nodeType')\" TypeError.\n\nThe dev-mode check that would normally catch this (validateMatchingNode)\nalready handles a missing node, but it's compiled out of production\nbuilds, so the crash only shows up outside of dev mode.\n\nAdd a null check ahead of the nodeType check that throws a coded\nRuntimeError using the existing HYDRATION_MISSING_NODE (NG0502) code,\nwith a descriptive message in dev mode and a cheap fallback in\nproduction. Also add a regression test that removes a server-rendered\nelement before hydration runs and asserts a coded RuntimeError is\nthrown instead of a raw TypeError.\n\n(cherry picked from commit c6e4a36be1246e0fbe06d591a69ac79cd74348d2)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "c6e4a36be1246e0fbe06d591a69ac79cd74348d2", "date": "2026-08-14T00:20:55+03:00", "subject": "fix(core): throw a descriptive error instead of crashing when a hydration node is missing", "body": "locateOrCreateElementNodeImpl looks up the DOM node for an element\nduring hydration and immediately checks its nodeType. If the\nclient-rendered DOM has fewer nodes than the server-rendered HTML,\nthe lookup returns null, and in production that null flows straight\ninto the nodeType check and crashes with a raw, uncoded\n\"Cannot read properties of null (reading 'nodeType')\" TypeError.\n\nThe dev-mode check that would normally catch this (validateMatchingNode)\nalready handles a missing node, but it's compiled out of production\nbuilds, so the crash only shows up outside of dev mode.\n\nAdd a null check ahead of the nodeType check that throws a coded\nRuntimeError using the existing HYDRATION_MISSING_NODE (NG0502) code,\nwith a descriptive message in dev mode and a cheap fallback in\nproduction. Also add a regression test that removes a server-rendered\nelement before hydration runs and asserts a coded RuntimeError is\nthrown instead of a raw TypeError.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "be0076b71b68f2e7cb2553af48aa8f8ecc8bda40", "date": "2026-08-13T16:35:07-05:00", "subject": "Fix authorization check for role membership changes.", "body": "Presently, check_role_membership_authorization() decides whether\nthe current user may grant or revoke membership in a role by\ncalling is_admin_of_role(), which recurses through all grants,\nwhile it chooses the grantor to record for the resulting entry by\ncalling select_best_admin(), which recurses only through inherited\ngrants. When the two disagree, the permission check passes and the\ngrantor lookup then comes up empty, so the user sees an internal\n\"no possible grantors\" error. ALTER GROUP ... ADD USER reaches the\nsame error through the separate check in AlterRole().\n\nTo fix, teach both checks to search the same way\nselect_best_admin() does via a new has_admin_privs_of_role(). The\nnew check passes exactly when the grantor lookup was going to\nsucceed, so nothing that works today starts failing; the internal\nerror simply becomes a proper permission error. Note that this\nleaves the other callers of is_admin_of_role() alone, so a role\nreachable only through a non-inherited grant can still be dropped,\nrenamed, or altered. Whether that ought to change as well is left\nas a future exercise.\n\nOversight in commit ce6b672e44.\n\nReported-by: ChangAo Chen \nAuthor: ChangAo Chen \nReviewed-by: Chao Li \nReviewed-by: Pretham \nReviewed-by: Robert Haas \nReviewed-by: Jacob Champion \nDiscussion: https://postgr.es/m/tencent_ADCE2B34B230A9B631854806104FEF40C105%40qq.com\nDiscussion: https://postgr.es/m/CAJUn_kN%2BMhbb8fYP5xxQCq1KEziOinM6HgYx4ts_pPDnQ2y1nQ%40mail.gmail.com\nBackpatch-through: 16", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "b3c78a50816af5685b2a6ed51166d08941c09946", "date": "2026-08-14T00:47:55+03:00", "subject": "fix(common): preserve literal key union in KeyValuePipe.transform()", "body": "Previously, when you passed an object typed like\nRecord<'a' | 'b', number> into the `keyvalue` pipe, TypeScript would\n\"forget\" that the keys could only ever be 'a' or 'b', and just tell\nyou the key was a plain `string` instead. So code like this used to\nfail to compile, even though it's correct:\n\n```ts\n const input: Record<'a' | 'b', number> = {a: 1, b: 2};\n const result = pipe.transform(input);\n const key: 'a' | 'b' = result[0].key; // error: string is not 'a' | 'b'\n```\n\nThis happened because the pipe has multiple overloaded versions of\ntransform(), and TypeScript checks them top to bottom, using the\nfirst one that matches. The \"number keys\" overload was listed first,\nand it happened to also match string-keyed objects by accident, so\nit \"won\" before the correct \"string keys\" overload ever got a\nchance to run.\n\nThe fix just reorders those two overloads so the string-keys one is\nchecked first. Nothing about runtime behavior changes — objects with\nactual numeric keys (e.g. Record<1 | 2, string>) still correctly\nreport their keys as plain `string`, matching what Object.keys()\nreally returns at runtime.\n\n(cherry picked from commit 46d2cb7ff049dc4549080f64dceb1541b3a87cee)", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "angular", "sha": "46d2cb7ff049dc4549080f64dceb1541b3a87cee", "date": "2026-08-14T00:47:55+03:00", "subject": "fix(common): preserve literal key union in KeyValuePipe.transform()", "body": "Previously, when you passed an object typed like\nRecord<'a' | 'b', number> into the `keyvalue` pipe, TypeScript would\n\"forget\" that the keys could only ever be 'a' or 'b', and just tell\nyou the key was a plain `string` instead. So code like this used to\nfail to compile, even though it's correct:\n\n```ts\n const input: Record<'a' | 'b', number> = {a: 1, b: 2};\n const result = pipe.transform(input);\n const key: 'a' | 'b' = result[0].key; // error: string is not 'a' | 'b'\n```\n\nThis happened because the pipe has multiple overloaded versions of\ntransform(), and TypeScript checks them top to bottom, using the\nfirst one that matches. The \"number keys\" overload was listed first,\nand it happened to also match string-keyed objects by accident, so\nit \"won\" before the correct \"string keys\" overload ever got a\nchance to run.\n\nThe fix just reorders those two overloads so the string-keys one is\nchecked first. Nothing about runtime behavior changes — objects with\nactual numeric keys (e.g. Record<1 | 2, string>) still correctly\nreport their keys as plain `string`, matching what Object.keys()\nreally returns at runtime.", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "openssl", "sha": "f9f6047e9c0c8011d3c875003d4b03f74e05f295", "date": "2026-08-13T16:00:15-06:00", "subject": "Allow pushing an empty ASN1_UTF8STRING onto a CMP freeText stack", "body": "Reviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:53 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d24764e778d92aff18be5798df1fa8fd65753244", "date": "2026-08-13T16:02:13-06:00", "subject": "Add test coverage for ASN1_UNIVERSALSTRING_to_string", "body": "Reviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "28995f051e72a1efae0e85d0ea43f80c38166657", "date": "2026-08-14T07:37:01+09:00", "subject": "Use explicit fetching of digests in channel binding (OpenSSL >= 3.0)", "body": "This commit touches both the libpq and backend-side code of channel\nbinding where respectively pgtls_get_peer_certificate_hash() and\nbe_tls_get_certificate_hash() are upgraded to retrieve digests using the\nmethod recommended by OpenSSL 3.0: no more direct EVP_sha256() or\nsimilar, just a EVP_MD_fetch() through EVP to get an algorythm type,\nbased on a name.\n\nThe pre-3.0 code is still required for LibreSSL and as long as we\nsupport OpenSSL 1.1.1.\n\nSimilar work has been done in b91f79cd08ab and 1f3b9bb109b8.\n\nAuthor: Mark Atwood \nCo-authored-by: Michael Paquier \nDiscussion: https://postgr.es/m/178596055149.1584250.13974609482797470185@reviewcommit.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "fc3f71fb7d5fdc7746576dcf4c896b94d54c4711", "date": "2026-08-13T18:38:39-05:00", "subject": "fix(messages): `inputlist()` offers the mouse when it cannot be used", "body": "Problem:\n`inputlist()` advertises \"click with the mouse\" purely because it\nimplements click selection, so under the default `'mouse'` of \"nvi\" it\noffers a click that command-line mode never receives.\n\nSolution:\nOnly offer the mouse when `'mouse'` covers command-line mode, the same\ncondition `:help inputlist()` already documents.", "style": "conventional", "type": "fix", "scope": "messages", "breaking": false} {"repo": "postgres", "sha": "f2ccb8e6c5915617466ccea2abcf2d443d67dab0", "date": "2026-08-14T09:07:54+09:00", "subject": "Refactor PgStat_TableStatus to new PgStat_RelationStatus", "body": "This new structure is split depending on the stats kind it deals with:\n- PGSTAT_KIND_RELATION, for tables.\n- PGSTAT_KIND_INDEX, for indexes.\n\nThis change designs a cleaner barrier for the handling of stats data\nrelated to tables and indexes, by being able to track precisely what are\nthe counters used by one or the other for pending data. Using a common\nground for both eases the tracking of Relations in the relcache, with a\nstructure based on a union separated by stats kind.\n\nThe code originally pointed to some counters that may not be used at\nall. For example, indexes have no need for the tracking of\nsub-transaction data, still the code implied that these could be touched\nfor an index. A consequence is that PgStat_TableXactStatus is not\nrenamed, as it is not used by indexes.\n\nSome asserts are added as an extra layer of protection to prevent the\nupdate of fields that should not be touched. find_tabstat_entry() and\nfind_tabstat_entry_kind() are removed, replaced by a single\nfind_relstat_entry_kind() able to work for indexes and the rest.\n\n72a6dad1c911 took care of the shared memory and on-disk part of the\nsplit. This focuses on the pending data part at backend level.\n\nAuthor: Michael Paquier \nReviewed-by: Bertrand Drouvot \nDiscussion: https://postgr.es/m/f572abe7-a1bb-e13b-48c7-2ca150546822@gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "dc2637592af536255bfe5756bed4e74c6a6b3f9e", "date": "2026-08-14T08:26:18+08:00", "subject": "docs: make AGENTS.md the canonical project guide", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "9dd7d164c0c6cb82cbfdec3eb93a0c80e849799b", "date": "2026-08-14T08:27:11+08:00", "subject": "site: add skeleton for homepage contributor avatars", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "3be9ea04ab3e42dd7156ed03dec30aefccea62c7", "date": "2026-08-14T08:31:00+08:00", "subject": "docs(List): update with Listy migration guidance", "body": "* docs(List): update with Listy migration guidance\n\n* update\n\n* Refine List deprecation notice in English docs", "style": "conventional", "type": "docs", "scope": "List", "breaking": false} {"repo": "postgres", "sha": "1f97b252dfee27b6300959b8403a4a9a2003046e", "date": "2026-08-14T09:45:24+09:00", "subject": "Fix thinko in error of algorithm type lookup for channel binding", "body": "The error printed would always use a NULL value, due to OBJ_nid2sn()\nknown to fail. This updates the error message to use the algorithm\nnumber instead.\n\nOversight in 28995f051e72, in the shape of copy-pasto failure.\n\nReported-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAHGQGwGPjENxX1SmkPrv9Rb-Fnij8jN4ginObmSTnWXqROS2+g@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "17f69cae0c25d921c3fe713a16aa7c91282542cf", "date": "2026-08-14T09:43:55+08:00", "subject": "fix(runtime-vapor): preserve child keys in nested fragments", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "cd8b8dafb0edc8979b3cb8fca269aa320b4f6a6f", "date": "2026-08-14T09:54:07+08:00", "subject": "fix(runtime-vapor): preserve cached input effects across nested branch teardown", "body": "Move isolated prop and slot source effects into a cache-owned scope so nested\nbranch teardown does not stop them. Pause the scope while deactivated, resume it\non activation, and stop it when the cached component is unmounted.", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "neovim", "sha": "2b232a9d6e790966dbdb3402eca96f00412b135d", "date": "2026-08-13T23:05:18-04:00", "subject": "build(vim-patch): v8.1.1628 docs are n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vitest", "sha": "fe7906275783ca44b1d02d4e93271cc709bdc564", "date": "2026-08-14T08:08:35+05:00", "subject": "fix(utils): don't crash on a malformed inline source map (fix #10892)", "body": "", "style": "conventional", "type": "fix", "scope": "utils", "breaking": false} {"repo": "neovim", "sha": "692e41bb660957af96bfaf392caff3c07c8907f2", "date": "2026-08-14T01:37:04-04:00", "subject": "build(vim-patch): v8.1.1713 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "postgres", "sha": "0fd73cdffc18d3914311122bf4fbb5acbae626f3", "date": "2026-08-14T14:44:48+09:00", "subject": "injection_points: Add support for string comparison in conditions", "body": "injection_points is extended so as a caller can attach a string value to\na point, checked with a value provided at runtime. There are a couple\nof use cases where this can be useful, to be able to target points for\nspecific object names.\n\nPID and string conditions can be combined together, so as a point can be\nallowed for a specific process and string value at the same time.\n\nThe condition string is capped at 255 characters. As the attach SQL\nfunction now takes an optional condition string, it is no longer STRICT,\nand rejects NULL inputs. Empty condition strings are also rejected, as\nthey don't make sense.\n\nAuthor: Sami Imseih \nReviewed-by: Bertrand Drouvot \nReviewed-by: Michael Paquier \nDiscussion: https://postgr.es/m/CAA5RZ0tsGHu2h6YLnVu4HiK05q+gTE_9WVUAqihW2LSscAYS-g@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9bbd6d585a164d2272bf7b354e80fb03bd5fccd7", "date": "2026-08-14T15:00:59+09:00", "subject": "test: keep QUIC RADIX diagnostics out of TAP pipe", "body": "quic_radix_test emits several MiB of diagnostics on stderr even when\nit succeeds. In #32128 the process was observed blocked in fwrite()\nwhile writing this output to its inherited stderr pipe.\n\nAs a targeted workaround, redirect stderr to a log during non-verbose\nharness runs. If the test fails, replay diagnostics from at most the\nfinal 32 KiB and report the log path. This preserves useful failure\ncontext without feeding the full multi-megabyte log back through the\nTAP pipe. Verbose and direct recipe runs remain unchanged.\n\nThis removes the observed writer-side backpressure point without\nchanging general harness behavior. The reason the harness reader\nstopped making progress remains unknown.\n\nRelated to #32128\n\nAssisted-by: Codex:gpt-5.6-sel\n\nReviewed-by: Neil Horman \nReviewed-by: Saša Nedvědický \nMergeDate: Mon Aug 17 20:19:49 2026\n(Merged from https://github.com/openssl/openssl/pull/32395)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "2b950f6b7fdb766d80349522d2c9b9ce17a40082", "date": "2026-08-14T14:01:13+08:00", "subject": "docs(Form): correct required default value", "body": "* docs(Form): correct required default value\n\n* docs(Form): clarify required behavior\n\n* docs(Form): align required descriptions", "style": "conventional", "type": "docs", "scope": "Form", "breaking": false} {"repo": "postgres", "sha": "eb983d0a94f666d91058552117d029939821d648", "date": "2026-08-14T15:13:30+09:00", "subject": "Add string-based injection point wait in 051_effective_wal_level.pl", "body": "This serves as an in-core example of the condition-string API of\ninjection_points introduced in 0fd73cdffc18, where a wait at the\nbeginning of a replication slot creation is tightened to happen only for\na specific slot name.\n\nAuthor: Sami Imseih \nDiscussion: https://postgr.es/m/CAA5RZ0tsGHu2h6YLnVu4HiK05q+gTE_9WVUAqihW2LSscAYS-g@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "c359db3cc6e2c9cb7b35198d28cf12ea05d1c938", "date": "2026-08-14T14:25:15+08:00", "subject": "site: add fallback for homepage contributor avatars", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "31c624e203b257fe8a934f2f81a53705aa5737d5", "date": "2026-08-14T09:07:04+02:00", "subject": "build: add webdriver build rule", "body": "Sets up a build rule to use Selenium Webdriver for e2e tests.\n\n(cherry picked from commit 4d2dc1fb08a1a587880fdb37ef5b74af8058f98c)", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "4d2dc1fb08a1a587880fdb37ef5b74af8058f98c", "date": "2026-08-14T09:07:04+02:00", "subject": "build: add webdriver build rule", "body": "Sets up a build rule to use Selenium Webdriver for e2e tests.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "48fb5647b767984efac5228614deb426a5dbe4b9", "date": "2026-08-14T09:10:52+02:00", "subject": "build: switch playground away from protractor", "body": "Reworks the `playground` tests not to depend on Protractor.\n\n(cherry picked from commit 3d43067e206eeb0a64772e79e0c9589153450860)", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "3d43067e206eeb0a64772e79e0c9589153450860", "date": "2026-08-14T09:10:52+02:00", "subject": "build: switch playground away from protractor", "body": "Reworks the `playground` tests not to depend on Protractor.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "003e35de814a61408ded69af6303931b5e106199", "date": "2026-08-14T09:13:11+02:00", "subject": "build: switch core tests away from protractor", "body": "Reworks the tests under `core` not to depend on Protractor.\n\n(cherry picked from commit 413c79c10b51f0a90120a1561e022518c6c807a5)", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "413c79c10b51f0a90120a1561e022518c6c807a5", "date": "2026-08-14T09:13:11+02:00", "subject": "build: switch core tests away from protractor", "body": "Reworks the tests under `core` not to depend on Protractor.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "3e02b0a2e5e16e16d9ccb9103b036170f7337640", "date": "2026-08-14T09:15:20+02:00", "subject": "build: switch examples away from protractor", "body": "Reworks the `examples` tests not to depend on Protractor.\n\n(cherry picked from commit 204265f9a3a754c9082733fb76bb90535cd16ab4)", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "204265f9a3a754c9082733fb76bb90535cd16ab4", "date": "2026-08-14T09:15:20+02:00", "subject": "build: switch examples away from protractor", "body": "Reworks the `examples` tests not to depend on Protractor.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "e120d2830b2b9d1b1e09436ed7d6a0939fde5d76", "date": "2026-08-14T09:23:04+02:00", "subject": "build: switch adev tests away from protractor", "body": "Reworks the tests in adev not to depend on Protractor.\n\n(cherry picked from commit 7b74bee5b4485b2a42198db115abcfd5de103c71)", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "7b74bee5b4485b2a42198db115abcfd5de103c71", "date": "2026-08-14T09:23:04+02:00", "subject": "build: switch adev tests away from protractor", "body": "Reworks the tests in adev not to depend on Protractor.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "curl", "sha": "a3aaca10d9acf4a51d2bf1fb9ea0b933e18e3f33", "date": "2026-08-14T09:48:46+02:00", "subject": "multi: cap expire times to INT_MAX internally", "body": "When the time-out value is passed to the outside world it needs to fit\nin a signed 32-bit variable (on Windows and 32-bit architectures)\nanyway. Also, this is 3.5 weeks and we should not knowingly set timeouts\nthat long anyway.\n\nThe previous cap introduced in 3089e7eec8b694ff was only partial.\n\nCloses #22579", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "4359fb0ceea1a6662c0c2f15e807766c4563e23d", "date": "2026-08-14T15:49:54+08:00", "subject": "refactor(runtime-vapor): simplify scope id propagation", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "039b332452c2597052d455b259d1fcf3a87b0b07", "date": "2026-08-14T16:10:49+08:00", "subject": "chore: improve treeshaking", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "angular", "sha": "0cd243d99ab2b73c88d70d595dbf003c8ab8334c", "date": "2026-08-14T10:11:35+02:00", "subject": "build: resolve CI failures", "body": "Resolves some e2e failures that only happened on CI.\n\n(cherry picked from commit 262fba7f7d1ec4ec5aba5295cda8fdb0474d65a9)", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "262fba7f7d1ec4ec5aba5295cda8fdb0474d65a9", "date": "2026-08-14T10:11:35+02:00", "subject": "build: resolve CI failures", "body": "Resolves some e2e failures that only happened on CI.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "fe191ed8af05cbcc46769f435de0f084c5dca172", "date": "2026-08-14T16:13:39+08:00", "subject": "chore: improve", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "postgres", "sha": "48965e14655a007c1250dbd4a1aba65e478ce929", "date": "2026-08-14T17:40:01+09:00", "subject": "postgres_fdw: Rename option \"restore_stats\" to \"import_stats\".", "body": "The option added by commit 28972b6fc was named \"restore_stats\", because\nit was calling pg_restore_relation_stats()/pg_restore_attribute_stats()\nin place of fetching a remote rowsample. However, those code paths have\nbeen replaced with calls to\nimport_relation_statistics()/import_attribute_statistics() (cf. commit\n54cd6fc83), so the option is now slightly misnamed.\n\nSwitch the option to \"import_stats\", which more closely reflects both\nthe functions being called internally, and the user's perception of what\nthe operation is doing.\n\nBack-patch to v19 where commit 28972b6fc went in.\n\nSuggested-by: Etsuro Fujita \nAuthor: Corey Huinker \nDiscussion: https://postgr.es/m/CADkLM%3DezuspTk3VD-2EM6XtFP-Dfz7cJKtnbn0cB6gocsAgu1g%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "d854ab4673c2f9d8048c7f0f6d164b7e4d5e0865", "date": "2026-08-14T11:27:09+02:00", "subject": "GHA/windows: install MSYS2 Perl manually on Windows 2025 runner", "body": "In GitHub runner image windows-2025-vs2026 v20260810.198.2, the default\nPerl binary changed from MSYS2 to Strawberry. The reason is that the MSYS2\ninstallation not longer has Perl preinstalled.\n\nAs seen in CMake configuration:\n```diff\n- -- Found Perl: C:/a/_temp/msys64/usr/bin/perl.exe (found version \"5.42.2\")\n+ -- Found Perl: C:/Strawberry/perl/bin/perl.exe (found version \"5.42.0\")\n```\n\ncurl's Perl scripts require an MSYS2 Perl. Fix by installing it manually\nin this one job.\n\nSymptom was `gencert.pl` no longer finding openssl.exe:\n```\nMissing or broken 'openssl' tool. openssl 1.0.2+ is required.\n```\nThen the script hanging while trying to trace it.\n\nAlso:\n- add distinct error message and list PATH elements if openssl is not\n found via `gencert.pl`.\n- tried falling back to Git for Windows Perl, which caused the test step\n to either hang or run too slowly to fit the time slot.\n\nRefs:\nhttps://github.com/actions/runner-images/pull/14541\nhttps://github.com/actions/runner-images/releases/tag/win25-vs2026%2F20260810.198\n\nBug: https://github.com/curl/curl/pull/22577#issuecomment-5286533641\nBug: https://github.com/curl/curl/pull/22577#issuecomment-5291468024\n\nCloses #22580", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "ec7606d059cd96766faf1f05660dd410b39c7a8d", "date": "2026-08-14T17:42:30+08:00", "subject": "fix(FloatButton): keep BackTop progress ring visible", "body": "* fix(FloatButton): keep BackTop progress ring visible\n\n* chore: retry CI", "style": "conventional", "type": "fix", "scope": "FloatButton", "breaking": false} {"repo": "postgres", "sha": "9924e2d900bf833c1d732d4bf877f51269b03157", "date": "2026-08-14T11:43:19+02:00", "subject": "Enforce WITH CHECK OPTION on DELETE FOR PORTION OF leftovers", "body": "DELETE FOR PORTION OF inserts temporal leftovers through ExecInsert(),\nbut the rewriter only attached view WCOs for INSERT/UPDATE. Leftover\nrows could therefore silently escape a WITH CHECK OPTION view, while\nthe equivalent UPDATE correctly raised an error.\n\nThis commit fixes it by attaching the WCOs for FOR PORTION OF deletes\ntoo.\n\nAuthor: Zsolt Parragi \nCo-authored-by: Paul A Jungwirth \nReviewed-by: solai v \nReviewed-by: Dean Rasheed \nDiscussion: https://www.postgresql.org/message-id/flat/CAN4CZFOuTyhGspG0Nyits8PiK2keoNXkLj-u3APzc66aRcWY9A%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "adcb2709e7318244e3505316b5d958d75e2ae5a8", "date": "2026-08-14T18:50:25+09:00", "subject": "refactor(ui): consolidate project config metadata derivation", "body": "", "style": "conventional", "type": "refactor", "scope": "ui", "breaking": false} {"repo": "vitest", "sha": "d8454652370ec130fa2fcb40fa372a820f6c2398", "date": "2026-08-14T18:50:46+09:00", "subject": "refactor(ui): consolidate ws handlers definition and remove unused code", "body": "", "style": "conventional", "type": "refactor", "scope": "ui", "breaking": false} {"repo": "nestjs", "sha": "4fbd6e620627824327ef490f9084ae5f8fc1ee67", "date": "2026-08-14T12:09:39+02:00", "subject": "Potential fix for pull request finding", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "554efab7f69dad4391372e9edc2214eff9ab0fe1", "date": "2026-08-14T12:49:15+02:00", "subject": "multi: drop the assert from large expires", "body": "Triggers the fuzzers\n\nCloses #22582", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "7627546075243af4d6d6668102e47be380a4ba44", "date": "2026-08-14T12:51:35+02:00", "subject": "feat(core): sse signal", "body": "", "style": "conventional", "type": "feat", "scope": "core", "breaking": false} {"repo": "nestjs", "sha": "6498f11a985bef0134102c132bb8597dc4a935fe", "date": "2026-08-14T13:55:33+02:00", "subject": "feat: signal should finalize in every case", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vitest", "sha": "bac009d7eeea32b1ee2ce5cc21c33daeb8179578", "date": "2026-08-14T14:09:30+02:00", "subject": "refactor(projects): cleanup comments, add more `debug` statements", "body": "", "style": "conventional", "type": "refactor", "scope": "projects", "breaking": false} {"repo": "vitest", "sha": "9b4df4b057b645eb65c89841586ef09dd0d341b1", "date": "2026-08-14T05:11:36-07:00", "subject": "fix(browser): disallow filter-only options in getBy* locators (fix #10295)", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "vitest", "sha": "cf9176bfc285093c1ef3b3f63db0dfad4bc60436", "date": "2026-08-14T21:12:32+09:00", "subject": "fix: revive global concurrency limit for test lifecycle", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "fe703df949864cbae594d102d43c4b9ad3e30ef0", "date": "2026-08-14T14:35:04+02:00", "subject": "lib: optimize struct layouts for reduced memory usage", "body": "struct SingleRequest tweaks\n\n- 'upgr101' from enum upgrade101 => 'unsigned char', saves three bytes\n- made some 'unsigned char' => uint8_t\n- moved 'io_flags' to fill a hole on 64-bit arch\n\nstruct UserDefined tweaks\n\n- Sort the fields by size. Larger to smaller. Helps avoding holes.\n- httpsig_algorithm moved\n- FTP uint8_t fields moved\n- new_file_perms moved\n- rtspreq moved\n- sort fields on size\n\n- urldata: drop 'struct Curl_data_priority'\n\n It only had a single struct member 'weight'. Use that directly instead\n to save indirections and struct alignments. Move field for size order.\n\n- urldata: move the RTSP fields in the UrlState struct to be ordered by\n size\n\nstruct PureInfo tweaks\n\n- make 'pxcode' a uint8_t\n- move 'conn_protocol' to fill a hole\n- sort 'struct PureInfo' fields by size\n\nCloses #22585", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a82e33a83d7cd39f3abd18c7b941625ca78d4c35", "date": "2026-08-14T14:53:50+02:00", "subject": "Make generate_queries_for_path_pattern_recurse() interruptible", "body": "In case of a very long path pattern with each element pattern being\nresolved to many graph elements\ngenerate_queries_for_path_pattern_recurse() may take very long time to\ngenerate all the possible graph paths. Make\ngenerate_queries_for_path_pattern_recurse() interruptible so that a\nuser may be able to cancel such a query if required and the interrupts\nare processed in timely manner.\n\nAuthor: Satyanarayana Narlapuram \nReviewed-by: Ashutosh Bapat \nDiscussion: https://www.postgresql.org/message-id/flat/CAHg%2BQDfDwcM4%3DDSiAV6Ly89YQ5EcMhzO1-9x%3DmGG1WJzODcAig%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "3d6d93a6beb9eeaca98ce20eeed7f64ebb0c9297", "date": "2026-08-14T14:55:49+02:00", "subject": "multi: timeout improvements", "body": "- Move expire timeout code from multi into splay.c\n- keep a \"time_base\" timestamp to calculate timediff_t for\n actual timeout values. Unfortunately this means our\n timeouts will go wrong after ~500,000 years of continuous\n operations...\n- use timediff_t as key in splay instead of curltime\n- use timediff_t in transfers expire times instead of curltime\n- re-comment splay.c for better understanding how it works\n- replace splay nodes double-linked \"same\" list with a single link,\n we almost never have duplicate keys\n- keep transfer `mid` in splay nodes instead of the transfer pointer\n- keep registered bit in splay node for tracking instead of separate\n bit in transfer\n- adapt unit1309.c to changes in timediff_t and mid\n\nCloses #22584", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "3a82a16314bd7ec543e11d7a007a3cedb33c6b4a", "date": "2026-08-14T16:09:49+03:00", "subject": "docs: use `transformedValue` in `Forms | Custom Controls | Value transformation` section", "body": "The \"Value transformation\" section taught readers to hand-roll\ntransformation with `linkedSignal()` and a manual parse method, even\nthough `@angular/forms/signals` ships `transformedValue()` for exactly\nthis case. Readers ended up with a weaker version of a feature the\nframework already provides — notably, no parse error reporting.\n\nRewrite the section around `transformedValue()` and document the parts\nthe manual pattern could not cover: returning `{error}` from `parse` to\nsurface parse errors on the field's `errors()`, and `reset()` clearing\nthem. This also makes good on the cross-reference from the validation\nguide, which pointed here for parse error details the section never\ncovered.\n\nFixes #70206\n\nCo-authored-by: Matthieu Riegler \n(cherry picked from commit 182c371d829375ac51d0b3f390c7da9c3c17e580)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "182c371d829375ac51d0b3f390c7da9c3c17e580", "date": "2026-08-14T16:09:49+03:00", "subject": "docs: use `transformedValue` in `Forms | Custom Controls | Value transformation` section", "body": "The \"Value transformation\" section taught readers to hand-roll\ntransformation with `linkedSignal()` and a manual parse method, even\nthough `@angular/forms/signals` ships `transformedValue()` for exactly\nthis case. Readers ended up with a weaker version of a feature the\nframework already provides — notably, no parse error reporting.\n\nRewrite the section around `transformedValue()` and document the parts\nthe manual pattern could not cover: returning `{error}` from `parse` to\nsurface parse errors on the field's `errors()`, and `reset()` clearing\nthem. This also makes good on the cross-reference from the validation\nguide, which pointed here for parse error details the section never\ncovered.\n\nFixes #70206", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6ac66ab20b8e178fb3e80ad07d80ce6cd5efefb5", "date": "2026-08-14T15:23:30+02:00", "subject": "Update corpora submodule", "body": "Reviewed-by: Jakub Zelenka \nReviewed-by: Milan Broz \nMerge-date: Tue Sep 15 15:00:02 2026\nMerged-from: https://github.com/openssl/openssl/pull/32342", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9eeafd3ad83533bdd8160dd35cf7230d8118e7ab", "date": "2026-08-14T15:24:08+02:00", "subject": "Bugfix in the file BIO - avoiding false negatives based on previous errno.", "body": "Base on the idea of @Rydier.\n\nFixes #32179\n\nReviewed-by: Richard Levitte \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 19 11:03:50 2026\nMerged-from: https://github.com/openssl/openssl/pull/32382", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "42517c886a25f9f2a9cd659f5987d78fde77be20", "date": "2026-08-14T15:24:08+02:00", "subject": "Bugfix in the file BIO - avoiding false negatives based on previous errno.", "body": "Base on the idea of @Rydier.\n\nFixes #32179\n\nReviewed-by: Richard Levitte \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 19 11:03:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/32382", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "64a301184ef39a0653c99c8edc1b63cd3fd16236", "date": "2026-08-14T09:30:31-04:00", "subject": "feat(input)!: CmdAtom event #41297", "body": "Problem:\nThere is no unified notion of a \"user action\".\n\nVim processes input by one-char-at-a-time, and mostly throws away any\nhints it might gather about the user's action, with one exception: it\nstores the last _edit_ action (the \"redo buffer\", encoded as\nunstructured `[\"x][v][count]body` bytes).\n\nPlugins can only observe individual keys (vim.on_key) and high-level\neffects (TextChanged, CursorMoved).\n\nSolution:\n- Users can subscribe to `CmdAtom` events to handle any user action.\n - Event is deferred; handlers cannot cancel or interfere with user\n actions.\n- Capture `CmdSpec` from the normal/insert/visual subsystems.\n - typeahead/readahead stay unstructured (`buffheader_T`): they are key\n streams, not commands.\n - the redo/record buffers become `StringBuilder`: fewer\n allocations/copies.\n- Repurpose the input/redo engine to accept `CmdSpec` objects.\n\n\"atom\": one repeatable unit of user input, as a resolved (post-mapping)\nkeysequence plus structured fields. Only user actions, not `:normal`,\nAPI calls, or non-\"t\" `feedkeys`.\n\nBREAKING: dot-repeat of an Insert session, replays the entire session\nincluding cursor-moves (:help ins-repeat).\n\nBREAKING: dot-repeat of a Visual operation, replays the selection\ninstead of operating on a fixed-size region.", "style": "conventional", "type": "feat", "scope": "input", "breaking": true} {"repo": "openssl", "sha": "2d9d7ac439d56097137036da321ba9490262d0b6", "date": "2026-08-14T17:01:14+02:00", "subject": "ci: Reduce mfail fuzz test to only few corpus samples", "body": "The fuzz_tests_mfail job ran malloc-failure injection over the full corpora.\nThe full run takes minutes, and with a bigger corpus, tens of minutes.\n\nCI should only confirm that the fuzzers still build and process inputs with mfail.\nReduce the run to just a few samples, which is enough to let the test run.\n\nAssisted-by: Claude:claude-opus-4-8\nReviewed-by: Jakub Zelenka \nReviewed-by: Richard Levitte \nReviewed-by: Neil Horman \nMerge-date: Tue Aug 18 12:17:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32384", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vitest", "sha": "10b2cd2017d20dfa62436aed5e43bb1ef078a019", "date": "2026-08-14T17:28:06+02:00", "subject": "fix(browser): encode iframeId in tester iframe URL [backport to v4]", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "angular", "sha": "ec678c0f8d449b5be81438738d09f9f503f3e1a0", "date": "2026-08-14T16:12:19Z", "subject": "build: update dependency ng-packagr to v22.2.0-next.3", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "3828ab7047ba8204c9d71cf606762e9b2c46c024", "date": "2026-08-14T12:14:24-04:00", "subject": "psql: count every COPY FROM STDIN when scanning a query string.", "body": "When SendQuery() is not told how many COPY FROM STDIN commands the\nquery string contains (as for -c, \\gexec, and \\watch), it scans the\nstring to count them itself. But it called psql_scan() only once,\nwhich stops at the first semicolon, so any COPY FROM STDIN past the\nfirst sub-command was not counted, causing failure of cases that used\nto work. Oversight in commit 3045a25ba.\n\nAuthor: Zsolt Parragi \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/CAN4CZFPqa6c+u4uX5jJ8LANHTQ4dxM3m4_8G9WmX_A4-2wuv2A@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "0af3b9827bc3c45b315e5ede3f358b4ced678ea3", "date": "2026-08-14T11:25:31-05:00", "subject": "feat(dir): user can sort/filter listings, DirReadPost event #41138", "body": "Problem:\nDirectory listing entries cannot be customized (filtered, reordered).\nListings are read by a BufReadCmd, which suppresses BufReadPost, so they\nare the only buffers with no post-read event to hook.\n\nSolution:\nIntroduce a post-render User autocmd `DirReadPost`, marking the dir\nbuffer writable for the duration and before the cursor is placed, so\nhandlers can sort or filter it with ordinary commands. Document common\nrecipes", "style": "conventional", "type": "feat", "scope": "dir", "breaking": false} {"repo": "postgres", "sha": "40af05bb6157670482ae31fd5b0a12ff2992c966", "date": "2026-08-14T11:29:21-05:00", "subject": "Retire PQfn().", "body": "PQfn() has been documented as somewhat obsolete since commit\nefc3a25bb0 (2003), and commit bd48114937 recently marked it unsafe\nbecause it may write past the end of result_buf when result_is_int\nis 0. Furthermore, searches of publicly available code turned up\nno callers, only language bindings that expose the function\nwithout using it, so removing it seems unlikely to cause too much\ntrouble.\n\nSince the symbol is exported, deleting it outright isn't an option,\nso instead this commit teaches PQfn() to fail unconditionally with\nan error that suggests alternatives. The documentation for the\nfast-path interface is replaced with a stub in the \"Obsolete or\nRenamed Features\" appendix. The server's fast-path support is\nretained for the benefit of older clients and the frontend large\nobject interface.\n\nThe frontend large object interface, the only in-tree caller, now\nuses PQnfn(), the private version of PQfn() added by commit\nbd48114937. I considered converting this code away from the\nfast-path protocol entirely, but prepared statements can be\ndeallocated out from under libpq (e.g., by DISCARD ALL), and\nPQexecParams() proved noticeably slower, so that is left as a\nfuture exercise.\n\nReviewed-by: Christoph Berg \nReviewed-by: Jacob Champion \nReviewed-by: Dagfinn Ilmari Mannsåker \nDiscussion: https://postgr.es/m/ahXE28klgxIJXBLq%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6f09efc78145d9bea126a2b2345e7fb0b280eb69", "date": "2026-08-14T18:52:58+02:00", "subject": "Consistently use ASN1_ITEM_rptr() for ASN1_ITEMs", "body": "Since #9285 this indirection is no longer really needed and perhaps\nit is time to mop up the internal uses of ASN1_ITEM_rptr(). I just\nwanted to point out that these are inconsistent with all the other\nuses of ASN.1 items, of which there are tons.\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Nikola Pajkovsky \nMerge-date: Tue Aug 18 07:16:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/32388", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "0cd635e9e2c409a045eaeaf920a226ef54aecc48", "date": "2026-08-14T12:08:18-05:00", "subject": "fix(http): cancel oversized fetch response bodies", "body": "Cancel the unread response body before reporting NG02825 when its declared Content-Length exceeds the configured buffer limit. Without cancellation, SSR can finish while the underlying connection remains open.\n\nAdd regression coverage for the declared-length rejection path.\n\n(cherry picked from commit 1a006a8f9794240a950c388387bd8b9a671801da)", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "angular", "sha": "1a006a8f9794240a950c388387bd8b9a671801da", "date": "2026-08-14T12:08:18-05:00", "subject": "fix(http): cancel oversized fetch response bodies", "body": "Cancel the unread response body before reporting NG02825 when its declared Content-Length exceeds the configured buffer limit. Without cancellation, SSR can finish while the underlying connection remains open.\n\nAdd regression coverage for the declared-length rejection path.", "style": "conventional", "type": "fix", "scope": "http", "breaking": false} {"repo": "angular", "sha": "84a210e8f326ec8fa1caf936ebab35c8ebe08327", "date": "2026-08-14T10:08:44-07:00", "subject": "refactor(router): add support for non-blocking router resources", "body": "Introduces support for route-level resources via the `resources` property on route definitions, enabled with `withRouterResources()`.\n\nRouter resources provide a reactive, signal-based alternative to resolvers, allowing routes to declare resources tied to route parameters and route lifecycle:\n- The `resources` function executes during navigation transitions within an injection context scoped to the route (`_localInjector`).\n- It receives a `ResourceContext` containing signals for `params`, `queryParams`, `fragment`, and `data`, alongside the static `snapshot`.\n- For newly created routes, `resources` runs once and attaches to `ActivatedRoute.resources`. For reused routes, parameter signals update reactively to trigger new data fetches while keeping resource references stable.\n- Wrapped resources (`routerResource`) provide transactional stability: snapshots are frozen during active navigations to prevent UI jitter, unfreezing on `NavigationEnd`.\n- On cancelled navigations or errors, rollback recovery retains the frozen snapshot until reverted signals settle, avoiding flashes of loading state. Manual reloads are rejected while frozen.\n- Local injectors are automatically cleaned up if navigations are cancelled or rolled back.\n\nNote that this commit only implements non-blocking resources (marked with `nonBlocking()`), deferring blocking resource resolution to future work to keep the initial changeset smaller and less complex.\n\nPR Close #70211", "style": "conventional", "type": "refactor", "scope": "router", "breaking": false} {"repo": "nestjs", "sha": "e2557553672a96df5b9ee9e0b2435eb07c07c337", "date": "2026-08-14T19:44:42+02:00", "subject": "chore: resolve conflicts, minor fixes", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "a326599e8c7451111747d328bdf98d452c0de677", "date": "2026-08-14T19:50:03+02:00", "subject": "fix(core): fix sse early-return", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "b06ff7537039ae16a87607014605f7bd0e3e31ce", "date": "2026-08-14T11:13:11-07:00", "subject": "refactor(compiler): support block-specific deferredImports mapping", "body": "Allow `@Component.deferredImports` to be an object mapping block names to arrays of dependencies, and support `@defer (name blockName)` syntax to configure block-specific dependency lists in both standard and local compilation modes.\n\nThis enables more targeted dependency chunk generation during local compilation instead of over-eagerly loading all deferred imports together.", "style": "conventional", "type": "refactor", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "7d9f55da11319da8f273d9edcd38ff2983bdbb0c", "date": "2026-08-14T11:13:21-07:00", "subject": "feat(compiler-cli): scope type-checking of keyed defer blocks", "body": "Enforces that components, directives, and pipes in @Component.deferredImports mapped to specific block keys are only used within their designated @defer blocks during template type checking, reporting out-of-band diagnostics when used eagerly or across mismatched blocks.", "style": "conventional", "type": "feat", "scope": "compiler-cli", "breaking": false} {"repo": "postgres", "sha": "45df9b21ae617ff124da4e8364ae76199ae5acf6", "date": "2026-08-14T13:41:32-05:00", "subject": "Add missing PGDLLIMPORT marker.", "body": "Oversight in commit ffca23839c.\n\nReported-by: Anton Voloshin \nAuthor: Anton Voloshin \nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5578e426ccd22135301379557683130d8bc30afe", "date": "2026-08-14T11:58:52-07:00", "subject": "Remove the F5 ClientHello-padding workaround", "body": "And remove the support for the padding extension (RFC7685).\n\nWith this change our client will cease sending the padding extension.\nIt was only ever sent, via SSL_OP_TLSEXT_PADDING, to work around a\nvery old bug in F5 middleboxes that choked on ClientHellos between 256\nand 511 bytes.\n\nThe fix shipped long ago from F5, and the affected hardware is long out of\nsupport, so nothing should still be running the problematic version.\n\nSSL_OP_TLSEXT_PADDING is now a no-op retained for compatibility and\nis no longer part of SSL_OP_ALL.\n\nIn case anyone wants to actually still use the padding extension\nfor some reason, as we no longer have a handler for this extension,\nthey can register a custom extension handler to do whatever they like\nwith it.\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Nikola Pajkovsky \nMerge-date: Tue Aug 18 08:39:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32389", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "b169d9376c9bd7edc7277d5e089894b3f6a43f34", "date": "2026-08-14T14:24:16-05:00", "subject": "test(dir_spec): fix failures with PUC Lua", "body": "", "style": "conventional", "type": "test", "scope": "dir_spec", "breaking": false} {"repo": "postgres", "sha": "de73e691e7f53a4fe08f09928a9f72331476a439", "date": "2026-08-14T21:52:06+02:00", "subject": "Use int64 for number of entries in pg_stat_statements", "body": "Commit 13b935cd changed hash_get_num_entries to return int64 instead\nof long. This fixes a few more users of hash_get_num_entries which\nwere missed in the original commit.\n\nPGSS_FILE_HEADER is left unchanged since this only affects as of yet\nunreleased versions of PostgreSQL. Backpatch to v19 where the int64\nchange was performed.\n\nAuthor: Karina Litskevich \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CACiT8iYTkc33YWaA2D3t51Y5s=GqBO7T1zX7bkpSmet2njcLLw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1b87dd4303a9a90f2f4004fb08a00af7d3d25207", "date": "2026-08-14T21:52:21+02:00", "subject": "Use correct data type for version identifier", "body": "PGSS_PG_MAJOR_VERSION is written as an uint32 field, but was read back\ninto an int32. Fix to make sure reading and writing use the same data\ntype.\n\nAuthor: Karina Litskevich \nDiscussion: https://postgr.es/m/CACiT8iYTkc33YWaA2D3t51Y5s=GqBO7T1zX7bkpSmet2njcLLw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "786db8dcf168bd9df8f55047337525ac19118b1c", "date": "2026-08-14T15:27:16-05:00", "subject": "doc: Clarify the logging collector's guarantees.", "body": "Presently, the documentation for logging_collector says that the\ncollector \"is designed to never lose messages,\" which reads as a\nstronger promise than we actually make. The collector does not\nfsync the log file or retry failed writes, so log messages can go\nmissing after an operating system crash, power loss, or a write\nerror. Reword that sentence and add a note about what is not\nguaranteed.\n\nAuthor: Daniel Bauman \nReviewed-by: Fujii Masao \nReviewed-by: Zhenwei Shang \nReviewed-by: Robert Treat \nDiscussion: https://postgr.es/m/CAMtj0_a86DdDKkW-ReVpQpqjndVS6GMrwXVpQY4G3-SGY7saMQ%40mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d1370730247b71428c7486a931095946103f49ed", "date": "2026-08-14T17:12:17-04:00", "subject": "build(vim-patch): v8.1.2195 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "angular", "sha": "aea121e53217a7f74d2a8e4a7a6bf0f3a96b7c49", "date": "2026-08-14T21:25:16Z", "subject": "fix(compiler-cli): do not flag callable objects with zero parameters in uninvoked track function check", "body": "In `@for` blocks, tracking callable objects by reference (e.g. signal forms `FieldTree`, signals, or custom callable objects) is a valid pattern when tracking by object identity. Previously, `UninvokedTrackFunctionCheck` (NG8115) flagged any property read whose type has call signatures, regardless of whether the target was an actual track function expecting arguments or a method reference.\n\nThis commit updates `UninvokedTrackFunctionCheck` to only emit a diagnostic when the target expression has call signatures that declare parameters (functions/methods expecting arguments like `(item)` or `(index, item)`) or is a method declaration. Callable objects without parameters accessed as properties are now recognized as tracked values and not flagged as uninvoked track functions.\n\nFixes #70207\n\n(cherry picked from commit f0a271c7bd4be000c52592034573cda9866f26bb)", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "f0a271c7bd4be000c52592034573cda9866f26bb", "date": "2026-08-14T21:25:16Z", "subject": "fix(compiler-cli): do not flag callable objects with zero parameters in uninvoked track function check", "body": "In `@for` blocks, tracking callable objects by reference (e.g. signal forms `FieldTree`, signals, or custom callable objects) is a valid pattern when tracking by object identity. Previously, `UninvokedTrackFunctionCheck` (NG8115) flagged any property read whose type has call signatures, regardless of whether the target was an actual track function expecting arguments or a method reference.\n\nThis commit updates `UninvokedTrackFunctionCheck` to only emit a diagnostic when the target expression has call signatures that declare parameters (functions/methods expecting arguments like `(item)` or `(index, item)`) or is a method declaration. Callable objects without parameters accessed as properties are now recognized as tracked values and not flagged as uninvoked track functions.\n\nFixes #70207", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "nestjs", "sha": "7f4f0b8008034d8d07a483da3d6e341a9e529135", "date": "2026-08-15T10:33:25+08:00", "subject": "docs: remove dead DEVELOPER.md link definition from CONTRIBUTING", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "c7ee6af777a9a0b1cd54817c646fe32fa7267c47", "date": "2026-08-15T00:14:06-04:00", "subject": "build(vim-patch): v8.1.2219 is n/a", "body": "Vim's TerminalWinOpen seems to be required because of Vim\nbuffer-job-popupwin implementation.\nBased on the patch, I'm puzzled why fzf needs this on Vim.\n\nNvim's TermOpen, TermEnter, and detection mechanisms to know\nif buffer is on a (active,visible) window should suffice to not port it.\n\nIf there was a feature request or issue without a merged fix,\nthen I can't find it.\n\nhttps://github.com/junegunn/fzf/pull/2000", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vue-core", "sha": "1a2de08c0b3ebfebe2fa07ed2b545df17073fea4", "date": "2026-08-15T13:48:53+08:00", "subject": "fix(runtime-vapor): preserve nested vdom slot content", "body": "Close #15303", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "be31004816cb8806f433909e79b9d6bf4b89b1af", "date": "2026-08-15T14:21:58+08:00", "subject": "fix(ColorPicker): keep color selection draggable after clearing", "body": "* fix(ColorPicker): keep color selection draggable after clearing\n\n* Remove test for clear button text selection\n\nRemoved test for preventing text selection from the clear button.\n\nSigned-off-by: 多格 <40968928+dogledogle@users.noreply.github.com>\n\n---------", "style": "conventional", "type": "fix", "scope": "ColorPicker", "breaking": false} {"repo": "ant-design", "sha": "e57c50affe02cd1e84e4cc029a64e96dbf62019f", "date": "2026-08-15T14:35:15+08:00", "subject": "docs(float-button): unify FloatButton translation", "body": "", "style": "conventional", "type": "docs", "scope": "float-button", "breaking": false} {"repo": "ant-design", "sha": "8b4c97310fa2d2bf3a7f2eb9f5ce5146706bbb3b", "date": "2026-08-15T14:44:14+08:00", "subject": "fix(locale): correct Form min placeholders", "body": "* fix(locale): validate Form message placeholders\n\nCorrect invalid min tokens in the Icelandic and Mongolian validation messages and add a cross-locale guard for missing or renamed Form placeholders.\n\nAI-assisted: Codex\nSigned-off-by: 陳廷安 <73953029+nrps9909@users.noreply.github.com>\n\n* test(locale): keep placeholder fix focused\n\nRemove the generalized cross-locale placeholder guard in response to maintainer feedback, leaving the PR scoped to the two invalid Form placeholder corrections.\n\nAI-assisted: Codex\nSigned-off-by: 陳廷安 <73953029+nrps9909@users.noreply.github.com>\n\n---------", "style": "conventional", "type": "fix", "scope": "locale", "breaking": false} {"repo": "curl", "sha": "e090cf59ad4c88be9da873fe37dae3487b7ad77e", "date": "2026-08-15T09:06:05+02:00", "subject": "build: drop `dirent.h` and `opendir()` detections on Windows", "body": "`opendir()` was used previously with mingw-w64 for `file://` directory\nlisting support. This was replaced by a native Win32 solution on all\nWindows platforms, leaving `opendir()` no longer used. Drop related\nfeature detections.\n\nFollow-up to c7c42afab91c46e6baaa0e2801626e13acd6c30a #22577\n\nCloses #22588", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "3f25c6ab36558867c9add1f79a6cdd0dcf860d34", "date": "2026-08-15T15:12:44+08:00", "subject": "fix: option position announcement for virtual mode", "body": "* chore: bump select/tree-select/cascader\n\n* fix: update snapshot\n\n* size-limit", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "26dadd9ea3be8c21909bd10e4677890af59011d9", "date": "2026-08-15T09:37:03+02:00", "subject": "GHA/windows: log preinstalled MSYS2 packages", "body": "CI jobs rely on defaults, sometimes to reduce log noise (possibly\nimprove performance a bit) or simplify, sometimes by accident. List them\nto help debug or optimize on changes.\n\nRefs:\nhttps://github.com/actions/runner-images/issues/14562#issuecomment-5301777198\nhttps://github.com/msys2/MSYS2-packages/commit/636489312cbc55011cb7ec493550598125e7716f\nhttps://github.com/msys2/MSYS2-packages/pull/6491\nhttps://github.com/msys2/MSYS2-packages/pull/6479\n\nFollow-up to d854ab4673c2f9d8048c7f0f6d164b7e4d5e0865 #22580\nFollow-up to c349bd668c91f2484ae21c0f361ddf497143093c #14097\n\nCloses #22589", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "91323de7f3ddfad2fb6860a695ffa805e9389afe", "date": "2026-08-15T10:32:30+02:00", "subject": "GHA/non-native: bump cross-build-actions, NetBSD 11.0, FreeBSD riscv64", "body": "New riscv64 job takes 4.5 minutes.\n\nAlso:\n- use prefill on riscv64 to save 2 minutes configure time.\n- disable manuals/perl on riscv64 to save ~1 minute (install, build).\n- cmake: fix `HAVE_GETHOSTBYNAME_R_6` prefill for FreeBSD.\n\nRef: https://github.com/cross-platform-actions/action/releases/tag/v1.4.0\n\nCloses #22590", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "82882fae51d13705ea05320379800afae4264144", "date": "2026-08-15T11:55:57+03:00", "subject": "refactor(devtools): add TTL and outline style to the highlighting mechanism", "body": "Add the ability to automatically destroy created highlights by a provided TTL;\nAdd the option for outline style of the highlight overlays;\nAdd `prefer-inset` label position;\n\nPR Close #70254", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "curl", "sha": "787cbd20330a6a3d9b024d47c2c35691c7b35cab", "date": "2026-08-15T11:07:15+02:00", "subject": "multi: shrink expire timer indices", "body": "The enum `exire_id` fits easily into a uint8_t. Use that type for\nstoring expire timeout list indices.\n\nCloses #22591", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "3ab3f33281f62bdcef1dd987a89fcfe7f082b6d8", "date": "2026-08-15T18:12:14+09:00", "subject": "Rename EXISTS-to-ANY converted subplan to exists_to_any", "body": "Simple EXISTS subplans can be converted to hashed ANY subplans as\nan alternative implementation. Previously, both the original EXISTS\nsubplan and the converted ANY subplan used names with the\nexists_ prefix, making EXPLAIN output harder to read and\npg_plan_advice targets less clear.\n\nUse the exists_to_any_ prefix for converted ANY subplans so that\ntheir names distinguish them from the original EXISTS alternative.\nThis only changes the names shown by EXPLAIN and used by plan advice;\nit does not affect planner behavior.\n\nAuthor: Yugo Nagata \nReviewed-by: Tom Lane \nReviewed-by: solai v \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/20260605165641.3950f99ace0aad8f807abe96@sraoss.co.jp\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "ebc0fbc0e86d8ef635b9578348bf21bb8a4f480a", "date": "2026-08-15T11:20:23+02:00", "subject": "connect: only set connect timer on first socket", "body": "Only collect connect/appconnect timer stats on the first socket.\n\nFixes #22587\nCloses #22592", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "d613411e37947047ef4ce0cca262f57a1fad80b9", "date": "2026-08-15T17:58:55+08:00", "subject": "fix(runtime-vapor): align scope id propagation with vdom", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "5b9946a106df6d923b3dbe37ebf632f756afa0d1", "date": "2026-08-15T12:47:34+02:00", "subject": "curl_threads: always use native threads/mutex on Windows", "body": "Syncing with build systems, which already enforce this.\n\nCloses #22593", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "4b42956cd1b4ee5c739d9bd73662d21c6b13cae8", "date": "2026-08-15T12:56:23+02:00", "subject": "build: drop detecting `gettimeofday()` on Windows", "body": "To remove a build difference between mingw-w64 and MSVC builds (where\nmingw-w64 has it, while MSVC does not). Also to simplify and save the\ncost of detection/verification in CI.\n\n`gettimeofday()` was used in `curl_threads.c` in codepath practically\nunreachable in Windows builds, and in `tool_writeout.c` that falls back\nto `time()` for targets without it.\n\nRef: #22593\n\nCloses #22594", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "neovim", "sha": "aef9356e341ed0df10dcbf48cff2c9efb9f0dcd2", "date": "2026-08-15T08:26:57-04:00", "subject": "build(vim-patch): fix compiled,popup-window regexp for '|' #41326", "body": "git \"-I\" regex seem to be \"extended\".\ngrep's default regex is \"basic\".\nFix regexps for switch/case \"runtime/doc/*.txt\"\nbased on the \"*.h\" switch/case.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "c059f726d706e311a3c76b544e6e4fc063646566", "date": "2026-08-15T22:17:44+09:00", "subject": "ocsp: Reject update times X509_cmp_time() cannot compare", "body": "OCSP_check_validity() accepts GeneralizedTime forms that\nX509_cmp_time() cannot compare. The latter requires the canonical RFC\n5280 form and returns 0 for numeric UTC offsets and fractional seconds.\nThat error was treated as an in-range result, allowing OCSP freshness\nchecks to be bypassed.\n\nTreat errors from the initial thisUpdate and nextUpdate comparisons as\nfield errors. Run the optional maximum-age check only after the initial\nthisUpdate comparison succeeds, avoiding a duplicate field error.\n\nAdd regression coverage for offset and fractional forms, including the\nmaximum-age path.\n\nFixes #32377\n\nAssisted-by: Codex:gpt-5.6-sol\nSigned-off-by: Mounir IDRASSI \nReviewed-by: Bob Beck \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 15:15:28 2026\nMerged-from: https://github.com/openssl/openssl/pull/32396\n(cherry picked from commit 6ceaf1dfe6729853961c6f53e97e635e82da9168)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6ceaf1dfe6729853961c6f53e97e635e82da9168", "date": "2026-08-15T22:17:44+09:00", "subject": "ocsp: Reject update times X509_cmp_time() cannot compare", "body": "OCSP_check_validity() accepts GeneralizedTime forms that\nX509_cmp_time() cannot compare. The latter requires the canonical RFC\n5280 form and returns 0 for numeric UTC offsets and fractional seconds.\nThat error was treated as an in-range result, allowing OCSP freshness\nchecks to be bypassed.\n\nTreat errors from the initial thisUpdate and nextUpdate comparisons as\nfield errors. Run the optional maximum-age check only after the initial\nthisUpdate comparison succeeds, avoiding a duplicate field error.\n\nAdd regression coverage for offset and fractional forms, including the\nmaximum-age path.\n\nFixes #32377\n\nAssisted-by: Codex:gpt-5.6-sol\nSigned-off-by: Mounir IDRASSI \nReviewed-by: Bob Beck \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 15:15:28 2026\nMerged-from: https://github.com/openssl/openssl/pull/32396", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9cca923ab4f9ddca4b817e989eba546ddba675e1", "date": "2026-08-15T10:13:32-04:00", "subject": "fix(extmarks): redo of a mark created during an edit #41324", "body": "Problem:\nA mark created by `nvim_buf_set_extmark()` while an undo block is open\nnever comes back on redo.\n\nAnalysis:\nUndo deletes the text it covers and collapses the range; redo replays\nthe splices, which re-insert the text but cannot re-expand the mark.\n`extmark_set()` records a position only for a mark it moves, not for one\nit creates.\n\nSolution:\nRecord the created position for redo; undo leaves the mark to the splice\nreplay, since it did not exist before the edit. Each side of a paired\nmark gets its own entry. Redo also revives a mark that undo invalidated,\nelse its position returns but its highlight does not.", "style": "conventional", "type": "fix", "scope": "extmarks", "breaking": false} {"repo": "postgres", "sha": "3d2f2eb1664e5c823b66a063d2de9a9787970d42", "date": "2026-08-15T23:15:31+09:00", "subject": "pgindent fix for commit 7b7c4a8dcc9", "body": "Per buildfarm member koel.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "1db93bdb7c1943132b13543539aaaf4866f31419", "date": "2026-08-15T16:18:48+02:00", "subject": "config-win32.h: limit use to MSVC IDE Project builds", "body": "This is the only purpose it's maintained for.\n\nDrop mingw-w64 and other logic falling outside of VS2010-2013 needs.\n\nmingw-w64/MSVC platform differences keep being maintained/documented as\npart of the CMake pre-fill logic in `CMake/win32-cache.cmake`.\n\nFollow-up to ba8752e5566076acc8bdec7ae4ec78901f7050f4 #12224\n\nCloses #22595", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "406edd036ae560dd1c37da327f213b2c1a82b654", "date": "2026-08-15T17:14:41+02:00", "subject": "filter: change time reporting", "body": "Replace the QUERY filter methods for connect and appconnect time with a\nnew control CF_CTRL_REPORT_STATS that is triggered when a connect ends\n(successful or not).\n\nFilters in the connection can then report their statistics. Socket and\nTLS filters do this only once. Subsequent CF_CTRL_REPORT_STATS will do\nnothing.\n\nThis prevents timers to be reported twice in STARTTLS scenarios.\n\nFixes #22587 (again)\nCloses #22596", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "dec546cde4b25c42f0260bd481872cb34f0c197d", "date": "2026-08-15T17:17:39+02:00", "subject": "tidy-up: formatting, messages and comments", "body": "- config-*.h: 'handcrafted'\n- tool_operate: rename `CURL *` variable to `curl` to match rest\n of code.\n- urldata.h: drop redundant parentheses from macro values.\n- curlinfo, dnsd: omit unused main arguments.\n- easy: sync function name in error message with source code.\n- cmake: drop period from version requirement messages.\n- tests/server/util: link three signal-related articles.\n- inline comments, fix indent, newlines, clang-format.\n\nCloses #22774", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "0c22fcffe3cc83df5d977b2bdeda4f3d894689d5", "date": "2026-08-15T17:28:32+02:00", "subject": "config-riscos.h: delete handcrafted RISC OS config header, in favor of autotools", "body": "According to the content of the latest downloadable builds available at\n(from 2022): https://www.riscos.info/packages/src/Network/, these builds\nused autotools, not the handcrafted header. Also the handcrafted header\ndid not see an update (that I could find) since\n0866c1b52dcff1f872860a42183397b4e8f6dedf (2002-04-08), other than\nsource-wide maintenance updates.\n\nCloses #22598", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "cd2db7913af68632efb1642b7dafd5cba24bf75c", "date": "2026-08-15T11:45:44-04:00", "subject": "fix(cwd): nvim_win_set_buf changes global CWD #41328", "body": "Problem:\nnvim_win_set_buf() on a non-current win, while the current win has\na win-local dir, changes the global CWD:\n\n :vsplit | lcd ..\n :call nvim_win_set_buf(other_win, buf)\n :wincmd l\n :verbose pwd\n [global] /parent \" expected: the initial cwd\n\nAnalysis:\n`globaldir` is where to return when no local dir applies; NULL means the\nprocess CWD is already there. Switching to a window with no local dir\nmakes update_cwd() chdir back to `globaldir` and clear it. kCtxKeepCwd\nrestores the process CWD but not that bookkeeping, so the restored\nwindow-local dir is mistaken for the global one.\n\nSolution:\nSave/restore `globaldir` with the CWD.", "style": "conventional", "type": "fix", "scope": "cwd", "breaking": false} {"repo": "postgres", "sha": "7b846015a40df872db572d92476d1e7c4e0c0066", "date": "2026-08-15T09:42:23-07:00", "subject": "Clarify logic in CreateSubscription().", "body": "No bug found in previous code, but it unnecessarily relied on grammar\nrules. Per complaint from Coverity.\n\nReported-by: Tom Lane \nDiscussion: https://postgr.es/m/1787286.1786328153@sss.pgh.pa.us\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "fb180287d0d3068b5873f4a7a39f506b4d52086d", "date": "2026-08-15T13:15:37-04:00", "subject": "fix(cwd): nvim_win_set_buf of :bcd buf, changes caller CWD #41329", "body": "Problem:\nSetting a :bcd buffer into another window, modifies the caller's CWD.\n\n local b = vim.api.nvim_create_buf(true, true)\n vim.api.nvim_buf_call(b, function() vim.cmd.bcd('..') end)\n vim.cmd('vsplit')\n vim.api.nvim_win_set_buf(vim.fn.win_getid(2), b)\n\n :echo haslocaldir(0) haslocaldir(-1,0) haslocaldir(-1,-1,0)\n 0 0 0\n :echo getcwd() ==# getcwd(-1,-1)\n 0\n\nAnalysis:\n`ctx_dirs_save` only saves CWD if it predicts the switch can change it.\nBut win_set_buf() replaces the target window's buffer *after* the\nswitch, which cannot be \"predicted\" from `ctx_dirs_save`.\n\nSolution:\nAlways snapshot whenever the switch enters another window.\nSkipping `os_dirname` was a micro-optimization.", "style": "conventional", "type": "fix", "scope": "cwd", "breaking": false} {"repo": "neovim", "sha": "37c670e68297264a5606ffd0e25d7fb4c585a379", "date": "2026-08-15T13:34:51-04:00", "subject": "fix(marks): undo reverts a mark set after the change #41330", "body": "Problem:\nA named mark updated after a change is moved back (treated as the\noriginal mark) by undo:\n\n :1mark d\n :$\n dw\n :2mark d \" 'd is on line 2\n :undo \" 'd is back on line 1\n\nThe undo header snapshots `b_namedm` when the change is recorded, and\n`u_undoredo()` restores that snapshot indiscriminately.\n\nSolution:\nUpdate the pending header's snapshot when a mark is set explicitly.\nMarks that the change itself moved go through mark_adjust(), not\nsetmark_pos(), so those are still reverted.\n\nSimilar to 2546741d1b36 (for extmarks): an explicit set inside an undo\nblock is confused with an edit-driven adjustment. But the extmarks case\nis dealing with mid-edit moves, whereas named/regular marks only need\nthe stale snapshot dropped.", "style": "conventional", "type": "fix", "scope": "marks", "breaking": false} {"repo": "curl", "sha": "fa21937ab0157325f426de40299f5fc5ac138190", "date": "2026-08-15T20:19:19+02:00", "subject": "GHA/non-native: sync BSD test exclusion logic with rest of workflows", "body": "Stop auto-skipping tests on emulated CPUs, replace it with explicit\n`skipall` tags instead. To make it easier to enable tests on emulated\nCPUs and to align with rest of workflows.\n\nCloses #22599", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "0c72958a67d8640351b1e3656e783e7effd31562", "date": "2026-08-15T23:00:46+02:00", "subject": "GHA/windows: bump stunnel to 5.80", "body": "Ref: https://www.stunnel.org/NEWS.html\n\nCloses #22600", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "8cbc8a17f05b3624742b8738d0afbc9f65a4f3f1", "date": "2026-08-16T07:25:04+08:00", "subject": "fix(lua): don't limit indexed vim.cmd positional argument count", "body": "(cherry picked from commit 2edb1c000958952c87edda596ed9ed53628084ed)", "style": "conventional", "type": "fix", "scope": "lua", "breaking": false} {"repo": "neovim", "sha": "2edb1c000958952c87edda596ed9ed53628084ed", "date": "2026-08-16T07:25:04+08:00", "subject": "fix(lua): don't limit indexed vim.cmd positional argument count", "body": "", "style": "conventional", "type": "fix", "scope": "lua", "breaking": false} {"repo": "curl", "sha": "bf594226d66dc2bbf62a18f3dea1ceabe5b26dcf", "date": "2026-08-16T01:57:00+02:00", "subject": "GHA/non-native: BSD build improvements", "body": "- show binary sizes and file types in 'curl -V' step, like rest of\n workflows.\n- make `build tests` step verbose.\n- disable typecheck on emulated CPUs.\n It results in 3-4x speed-up for riscv64, and ~2x for ARM64 when\n building tests (in particular the libtests binary.)\n Ref: 9e6f1c5efb7a70e1f33e467a738f3e3f652f3174 #19637\n- enable building tests in an ARM64 job.\n- move examples build test from ARM64 to riscv64 (to spread job times).\n\nCloses #22601", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "992685cb1c096050cfb2d87f918d85992466cb67", "date": "2026-08-16T08:37:43+08:00", "subject": "perf(runtime-vapor): reduce redundant scope id work", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "7871ff6e28478d0744292d77d4cbe8d22b43cfe7", "date": "2026-08-16T09:09:46+08:00", "subject": "docs(Alert): label visibility switch for accessibility", "body": "", "style": "conventional", "type": "docs", "scope": "Alert", "breaking": false} {"repo": "ant-design", "sha": "80238d400b6941e5d3d7afc67900c19ca17a1308", "date": "2026-08-16T09:10:51+08:00", "subject": "test(Rate): enable listitem accessibility checks", "body": "", "style": "conventional", "type": "test", "scope": "Rate", "breaking": false} {"repo": "ant-design", "sha": "fa1d88699ed3257bd8d446575c7cd04890be6821", "date": "2026-08-16T09:21:05+08:00", "subject": "docs: improve complex Form label association", "body": "* docs(Form): associate complex labels with controls\n\nDocument and test explicit htmlFor/id pairing when a layout-only Form.Item wraps a nested field.\n\nAI-assisted: Codex\nSigned-off-by: 陳廷安 <73953029+nrps9909@users.noreply.github.com>\n\n* test(form): refine complex control accessibility example\n\nUse valid JSX comments in the copyable documentation snippets and snapshot the labeled input in both test runners.\n\nAI-assisted: Codex\nSigned-off-by: 陳廷安 <73953029+nrps9909@users.noreply.github.com>\n\n---------", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "8faf7561a5040265a2c281c7124aee411bd4c9f8", "date": "2026-08-16T09:45:22+08:00", "subject": "perf(runtime-vapor): tree-shake slotted scope id propagation", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "neovim", "sha": "cd71a1358902ac640b0feeda649c335af7a983a8", "date": "2026-08-15T22:28:08-04:00", "subject": "build(vim-patch): auto-n/a refactor patch like v8.2.0514", "body": "Follow-up refactor patches, mixed with unrelated patches in the middle,\nis normal in vim-dev when the original patch is not fully tested\nacross all builds via CI or reviewed by others.\n\nUnless core maintainers push \"vim-patch:\" directly to master/main branch\nwithout running the full test-suite,\nthere is little reason to merge incomplete ports that will fail\non Nvim's CI or code review.\n\nRelevant changes in patches like v8.2.0514 are either ported\nor (will) become N/A.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "6e228b1cdaa37a8a67e5690c2c5830b6b59b6c18", "date": "2026-08-16T14:24:03+09:00", "subject": "ocsp: Report expired nextUpdate with the correct reason", "body": "OCSP_check_validity() already rejects an expired nextUpdate, but it\nreports the not yet valid reason. Raise OCSP_R_STATUS_EXPIRED instead.\n\nFixes #32399\n\nCLA: trivial\nReviewed-by: Tomas Mraz \nReviewed-by: Richard Levitte \nReviewed-by: Bob Beck \nMerge-date: Wed Aug 19 10:50:59 2026\nMerged-from: https://github.com/openssl/openssl/pull/32400", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "13e3ecbcb94b39d05815e301845e4490278f842a", "date": "2026-08-16T14:24:03+09:00", "subject": "ocsp: Report expired nextUpdate with the correct reason", "body": "OCSP_check_validity() already rejects an expired nextUpdate, but it\nreports the not yet valid reason. Raise OCSP_R_STATUS_EXPIRED instead.\n\nFixes #32399\n\nCLA: trivial\nReviewed-by: Tomas Mraz \nReviewed-by: Richard Levitte \nReviewed-by: Bob Beck \nMerge-date: Wed Aug 19 10:50:46 2026\nMerged-from: https://github.com/openssl/openssl/pull/32400", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "ae2b1ab79620e51335b0b32fa4e60b8fede39b46", "date": "2026-08-16T07:41:29+02:00", "subject": "Add AI policy", "body": "* Add AI policy\n\n* Update AGENTS.md\n\nCo-authored-by: Huáng Jùnliàng \n\n---------", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "5a552db9a8ebc801300c759cdd5f18a54b2834e9", "date": "2026-08-16T16:58:58+08:00", "subject": "feat(Dropdown): add menu Semantic DOM", "body": "", "style": "conventional", "type": "feat", "scope": "Dropdown", "breaking": false} {"repo": "eslint", "sha": "2ba302554e7a24e9909bbdd026fd0c2d1d0d8638", "date": "2026-08-16T13:05:46+03:00", "subject": "fix: prevent unsafe `no-var` autofix with hoisted functions", "body": "* fix: prevent unsafe `no-var` autofix with hoisted functions\n\n* apply suggestion\n\n* apply suggestion", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "angular", "sha": "83450d292443cae811e87ab61020f3ed72d27d77", "date": "2026-08-16T12:26:25+02:00", "subject": "fix(forms): report forbidden 2way bindings on when `FormField` is applied", "body": "We were already reporting regular bindings but not 2way.\n\nfixes #70219\n\n(cherry picked from commit c819880b91bf5ff1c5aa71afd314ae7cae241065)", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "c819880b91bf5ff1c5aa71afd314ae7cae241065", "date": "2026-08-16T12:26:25+02:00", "subject": "fix(forms): report forbidden 2way bindings on when `FormField` is applied", "body": "We were already reporting regular bindings but not 2way.\n\nfixes #70219", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "curl", "sha": "aa66b02dd8b74b4c72484c240d39399387054ea3", "date": "2026-08-16T12:36:24+02:00", "subject": "tidy-up: misc C and CMake nits", "body": "- CMake/win32-cache.cmake: sort and sync MSVC pre-fills with mingw-w64.\n\n- curl_sasl: use `uint32_t` for local `httpauth` value. To match struct.\n Follow-up to e3691612ea1acb28e77fee6b4b2b9da0ac2c8f8b #20209\n\n- libstests: move literal to the right-hand side of comparisons.\n\n- windows: drop `!= 0`/`== 0` from `WSAEventSelect()` result checks,\n to keep it short.\n\n- tool_cfgable: replace `CURL_GLOBAL_ALL` with `CURL_GLOBAL_DEFAULT` in\n debug builds to match the macro used in non-debug and libcurl init.\n (Both macros map to the same literal value; there is no code change.)\n\n- perf/urlparser: use double-const in test array.\n\n- examples: `CURL_GLOBAL_DEFAULT` -> `CURL_GLOBAL_ALL` to match rest of\n examples.\n\n- libtests: `CURL_GLOBAL_DEFAULT` -> `CURL_GLOBAL_ALL` to match rest of\n tests.\n\nCloses #22783", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "8eb978b9c212068371c39b37be1522904a790811", "date": "2026-08-16T12:40:05+02:00", "subject": "GHA/non-native: drop leftover perl5 install from riscv64 job [ci skip]", "body": "Follow-up to bf594226d66dc2bbf62a18f3dea1ceabe5b26dcf #22601", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "d00000673d05fdf6ceee8941ccd63ad2906257e4", "date": "2026-08-16T12:42:20+02:00", "subject": "GHA/non-native: sync BSD tflags logic with other workflows", "body": "To make it easier to customize tests for individual matrix jobs.\n\nAlso:\n- tested running riscv64 tests, but it takes too long (8.5-12 minutes,\n for ~1000 HTTP tests.)\n\nFollow-up to bf594226d66dc2bbf62a18f3dea1ceabe5b26dcf #22601\n\nCloses #22602", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "83730db6477484c773f39c8cc18c30e09dd76f4c", "date": "2026-08-16T08:29:58-04:00", "subject": "perf(marktree): binary search the node intersect array #41331", "body": "Problem:\nUndo of a change spanning many paired marks is quadratic. A node's\n\"intersect\" array holds every pair crossing that node, and both\nintersect_node() and unintersect_node() walked it linearly. Undoing an\nedit over 1M paired marks spends 68% of its time in unintersect_node()'s\nscan alone.\n\nSolution:\nThe array is sorted, so binary search it.\n\n marks undo before after\n 200k 831ms 385ms\n 1M 14006ms 3820ms\n\nRedo is unaffected: it is dominated by marktree_move() actually\nrepositioning the marks.", "style": "conventional", "type": "perf", "scope": "marktree", "breaking": false} {"repo": "neovim", "sha": "9143da2a05f4f4c70bf4491639d415e8056d8f84", "date": "2026-08-16T08:52:51-04:00", "subject": "build(vim-patch): n/a vim9 patch 9.0.1821", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "nestjs", "sha": "e03cf5c865d39cc45a0134e82d8f79f865f05d55", "date": "2026-08-16T15:24:25+02:00", "subject": "fix(express,fastify): apply falsy status codes in reply()", "body": "The truthiness guard in reply() is there to skip the status call when no status is\npassed, but it also drops codes that were provided and happen to be falsy. NaN and 0\nnever reach response.status(), so the status set before the handler ran survives and\nthe exception goes out with a 200. Truthy invalid codes such as -1 or 99 already end\nup as a 500, because express and fastify reject them.\n\nSkip the call only when no status is given, so a provided code always reaches the\nframework that can validate it.", "style": "conventional", "type": "fix", "scope": "express,fastify", "breaking": false} {"repo": "neovim", "sha": "214bcf24cc7d70d8fd8fc7f172251abe4bd5ed35", "date": "2026-08-16T10:24:24-04:00", "subject": "fix(undo): crash on corrupted undo file #41339", "body": "Problem:\n`:rundo` on a corrupted undo file crashes or hangs, instead of failing\nwith E825. Patching one 4-byte field is enough:\n\n ue_size = 0xFFFFFFFF \" walks a NULL ue_array\n ue_size = 0x7FFFFFF0 \" 17 GB xmalloc + memset, then preserve_exit()\n ue_top = 0xFFFFFFFB \" negative lnum reaches ml_delete()\n\nAnalysis:\nEvery count in the file is read with `undo_read_4c()` and then checked,\ndifferently at each site. None bounds the value by what the file can\nhold, so a 2 GB count reaches `xmalloc()`.\n\nNote:\n- Vim doesn't have `bi_fsize` because it checks `U_ALLOC_LINE` result\n everywhere (thus doesn't crash, but may thrash...); those checks were\n dropped when Nvim moved to `xmalloc()`, and the `ue_size` loop counter\n became unsigned.\n- Vim *does* have the negative line numbers bug: `u_undoredo()` checks\n `top > ml_line_count || top >= bot || bot > ml_line_count + 1`, which\n rejects none of them.\n\nSolution:\n- Introduce `undo_read_len()` and use it to fail early instead of\n continuing with nonsense.\n- Validate `ue_top`/`ue_bot`/ `ue_lcount`.\n- Use `xcalloc()`, so no site can proceed with a NULL array.\n- Report a truncated \"U\" line, distinguish EOF from a 0xFFFFFFFF field,\n and free the header on the extmark error path.", "style": "conventional", "type": "fix", "scope": "undo", "breaking": false} {"repo": "neovim", "sha": "7b6f344627f7d8af2016cd9e89ac145f1138643d", "date": "2026-08-16T07:34:17-07:00", "subject": "fix(restart): preserve global cwd on :restart #41304", "body": "Problem:\nOn :restart, the new Nvim may \"inherit\" a local dir as its global CWD.\n\nSolution:\nInherit the global CWD explicitly.", "style": "conventional", "type": "fix", "scope": "restart", "breaking": false} {"repo": "neovim", "sha": "5d1901b904401b929e11ff3c332d8097db66806c", "date": "2026-08-16T10:39:03-04:00", "subject": "build(vim-patch): v9.1.0349 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "9071cd2874ca4f2c35ab6eedf2d90fe7811f61ce", "date": "2026-08-16T08:16:15-07:00", "subject": "Make the genec test not test process fork and teardown.", "body": "The genec test tested 912 curves with a spawn and teardown of an\nrun of the app each - this takes a long time, and is largely\nbound on how fast you can start up and tear down the app.\n\nInstead, let's test all the curves in a C program doing the same\nthing, which takes almost no time at all, and test the representative\nparameters in the app to ensure they also get test coverage. Doing\nthose same parameters 912 times in the app to make sure they function\n912 times buys us nothing for testing the app. So this will give\nus coverage while making the test run faster, and being less\ntrouble for OS zoo to run testing fork exec teardown instead\nof what we want to test.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:17:15 2026\nMerged-from: https://github.com/openssl/openssl/pull/32404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "5e69aba6540019bce86201ceb34311d45d909e94", "date": "2026-08-16T17:32:16+02:00", "subject": "fix: own the post-restart rerun, keep `process.exit` disabled in workers", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "neovim", "sha": "aaf57a053d3d623515a409ee3a5cc3539c94a936", "date": "2026-08-16T11:37:51-04:00", "subject": "fix(coverity): false positives in kvec usages #41341", "body": "Coverity can't follow kv_ensure_space()'s `kv_roundup32()` bit math, so\nevery kv_concat_len() looks like an overrun; and it doesn't know\n`kv_push()` allocates when `size == capacity`.\n\n _____________________________________________________________________________________________\n CID 653191: Memory - illegal accesses (OVERRUN)\n /src/nvim/input.c: 3536 in paste_store()\n 3530\n 3531 if (s > start) {\n 3532 if (need_redo) {\n 3533 kv_concat_len(redobuff.cur.keys, start, (size_t)(s - start));\n 3534 }\n 3535 if (need_record) {\n >>> CID 653191: Memory - illegal accesses (OVERRUN)\n >>> Overrunning dynamic array \"recordbuff.items\" at offset corresponding to index variable \"recordbuff.size\" through dereference in call to \"memcpy\".\n 3536 kv_concat_len(recordbuff, start, (size_t)(s - start));\n 3537 }\n 3538 }\n 3539\n 3540 if (s < str_end) {\n 3541 int c = (uint8_t)(*s++);\n\n _____________________________________________________________________________________________\n CID 653190: Memory - illegal accesses (OVERRUN)\n /src/nvim/input.c: 730 in redo_append_spec()\n 724 return;\n 725 }\n 726\n 727 while (*s != NUL) {\n 728 if ((uint8_t)(*s) == K_SPECIAL && s[1] != NUL && s[2] != NUL) {\n 729 // Insert special key literally.\n >>> CID 653190: Memory - illegal accesses (OVERRUN)\n >>> Overrunning dynamic array \"redobuff.cur.keys.items\" at offset corresponding to index variable \"redobuff.cur.keys.size\" through dereference in call to \"memcpy\".\n 730 kv_concat_len(redobuff.cur.keys, s, 3);\n 731 s += 3;\n 732 } else {\n 733 sb_add_char(&redobuff.cur.keys, mb_cptr2char_adv(&s));\n 734 }\n 735 }\n\n CID 653189: (OVERRUN)\n /src/nvim/input.", "style": "conventional", "type": "fix", "scope": "coverity", "breaking": false} {"repo": "neovim", "sha": "732da013c70124b6ef66c696091e736d0f07c33b", "date": "2026-08-16T18:38:38+03:00", "subject": "ci: bundle wasm build output with demo into a single release zip", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "tmux", "sha": "40cd869d770732fb4f7aad705360600ae6dc9b43", "date": "2026-08-16T17:09:11+01:00", "subject": "Readability and clarity.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9c15051a9fe23aae617c5604b81081dbb1ece4d1", "date": "2026-08-16T09:55:28-07:00", "subject": "Make test framework output records atomic across threads", "body": "A single TEST_info() message is composed of several writes, each\ntaking the output lock independently, so concurrent callers\ninterleaved mid-line. This makes it annoying when we run multiple\nthreads in a test job and get interleaved output that is hard to\ndiagnose.\n\nTrack the lock depth per thread so it can be held across a whole\nrecord. A thread with no thread-local data cannot track the depth and\nfalls back to the previous per-write locking.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:17:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/32404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e0e2f978a019c4973f9c91e7a598e92e913393f2", "date": "2026-08-17T01:29:34+08:00", "subject": "feat(vim.fs): slug() supports URI #41241", "body": "Problem:\n`vim.fs.slug()` does not handle URIs like `term://foo//123:bash`,\nso callers (e.g. terminal persistence) must strip the scheme before\ncalling `slug()`.\n\nSolution:\nDetect `scheme://` from the raw input before `normalize()` and\nreplace it with a `=uri--` prefix.", "style": "conventional", "type": "feat", "scope": "vim.fs", "breaking": false} {"repo": "openssl", "sha": "83cb4e0bc3989a5d4252ddb8ebf0d5736d3a44de", "date": "2026-08-16T10:40:59-07:00", "subject": "Size the multithreaded lhash test to the coverage it needs", "body": "The iteration and worker counts were not derived from what the test has\nto cover. What matters is each ordered pair of behaviours being\nexercised concurrently on the same key, and the chance of missing one\nfalls off exponentially with the iteration count, so 150000 iterations\ncovers that with a wide margin. The derivation is documented at the\nconstant.\n\nCap the workers at four rather than taking a quarter of HARNESS_JOBS.\nFour is enough to interleave more than a pair of threads on a key, which\ntwo cannot do, without tying up a machine.\n\n02-test_lhash was the longest recipe in the suite and set the floor for\nthe whole run. On a MacBook Pro with HARNESS_JOBS=14 it goes from 111\nseconds to 5.4 seconds.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:17:17 2026\nMerged-from: https://github.com/openssl/openssl/pull/32404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "1c2b0575ed03cf4a1d458df2a0129fac29905643", "date": "2026-08-16T21:32:59+03:00", "subject": "feat(docs-infra): add Angie to the roadmap illustration", "body": "The roadmap header draws a road between a start pin and a destination\npin, but nothing travels it. Angie already appears on the 404 page, the\nembedded editor states and the error snack bar, so put her on the road\ntoo.\n\nDecorative headers are inlined into the page so the road can use CSS\nvariables for dark mode, which means the pose has to be embedded as\npaths, with its classes flattened and its ids prefixed to keep them out\nof document scope.\n\nNote this asset is also the essentials Next Steps header.", "style": "conventional", "type": "feat", "scope": "docs-infra", "breaking": false} {"repo": "neovim", "sha": "0e436350a589a5ac3f40f8a622470fd247e8fd88", "date": "2026-08-16T16:14:33-04:00", "subject": "refactor(cmdatom): atom_redo_keys #41345", "body": "Some names/comments are misleading.\nAlso add some asserts.", "style": "conventional", "type": "refactor", "scope": "cmdatom", "breaking": false} {"repo": "curl", "sha": "4468b8e88c7df13f11b2f4b7674f0d58b0e4aacd", "date": "2026-08-16T23:15:58+02:00", "subject": "openssl: add 1.1.1 support", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c7525d3628d10008a7e2cce66998e4d34e0eb19f", "date": "2026-08-16T23:24:18+02:00", "subject": "CVE-2026-8924: trailing dot domain super cookie", "body": "The cookie change from 51beed175dbfc37da (skipped the test part)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "9275aca5b123049ed4d566bc1625e10183954337", "date": "2026-08-16T23:31:06+02:00", "subject": "CVE-2026-8927: env-set cross-proxy Digest auth state leak", "body": "Cherry-picked 5c225384b8d52c6, except the test parts", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "9e948bd3d74e58b204060cea845ded9902556969", "date": "2026-08-16T23:38:45+02:00", "subject": "CVE-2026-9079: stale proxy password leak", "body": "Cherry-picked from 88c7e16cceec (except the test changes)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a857515a0984d869d761a9db194b06e8e8be303f", "date": "2026-08-16T23:41:28+02:00", "subject": "Fix bugs in OSSL_STORE_attach", "body": "Unify error handling in OSSL_STORE_attach with OSSL_STORE_open_ex.\n\nIf setting loader params fails loader_ctx is closed but not set to NULL.\nWhich leads to successful return with invalid object.\nFixed by setting loader_ctx to NULL triggering error return path.\n\nIf allocation OSSL_STORE_CTX fails after the loader context creation\nclose the loader context and release the fetched loader.\nThis covers both OSSL_STORE_CTX allocation failure and UI method setup failure.\n\nFound by : store fuzzer (MFAIL-test)\n\nAssisted-by: ChatGPT:gpt-5.6\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Neil Horman \nMergeDate: Tue Sep 15 14:20:01 2026\n(Merged from https://github.com/openssl/openssl/pull/32413)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3b2da7a507f362f9d25b16e989bc7f33df2b9714", "date": "2026-08-16T14:44:59-07:00", "subject": "Size test_mod_exp_x2 to the coverage it needs", "body": "A defect either shows on a large fraction of random inputs or is not\nreachable by sampling in one run at all. Twenty rounds per factor size\ncovers the former.\n\nDrive the sizes from a table rather than a ladder of index ranges, which\nalso makes the bands equal.\n\nOn a MacBook Pro the exptest binary goes from 24.1 to 5.3 seconds.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:17:18 2026\nMerged-from: https://github.com/openssl/openssl/pull/32404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "71d721d3574b9783cd0149026b4a52f67794506a", "date": "2026-08-16T14:59:52-07:00", "subject": "Stop running the entire ecparam corpus through the app", "body": "Checking 136 parameter files with ecparam and pkeyparam cost a process\nper file per check, 641 in all, which is process startup rather than\ncoverage. Test all of these in one process per set in C instead.\n\nThe app keeps it's own test, covering the options and a few\nfiles through the output path, in 20-test_app_ecparam.t.\n\nTogether the two now take 1.0 seconds where the one took 24.5.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:17:19 2026\nMerged-from: https://github.com/openssl/openssl/pull/32404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "581ce0b3dad3d8dcb7fdfd872f2d447f6c75dc1c", "date": "2026-08-16T18:00:55-04:00", "subject": "fix(cmdatom): mappings #41347", "body": "Problem:\n`` mappings do not emit `CmdAtom.text`.\n`` and Lua-callback mappings that edit the buffer apply only at the\nprimary cursor, not cascaded (multicursor).\n\nSolution:\nCapture the `` command in getcmdkeycmd().\nAdd kKeyOpaque (\"no capturable keys\"); narrow kKeySynthetic (\"not\na keystroke\") to K_EVENT/K_IGNORE, so an opaque mapping's edit still\nsets `map_edit` and cascades via LHS-replay.", "style": "conventional", "type": "fix", "scope": "cmdatom", "breaking": false} {"repo": "openssl", "sha": "a5031f7e8cedb975628901130ab18ee9a212e655", "date": "2026-08-16T15:51:27-07:00", "subject": "Make the low contention torture writers write", "body": "OSSL_sleep() takes milliseconds, so the contention == 0 path slept a\nsecond between writes and the low contention tests managed six writes\neach in their four second budget. The readers spun tens of millions of\ntimes over a value that was not changing.\n\nSleep a millisecond instead. In the same four seconds torture_rw_low\ngoes from 6 writes to 4940, and torture_rcu_low from 6 to 6344.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:17:19 2026\nMerged-from: https://github.com/openssl/openssl/pull/32404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "26068bb453829fc7f0b49e885f3e2744f8a8e8d4", "date": "2026-08-16T16:22:57-07:00", "subject": "Bound the threadstest torture writers by writes rather than by time", "body": "The five torture tests ran their writers for four seconds each, so the\nwork done varied with the machine while the cost did not: this one\nperforms 8.8 million writes in torture_rcu_high where a loaded VM might\nmanage a hundredth of that, and both spend the same twenty seconds.\n\nGive the writers a number of writes to perform instead, so the sampling\nis the same everywhere and the time falls out of it. The elapsed time\ncheck stays as a backstop, so a machine too slow to reach the target\nstops where it stopped before and nothing gets slower than it was.\n\nOn a MacBook Pro 90-test_threads goes from 23.7 to 7.6 seconds, and from\n169 to 49 CPU seconds.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:17:20 2026\nMerged-from: https://github.com/openssl/openssl/pull/32404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7065655125e80d3026adcb66b1bc60894e4d4d7b", "date": "2026-08-16T16:44:48-07:00", "subject": "Poll the TLSProxy test sockets ten times a second", "body": "The forwarding loop waited a second at a time for data and gave up\nafter ten idle iterations. A TLS connection leaves the loop when the\nsockets close, so it never waits, but DTLS runs over UDP where there\nis no EOF, and a DTLS test whose handshake is not expected to complete\ncan only leave by exhausting the whole ten second budget. Two tests in\n70-test_sslrecords did exactly that.\n\nPoll at a tenth of a second instead. The iteration limit is unchanged,\nso a second of complete silence now ends the connection rather than ten,\nand a run that is only waiting for its exit condition notices ten times\nsooner.\n\n70-test_sslrecords goes from 32.3 to 13.7 seconds. The other TLSProxy\nrecipes are unaffected, as their loops exit on socket close without ever\nreaching the timeout.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:17:21 2026\nMerged-from: https://github.com/openssl/openssl/pull/32404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "bbe1a8950270c3b28bf15251312707d7c6b47373", "date": "2026-08-17T07:48:22+08:00", "subject": "fix(runtime-vapor): align scope id propagation across interop boundaries", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "2e5e63e9a708572134176e0bd766104f4213ccff", "date": "2026-08-16T23:49:25Z", "subject": "[autofix.ci] apply automated fixes", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0391fee2dc30c801fc17a1eeb6a82448d359e052", "date": "2026-08-17T08:53:29+09:00", "subject": "Fix ASAN failure after flex errors in GUC file parsing", "body": "As detected by ASAN, the scanner value used when parsing GUC files can\nbe indeterminate when the flex error handler sigjumps to the old cleanup\npath, before yylex_init() is called.\n\nThe flex scanner state is now made volatile in ParseConfigFp(), since\nits value is assigned after sigsetjmp() and cna be accessed after\nsiglongjmp(). yylex_init() cannot use a volatile pointer; a temporary\nvariable is used before assigning the result of yylex_init() to it.\n\nWhile on it, yy_create_buffer() is changed to detect the case where it\nreturns a NULL value. Based on my read of the flex code, this cannot be\nreached currently. Future upstream changes or changes in the error\nlogic of the GUC file parsing could make that reachable, and it is four\nextra lines of code.\n\nOversight in d663f150b5ed.\n\nReported-by: Ilia Kashintsev \nReviewed-by: Tom Lane \nReviewed-by: Andrey Rachitskiy \nDiscussion: https://postgr.es/m/19612-24ccb4fc6da7786f@postgresql.org\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e382876248092d281a5250b9de0ea76af29e85a1", "date": "2026-08-16T17:07:36-07:00", "subject": "Hold DTLS s_server stdin open with a pipe rather than a timer", "body": "TLSProxy gave a DTLS s_server a stdin that was the read end of a pipe\nfrom a \"perl -e sleep(10)\" process, so that it would not see EOF while\nthe test ran. Nothing ever closed that early, so the recipe paid out\nthe remainder of the ten seconds before it could exit.\n\nUse a plain pipe and keep the write end in the proxy, closing it in the\nteardown that already reaps s_server. Stdin then stays open exactly as\nlong as the test needs and no longer, with no timer to tune.\n\n70-test_sslrecords goes from 13.7 to 4.0 seconds on my mac\n\nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:17:22 2026\nMerged-from: https://github.com/openssl/openssl/pull/32404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "baa7b142aace6821ce085906f314a75bcc4d95c8", "date": "2026-08-17T09:21:08+09:00", "subject": "psql: Fix \\d+ display of REPLICA IDENTITY NOTHING", "body": "Commit 18954ce7f69 has replaced hardcoded relreplident values in\ndescribe.c with CppAsString2 macros, but used a DEFAULT instead of a\nNOTHING at one location. This caused \\d+ on a table with REPLICA\nIDENTITY NOTHING to show incorrect data.\n\nAuthor: Shinya Kato \nDiscussion: https://postgr.es/m/CAOzEurRHWt+9XLBrbXQOSVkE45NksJ4F28twuqgZ7veW0RzpUQ@mail.gmail.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "fe5cf3259cabde7c66bbcb67fb5c2b317a038197", "date": "2026-08-16T20:30:18-04:00", "subject": "build(vim-patch): v9.1.2033 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "neovim", "sha": "1e153b9a1c041feb329ff3004624cd6f2e13a157", "date": "2026-08-16T20:52:24-04:00", "subject": "build(vim-patch): v9.2.0026 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "d81896297f1a09bf86f532762d11b2066553f565", "date": "2026-08-16T17:53:11-07:00", "subject": "Assert on the correct size.", "body": "Reviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:17:22 2026\nMerged-from: https://github.com/openssl/openssl/pull/32404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9ca2185a2b099a17e94e2cac815aba714625e858", "date": "2026-08-16T17:56:09-07:00", "subject": "Pass the ecparam corpus in a file rather than as arguments", "body": "The recipe passed all 102 valid parameter files as arguments, but a test\ncan be given at most 100 before test_get_argument() runs off the end of\nits bookkeeping array. The address sanitizer build caught it; on a\nnormal build it silently wrote past the end.\n\nWrite the file names to a list file and pass that instead, so the corpus\ncan grow without bound. The globbing and the sm2 filtering stay in the\nrecipe.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:17:23 2026\nMerged-from: https://github.com/openssl/openssl/pull/32404", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "833e58341c0e99e364c64c3cc59f15afbb8b3fb3", "date": "2026-08-16T21:12:14-04:00", "subject": "build(vim-patch): n/a \":terminal\" opts", "body": "Target v8.1.2195 .\n\nNvim did not port Vim's \":terminal\" opts.\nIncompatible implementations.\nEx-command was ported from C to Lua.\nVim needs them partly because of splitting the current window.\n\nI keep forgetting `++close` option so I either run \":qall!\"\nor kill the parent process (ie. terminal emulator).\n\nUnsatisfied users should create their Ex-command that runs jobstart().", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "ant-design", "sha": "4b323c453dacb31a4edecdb380ec94575b5c14ce", "date": "2026-08-17T09:18:55+08:00", "subject": "fix(radio): trigger option onChange when using options in radio-group", "body": "", "style": "conventional", "type": "fix", "scope": "radio", "breaking": false} {"repo": "vue-core", "sha": "1129f0cfa8440e884e43d7f10f09a9ce47d03e52", "date": "2026-08-17T09:41:21+08:00", "subject": "fix(runtime-vapor): preserve root bindings across attrs fallthrough", "body": "Keep separate local and inherited ownership for component root bindings so\nparent attrs can override child values without discarding the latest local\nstate. Restore ordinary props when inherited keys are removed, merge class\nand style caches, and dispatch local and inherited dynamic listeners through\none stable native listener.\n\nInstall fallthrough effects for initially empty dynamic sources, apply the\nfunctional-component whitelist to component roots, and avoid capturing\nv-show's transient display state as local style.", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "d246a7000aeae4f4e67818083a27b97e51726100", "date": "2026-08-17T09:45:34+08:00", "subject": "docs: add v6 upgrade guidance to maintenance notice", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "447589c6720ea578b5729f4f35444b9d2d21b997", "date": "2026-08-17T10:00:33+08:00", "subject": "docs: update v4 Upload demo endpoint", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "205fa09f90726d3ed9938ebfc87734314ea5d9ab", "date": "2026-08-17T10:03:57+08:00", "subject": "fix(runtime-vapor): preserve slotted scope id context", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "90ebf247a82aecaee21e49fa98078f5abf200dc7", "date": "2026-08-17T11:04:24+09:00", "subject": "fix(Table): apply first-row radius when header is hidden", "body": "", "style": "conventional", "type": "fix", "scope": "Table", "breaking": false} {"repo": "vue-core", "sha": "f86f54ebea7e8c16d67d0052bba122bbb4588925", "date": "2026-08-17T10:08:40+08:00", "subject": "fix(runtime-vapor): preserve once listeners across fallthrough updates", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "4fe8ad0115d67d077a9ad41749bf22ca67b5e3d0", "date": "2026-08-17T10:11:30+08:00", "subject": "fix(DatePicker): hide decorative range separator", "body": "", "style": "conventional", "type": "fix", "scope": "DatePicker", "breaking": false} {"repo": "ant-design", "sha": "460b5fc504ee18176ee569acd52b0d46b4412e31", "date": "2026-08-17T10:11:57+08:00", "subject": "feat(Carousel): localize default arrow aria-labels via ConfigProvider locale", "body": "", "style": "conventional", "type": "feat", "scope": "Carousel", "breaking": false} {"repo": "ant-design", "sha": "be8e187c6083bceaa3c1f84313b63982d3ca4087", "date": "2026-08-17T10:12:34+08:00", "subject": "fix(Splitter): localize collapse control labels", "body": "", "style": "conventional", "type": "fix", "scope": "Splitter", "breaking": false} {"repo": "neovim", "sha": "19a7f26a9cad60a695839562e8a718fa60f7d4c3", "date": "2026-08-16T22:12:42-04:00", "subject": "build(vim-patch): n/a docs for v8.1.2219", "body": "Ignoring incompatible implementation,\nTerminalOpen and TermOpen events are not 1-1.\n\"TerminalWinOpen\" was accepted as N/A in\ncommit c7ee6af777a9a0b1cd54817c646fe32fa7267c47 .\n\nI planned to not do this to have more test cases\nafter manipulating the hunks header to filter out more hunks\nbut Justin is eager to just mark these N/A\nto bump the Vim major.minor version.\n\nTime to move on from v8.1.x.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "ant-design", "sha": "61140f857a102295139f8acd45284d474b611f19", "date": "2026-08-17T10:12:58+08:00", "subject": "fix(Collapse): hide decorative expand icon", "body": "", "style": "conventional", "type": "fix", "scope": "Collapse", "breaking": false} {"repo": "ant-design", "sha": "6b28e00b5166c4de62b10b2ee61eb257a7e49dec", "date": "2026-08-17T10:19:56+08:00", "subject": "fix(Transfer): avoid duplicate onSearch calls when clearing", "body": "", "style": "conventional", "type": "fix", "scope": "Transfer", "breaking": false} {"repo": "angular", "sha": "ef2ce9a09802faa5ffff0860f473c4018f25f2ac", "date": "2026-08-16T22:43:09-04:00", "subject": "fix(core): expose debuggableFn for non-computed signal graph nodes", "body": "Before: getSignalGraph only exposed debuggableFn on computed and template\nnodes, preventing DevTools from linking to source code for other node types.\n\nAfter: debuggableFn is also populated for linkedSignal, effect, and\nafterRenderEffectPhase nodes.\n\n(cherry picked from commit 0ddbc47e7f6a4eced3a1a986b260747dc37ee9c9)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "0ddbc47e7f6a4eced3a1a986b260747dc37ee9c9", "date": "2026-08-16T22:43:09-04:00", "subject": "fix(core): expose debuggableFn for non-computed signal graph nodes", "body": "Before: getSignalGraph only exposed debuggableFn on computed and template\nnodes, preventing DevTools from linking to source code for other node types.\n\nAfter: debuggableFn is also populated for linkedSignal, effect, and\nafterRenderEffectPhase nodes.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "ant-design", "sha": "c9aa0cdb503636444bd5440e11fab4ce28b7ba8f", "date": "2026-08-17T10:57:09+08:00", "subject": "fix(Menu): prevent tooltip flash when collapsing while hovered", "body": "* fix(Menu): prevent tooltip flash when collapsing while hovered\n\nCo-authored-by: Cursor \n\n* fix(Menu): respect controlled tooltip.open when collapsed\n\nCo-authored-by: Cursor \n\n---------", "style": "conventional", "type": "fix", "scope": "Menu", "breaking": false} {"repo": "ant-design", "sha": "5957ad9b11441b79f1259becab1584e9c107f489", "date": "2026-08-17T05:12:00+02:00", "subject": "fix: preserve Form labelCol offset in vertical layout", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "3afd59fb49c1441d1cd157607ba9fe0866da3ef9", "date": "2026-08-17T11:13:34+08:00", "subject": "fix(runtime-vapor): preserve scope ids for cached and teleported roots", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "neovim", "sha": "8c0bf183745928f71fbd58f4b73afafbd6a045d4", "date": "2026-08-17T08:46:11+05:30", "subject": "fix(spell): avoid invalid window state after async spell select", "body": "Problem:\nWhen `z=` delegates to `vim.ui.select()`, the picker may change the\ncurrent window before returning. `spell_suggest()` then continues to the\ncursor restoration branch with the new window and assigns `prev_cursor`,\nwhich belongs to the original window. This can leave Normal mode with an\ninvalid cursor position and produce E315.\n\nSolution:\nClean up the spell suggestion state and return immediately after handing\ncontrol to `vim.ui.select()`.", "style": "conventional", "type": "fix", "scope": "spell", "breaking": false} {"repo": "neovim", "sha": "e96be45bba4e209629bcf053113287897fb02d5c", "date": "2026-08-16T23:28:43-04:00", "subject": "build(vim-patch): v8.2.4153 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vite", "sha": "56f75f68d403b9c8d1860572c12c5f5c08332add", "date": "2026-08-17T12:35:52+09:00", "subject": "docs(deps): update vitepress to 1.0.0-alpha.19 and theme to 5.0.6", "body": "", "style": "conventional", "type": "docs", "scope": "deps", "breaking": false} {"repo": "ant-design", "sha": "68717476f4838abbf199f709fa2863129998cf7a", "date": "2026-08-17T11:47:52+08:00", "subject": "fix(Menu): prevent icon jitter during inline collapse", "body": "", "style": "conventional", "type": "fix", "scope": "Menu", "breaking": false} {"repo": "ant-design", "sha": "a76aa2353a55d76be14ee883dd97dc3ad7e53413", "date": "2026-08-17T11:58:18+08:00", "subject": "fix(form): prevent height fluctuation during rapid validation changes", "body": "", "style": "conventional", "type": "fix", "scope": "form", "breaking": false} {"repo": "angular", "sha": "a3a77649754cebee0b58309e9bade56f551599ba", "date": "2026-08-17T00:06:34-04:00", "subject": "fix(devtools): prevent duplicate highlightComponent listener in injector tree", "body": "Move the message bus subscription out of afterRenderEffect into constructor initialization and clean up the listener on destroy using DestroyRef.", "style": "conventional", "type": "fix", "scope": "devtools", "breaking": false} {"repo": "vue-core", "sha": "3c1ce9add484ec3aaae65552ea171e5f432f76ca", "date": "2026-08-17T13:44:51+08:00", "subject": "fix(runtime-vapor): scope forwarded slot ids to direct roots", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "neovim", "sha": "ea6abf15fe1f570a10c54211bf5209e9f9d38707", "date": "2026-08-17T01:55:14-04:00", "subject": "refactor(cmdatom): do \"redo prep\" in one place #41348", "body": "Problem:\nRedo prep is scattered/duplicated.\n- `do_pending_operator()` has 3 prep blocks whose conditions must be in\n sync with `atom_capture_op()`.\n- insert.c, spell_suggest() hand-roll `redo_new()` + `redo_append_xx()`\n sequences.\n- prep_redo() has 2 roles, decided by `keys != NULL`.\n\nSolution:\n- `atom_capture_op()` is the \"operator\" entry point: capture, then\n prep.\n- Extract `prep_redo_visual()`, `atom_capturable()`.", "style": "conventional", "type": "refactor", "scope": "cmdatom", "breaking": false} {"repo": "nestjs", "sha": "bcc0428072caefa404c55cd13a0f0a7fac4253ee", "date": "2026-08-17T13:59:26+08:00", "subject": "docs(changelog): point sample commit scope at sample/", "body": "", "style": "conventional", "type": "docs", "scope": "changelog", "breaking": false} {"repo": "nestjs", "sha": "f5bf4ddeabd46f43af1f79d0dd8045aaaf2e6ab6", "date": "2026-08-17T14:01:34+08:00", "subject": "docs(core): update Mercure SSE header comment to current path", "body": "", "style": "conventional", "type": "docs", "scope": "core", "breaking": false} {"repo": "vue-core", "sha": "4b1d20c7c16e2d9b2642036676bd4822d44934e0", "date": "2026-08-17T14:06:07+08:00", "subject": "fix(runtime-vapor): reconcile root bindings with fallthrough attrs", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "5406d2627c07dc2ed3d77650c00c97afdd2d9502", "date": "2026-08-17T08:25:54+02:00", "subject": "urldata: drop four strings from the aptr struct", "body": "- 'rtsp_transport'\n\n There is no need to keep this data around once the RTSP request has been\n issued.\n\n- 'accept_encoding'\n\n Only needed when creating the RTSP request.\n\n- 'uagent'\n\n The header is generated on demand from set.str[STRING_USERAGENT]\n\n- 'ref'\n\n Not necessary to keep around\n\nCloses #22603", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "ec34ff52cc3aeddac6d68c92ede60d05ea261cf3", "date": "2026-08-17T06:45:16Z", "subject": "Do not copy too many positions when deleting, GitHub issue 5478 from Uzair Aftab.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "5c1c6c609718303202832f706884192e1f1e9223", "date": "2026-08-17T14:51:24+08:00", "subject": "refactor: use named regex constants over inline literals", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "d24b4c12befcd0cf7917c421392aee7314768877", "date": "2026-08-17T14:54:50+08:00", "subject": "fix(runtime-vapor): separate live and inherited scope ids", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "nestjs", "sha": "35a79dad348b3d7a1974d53339da8d46e843911f", "date": "2026-08-17T08:54:57+02:00", "subject": "chore: upgrade vitest", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "tmux", "sha": "a0d3585c6c5200ebce6e48b1dee06cc4b5467ff5", "date": "2026-08-17T07:04:45Z", "subject": "Attach window to session earlier so resize cannot cause customize mode to blow up. Reported by Marcel Partap.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0ac8138ce388624641b495386b61dba75a1af5c9", "date": "2026-08-17T08:06:09+01:00", "subject": "Add a break-pane test for customize mode.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "51c43a5dbd86ad8461544e54c00bd9f487abfacd", "date": "2026-08-17T16:15:45+09:00", "subject": "Use safer allocation routines in more places of the tree", "body": "This is part of a more global effort to modernize the code tree with\nsafer allocation functions, which offer more protection in terms of:\n- stronger type safety guarantees\n- size overflow checks, particulary for arrays.\n\nThis set of updates includes allocation changes based on\npalloc_object(), palloc_array() and repalloc_array(). This work is\nsimilar to 1b105f9472bd, expanding more the new APIs in the tree in an\nincremental fashion.\n\nAuthor: Tristan Partin \nDiscussion: https://postgr.es/m/DKJB5887LGX9.359DV1UMZ1URK@partin.io", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "5ec2e5823659785ff5f34e690b9d2cb20e10e0d1", "date": "2026-08-17T09:22:47+02:00", "subject": "Update io-adapter.ts", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "11a07c086ab266e425346cd8663d5d475a94ba5b", "date": "2026-08-17T09:24:37+02:00", "subject": "fix(Anchor): avoid duplicate onChange with getCurrentAnchor", "body": "* fix(Anchor): avoid duplicate onChange with getCurrentAnchor\n\n* fix(Anchor): avoid remapping getCurrentAnchor results\n\n* fix(Anchor): preserve raw click identity\n\n* test(Anchor): use named demo props interface\n\n---------", "style": "conventional", "type": "fix", "scope": "Anchor", "breaking": false} {"repo": "ant-design", "sha": "c452c62a1fb7bc6f8f533b3653304c196aa76bc0", "date": "2026-08-17T15:25:44+08:00", "subject": "docs: clarify Typography editable onChange trigger description", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "ef96e75461248b745b2b08d3e5f5018009cb0e14", "date": "2026-08-17T15:27:26+08:00", "subject": "test(Tree): restore line demo accessibility coverage", "body": "", "style": "conventional", "type": "test", "scope": "Tree", "breaking": false} {"repo": "ant-design", "sha": "a8443b72e88963365510e59fc8536914e6b84a16", "date": "2026-08-17T15:28:07+08:00", "subject": "fix(Upload): preserve new files during async removal", "body": "* fix(upload): preserve files added during async removal\n\n* refactor(Upload): read latest file list with useEvent", "style": "conventional", "type": "fix", "scope": "Upload", "breaking": false} {"repo": "ant-design", "sha": "164a310f742f801c49b3ff748b2823d5596104bd", "date": "2026-08-17T15:30:45+08:00", "subject": "docs: update changelog for 6.6.1", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "tmux", "sha": "d384557b636e51bd8b2cab63b52c2a748ee5a134", "date": "2026-08-17T07:33:55Z", "subject": "Remove flags from both keys for menus, GitHub issue 5484.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "d1e0b1d70551f2460afca477b79ec41794ef21b7", "date": "2026-08-17T15:34:55+08:00", "subject": "feat(Space): support stretch alignment", "body": "", "style": "conventional", "type": "feat", "scope": "Space", "breaking": false} {"repo": "angular", "sha": "079a846263599daf38cc1903c46ccea99c5758f8", "date": "2026-08-17T09:38:45+02:00", "subject": "fix(compiler-cli): Produce correct tcb expression for optional chaining", "body": "The semantics of optional chaining changes when there are extra parenthesis. We need to make sure that we do not introduce some unnecessary ones.\n\nfixes #70143\n\n(cherry picked from commit e9ba39d671a3645ef5ca8233c96bacc94f02d155)", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "e9ba39d671a3645ef5ca8233c96bacc94f02d155", "date": "2026-08-17T09:38:45+02:00", "subject": "fix(compiler-cli): Produce correct tcb expression for optional chaining", "body": "The semantics of optional chaining changes when there are extra parenthesis. We need to make sure that we do not introduce some unnecessary ones.\n\nfixes #70143", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "vitest", "sha": "438eda13f6106604bb37c5a5fb1ffe5294020a65", "date": "2026-08-17T16:40:12+09:00", "subject": "fix(ui): fix code editor error and annotations gadgets layout shifts", "body": "", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "tmux", "sha": "c0e38f3748b2162aaa8722d3e55f32f5175a64be", "date": "2026-08-17T08:42:29+01:00", "subject": "Cleanup and simplification.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "5a6fefb427bd78f27c8d530974c03cccaa4cfef9", "date": "2026-08-17T08:48:08+01:00", "subject": "image: simplify geometry cleanup", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "9fdd04ae61db00dbf754289ecceb70a84af4e8c1", "date": "2026-08-17T07:52:16Z", "subject": "Add default terminal features for Rio, GitHub issue 5489 from Raphael Amorim.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "1729bb8e8af4a1deddd772761e8c058092cd4cfe", "date": "2026-08-17T07:56:56Z", "subject": "If writing a file fails, propagate the error to the server via a new message. Use a client flag rather than bumping the protocol version. GitHub issue 5451.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "036557393f368c7d91badcb06e57513975776884", "date": "2026-08-17T15:58:34+08:00", "subject": "test(Badge): enable button-name accessibility checks", "body": "Label the two unnamed demo switches and restore axe button-name enforcement across every Badge demo.\n\nAI-assisted: Codex", "style": "conventional", "type": "test", "scope": "Badge", "breaking": false} {"repo": "ant-design", "sha": "fe3c108262e54966df6b2df369d76baeb1ecad1f", "date": "2026-08-17T16:12:53+08:00", "subject": "fix(Table): avoid duplicate filter close callbacks", "body": "", "style": "conventional", "type": "fix", "scope": "Table", "breaking": false} {"repo": "vitest", "sha": "78a82b0792777a343e42abbe68fa1d124a56f07a", "date": "2026-08-17T12:00:11+03:00", "subject": "test: add module diagnostic coverage", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vitest", "sha": "30c21e6681215877fbe64e6ae6ee6e3c0ba88cc1", "date": "2026-08-17T18:01:30+09:00", "subject": "fix(ui): synchronize watch-run UI state on file removal", "body": "", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "vitest", "sha": "9851dbc41c286a30abfb6b29cce65f3e5b7b40a1", "date": "2026-08-17T18:02:24+09:00", "subject": "fix(browser): trigger playwright/chromium gc on lower disk availability [backport to v4]", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "vitest", "sha": "a8aa1dfc970896a27e68ddaf444a5fea182598a5", "date": "2026-08-17T11:03:11+02:00", "subject": "chore: update agents file", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "f7c5bde90efdf65d2576709c8ebc65ce35af1369", "date": "2026-08-17T17:06:39+08:00", "subject": "fix(Table): improve row selection accessibility", "body": "", "style": "conventional", "type": "fix", "scope": "Table", "breaking": false} {"repo": "vitest", "sha": "c496c2e333f5b06f8ec74bff83697dc6eb851c7f", "date": "2026-08-17T11:08:34+02:00", "subject": "fix(browser): fail instead of hanging when the browser stops responding", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "vitest", "sha": "67e8485645a5e272ab0044497b57212cc3486646", "date": "2026-08-17T11:09:00+02:00", "subject": "fix(browser): exit gracefully when the browser disconnects during cancellation", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "vitest", "sha": "210ba00cbb399bbc15791ab43e7826990df86aa9", "date": "2026-08-17T12:32:27+03:00", "subject": "fix: preserve correct root during resolve config", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "neovim", "sha": "5f07fc91c23196cac1ff39961d8f60f0dbd57749", "date": "2026-08-17T05:37:06-04:00", "subject": "refactor(cmdatom): drop RedoBuf #41351", "body": "Problem:\n`RedoBuf` is mostly indirection. It has a mild benefit as an \"ownership\"\nsignal but it counteracts the general goal of unifying how \"redo state\"\nis passed throughout the system, tends to sprout redundant interfaces,\nand reduces clarity.\n\nSolution:\nAdd `CmdSpec.body` to hold the \"prefixless\" key sequence.\nReuse `CmdSpec` to represent a \"redo\" buf.", "style": "conventional", "type": "refactor", "scope": "cmdatom", "breaking": false} {"repo": "vitest", "sha": "5622f0959bcbccb9fe4757dc3bb694153998be65", "date": "2026-08-17T12:19:30+02:00", "subject": "fix(ui): require auth for coverage report and all UI subtree requests", "body": "", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "angular", "sha": "01fa2431e469532642eb9aaf428498c651584156", "date": "2026-08-17T12:21:02+02:00", "subject": "ci: update github nick", "body": "ethan changed his nick, this was causing some pullapproves errors when a pr hit the primitives-shared scope.\n\n(cherry picked from commit 28f323a57810ae4da3fafab150f1ef355c8f44a6)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "28f323a57810ae4da3fafab150f1ef355c8f44a6", "date": "2026-08-17T12:21:02+02:00", "subject": "ci: update github nick", "body": "ethan changed his nick, this was causing some pullapproves errors when a pr hit the primitives-shared scope.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vitest", "sha": "59a037354f9213f955fbb4592656f71af676f793", "date": "2026-08-17T12:30:34+02:00", "subject": "fix(pool): report the worker exit instead of a pipe error that raced it", "body": "", "style": "conventional", "type": "fix", "scope": "pool", "breaking": false} {"repo": "curl", "sha": "c437d28c76485563b2570ee48ee2992e68ae7323", "date": "2026-08-17T13:08:42+02:00", "subject": "urldata: cleanups", "body": "- sockindex: int -> int8_t\n- move connectdata members around for fun and profit\n- remove data->state.current_speed as unused\n- remove data->state.trailers_bytes_sent as unsed\n- remove data->state.trailers_buf as unsed\n\nThe sockindex type change spread a little.\n\nCloses #22604", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "722362d87bad5df80ba76dbf1ccd578ec7b19e75", "date": "2026-08-17T13:29:34+02:00", "subject": "scorecard: fix argument handling when count and parallel are both 1", "body": "Closes #22605", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "8ff9b9a9efca7c6cfd5243569440de8d7a33aec4", "date": "2026-08-17T13:41:20+02:00", "subject": "fix(mocker): restrict redirect mocks to the fs allowlist", "body": "", "style": "conventional", "type": "fix", "scope": "mocker", "breaking": false} {"repo": "vitest", "sha": "fe5a11d3ceac5ec10d6d7d21a46d4caca132c48f", "date": "2026-08-17T13:41:36+02:00", "subject": "fix(mocker): restrict redirect mocks to the fs allowlist [backport to v4]", "body": "", "style": "conventional", "type": "fix", "scope": "mocker", "breaking": false} {"repo": "neovim", "sha": "98053bfecf35901c041acee349e744ce9e951edf", "date": "2026-08-17T07:55:47-04:00", "subject": "perf(input): drop buffheader_T #41353", "body": "Problem:\n`buffheader_T` is a linked list (Vim's favorite data structure) with\ncomplex bookkeeping, optimized for ancient hardware:\n\n1. Memory topology: On old platforms (Amiga 512KB without MMU, MS-DOS\n 64KB segments), large contiguous allocations were hazardous (no VM).\n2. Never-move appends: `add_buff()` fills spare space or links a new\n block.\n3. OOM \"recovery\": a failed block allocation only loses one append.\n\nTo avoid OOM, it prefers to allocate small, linked chunks, instead of\nresizing one continguous slice. Each stuff/drain cycle costs\na malloc+free.\n\nSolution:\nReplace Vim's favorite data structure with Nvim's favorite data structure.\n\nNvim doesn't have granular handling of OOM (`xmalloc`), and hardware has\nchanged: caches favor contiguous memory, allocators handle\nfragmentation. And these buffers are ~kb scale, so OOM is irrelevant on\nany system that can run Nvim.\n\n- Use a flat `StringBuilder` + `read`/`insert` offsets.\n- Keeps capacity (does not shrink) until `free_buff`.\n- \"Steady state\" allocates nothing: 1 fewer malloc+free per dot-repeat.", "style": "conventional", "type": "perf", "scope": "input", "breaking": false} {"repo": "neovim", "sha": "a4aa0417cf5553f8c154d5f6c92ab90a677542a3", "date": "2026-08-17T13:03:29+01:00", "subject": "feat(ui2): drop default enter-pager-via-CR mapping #40993", "body": "", "style": "conventional", "type": "feat", "scope": "ui2", "breaking": false} {"repo": "curl", "sha": "8e2eced23436c832a858802423b4ea8b45cf403a", "date": "2026-08-17T14:18:50+02:00", "subject": "multi: initialisation order", "body": "When creating a multi handle, global_init() may not have been\ninvoked that. There is a lazy check for this in easy_init()\nthat trigger global_init().\n\nCreate the admin handle for a multi first, before any further\noperations, so easy_init() can trigger a global_init() when\nnecessary.\n\nCloses #22606", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "aeb67209596883cbdce93b4ce179a26bc3382633", "date": "2026-08-17T15:10:43+02:00", "subject": "fix(browser): don't redirect vitest in client environment", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "vitest", "sha": "49e24fa38e263c818cd7e0f0f2102ca72864f953", "date": "2026-08-17T16:13:42+03:00", "subject": "fix(reporters): display `test.name` in GitHub Actions reporter summary header", "body": "", "style": "conventional", "type": "fix", "scope": "reporters", "breaking": false} {"repo": "angular", "sha": "6edbae2b17b829fa412b58c23d9cca56b8d63ada", "date": "2026-08-17T16:17:21+03:00", "subject": "feat(docs-infra): add Angie to the routing illustration", "body": "The routing header draws a serpentine route with map pins on it but\nnobody travelling it. Angie now stands on the top segment, the same way\nshe stands on the road in the roadmap header.\n\nShe needs more headroom than the canvas had, so the viewBox gains 12\nunits at the top and the illustration carries its own max-height, which\nkeeps the rest of the drawing at the size it rendered before.\n\n(cherry picked from commit 064530447e7834ac672a33c4dff612ea7dc2d46a)", "style": "conventional", "type": "feat", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "064530447e7834ac672a33c4dff612ea7dc2d46a", "date": "2026-08-17T16:17:21+03:00", "subject": "feat(docs-infra): add Angie to the routing illustration", "body": "The routing header draws a serpentine route with map pins on it but\nnobody travelling it. Angie now stands on the top segment, the same way\nshe stands on the road in the roadmap header.\n\nShe needs more headroom than the canvas had, so the viewBox gains 12\nunits at the top and the illustration carries its own max-height, which\nkeeps the rest of the drawing at the size it rendered before.", "style": "conventional", "type": "feat", "scope": "docs-infra", "breaking": false} {"repo": "vitest", "sha": "af83d1b1933b5d74c421d30849717369b828de0b", "date": "2026-08-17T15:17:32+02:00", "subject": "chore: release v5.0.0-rc.2", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "1d8f54823e2d0b1e4f071059ed0f14b03ad626a3", "date": "2026-08-17T15:45:01+02:00", "subject": "Happy Eyeballing v3: resolution delay of 25ms", "body": "After telemetry data from Mozilla reported the P99 DNS resolution\ndifferences between A and AAAA responses to by 21ms, reduce curl's\nresolution delay to 25ms.\n\nWhat could possibly go wrong?\n\nCloses #22612", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dbc66697a36a2c2b6c025c2fd6998a40f0882f9a", "date": "2026-08-17T10:18:58-04:00", "subject": "DTLS 1.3 Set the listener demo to non-blocking mode", "body": "A DTLS listener, and the connections it accepts, now default to\nblocking mode. The dtlslistenerecho demo drives the listener and its\nconnections with SSL_poll() and uses SSL_ACCEPT_CONNECTION_NO_BLOCK,\nso call SSL_set_blocking_mode() on the listener to put it back into\nnon-blocking mode. The accepted connections inherit that mode.\n\nAlso update the DTLS 1.3 guide (ossl-guide-dtlsv13) to mention\nSSL_set_blocking_mode() and SSL_get_blocking_mode() for DTLS\nlisteners and to note that a listener is blocking by default.\n\nReviewed-by: Tomas Mraz \nReviewed-by: Jakub Zelenka \nReviewed-by: Matt Caswell \nMerge-date: Thu Aug 20 09:50:38 2026\nMerged-from: https://github.com/openssl/openssl/pull/31983", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "81aa0a2fa7c4d35511e68daf1ba47bdfa3ec5474", "date": "2026-08-17T16:36:40+02:00", "subject": "Replace printf format %i by %d", "body": "as is PostgreSQL standard", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "a18d4e00679e02d464cd284a9cb6a1453d8a3b26", "date": "2026-08-17T14:47:41Z", "subject": "Allow features to be disabled using @ suffix in terminal-features, from Michael Grant.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a01a24deaf7dee72e2a22f0d0f0a1e9db08f92aa", "date": "2026-08-17T18:43:28+02:00", "subject": "curlx_inet_pton: drop setting `errno` on error", "body": "Follow-up to bb406386d9e088873f5f640c55e14a69390abd01 #22229\n\nCloses #22607", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0499c6ccf5ce510ad8c57211ee8c9414e92c71b7", "date": "2026-08-17T12:16:03-05:00", "subject": "Remove extract_autovac_opts().", "body": "extract_autovac_opts() returns a palloc'd copy of just the\nAutoVacOpts portion of a relation's reloptions, but upcoming work\nneeds the other StdRdOptions fields, too. Remove it in favor of\ncalling extractRelOptions() directly. av_relation now caches the\nmain table's whole StdRdOptions.\n\nReviewed-by: Michael Paquier \nReviewed-by: Sami Imseih \nTested-by: Greg Burd \nTested-by: solai v \nDiscussion: https://postgr.es/m/aFRxC1W_kZU9OjJ9%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "b107154ba243c12c076a5996058c606b276de251", "date": "2026-08-17T13:37:08-04:00", "subject": "docs: misc, cwd, vimscript.txt #41356", "body": "Extract vimscript.txt from repeat.txt", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f8c7d65c6cea5710acf55e769ee37988c416217a", "date": "2026-08-17T13:50:37-04:00", "subject": "Fix time_t / integer compares in tls13tickettest", "body": "Coverity flagged a few instances of the above test calling\nSSL_SESSION_set_time_ex and casting the result to an int for comparison,\nbut ints are smaller than time_t's so we may get faulty comparisons.\n\nFix it by using the proper TEST_time_t_gt check instead of TEST_int_gt\n\nFixes https://scan5.scan.coverity.com/#/project-view/60762/10222?selectedIssue=1700237\nFixes https://scan5.scan.coverity.com/#/project-view/60762/10222?selectedIssue=1700236\nFixes https://scan5.scan.coverity.com/#/project-view/60762/10222?selectedIssue=1700235\nFixes https://scan5.scan.coverity.com/#/project-view/60762/10222?selectedIssue=1700234\n\nReviewed-by: Daniel Kubec \nReviewed-by: Bob Beck \nReviewed-by: Paul Dale \nReviewed-by: Andrew Dinh \nMergeDate: Wed Aug 19 03:31:50 2026\n(Merged from https://github.com/openssl/openssl/pull/32410)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "3d8f71a370f0ad93762f140f6e69c2abe05b229a", "date": "2026-08-17T13:17:12-05:00", "subject": "Make autovacuum_enabled a ternary reloption.", "body": "This commit reimplements autovacuum_enabled as a ternary, using the\nsupport added in commit 4d6a66f675 and following the example of\nvacuum_truncate. This changes only the internal representation: an\nunset value still behaves as enabled, and the option accepts the\nsame input as before.\n\nThis is preparatory work for a follow-up commit that will make use\nof the new \"unset\" state.\n\nReviewed-by: Michael Paquier \nReviewed-by: Greg Burd \nReviewed-by: Sami Imseih \nTested-by: solai v \nDiscussion: https://postgr.es/m/aFRxC1W_kZU9OjJ9%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "af659efcb3a15dad690e9c9666ed82ed350e5522", "date": "2026-08-17T15:06:02-04:00", "subject": "Make plperl's handling of Perl arrays safer and more consistent.", "body": "plperl_func_handler()'s stanza for handling an arrayref result in\na SETOF function could loop forever (or at least till OOM) when\ngiven a tied array, since av_fetch won't necessarily ever return\na null pointer in that case. Be consistent with the other places\nwhere we traverse a perl array: call av_len() once and use len+1\nas the loop limit, silently ignoring any null pointers we get back\nfrom that range of subscripts.\n\nBut actually, Perl's preferred locution for this seems to be to\nuse av_count() not av_len()+1. av_count() seems better since\nthere's less risk of forgetting to add 1. Also, both of those\nfunctions return Size_t (or SSize_t) not int, creating at least\na theoretical overflow hazard. While we're modernizing this,\nlet's use the correct variable type where we can, and include an\noverflow check where we can't.\n\nReported-by: Claude Code (via Noah Misch)\nAuthor: Tom Lane \nReviewed-by: Andrey Rachitskiy \nDiscussion: https://postgr.es/m/569769.1786901901@sss.pgh.pa.us\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c66e7531e4141d5fa4accaeafaec4e7c2f638da0", "date": "2026-08-17T14:41:59-05:00", "subject": "Add an \"unset\" value for vacuum_index_cleanup.", "body": "This commit adds a new value to StdRdOptIndexCleanup to distinguish\nwhether it is explicitly set, similar to ViewOptCheckOption's\nVIEW_OPTION_CHECK_OPTION_NOT_SET. This changes only the internal\nrepresentation; an unset value still defaults to AUTO, and the\noption accepts the same input as before.\n\nThis is preparatory work for a follow-up commit that will make use\nof the new \"unset\" state.\n\nReviewed-by: Michael Paquier \nReviewed-by: Sami Imseih \nTested-by: Greg Burd \nTested-by: solai v \nDiscussion: https://postgr.es/m/aFRxC1W_kZU9OjJ9%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "b2de803871e579565a6f79296f386115142593ed", "date": "2026-08-17T20:04:00Z", "subject": "Flush output before ending sync. GitHub issue 5495 from xiangzhedev at gmail dot com.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "30abbe36a818632a6099347fcd20be35a3d53785", "date": "2026-08-17T21:04:26+01:00", "subject": "Regress for sync, from xiangzhedev@gmail.com.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "db194ced5b41b35107e58933c77d02a22d2ccb9a", "date": "2026-08-17T20:14:31Z", "subject": "Log unknown message types.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ed6775db84db7d28c787bbe4b9e8989f3e8f107a", "date": "2026-08-17T22:40:22+02:00", "subject": "Make data checksums launcher cancel its worker at SIGINT", "body": "Make sure the launcher cancels the currently running worker by calling\nTerminateBackgroundWorker when it receives SIGINT. In order to handle\ncases where SIGINT arrives while tje launcher is waiting for a worker\nto start or exit, implement a version of WaitForBackgroundWorkerStartup\nand WaitForBackgroundWorkerShutdown which checks the abort_requested\nsignalling. A new test which kills processing with SIGINT is added.\n\nBackpatch to v19 where online checksums were introduced.\n\nAuthor: Fujii Masao \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CAHGQGwEQ1-+iPQnUpTXYiHmzSz9ufFVkOK4kL_uyTdYt7jgg0Q@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "dd13dd39dbd3eec00ec1bff857b14bbc77fff86f", "date": "2026-08-17T22:40:26+02:00", "subject": "Reorder function prototypes to match definition order", "body": "While purely aesthetic, it makes reading the code easier to rearrange\nbefore the feature has shipped since doing it after risk backpatching\nconflicts. Also add missing prototypes.\n\nBackpatch to v19 where online checksums were introduced.\n\nAuthor: Fujii Masao \nDiscussion: https://postgr.es/m/CAHGQGwEQ1-+iPQnUpTXYiHmzSz9ufFVkOK4kL_uyTdYt7jgg0Q@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "9d8aa3a829c5782d7af2c2bd1ee5824d947c43b0", "date": "2026-08-17T15:58:11-05:00", "subject": "refactor(compiler-cli): add error guide links to diagnostics", "body": "Add the error guide URL when a compiler diagnostic uses a negative\nmarked error code.\n\n(cherry picked from commit 45ebb127e32182e2635cc441c56a7adb94aca8b8)", "style": "conventional", "type": "refactor", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "45ebb127e32182e2635cc441c56a7adb94aca8b8", "date": "2026-08-17T15:58:11-05:00", "subject": "refactor(compiler-cli): add error guide links to diagnostics", "body": "Add the error guide URL when a compiler diagnostic uses a negative\nmarked error code.", "style": "conventional", "type": "refactor", "scope": "compiler-cli", "breaking": false} {"repo": "openssl", "sha": "53287d05507e3e135b8b6ec0709a99029bc0b56c", "date": "2026-08-17T23:01:20+02:00", "subject": "Add OSSL_STORE store regression tests", "body": "Exercise allocation failures across OSSL_STORE_attach() and\nOSSL_STORE_load().\n\nAlso test that rejected loader parameters make OSSL_STORE_attach()\nreturn NULL instead of returning a context containing a dangling\nloader context.\n\nAssisted-by: ChatGPT:gpt-5.6\n\nReviewed-by: Jakub Zelenka \nReviewed-by: Neil Horman \nMergeDate: Tue Sep 15 14:20:06 2026\n(Merged from https://github.com/openssl/openssl/pull/32413)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f7cc6fd76007c5a051109cf8bbd7e1d391bd45a8", "date": "2026-08-17T17:31:12-04:00", "subject": "GiST: Deprecate F_TUPLES_DELETED opaque area flag.", "body": "This flag hasn't been useful since the removal of old-style VACUUM FULL,\nso remove it now (this includes removing vestigial code that\nunnecessarily set the flag).\n\nAlso add test coverage of gistprunepage(). Had that test case been\navailable before now, the issue fixed by this commit would have been\ndetected by wal_consistency_checking buildfarm animals.\n\nAuthor: Peter Geoghegan \nReviewed-by: Michael Paquiër \nDiscussion: https://postgr.es/m/CAH2-WznbTsQCrjmd=eSawfPqcxCjSFUkk6Qzd3z+gpNte5i03Q@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "3ddcb1a10168ead616e523bbe044f4c39ada4cca", "date": "2026-08-17T16:59:50-05:00", "subject": "fix(platform-browser): disallow event handler attributes in Meta", "body": "Prevent arbitrary MetaDefinition properties from writing on* handlers directly to meta elements. Browser events can execute these handlers, including on meta elements rendered in the document body.\n\n(cherry picked from commit 6f9a6bea50944fff894eb5629075b13f8c3090d7)", "style": "conventional", "type": "fix", "scope": "platform-browser", "breaking": false} {"repo": "angular", "sha": "6f9a6bea50944fff894eb5629075b13f8c3090d7", "date": "2026-08-17T16:59:50-05:00", "subject": "fix(platform-browser): disallow event handler attributes in Meta", "body": "Prevent arbitrary MetaDefinition properties from writing on* handlers directly to meta elements. Browser events can execute these handlers, including on meta elements rendered in the document body.", "style": "conventional", "type": "fix", "scope": "platform-browser", "breaking": false} {"repo": "postgres", "sha": "05556bdf938db0f4b87f3852cd5802fb71a13263", "date": "2026-08-17T18:09:07-04:00", "subject": "Tighten up tsqueryrecv().", "body": "tsqueryrecv() accepted zero-length lexemes, which tsqueryin() doesn't.\nIt also accepted phrase distance values larger than MAXENTRYPOS,\nwhich tsqueryin() doesn't. While neither of these omissions are\nvery harmful in themselves, they do allow accepting tsquery values\nthat will fail in a subsequent textual dump/reload.\n\nCommit 23d9ad771 performed similar tightening of tsvectorrecv(),\nbut I left off these changes at the time because they didn't seem\nto have security implications.\n\nReported-by: Claude Code (via Noah Misch)\nAuthor: Tom Lane \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/455079.1786897319@sss.pgh.pa.us\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "cce1fa9dbcc134281e6d17f34d2f30cd51345184", "date": "2026-08-17T18:17:51-04:00", "subject": "Doc: clean up documentation about text search datatype limits.", "body": "textsearch.sgml neglected to mention that the MAXSTRPOS total-length\nlimit applies to tsquery as well as tsvector. It also claims that\nthere is a 32K limit on the total number of nodes in a tsquery, which\nis wrong. (I suspect that QueryOperator.left may once have been\nint16, which would give rise to such a limit. But it's uint32 now,\nso you'd hit the 1GB varlena limit well before overflowing that.)\n\nWhile at it, re-order the bullet points into an order that makes\nmore sense, to me anyway.\n\nReported-by: Claude Code (via Noah Misch)\nAuthor: Tom Lane \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/455079.1786897319@sss.pgh.pa.us\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "8a11d2380ee94c25e8ea390a1f60a6b9232e154b", "date": "2026-08-18T08:15:46+08:00", "subject": "docs: sync Transfer `local.notFoundContent` default value", "body": "* docs(transfer): sync `local.notFoundContent` default value\n\n* chore: rerun CI\n\n---------", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "27e2afb492cd553093c92ea917d24651b5a9ae16", "date": "2026-08-17T17:55:42-07:00", "subject": "pg_locale.c: Fix REL_18_STABLE for srclen < 0.", "body": "Ensure that srclen >= 0 before calling strlower_c(), etc.\n\nCommit 5f003855e7 backported strlower_c(), etc., to 18, but did not\naccount for the srclen < 0 behavior that existed in 18 (and was\nremoved later in version 19 with commit 6d22c67c3b).\n\nNo callers in core use the srclen < 0 behavior, but it's possible\n(though unlikely) that extensions do.\n\nThis applies to REL_18_STABLE only.\n\nDiscussion: https://postgr.es/m/v36ssaygf7grb3qzfsjhtdzi7kqd45ds56nyuf7gi5qjml4qbb@ezmfqzmhlrs2", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "1cbf7d63e14845541dcbbc8e393cfd82a0cb25b6", "date": "2026-08-18T10:30:29+08:00", "subject": "fix(create-vite): update moved eslint-react plugin links in generated README", "body": "", "style": "conventional", "type": "fix", "scope": "create-vite", "breaking": false} {"repo": "ant-design", "sha": "62833fbb5f4d6ddcc7e8df611de9757af2db2b35", "date": "2026-08-18T10:37:34+08:00", "subject": "docs: restore DatePicker showNow version", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "c6e0c4b2f0a63392bb34178da1761581be703915", "date": "2026-08-18T10:56:06+08:00", "subject": "docs: sync TimePicker API types", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "3b70fa6f3d32ac832ed9ddd566c11b05e575f2c7", "date": "2026-08-18T12:16:02+09:00", "subject": "Fix cross-type foreign keys in the batched fast-path FK check", "body": "ri_FastPathFlushArray() rechecked a concurrently updated PK tuple with a\nscan key it built itself, putting found_val, the key of the tuple it had\njust locked, into sk_argument, and passing that same slot to\nrecheck_matched_pk_tuple(). Both operands therefore came from the locked\ntuple, and since sk_argument is the operator's right-hand input, the PK\nvalue was read as an FK value. For a foreign key using a cross-type\nequality operator, such as a \"date\" primary key referenced by a \"timestamp\"\ncolumn, that compares days against microseconds, so the recheck always\nfailed and a batch that had to follow an update chain reported a violation\neven though the version it locked still had the key.\n\nRemove the recheck. For a same-type key it compared the tuple against\nitself and so never rejected anything, which was harmless only because the\nloop a few lines below does the real work: it already compares found_val,\nread after the chain has been followed, against every buffered FK value,\nwith the arguments in the order the operator expects. That makes the\nrecheck redundant as well as wrong. Detection is not weakened by dropping\nit, since the buffered value that led the scan to a tuple can be matched by\nno other row visible to our snapshot.\n\nri_FastPathProbeOne() passes its original scan key, with the FK value still\nin sk_argument, and was never affected; nor were single-row statements or\nmulti-column foreign keys, which go through it.\n\nAdd an isolation test covering the cross-type case, a permutation where\nthe key really does move away, and a same-type permutation that should\nbehave identically.\n\nReported-by: Peter Geoghegan \nCo-authored-by: Peter Geoghegan \nDiscussion: https://postgr.es/m/CAH2-WznQjX3GByh_Ju7unuzMcik_5PJ5D7i_=qhwk=gPEkhfVQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "efcc29c80b7694fe4e0df48d02b5f36bc4849b69", "date": "2026-08-18T11:27:14+08:00", "subject": "test(Card): enable button-name accessibility checks", "body": "Label the loading demo switch and restore axe button-name enforcement across every Card demo.\n\nAI-assisted: Codex", "style": "conventional", "type": "test", "scope": "Card", "breaking": false} {"repo": "postgres", "sha": "92e20d4a4382810a30bdb394473d6ee891edab73", "date": "2026-08-18T13:44:04+09:00", "subject": "hstore: Rework module to use Size and {add,mul}_size for allocation sizes", "body": "hstore has been relying on a set of int/int32 to count the total length\nof all the keys and values constructed, relying on the maximum values of\nthe arguments rather than proper bound checks. As proved in the\nreported thread, these counters could overflow and wrap, leading to\nincorrect allocations.\n\nThis commit replaces the int/int32 values with a set of Size values,\nwith add_size() and mul_size() in charge of checking for overflows when\nconstructing hstore values before doing any allocation.\n\nThe changes that matter in terms of the report are mostly in\nhstore_io.c. hstore_compat.c is adjusted for consistency, to keep the\nallocation maths uniform across the module.\n\nThis is an old bug. A backpatch should perhaps be done but I cannot get\nexcited about a change like that in stable branches.\n\nReported-by: Man Zeng \nAuthor: Tender Wang \nAuthor: Michael Paquier \nDiscussion: https://postgr.es/m/CAHewXNmfD6-9MDxgpRfTzjVj0mhBw3JMoNOHeTiPOjZGhNzd=g@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "02cffa9e2d38d5d8f12e4043ee9d0f7abb1471e2", "date": "2026-08-18T10:26:25+05:30", "subject": "fix: reduce Windows 8.3-short-name detection false-positives", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "38c5ff6bfa67962b4f1724fd38c1526b2de67297", "date": "2026-08-17T22:21:27-07:00", "subject": "tests: fix Content-Length mismatch in test 2064", "body": "Closes #22611", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "e9235d91cc318f220016a5ebab4fd9601042b5f7", "date": "2026-08-18T13:37:10+08:00", "subject": "test(bundled-dev): enable worker playground specs", "body": "", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "vue-core", "sha": "f2b8e60c177f90b2ac98309e7f368e74db70b178", "date": "2026-08-18T13:45:49+08:00", "subject": "refactor(runtime-vapor): creation-time scope id model", "body": "Replace the post-render scope id walkers with a creation-time ambient\nmodel aligned with VDOM semantics. Scope ids are compile-time\nconstants: elements receive their attributes when created.", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "263c44a590a9640dc074516916bc4d4f410043ec", "date": "2026-08-18T14:02:50+08:00", "subject": "fix(Upload): forward focus events", "body": "Expose Upload onFocus at the component root so focus entering select, drag, and picture-card layouts reaches consumers exactly once.\n\nAI-assisted: Codex", "style": "conventional", "type": "fix", "scope": "Upload", "breaking": false} {"repo": "curl", "sha": "dbeb551a11d9ae199c7923235723650ec3b22306", "date": "2026-08-18T08:19:08+02:00", "subject": "CVE-2026-9546: sending old referer", "body": "Cherry-picked from 862e8a74a84478d (excluding the tests)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "10d79a86b3baaf9a07b7da87074a0208e6c7cbf8", "date": "2026-08-18T08:27:31+02:00", "subject": "CVE-2026-11586: WS Auto-PONG memory exhaustion", "body": "Cherry-picked from 849317ff5c5a5e13f50ec3d0", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "4fcfa07d9c71a1d3af9fac6a246306cd50f55d1e", "date": "2026-08-18T08:30:37+02:00", "subject": "CVE-2026-12064: proto-default skips SSH verification", "body": "Cherry-picked from ab3bb8cd8be8f9d4ac", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "db06f97899cc286cc37421d41b97f16a02fa8411", "date": "2026-08-18T15:33:02+09:00", "subject": "doc: Fix description of log_line_prefix %L in release notes", "body": "%L was described as showing in output the client IP, but it shows the\nlocal IP of the server connected to.\n\nAuthor: Keisuke Kuroda \nReviewed-by: Greg Sabino Mullane \nDiscussion: https://postgr.es/m/CANDwggK8AZJx+tGKY+h_8a90rqZ_+4dvGOFvDuRVVSjQvAafbg@mail.gmail.com\nBackpatch-through: 18, only", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "2799990eb01f0dac907356c3f102b7110283a0f2", "date": "2026-08-18T12:47:12+06:00", "subject": "docs: fix pnpm overrides link and instructions", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "afec5d7374cbdad951350be124175b59b70894ac", "date": "2026-08-18T15:48:33+09:00", "subject": "refactor(ui): remove explorer `queueMicrotask` logic", "body": "", "style": "conventional", "type": "refactor", "scope": "ui", "breaking": false} {"repo": "vitest", "sha": "4595184984b04d0475a12d2c07f5f4a80609e154", "date": "2026-08-18T15:49:50+09:00", "subject": "test(snapshot): restore `printBasicPrototype` test", "body": "", "style": "conventional", "type": "test", "scope": "snapshot", "breaking": false} {"repo": "vite", "sha": "3ba026dade4af56df08815310d3458fa110f5c5c", "date": "2026-08-18T15:51:12+09:00", "subject": "fix(bundled-dev): handle lazy request error", "body": "", "style": "conventional", "type": "fix", "scope": "bundled-dev", "breaking": false} {"repo": "vitest", "sha": "9123111b3787ed10c085117967742f4ab42b8bd5", "date": "2026-08-18T15:52:12+09:00", "subject": "fix(ui): improve suite interaction and layout in explorer tree", "body": "", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "vitest", "sha": "17646ac7294e02200082bc40225d3158d5e3451b", "date": "2026-08-18T15:59:56+09:00", "subject": "fix(snapshot): reject when `toMatchFileSnapshot` uses same snapshot file as `toMatchSnapshot`", "body": "", "style": "conventional", "type": "fix", "scope": "snapshot", "breaking": false} {"repo": "vite", "sha": "354adfecaee79bf23c60ba782d7ea3dec3a995d5", "date": "2026-08-18T08:01:08+01:00", "subject": "docs(cli): add experimentalBundle option", "body": "", "style": "conventional", "type": "docs", "scope": "cli", "breaking": false} {"repo": "tmux", "sha": "19a085a6727650ae660196a38e1a412f1b386615", "date": "2026-08-18T08:23:16+01:00", "subject": "Fix socket names for macOS.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "6238793ad71f0bc0a6062c6575d866c180b18d08", "date": "2026-08-18T07:24:49Z", "subject": "Restore umask if bind fails, from nachalsa at naver dot com.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "3dbddefaafc091a879b06f9279296f776691e455", "date": "2026-08-18T15:26:15+08:00", "subject": "fix(bundled-dev): hot update through circular imports instead of reloading", "body": "", "style": "conventional", "type": "fix", "scope": "bundled-dev", "breaking": false} {"repo": "vite", "sha": "7822e148733bca9b31d8b4fc035ee7ba256ae5d7", "date": "2026-08-18T15:26:15+08:00", "subject": "test(bundled-dev): enable hmr playground", "body": "", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "tmux", "sha": "72c5037f06625f8dff165ea42c660ee696e9015e", "date": "2026-08-18T07:32:09Z", "subject": "Correctly clear clients referring to a session with destroy-unattached, GitHub issue 5497 from Jeong, Heon.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "293ca1ca3c8f49afcda69356bee5689b1d552488", "date": "2026-08-18T00:36:00-07:00", "subject": "fix(runtime-vapor): align attrs fallthrough semantics with vdom", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vitest", "sha": "f5eee72b208d80ae5b58b64470831f1fb428774b", "date": "2026-08-18T09:40:36+02:00", "subject": "chore: add hidden message for llms", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "tmux", "sha": "4714f478fed87ae2c510ecd035aa89c5a90dedf8", "date": "2026-08-18T07:43:44Z", "subject": "When a pane is respawned, reset the control mode offsets. GitHub issue 5498 from Alex Rattray.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "da641e4e8d2b0547487e8c6f28aca0e5981d5a92", "date": "2026-08-18T08:45:11+01:00", "subject": "Add test for GitHub issue 5498 from Alex Rattray.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "be7157e7076a26263ec41637b7229ded3e697d6e", "date": "2026-08-18T00:55:42-07:00", "subject": "fix(runtime-vapor): branch-scope ownership for fallthrough attr effects", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "ef8379038774e8c59afc6df1d647d848f97dec60", "date": "2026-08-18T01:03:38-07:00", "subject": "fix(runtime-vapor): full fallthrough for functional components with declared props", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "tmux", "sha": "b170c192742761a83ec3c8e1490692414d298860", "date": "2026-08-18T08:05:05Z", "subject": "Change correct depth counter for source-file when used from hooks, GitHub issue 5437.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ec51b80b348fe34da15cfb991191ef0bdc4838c6", "date": "2026-08-18T17:07:02+09:00", "subject": "Fix RI fast-path race with REINDEX CONCURRENTLY", "body": "The RI fast path reads pg_constraint.conindid before taking\nRowShareLock on the referenced table. REINDEX CONCURRENTLY can\nrepoint the constraint and mark the old index dead, or drop it,\nbetween those operations. A backend in that window does not yet\nhold a relation lock, so it is not covered by REINDEX CONCURRENTLY's\nwaits for lockers.\n\nOpening an index that has already been dropped produces \"could not open\nrelation with OID\". Opening one that has only been marked dead can\nproduce wrong answers: the index is no longer maintained or vacuumed,\nso a scan can miss a referenced row or follow a stale entry to a reused\nheap line pointer.\n\nAfter locking the referenced table, reload the constraint and use its\ncurrent conindid. LockRelationOid() processes invalidation messages\nafter acquiring the lock, so the reload sees a committed index swap.\nIf the lock was already held, REINDEX CONCURRENTLY cannot mark the old\nindex dead or drop it until the transaction releases that lock, so\ncontinuing to use the old conindid is safe.\n\nDo this at both RI fast-path call sites. Add injection-point coverage\nfor old indexes that have either been dropped or marked dead.\n\nAuthor: Mihail Nikalayeu \nDiscussion: https://postgr.es/m/CADzfLwUJiVuv69uwuF5z4TrMhNkVwQUXW03q+uVNwmYFLtjEhw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "750c665ab38e7a5ff59fc6206740f82bbfab5fda", "date": "2026-08-18T01:16:35-07:00", "subject": "build(zig): keep all symbols for native Lua modules #41361", "body": "Native Lua modules loaded with `require` or `package.loadlib` resolve Lua\nC API symbols from the nvim executable itself. Those symbols must stay in\nthe binary even when nothing inside nvim references them.\n\n`rdynamic` is enough on Linux. On macOS the linker dead-strips\nunreferenced symbols before rdynamic exports them, so `_lua_tothread` is\nmissing from the executable and loading such a module fails with:\n\n symbol not found in flat namespace '_lua_tothread'\n\nDisable section garbage collection so all non-static symbols are kept.", "style": "conventional", "type": "build", "scope": "zig", "breaking": false} {"repo": "vue-core", "sha": "70cefe810039f930092f64e89fa222ab1b4cdc78", "date": "2026-08-18T01:20:48-07:00", "subject": "refactor(runtime-vapor): single-pass fallthrough root resolution", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vite", "sha": "c0f2fc607ee97ee4499337b04826420c00654065", "date": "2026-08-18T16:24:25+08:00", "subject": "fix(vite): update outdated upstream file links in license comments", "body": "", "style": "conventional", "type": "fix", "scope": "vite", "breaking": false} {"repo": "angular", "sha": "44137117b361f5bbad64e3d488e9aefd2f9ec5cb", "date": "2026-08-18T10:52:19+02:00", "subject": "fix(core): replace all hasOwnProperty usages with Object.hasOwn", "body": "We keep getting PRs that target single usages of `hasOwnProperty` and we have ~100 of them. These changes aim to address the issue centrally by swapping out all the instances and adding a lint rule against introducing new ones.\n\n(cherry picked from commit 732e5050180190df24768ace7aa3e785ca0dcc0a)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "732e5050180190df24768ace7aa3e785ca0dcc0a", "date": "2026-08-18T10:52:19+02:00", "subject": "fix(core): replace all hasOwnProperty usages with Object.hasOwn", "body": "We keep getting PRs that target single usages of `hasOwnProperty` and we have ~100 of them. These changes aim to address the issue centrally by swapping out all the instances and adding a lint rule against introducing new ones.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "e58994c30199258c7abfdfd65140a699faac0a1c", "date": "2026-08-18T10:52:24+02:00", "subject": "Fix some -Wcast-qual warnings", "body": "Small rearrangements to avoid having to cast away qualifiers. In some\ncases, the whole cast was unnecessary.\n\nReviewed-by: Heikki Linnakangas \nDiscussion: https://www.postgresql.org/message-id/flat/459217ca-3840-4452-8cd1-8f9b47e8d479%40eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "10f666e67c39cc8781acb647b4f58302ff173daa", "date": "2026-08-18T02:00:15-07:00", "subject": "fix(runtime-vapor): stop fallthrough resolution at slot outlets", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "tmux", "sha": "e94fcd96f9384808a404995b935d722fe3ed4358", "date": "2026-08-18T09:01:20Z", "subject": "Use DECRQM to detect mode 2026, from Michael Grant.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b8b1dc0a0d44e565054382d3eced64713992333e", "date": "2026-08-18T11:06:35+02:00", "subject": "config-win32.h: drop UWP, c-ares, simplify more", "body": "UWP requires VS2015+. Project files have no support for c-ares.\nDrop redundant guard, merge `#if` blocks.\n\nFollow-up to 1db93bdb7c1943132b13543539aaaf4866f31419 #22595\n\nCloses #22613", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d9b4fb1273fd4649593926451121d416154738b1", "date": "2026-08-18T10:16:16+01:00", "subject": "feat(lsp): fallback to textDocument/formatting from vim.lsp.formatexpr #40079", "body": "Problem:\nWith the addition of the `:help al` text object, you can now easily\nformat the whole buffer with `gqal`. However, `vim.lsp.formatexpr` only\nuses `textDocument/rangeFormatting` which some language servers (like\ngopls) don't support.\n\nSolution:\n- Fall back to `textDocument/formatting` if the whole buffer is being formatted\n and the server doesn't support `textDocument/rangeFormatting`.\n - In theory, these two methods should return the same response if the whole\n buffer is being formatted, but I preserved the existing behaviour of\n prioritising `textDocument/rangeFormatting` in case that does not hold (i.e.\n LS bug).\n- Also: `vim.lsp.formatexpr` had no tests at all, so actually add tests for it.", "style": "conventional", "type": "feat", "scope": "lsp", "breaking": false} {"repo": "curl", "sha": "b729003d837a3f0ab4a7606494724a490ba9b7ea", "date": "2026-08-18T11:17:20+02:00", "subject": "TODO: remove \"Support MQTTS\"", "body": "implemented\n\nFollow-up to 6c31df453b02597e90c27c2e0bc69bb114105005", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "64ea25cfe6e65c450ffc085adcaba8bb2db2fcc1", "date": "2026-08-18T11:22:30+02:00", "subject": "KNOWN_BUGS: remove old entries and NTLM related", "body": "- The ancient Negotiate ones are just outdated\n- NTLM is going away, no need to list the bugs anymore\n\nCloses #22615", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "fad263fe8b4b77076dad08637ef4fc1d6352a0bc", "date": "2026-08-18T11:25:52+02:00", "subject": "perf: initialize trace view in the orchestrator once per worker", "body": "", "style": "conventional", "type": "perf", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "5073eb5bcb9b5ea6b3371d63413536d5bc2bc26c", "date": "2026-08-18T02:40:33-07:00", "subject": "fix(runtime-vapor): fallthrough attrs onto interop vnode roots", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vitest", "sha": "66c1cbb53a1bf2a59888df2fa3c002958d21394c", "date": "2026-08-18T11:56:42+02:00", "subject": "fix: combine multiple --project filters correctly", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "e669179681044514099ca726dbaa55653528f4fb", "date": "2026-08-18T12:01:51+02:00", "subject": "runtests: introduce a subset option", "body": "To make it easier to run different subsets of the tests in separate\ninvokes.\n\nCloses #22616", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "79132a1daf8f3b2b6db101be386fa9e8cb3f73f5", "date": "2026-08-18T12:16:32+02:00", "subject": "GHA: use `--subset` runtests option, run random subset in riscv64 job", "body": "Also:\n- drop most hand-tuned `--min` options, in favor of defaults.\n- enable tests on emulated riscv64.\n Suggested-by: Dan Fandrich\n Ref: https://github.com/curl/curl/pull/22590#issuecomment-5303238102\n- drop building examples on emulated riscv64.\n\nAdding around 3.5 minutes to the riscv64 job:\n0.5 Perl install, 2 minutes to build tests, 1 minute to run 100 tests.\n\nFollow-up to 66486691ee1c5c768c74e402b8b95902b88d90fb #22619\nFollow-up to e669179681044514099ca726dbaa55653528f4fb #22616\nFollow-up to 58cb1e2f1fa8d8f1c71596918707e566c6353cb2 #22618\nFollow-up to d00000673d05fdf6ceee8941ccd63ad2906257e4 #22602\n\nCloses #22617", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "c5ea9ca2ad9e519ac662bd3e0846610264bf6e72", "date": "2026-08-18T06:23:59-04:00", "subject": "fix(zip): missed \"password:\" prompt if output trails it #41364", "body": "Problem:\nFlaky test:\n\n RUN T1158 nvim.zip reports an incorrect archive password: 11092.84 ms FAIL\n\nThe prompt is detected only if the pty output *ends* with \"password: \",\nbut a read may return the prompt plus following bytes.\n\nSolution:\n- Match anywhere in the output since the last password was sent. The\n buffer is cleared before each send, so won't match stale text.\n- Assert on the reported message, so a failure shows what was reported.", "style": "conventional", "type": "fix", "scope": "zip", "breaking": false} {"repo": "curl", "sha": "58cb1e2f1fa8d8f1c71596918707e566c6353cb2", "date": "2026-08-18T12:28:05+02:00", "subject": "runtests: allow resetting `CURL_TEST_MIN` via `--min=0`", "body": "Set the default to `CURL_TEST_MIN` if set, first. Then process the\ncommand-line options to override it. Previously, the env took over if\nthe command-line explicitly tried to set it to zero.\n\nRef: #22617\n\nFollow-up to 3f1cd809eeae05f39fec72fe780f3a69d21972fb #19942\n\nCloses #22618", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "9ef407a3b051e74f50dc7fb8914e2bd89b3e5e53", "date": "2026-08-18T19:56:35+09:00", "subject": "docs: use eslint.config.* wherever config file names are listed", "body": "The CLI help text and the missing-config error each listed only\neslint.config.js, eslint.config.mjs, and eslint.config.cjs, omitting\nthe TypeScript config files that `FLAT_CONFIG_FILENAMES` already\nincludes. The glob covers all six without enumerating them, and it\nmatches the `config` typedef in `lib/options.js`, which already uses\n`eslint.config.*`.\n\n`-c` now says it replaces the lookup rather than a single file, since\nmultiple default config files can apply in one run and `-c` disables\nthat lookup entirely.\n\nFixes #21111", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "tmux", "sha": "b094564203899688004f7bb1ae81505c0de3f40c", "date": "2026-08-18T12:03:19+01:00", "subject": "build: rename image support option", "body": "", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7864a14c4370b11cac74bcdf4cae5bb6a6ae867b", "date": "2026-08-18T14:24:21+03:00", "subject": "Add wait-for-lsn process-exit cleanup callback", "body": "WaitLSNCleanup() was called from ProcKill(), but not from\nAuxiliaryProcKill(), even though xlogwait.c sizes its shared memory to\ninclude NUM_AUXILIARY_PROCS and thus accepts calls from auxiliary\nprocesses. Such a process exiting while waiting would leave its entry in\nthe heap.\n\nRegister an on_shmem_exit callback lazily before a process enters a\nwait-for-lsn heap instead, as suggested by Noah: that keeps the cleanup\nlocal to xlogwait.c and makes it harder to miss a caller that needs it,\nrather than having to remember every process-kill path.\n\nReported-by: Noah Misch \nAuthor: Xuneng Zhou \nReviewed-by: Alexander Korotkov \nDiscussion: https://postgr.es/m/20260706012642.f9.noahmisch%40microsoft.com\nDiscussion: https://postgr.es/m/CABPTF7UtW_cAa%3DQh4RDfKiUqu3pJJE22ai9tbWJVERbeRyssLw%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d2eb714c7ae70afc1837da5db4393b0f50dcc77f", "date": "2026-08-18T14:25:03+03:00", "subject": "Fix WAIT FOR LSN documentation examples", "body": "Pad example LSNs to match pg_lsn_out() output, and use\n\"standby_replay LSN\" in the timeout error to match the server message.\n\nAuthor: Xuneng Zhou \nReviewed-by: Alexander Korotkov \nDiscussion: https://postgr.es/m/CABPTF7UtW_cAa%3DQh4RDfKiUqu3pJJE22ai9tbWJVERbeRyssLw%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "58c48b4ec42583c69b3a7e9c6cf7649b38fc8151", "date": "2026-08-18T14:25:53+03:00", "subject": "Avoid locking when an LSN waiter is already removed", "body": "WaitLSNWakeup() removes each selected waiter from its heap and clears\nits inHeap flag before setting its latch. When such a waiter later\ncalls deleteLSNWaiter(), it acquires WaitLSNLock exclusively only to\ndiscover that there is nothing left to remove. Waking many waiters can\ntherefore make them serialize on the lock for no useful work.\n\nCheck inHeap before acquiring WaitLSNLock. A lockless false value is\nconclusive because only the owning backend can change inHeap from false\nto true. A concurrent waker can only clear it. A stale true value\nfalls through to the existing recheck under the lock.\n\nWaitLSNCleanup() performed the same lockless check before calling\ndeleteLSNWaiter(). Drop it there, as it is now redundant.\n\nAuthor: Xuneng Zhou \nReviewed-by: Alexander Korotkov \nDiscussion: https://postgr.es/m/CABPTF7UtW_cAa%3DQh4RDfKiUqu3pJJE22ai9tbWJVERbeRyssLw%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f2d6392e199256dd4a37bebbc92a2d28fac45b5f", "date": "2026-08-18T14:26:38+03:00", "subject": "Clarify LSN waiter cleanup after wakeup", "body": "WaitLSNWakeup() can be called by several processes, not only the startup\nprocess. Update the cleanup comment to explain that another process may\nremove the waiter before waking it and that inHeap prevents double\ndeletion.\n\nAuthor: Xuneng Zhou \nReviewed-by: Alexander Korotkov \nDiscussion: https://postgr.es/m/CABPTF7UtW_cAa%3DQh4RDfKiUqu3pJJE22ai9tbWJVERbeRyssLw%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0ec36f4ba325062882eb7a36b8f815e85ff8c123", "date": "2026-08-18T14:27:23+03:00", "subject": "Re-register LSN waiters after stale wakeups", "body": "WaitLSNWakeup() removes a waiter from the heap before setting its latch.\nIf the position that caused the wakeup moves backwards before the waiter\nrechecks it, as can happen when WAL streaming restarts, the waiter may\nsleep again while no longer registered. Subsequent WAL progress then\ncannot wake it.\n\nWhen an unmet waiter finds that it is no longer in the heap, add it back\nand restart the loop. Rereading the position after registration also\nprevents missing an advance between the previous read and the re-add.\n\nProcess interrupts before re-registering rather than after, so that a\nwakeup which goes stale again cannot postpone cancellation, however often\nit repeats. As a side effect, a pending cancel now wins over an expired\ntimeout, which previously reported a timeout instead.\n\nAdd deterministic TAP coverage that simulates a stale standby_write\nwakeup without advancing the actual write or replay positions.\n\nAuthor: Xuneng Zhou \nReviewed-by: Alexander Korotkov \nDiscussion: https://postgr.es/m/CABPTF7UtW_cAa%3DQh4RDfKiUqu3pJJE22ai9tbWJVERbeRyssLw%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "66486691ee1c5c768c74e402b8b95902b88d90fb", "date": "2026-08-18T13:37:48+02:00", "subject": "runtests: divide minimum required tests by subsets", "body": "When the minimum comes from `CURL_TEST_MIN`, to improve the default\nbehavior. It may still need manual adjustment in cases.\n\nFollow-up to e669179681044514099ca726dbaa55653528f4fb #22616\nFollow-up to 58cb1e2f1fa8d8f1c71596918707e566c6353cb2 #22618\nCherry-picked from #22617\n\nCloses #22619", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f6bf814378c6e51aeecb8333040a78666b543392", "date": "2026-08-18T07:50:27-04:00", "subject": "fixfix(cmdatom): repeat Visual ; CmdFrame + stage #41355", "body": "Problem:\n- Dot-repeat of a Visual selection prepared by a `` mapping,\n results in E1255 and leaves Visual mode active.\n- Visual-mode capture is implemented as a second, parallel \"capture\n engine\": it re-composes a CmdSpec from cmdarg_T per key and decides\n replayability from its own table of \"void\" key classes.\n\nSolution:\n- Model nested `normal_execute()` as a `CmdFrame` stack,\n instead of a single module-scoped `stage`.\n - Capture `` \":norm …\" commands as subatoms from its nested frames.\n- Produce every command exactly once; atom_push_raw() routes the atom to\n the Visual composite while a selection is open, like it already does\n for mapping composites.\n - `v/patd` is now repeatable and cascades.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "c4365c7cd1fd5bfe97917f0a1f3d45fb529af98c", "date": "2026-08-18T19:55:05+08:00", "subject": "docs: align notFoundContent default values across components", "body": "Update default notFoundContent documentation to reflect actual Empty component defaults:\n- English: 'No data' (not 'Not Found')\n- Chinese: '暂无数据' (not 'Not Found')\n\nAffected components: Select, TreeSelect, Cascader, Mentions, Transfer\n\nAll these components use DefaultRenderEmpty which renders the Empty component with locale-aware defaults.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "1eae512285e7131f4bf48416f9d4b8b1124e223a", "date": "2026-08-18T08:30:42-04:00", "subject": "fix(lifecycle): Lua state not freed if free_all_mem() is re-entered #41366", "body": "Problem:\nLeak reported by ASAN CI, always at Lua state init.\n\n ERROR: LeakSanitizer: detected memory leaks\n strbuf_init src/cjson/strbuf.c:62:22\n json_create_config src/cjson/lua_cjson.c:635:5\n lua_cjson_new src/cjson/lua_cjson.c:2060:5\n nlua_state_add_stdlib src/nvim/lua/stdlib.c:840:3\n nlua_state_init src/nvim/lua/executor.c:973:3\n ...\n SUMMARY: AddressSanitizer: 1055 byte(s) leaked in 2 allocation(s).\n\nAnalysis:\n`getout()` may run during `free_all_mem()` (that's why it checks\n`!entered_free_all_mem`). If that happens, the second `os_exit()`\nreturns early from `free_all_mem()` then `exit()` is called and Lua\nstate is never closed.\n\nSolution:\nAlways call `nlua_free_all_mem()` in the early-return case.", "style": "conventional", "type": "fix", "scope": "lifecycle", "breaking": false} {"repo": "neovim", "sha": "ab82c2c6b087a7740e3e9bc1c02732d2ce3d5511", "date": "2026-08-18T20:35:22+08:00", "subject": "feat(extmark): virt_lines can highlight until EOL #41289", "body": "", "style": "conventional", "type": "feat", "scope": "extmark", "breaking": false} {"repo": "curl", "sha": "327a4bb95cd5fc98104023eb839a95fc7e330391", "date": "2026-08-18T14:45:40+02:00", "subject": "CVE-2025-14524: bearer token leak on cross-protocol redirect", "body": "Implemented differently that official fix", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "10a6960632d4299324aaac69b25a9d9652b1fa0d", "date": "2026-08-18T09:06:05-04:00", "subject": "test_json_parser: Fix broken file ref and inverted result check", "body": "A comma following a file ref in a perl print statement makes the\nstatement just print the file's GLOB rather than redirecting the\nfollowing arguments to the file. This meant the test was not testing any\ncontents and the logs were instead bulked up with what the test was\nsupposed to be testing.\n\nAlso, run_log() returns 1 for success, not 0, so the tests for the\nreturn values were wrong. (We were getting 0 because of the comma error\nabove.)\n\nBackpatch-thru: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "5146df80b1f6adb4af70dd06a0ada9e88250cfe9", "date": "2026-08-18T15:15:08+02:00", "subject": "fix: revive global concurrency limit for test lifecycle [backport to v4]", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "neovim", "sha": "0c091cedc291ee4dd13ccae192057d92878aee0e", "date": "2026-08-18T09:22:05-04:00", "subject": "test(tui): unreliable \"TUI exits immediately when stdin is closed\" #41367", "body": "Problem:\nUnreliable test on slow CI (ASAN/TSAN):\n\n FAILED .../tui_spec.lua @ 3054: TUI exits immediately when stdin is closed\n retry() attempts: 1\n Expected: vim.NIL\n Actual: { name = \"nvim\", pid = 33201, ppid = -1 }\n\nThe test asserts \"immediate\" exit of the Nvim process, but this may be\nsubject to OS delays outside of our control.\n\nSolution:\nMake the \"timed out waiting for DA1\" log conditional on the actual\ntimeout, and assert the logs in the test.", "style": "conventional", "type": "test", "scope": "tui", "breaking": false} {"repo": "postgres", "sha": "ffb7e3af79b307cf93a5a26e0ddc4e96f420befc", "date": "2026-08-18T10:43:33-04:00", "subject": "doc PG 19: update to current", "body": "Backpatch-through: 19 only", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "558153c1b597723b4bb9fdd60761efb3acd31805", "date": "2026-08-18T16:52:39+02:00", "subject": "CVE-2026-5773: wrong reuse of SMB connection", "body": "Refuse SMB connections to do connection reuse. Different fix than the\nofficial one.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f238788601eb068d6a95cf775ac4e85c407db2cd", "date": "2026-08-18T11:38:52-04:00", "subject": "fix(tui): ui may attempt reattach during shutdown #41368", "body": "Problem:\nTSan CI, always at exit:\n\n WARNING: ThreadSanitizer: use of an invalid mutex (e.g. uninitialized or destroyed)\n #1 uv_mutex_lock\n #2 tui_flush src/nvim/tui/tui.c:1708\n #3 ui_client_event_flush\n #4 parse_msgpack src/nvim/msgpack_rpc/channel.c:259\n ...\n #10 event_teardown src/nvim/main.c:187\n #11 os_exit src/nvim/main.c:720\n ...\n #19 tinput_done_event src/nvim/tui/input.c:184\n\nAnalysis (guess):\n`rpc_close_event()` may run during `event_teardown()` and reattach to\na restarted server, then `ui_client_attached=true` gets set and redraws\nresume in an already-stopped TUI.\n\nSolution:\nCheck `!exiting` in `rpc_close_event`, like we already do in\n`channel_proc_exit_cb()`.", "style": "conventional", "type": "fix", "scope": "tui", "breaking": false} {"repo": "neovim", "sha": "c184ca2c68abfc6864c7fc64d7b3c3e657afd7a8", "date": "2026-08-18T17:50:28+02:00", "subject": "fix(lifecycle): ignore signals during os_exit", "body": "Problem:\nSIGTERM received during `os_exit()` can cause it to \"nest\":\n\n os_exit src/nvim/main.c:720\n preserve_exit src/nvim/main.c:910\n deadly_signal src/nvim/os/signal.c:205\n on_signal src/nvim/os/signal.c:240\n signal_event src/nvim/event/signal.c:45\n multiqueue_process_events src/nvim/event/multiqueue.c:159\n event_teardown src/nvim/main.c:178\n os_exit src/nvim/main.c:720\n\nSolution:\nCall `signal_reject_deadly()` at the start of `os_exit()` like\n`preserve_exit()` already does.", "style": "conventional", "type": "fix", "scope": "lifecycle", "breaking": false} {"repo": "angular", "sha": "49f4e1573bd767358db2cee7cce8dc6ae00bc53c", "date": "2026-08-18T18:05:25+02:00", "subject": "ci: exclude markdown files from requiring dev-infra review", "body": "613c51e wasn't enough to exclude dev-infra\n\n(cherry picked from commit d364c83c44309fa2a1adfd6fb3ccd9f933e0472c)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "d364c83c44309fa2a1adfd6fb3ccd9f933e0472c", "date": "2026-08-18T18:05:25+02:00", "subject": "ci: exclude markdown files from requiring dev-infra review", "body": "613c51e wasn't enough to exclude dev-infra", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f6708d3f437d25c2c6855846d71ca235ae6f6b0b", "date": "2026-08-18T12:12:36-04:00", "subject": "DTLS SSL Listener default address validation", "body": "Replace SSL_LISTENER_FLAG_REQUIRE_HVR and SSL_LISTENER_FLAG_REQUIRE_HRR\nwith a single SSL_LISTENER_FLAG_ADDRESS_VALIDATION, and make address\nvalidation the default. A DTLS listener now validates (HVR for DTLS\n1.0/1.2, HRR cookie for DTLS 1.3) unless SSL_LISTENER_FLAG_NO_VALIDATE\nis set. If both NO_VALIDATE and ADDRESS_VALIDATION are given, the\nlistener fails safe and validates.\n\nUpdate the tests, the dtlslistenerecho demo, util/other.syms and the\nlistener documentation to match.\n\nAssisted-by: Claude:claude-opus-4-8\nReviewed-by: Tomas Mraz \nReviewed-by: Jakub Zelenka \nReviewed-by: Matt Caswell \nMerge-date: Thu Aug 20 09:50:39 2026\nMerged-from: https://github.com/openssl/openssl/pull/31983", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "391c378b7507979864fb9912389f5dfd8d5f9142", "date": "2026-08-18T11:14:31-07:00", "subject": "refactor(core): pass AbortSignal to WebMCP tools to handle cancellation", "body": "This forwards the `AbortSignal` natively supplied by Chrome 153.0.8009.0's update to `registerTool` executions down into the Angular context. This helps tools gracefully handle execution cancellations initiated by agents or users, preventing unnecessary background work and leaking resources.\n\nWe combine the injector/component destruction abort signals with the native WebMCP execution cancellation signals so the underlying tools correctly cancel long-running requests without producing invalid states.\n\nJustification:\nhttps://groups.google.com/a/chromium.org/g/chrome-ai-dev-preview/c/9291sjhIRz0/m/RFuyRrs5AAAJ\n(cherry picked from commit e041f9483e577cbc8cf32afe8aaebdef0903e082)", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "e041f9483e577cbc8cf32afe8aaebdef0903e082", "date": "2026-08-18T11:14:31-07:00", "subject": "refactor(core): pass AbortSignal to WebMCP tools to handle cancellation", "body": "This forwards the `AbortSignal` natively supplied by Chrome 153.0.8009.0's update to `registerTool` executions down into the Angular context. This helps tools gracefully handle execution cancellations initiated by agents or users, preventing unnecessary background work and leaking resources.\n\nWe combine the injector/component destruction abort signals with the native WebMCP execution cancellation signals so the underlying tools correctly cancel long-running requests without producing invalid states.\n\nJustification:\nhttps://groups.google.com/a/chromium.org/g/chrome-ai-dev-preview/c/9291sjhIRz0/m/RFuyRrs5AAAJ", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "caa463e3f8d2b9c1441e6a59c50f90b8598439f3", "date": "2026-08-18T11:56:55-07:00", "subject": "Fix stream abort for a transaction that was never streamed.", "body": "Commit 072ee847ad4 taught logical decoding to discard the changes of a\ntransaction that is already known to be aborted when it is picked for\neviction. That path reuses ReorderBufferTruncateTXN(), which marks\nevery subtransaction that still has in-memory changes as\nstreamed. Since nothing is streamed in that path, and the top-level\ntransaction is never marked, a subtransaction ends up flagged as\nstreamed even though the output plugin has never seen it. Decoding the\nsubsequent abort record then makes ReorderBufferAbort() invoke the\nstream_abort callback for that subtransaction.\n\nFor pgoutput this sends a Stream Abort ('A') message to a subscriber\nthat requested streaming = off, and it does so regardless of the\nnegotiated protocol version, so even a client speaking a version that\npredates transaction streaming receives a message it cannot\nparse. test_decoding dereferences a NULL pointer and crashes, since it\nallocates its per-transaction state in the begin or stream start\ncallback, neither of which runs for a transaction discarded as\naborted.\n\nThis commit fixes this by marking a subtransaction as streamed only\nwhen it has changes and its top-level transaction is already marked as\nstreamed. All streaming call sites mark the top-level transaction\nbefore truncating it, so their behavior is unchanged, while the\nabort-discard path never marks the top-level transaction and therefore\nnow leaves its subtransactions unmarked.\n\nBackpatch to v18, where commit 072ee847ad4 was introduced.\n\nBug: #19616\nReported-by: Tyler Smart \nAuthor: Andrey Rachitskiy \nReviewed-by: Hayato Kuroda \nReviewed-by: Fujii Masao \nReviewed-by: Masahiko Sawada \nDiscussion: https://postgr.es/m/19616-f6153af509910853@postgresql.org\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "73e766f8f6ec248bd2bf023af32ef45f40d05d39", "date": "2026-08-18T21:30:05+02:00", "subject": "fix(lifecycle): json_create_config leak", "body": "Problem:\nAs it turns out, when you have 17 different ways to \"exit\", you need to\nplug 17 different holes.\n\nLeak still reported by ASAN CI, even after 1eae512285e7 f238788601eb.\n\n ERROR: LeakSanitizer: detected memory leaks\n strbuf_init src/cjson/strbuf.c:62:22\n json_create_config src/cjson/lua_cjson.c:635:5\n lua_cjson_new src/cjson/lua_cjson.c:2060:5\n nlua_state_add_stdlib src/nvim/lua/stdlib.c:840:3\n nlua_state_init src/nvim/lua/executor.c:973:3\n\nAnalysis:\n`os_breakcheck()` may process a `exit_event` placed by\n`exit_on_closed_chan()` during shutdown. This would re-enter\n`preserve_exit()`, which then skips everything because it sees\n`really_exiting=true`.\n\nSolution:\nSet `exiting` in `preserve_exit()`, not only in `getout()`.", "style": "conventional", "type": "fix", "scope": "lifecycle", "breaking": false} {"repo": "openssl", "sha": "c8ba7c89289074ca38b997b6559b664f52cd2387", "date": "2026-08-18T13:49:34-06:00", "subject": "Free connections before listeners in the radix test cleanup", "body": "A connection's assist thread reads from its network BIO, which for a\ndgram BIO pair belongs to the linked listener. Freeing a listener while\nthe connection's assist thread is still running is a use-after-free.\nFreeing connections first joins their assist threads, so no thread is\nreading when the listener BIOs are freed.\n\nFixes #32420\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Fri Aug 21 11:07:27 2026\nMerged-from: https://github.com/openssl/openssl/pull/32421", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6af84c944c89ceaf0acafdd792d479045cfeb413", "date": "2026-08-18T13:49:34-06:00", "subject": "Free connections before listeners in the radix test cleanup", "body": "A connection's assist thread reads from its network BIO, which for a\ndgram BIO pair belongs to the linked listener. Freeing a listener while\nthe connection's assist thread is still running is a use-after-free.\nFreeing connections first joins their assist threads, so no thread is\nreading when the listener BIOs are freed.\n\nFixes #32420\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Fri Aug 21 11:07:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/32421", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a6351c1dd7626d89eff618301de8d1f10188bcc3", "date": "2026-08-18T13:49:34-06:00", "subject": "Free connections before listeners in the radix test cleanup", "body": "A connection's assist thread reads from its network BIO, which for a\ndgram BIO pair belongs to the linked listener. Freeing a listener while\nthe connection's assist thread is still running is a use-after-free.\nFreeing connections first joins their assist threads, so no thread is\nreading when the listener BIOs are freed.\n\nFixes #32420\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Fri Aug 21 11:07:04 2026\nMerged-from: https://github.com/openssl/openssl/pull/32421", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "31c32ca387a01b1134b459941d5111ab977b8197", "date": "2026-08-18T13:49:34-06:00", "subject": "Free connections before listeners in the radix test cleanup", "body": "A connection's assist thread reads from its network BIO, which for a\ndgram BIO pair belongs to the linked listener. Freeing a listener while\nthe connection's assist thread is still running is a use-after-free.\nFreeing connections first joins their assist threads, so no thread is\nreading when the listener BIOs are freed.\n\nFixes #32420\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Fri Aug 21 11:06:40 2026\nMerged-from: https://github.com/openssl/openssl/pull/32421", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "02a2137c1ae521c09050e3d1614dd5d79c4107db", "date": "2026-08-18T13:19:59-07:00", "subject": "pg_locale.c: comment improvements.", "body": "Add missing comments and fix outdated/incorrect comments in\npg_locale.c and related files.\n\nSuggested-by: Andres Freund \nDiscussion: https://postgr.es/m/v3nniwcrxejmcfvz56xbd22hphprqleuornd6hqkmw2bl7kgmz@cnytz2ee5ltk\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "aaf8b9989f7aaeb6594122caee1eea6c87e10655", "date": "2026-08-18T22:25:42+02:00", "subject": "Record initial state of data checksums in controlfile", "body": "The controlfile records the current state of data checksums, which\nalso used to be the initial state from initdb when checksums could\nnot be altered after initialization. pg_control_init is documented\nto return information about cluster initialization state, which it\nno longer will if data checksums have been changed either using the\noffline tool or with online processing.\n\nFix by adding a new field in the control file which tracks the init\nvalue of data checksums, and is left read only after initialization.\n\nWhile this is a regression dating back to when changing checksum\nstate was made possible offline with pg_checksums, it is a control\nfile change so it cannot be backpatched.\n\nBackpatch to v19 where online checksums were introduced.\n\nAuthor: Daniel Gustafsson \nReviewed-by: Bertrand Drouvot \nDiscussion: https://postgr.es/m/B87ABFBE-A304-4839-8706-C80D73E6BF5C@yesql.se\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "397f0fd06edc3a5b114449164c5d39c59a468b61", "date": "2026-08-18T22:25:49+02:00", "subject": "Add data_page_checksum_version to pg_control_checkpoint", "body": "Commit f19c0eccae added the data_checksum_version to the pg_controldata\noutput, but omitted a corresponding change to the pg_control_checkpoint\nSQL function, which reports the same checkpoint information. The field\nis named to match what pg_control_init already reports for consistency.\n\nThe integer version reported is an implementation detail which bleeds\nthrough, but it is quite widely used and a more holistic approach to\nimproving this is left as an excercise for the next major version. The\nmapping between states and versions is added to the documentation to\nmake it easier for users.\n\nBackpatch to v19 where online checksums were introduced.\n\nAuthor: Ian Barwick \nCo-authored-by: Daniel Gustafsson \nReviewed-by: Fujii Masao \nReviewed-by: Chao Li \nReviewed-by: Bertrand Drouvot \nDiscussion: https://postgr.es/m/CAB8KJ=hb765sE8bKC-6sh=Yp3sCjN8xs474yuBrkwyoTM2pgZA@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0907112d3882192f3c14d5ce0735790341e073ed", "date": "2026-08-18T22:25:52+02:00", "subject": "basebackup: do not verify checksums on pages from before enabling", "body": "Enabling data checksums in a running cluster changes the state to \"on\"\nbefore the checkpoint which flushes the pages the worker rewrote. A\nbase backup which started before that transition absorbs the barrier\nmid-run and starts verifying pages whose on-disk copies legitimately\nlack checksums, and whose LSNs predate the backup start, so the LSN\ncheck does not skip them either. The backup fails with bogus\ncorruption warnings. The same applies to checksums being disabled and\nre-enabled while the backup runs: hint bits set while checksums were\noff reach disk without a checksum update and without moving the page\nLSN, tripping verification once the re-enabling completes.\n\nTo fix, verify checksums only while they have been continuously\nenabled since the checkpoint the backup started from: track the\nlocation of the last XLOG2_CHECKSUMS record inserted or replayed, and\nverify only when the state is \"on\" and the last change predates the\nbackup start. The starting checkpoint then guarantees that every page\nflushed before it has a checksum written, and any later change\ndisables verification for the rest of the backup.\n\nA standby loses the tracked location when restarting, while pg_control\nalready carries the new state, so it could reach consistency below the\nrecord and serve base backups with the location unknown. To prevent\nthis, replaying XLOG2_CHECKSUMS advances minRecoveryPoint to the\nrecord, like XLOG_PARAMETER_CHANGE does.\n\nThe tests hold the enabling between the state change and its final\ncheckpoint with injection points, straddling it with backups on the\nprimary and across a standby crash-restart.\n\nAuthor: Zsolt Parragi \nReviewed-by: Bertrand Drouvot \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CAN4CZFP=-cVVVPue+e8qqPtDfuLuQn=ZB4Mw_C9-Ncru2wqAsQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c61c3ec40cb8e08fe54573de7cf1a27fa4d10f9f", "date": "2026-08-18T22:25:55+02:00", "subject": "Minor test suite cleanup", "body": "A few catalog queries were missing proper schema qualification in the\ntest_checksums module test suites, and one suite contained a disable\ncall right before tearing down the test which can be removed.\n\nBackpatch to v19 where the test suite was added.\n\nAuthor: Daniel Gustafsson \nDiscussion: https://postgr.es/m/8CF9B235-AEE9-4E68-93DD-DF4F29E2FCE5@yesql.se\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1d28812160d4e7ec06ebb2223d2c67b0dec74720", "date": "2026-08-18T22:25:58+02:00", "subject": "Stabilize the FORCE drop test for online data checksums", "body": "Commit 51f55b13a4d added a test where DROP DATABASE ... WITH (FORCE)\nterminates a session holding a temporary table in the target database.\nWhile exiting, the terminated session drops its temporary table and\ncommits, and the commit waits for a WAL flush behind the backlog\ngenerated by the checksum workers. On machines with slow storage this\ncan exceed the five seconds DROP DATABASE waits for terminated backends\nto exit, making the test fail with \"database \"dropmeforce\" is being\naccessed by other users\", as observed on buildfarm member turaco.\n\nTo fix, use asynchronous commit in the terminated session, so that its\nexit does not wait for a WAL flush, and checkpoint before the drop so\nthat the exit-time WAL records do not queue up behind the backlog.\n\nAuthor: Zsolt Parragi \nReported-by: Alexander Lakhin \nDiscussion: https://postgr.es/m/361531e2-52b5-499c-a126-815f277bbef2@gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "b524338261841e1e1f15a3e9cb3b644fd66ec0b8", "date": "2026-08-19T04:27:14+08:00", "subject": "fix(l10n): refresh Traditional Chinese message catalog #41273", "body": "Problem:\nThe zh_TW catalog was stale, omitted 743 messages from the current source, and contained empty, fuzzy, and format-unsafe translations.\n\nSolution:\nRegenerate the catalog with the official translation build, preserve the focused translation repairs, and translate the reported E5200 diagnostic.", "style": "conventional", "type": "fix", "scope": "l10n", "breaking": false} {"repo": "postgres", "sha": "755f7e2c921f16e6bc09bdc8d286c495b2cee93b", "date": "2026-08-18T13:55:10-07:00", "subject": "test_decoding: Don't print virtual generated columns.", "body": "Virtual generated columns are stored in tuples as null values, so\nheap_getattr() in tuple_to_stringinfo() always returns null for them.\ntest_decoding printed such a column as null, which is\nindistinguishable from a column that genuinely holds null.\n\nThis commit skips virtual generated columns in tuple_to_stringinfo(),\nthe only place that prints a tuple. Stored generated columns continue\nto be printed, since their values do live in the heap tuple.\n\nThis is a behavior change rather than a bug fix, and it changes\ntest_decoding's output for tables that have virtual generated columns,\nwhich could affect consumers that parse it. Therefore it is not\nback-patched.\n\nAuthor: SATYANARAYANA NARLAPURAM \nCo-authored-by: Bharath Rupireddy \nReviewed-by: Euler Taveira \nReviewed-by: Masahiko Sawada \nDiscussion: https://postgr.es/m/CAHg+QDfTh3UbB-Ed--o2Bd=SBDJoEiG-qp3C0+ETDibF63y=dw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b9e34da583370a48cae42267e1191b961c095a5e", "date": "2026-08-18T21:03:01Z", "subject": "doc: add missing const to PEM write SYNOPSIS", "body": "The write macros in include/openssl/pem.h take const TYPE *x, but\nPEM_read_bio_PrivateKey.pod still documented several write functions\nwithout const. Match the public headers.\n\nFixes: https://github.com/openssl/openssl/issues/31169\n\nAssisted-by: Hermes Agent:gpt-5.6-sol\nReviewed-by: Jakub Zelenka \nReviewed-by: Frederik Wedel-Heinen \nMerge-date: Fri Sep 4 17:37:17 2026\nMerged-from: https://github.com/openssl/openssl/pull/32422\n(cherry picked from commit 4ddd1002cd41750ee644478a84306be2c59d775a)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4ddd1002cd41750ee644478a84306be2c59d775a", "date": "2026-08-18T21:03:01Z", "subject": "doc: add missing const to PEM write SYNOPSIS", "body": "The write macros in include/openssl/pem.h take const TYPE *x, but\nPEM_read_bio_PrivateKey.pod still documented several write functions\nwithout const. Match the public headers.\n\nFixes: https://github.com/openssl/openssl/issues/31169\n\nAssisted-by: Hermes Agent:gpt-5.6-sol\nReviewed-by: Jakub Zelenka \nReviewed-by: Frederik Wedel-Heinen \nMerge-date: Fri Sep 4 17:37:17 2026\nMerged-from: https://github.com/openssl/openssl/pull/32422", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "731b959a810a3cef8436ebea483397f9dcdaea78", "date": "2026-08-18T16:27:10-05:00", "subject": "fix(vscode-extension): prevent JSDoc link provider injection", "body": "Treat link targets as file paths so crafted documentation cannot select arbitrary VS Code resource providers.\n\nFixes https://github.com/angular/angular/issues/70512\n\n(cherry picked from commit 69c87301aa545adb2c842cac31f062848c48ede9)", "style": "conventional", "type": "fix", "scope": "vscode-extension", "breaking": false} {"repo": "angular", "sha": "69c87301aa545adb2c842cac31f062848c48ede9", "date": "2026-08-18T16:27:10-05:00", "subject": "fix(vscode-extension): prevent JSDoc link provider injection", "body": "Treat link targets as file paths so crafted documentation cannot select arbitrary VS Code resource providers.\n\nFixes https://github.com/angular/angular/issues/70512", "style": "conventional", "type": "fix", "scope": "vscode-extension", "breaking": false} {"repo": "postgres", "sha": "fe06b9b5bd61dc04ab788091c56a0cb9ef44b3d2", "date": "2026-08-18T17:33:44-04:00", "subject": "Defend against null \"SV *\" pointers in plperl modules.", "body": "Tied hashes, and probably tied arrays, are capable of returning Perl\nvalue pointers that are actually NULL, not the usual pointer to an\nundef SV. We were not defending against that everywhere, leading\nto possible SIGSEGV. Fix the code to consistently treat a null\npointer returned from hv_iternext or av_fetch like a !SvOK one.\n(Note that the large diff in SV_to_JsonbValue is actually quite\ntrivial, but it required reindenting a chunk of existing code.)\n\nClaude Code found the instance in hstore_plperl, and I found the\nothers by code auditing. Perhaps the other instances aren't\nactually reachable, but I see little reason to assume that.\n\nThe known test cases for these errors require perl's Tie modules,\nwhich may not be present, so it doesn't seem worth the trouble\nto create regression test cases that would cover them.\n\nReported-by: Claude Code (via Noah Misch)\nAuthor: Tom Lane \nDiscussion: https://postgr.es/m/569769.1786901901@sss.pgh.pa.us\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5bf8fee543a19ca9ffa5aa99046adae34ba6a864", "date": "2026-08-18T17:47:13-04:00", "subject": "Make plperl's handling of Perl hashes more consistent.", "body": "Make hek2cstr() available in plperl.h, so that it can be used\nin hstore_plperl and jsonb_plperl. Those modules were previously\nusing different coding techniques that probably don't get\nconversion from Perl strings to the database encoding quite right.\n(I'd prefer to make hek2cstr() non-inline, but cross-extension calls\nare messy and plperl has avoided them up to now, so stick with the\nexisting approach.)\n\nConsistently use hv_iternext, hek2cstr, and HeVAL for hash iterations,\nwith one exception in plperl_trusted_init: there, hv_iternextsv is\nfine since we don't actually care about the hash keys. (A later\npatch will remove the HeVAL calls again, but for now we just want\nconsistency.)\n\nRemove unnecessary extra calls of hv_iterinit.\n\nRemove duplicative pstrdup's in plperl_to_hstore.\n\nThe meat of this change is to use hek2cstr() in the contrib modules,\nwhich makes a user-visible change in encoding conversion behavior.\nWhile it's certainly a bug fix, we've had no field complaints about\nthose modules, so I'm hesitant to make this change in the back\nbranches. Hence, apply to master only.\n\nAuthor: Tom Lane \nReviewed-by: Andrey Rachitskiy \nDiscussion: https://postgr.es/m/569769.1786901901@sss.pgh.pa.us", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "a80e2d1af7a49af2767e76467764bf3fdb7de659", "date": "2026-08-18T22:57:27+01:00", "subject": "sixel: preserve image aspect across client geometries", "body": "Scale SIXEL rasters uniformly when clients have different cell dimensions, while retaining the original logical cell canvas as blank padding.\n\nMap crops using real cell boundaries so partial final cells are not stretched or compressed.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "695aa15743685a9d46c4c41bb9c95221d4659541", "date": "2026-08-18T23:58:16+02:00", "subject": "urldata: remove the aptr struct", "body": "Move the two remaining fields into the state struct instead, within the\nHTTP ifdef.\n\nAlso: make the AWS sigv4 code always rely on the http_host value. It\nwas previously conditionally trying to also work if it was not set, but\nthat logic was not creating an identical copy so it would fail.\n\nCloses #22620", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "f53976a266567c14baf3aca71cbe35c61cc8a61f", "date": "2026-08-18T23:59:32+01:00", "subject": "image: move basic fallback into image.c", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f6c146ffab8ffc17b3f4e3fd6061e6c8c5fe8d56", "date": "2026-08-18T18:35:03-06:00", "subject": "Add a watchdog timeout to the C tests built with testutil/main", "body": "This adds a timeout that will self-terminate a test with a\nsigabrt after a timeout - in the hope that if a test is wedged\nwe now get diagnostic information (like a core dump or stack trace)\ninstead of the CI job timing out globally\n\nThe default timeout for a single C test is 30 minutes, which should\nbe plenty, and it is configurable via environment variable\ndocumented in the test README.md file\n\nThis does nothing to the invocation yet to ensure we can get a\ncore dump.\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Neil Horman \nMergeDate: Fri Sep 11 19:12:58 2026\n(Merged from https://github.com/openssl/openssl/pull/32424)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "19d900358fd6f200ada6aabfa4d44d83c16a3f0a", "date": "2026-08-19T09:52:01+09:00", "subject": "Report single-page checksum failures in pg_stat_database", "body": "Base backups reported checksum failures to pg_stat_database only for\nfiles with more than one failing page. Commit 6b9e875f728 placed the\nreport inside the block emitting the per-file summary WARNING, which\nwas skipped for a single failure. As a result, a backup failing on\nfiles with one corrupted page each left checksum_failures untouched.\n\nTo fix, emit the per-file summary and the pgstat report for any\nnon-zero failure count. The end-of-backup total WARNING had the same\noff-by-one and is now also emitted for a single failure.\n\nAuthor: Zsolt Parragi \nReviewed-by: Nazir Bilal Yavuz \nDiscussion: https://postgr.es/m/CAN4CZFN+Bi6XmaH8zOdMWjoycYFx9nKtOr+dzQf0o-UQ+Rdqmw@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "668dc66f663729e13988da2222883e09be50b668", "date": "2026-08-18T18:55:16-06:00", "subject": "Add core dumps and matching binary to CI artifacts on linux", "body": "If we manage to have a test crash, let's get the core and\nbinary in the artifacts\n\nI mostly want this for https://github.com/openssl/openssl/pull/32424\nbut it has other uses.\n\nReviewed-by: Neil Horman \nReviewed-by: Norbert Pocs \nMergeDate: Fri Sep 11 19:22:23 2026\n(Merged from https://github.com/openssl/openssl/pull/32425)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3e0cfe3ae342369097a99d2f5e0dbd7174a36f6f", "date": "2026-08-18T21:20:48-04:00", "subject": "Skip the legacy session ID check for DTLS 1.3 resumption", "body": "Commit 4e8593e7ebbb added a session ID match in ssl_get_prev_session()\nguarded by !SSL_CONNECTION_IS_TLS13(s). That macro excludes DTLS, so\nthe pre-1.3 check wrongly ran for DTLS 1.3, where legacy_session_id is\na meaningless compatibility field.\n\nOpenSSL DTLS 1.3 clients send an empty legacy_session_id, so this went\nunnoticed. Peers that send a non-empty one (e.g. wolfSSL) tripped the\nmismatch and were refused resumption. Use SSL_CONNECTION_IS_VERSION13(s),\nwhich covers both TLS 1.3 and DTLS 1.3.\n\nAdd a DTLS 1.3 TLSProxy test that injects a mismatching\nlegacy_session_id and confirms the server still resumes.\n\nAssisted-by: Claude:claude-opus-4-8\nReviewed-by: Matt Caswell \nReviewed-by: Tomas Mraz \nMerge-date: Fri Aug 21 18:24:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32435", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a812cf647a828fc6157ca1f864f74067f088b580", "date": "2026-08-19T10:22:32+09:00", "subject": "Reduce LEFT JOIN to ANTI JOIN using quals within the RHS subtree", "body": "reduce_outer_joins() turns a LEFT JOIN into an ANTI JOIN when some Var\nthat an upper qual requires to be NULL is actually non-nullable in any\nmatching row. When that holds, only null-extended (unmatched) rows\ncan satisfy the upper qual, which is exactly anti-join semantics.\nUntil now we recognized such a Var as non-nullable only when the\njoin's own clauses were strict for it, or when it was defined NOT NULL\nby table constraints.\n\nThis patch allows strict quals applied within the RHS subtree to serve\nas the proof as well. Because such quals hold for every row the RHS\nemits, they hold for every matching row, so a Var they force non-null\ncan become NULL above the join only by null-extension.\n\nTo avoid re-walking the jointree at decision time, the first pass of\nthe reduce-outer-joins process gathers these proving quals into its\nper-subtree state, alongside nullable_rels. The second pass then\nproves non-nullness from the RHS subtree's collected quals together\nwith the join's own ON quals. As before, the reduction fires only\nwhen the proven, forced-null Var belongs to the RHS of the join.\n\nAuthor: Richard Guo \nReviewed-by: wenhui qiu \nDiscussion: https://postgr.es/m/CAMbWs49H9khf+1GwyzD0TYaks6cE-OU+1mbUiOn3gZSoBiO7zg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2ce5a57bc0d93c6b14f7cd5ea5eb105a987bd46f", "date": "2026-08-19T10:23:09+09:00", "subject": "Reduce FULL JOIN to ANTI JOIN", "body": "reduce_outer_joins() already recognizes that a LEFT JOIN is really an\nanti-join when an upper qual forces a nullable-side Var to be NULL but\nthat Var is provably non-null in every row the nullable side emits.\nOnly null-extended rows can then satisfy the qual, so the matched rows\nall drop out and JOIN_LEFT becomes JOIN_ANTI.\n\nThe same reasoning applies to a FULL JOIN. If a forced-null Var on\none side is proven non-null, every row where that side is present,\nmatched or not, has the Var non-null and is dropped by the qual. Only\nthe rows where that side was null-extended survive, which is an\nanti-join that keeps the other side's unmatched rows.\n\nWhen the proven Var is on the RHS, the surviving LHS rows are already\nthe left input, so this is a plain JOIN_ANTI. When it is on the LHS,\nthe surviving RHS rows must become the left input, so we tag the join\nJOIN_RIGHT_ANTI and let the existing input-switching step, the one\nthat flips JOIN_RIGHT to JOIN_LEFT, normalize it to JOIN_ANTI. Unlike\nthe LEFT JOIN case, the join's own ON quals cannot serve as proof\nhere, because they do not hold for the unmatched rows the proof must\ncover.\n\nReducing the full join this way also lets qual constraints reach its\ninputs. reduce_outer_joins_pass2() passes nothing down through a\nJOIN_FULL, but it does pass the join's own quals down through the\nresulting JOIN_ANTI, so outer joins below it can now be reduced too.\nThat is sound for the same reason it is for any anti-join: a row that\na lower join null-extends cannot satisfy those quals, so it can never\nmatch, and removing it does not change which rows the anti-join emits.\n\nThe proof that a forced-null Var is non-null, from the quals that hold\nfor every row a subtree emits (optionally plus extra quals the caller\nsupplies) or from a NOT NULL constraint, is factored into\nforced_null_var_is_nonnullable() and shared by the LEFT and FULL\npaths.\n\nAuthor: Richard Guo \nReviewed-by: wenhui qiu \nDiscussion", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "02119d6c87e20bd17f9b94b0420c2713d62275bc", "date": "2026-08-19T10:23:44+09:00", "subject": "Reduce outer joins to anti joins for whole-row IS NULL tests", "body": "reduce_outer_joins() recognizes \"WHERE b.z IS NULL\" above an outer\njoin as an anti-join condition, but not the whole-row \"WHERE b IS\nNULL\", which is a natural way to ask for an anti-join without naming a\nspecific column. Teach it to recognize the whole-row form too.\n\nA row that the join null-extends has all of b's columns set to NULL,\nso it satisfies \"b IS NULL\". A matched row satisfies the test only if\nits columns happen to be all NULL. Hence proving any one column of b\nnon-null in matching rows rules out every matched row, leaving only\nnull-extended rows: exactly anti-join semantics. This mirrors the\nsingle-column case, and the same proofs apply: a NOT NULL table\nconstraint, a strict join clause (for LEFT joins), or strict quals\nwithin the relation's subtree. Because any one column suffices, the\nwhole-row test reduces in a strict superset of the cases a\nsingle-column test does.\n\nTo implement this, find_forced_null_vars() now reports a whole-row Var\ntested with row-format IS NULL as a varattno-zero entry meaning that\nall of the relation's columns are forced null. Row-format tests on\nordinary composite-type columns remain excluded, since such a test\ndoes not force the column null: it is also true when the column is a\nnon-null row whose fields are all NULL. The proof functions in\nreduce_outer_joins() treat a whole-row entry accordingly: it is\nrefuted by proving any one column of its relation non-null. A match\non the whole-row attribute itself proves nothing, because a non-null\ncomposite datum can still have all columns NULL, so the per-column\nmatching now explicitly excludes that attribute.\n\nAuthor: Richard Guo \nReviewed-by: wenhui qiu \nDiscussion: https://postgr.es/m/CAMbWs49H9khf+1GwyzD0TYaks6cE-OU+1mbUiOn3gZSoBiO7zg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f863be3e2dbeed457885a109904b30bb8edcfcde", "date": "2026-08-19T10:24:12+09:00", "subject": "Relax strictness detection for row-format IS NOT NULL tests", "body": "find_nonnullable_rels() and find_nonnullable_vars() declined to look\nthrough a NullTest with argisrow set, treating row-format IS NOT NULL\nas proving nothing. But such a test returns FALSE, not TRUE, both\nwhen the composite datum is NULL and when any of its fields is NULL,\nso its truth implies a non-null input just as the plain test does.\nThat is the only property strictness detection relies on, so the\nargisrow restriction can simply be dropped.\n\nThis lets \"a LEFT JOIN b ... WHERE b IS NOT NULL\" reduce to an inner\njoin, the IS NOT NULL counterpart of the whole-row anti-join reduction\nin the preceding commit. make_outerjoininfo() likewise picks up such\ntests when it computes join strictness for outer-join ordering.\nRow-format tests on composite-type columns now also prove those\ncolumns non-null, which can feed the anti-join proofs.\n\nThe stronger implication of a row-format test, that every field of the\nrow is non-null, remains unexploited: a whole-row Var reported by\nfind_nonnullable_vars() promises only a non-null datum, since the same\nentry can arise from contexts that are merely strict at the datum\nlevel, such as record comparisons.\n\nAuthor: Richard Guo \nReviewed-by: wenhui qiu \nDiscussion: https://postgr.es/m/CAMbWs49H9khf+1GwyzD0TYaks6cE-OU+1mbUiOn3gZSoBiO7zg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a98858651ee5abd2bc68226c8d0ad49fef324f04", "date": "2026-08-18T21:36:27-04:00", "subject": "build(vim-patch): n/a docs from 9.0.1481, keep hunks without header", "body": "switch/case and goto labels v9.0.1481 and similar patches\nfrom auto-N/A.\nBlacklist as much as possible to auto-N/A later crypt patches.\n\nHunks without header in \"*.h\" is valid if only littered with #ifdefs\nor change is isolated to 1st line\nbut \"src/feature.h\" should be mostly N/A\nafter Nvim removed \"FEAT_\" guards.\nWonder if certain patches with only \"*.h\" changes were mistakenly\nmarked N/A.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vue-core", "sha": "75c1cd0855d6db4600634abc094ba8088913faf5", "date": "2026-08-18T19:08:57-07:00", "subject": "refactor(vapor): simplify template ref entry points", "body": "", "style": "conventional", "type": "refactor", "scope": "vapor", "breaking": false} {"repo": "vite", "sha": "d9b10a98db1c293ee64300bd75d568b44c8ae931", "date": "2026-08-19T10:11:18+08:00", "subject": "fix(module-runner): exclude completed modules from in-flight cycle detection (fix #22999)", "body": "", "style": "conventional", "type": "fix", "scope": "module-runner", "breaking": false} {"repo": "openssl", "sha": "8c13b56e862595fb801c9eae79bcdf9982bb39a0", "date": "2026-08-19T11:21:49+09:00", "subject": "test: cover CCM empty Final without a payload Update", "body": "Declare a zero payload length and supply AAD while deliberately omitting\nthe payload Update. Check that streaming and one-shot Final generate the\nsame tag, accept the correct tag, and reject a modified tag for every\navailable built-in CCM cipher.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 21 10:04:28 2026\nMerged-from: https://github.com/openssl/openssl/pull/32427", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4aa0fb3a5a858adb139650c56899a5f19ea7a9eb", "date": "2026-08-19T11:21:49+09:00", "subject": "test: cover CCM empty Final without a payload Update", "body": "Declare a zero payload length and supply AAD while deliberately omitting\nthe payload Update. Check that streaming and one-shot Final generate the\nsame tag, accept the correct tag, and reject a modified tag for every\navailable built-in CCM cipher.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 21 10:00:13 2026\nMerged-from: https://github.com/openssl/openssl/pull/32427", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e44fbf2ca18146cbd4f4d71c6078c9738964ec12", "date": "2026-08-19T11:21:49+09:00", "subject": "test: cover CCM empty Final without a payload Update", "body": "Declare a zero payload length and supply AAD while deliberately omitting\nthe payload Update. Check that streaming and one-shot Final generate the\nsame tag, accept the correct tag, and reject a modified tag for every\navailable built-in CCM cipher.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 21 09:57:47 2026\nMerged-from: https://github.com/openssl/openssl/pull/32427", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "79239c8c3377f01b9d04d116f6557e94669e0bdd", "date": "2026-08-19T11:21:49+09:00", "subject": "test: cover CCM empty Final without a payload Update", "body": "Declare a zero payload length and supply AAD while deliberately omitting\nthe payload Update. Check that streaming and one-shot Final generate the\nsame tag, accept the correct tag, and reject a modified tag for every\navailable built-in CCM cipher.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 21 09:54:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32427", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "35644accde0cf699d1af3435b694371702919d10", "date": "2026-08-19T11:21:49+09:00", "subject": "test: cover CCM empty Final without a payload Update", "body": "Declare a zero payload length and supply AAD while deliberately omitting\nthe payload Update. Check that streaming and one-shot Final generate the\nsame tag, accept the correct tag, and reject a modified tag for every\navailable built-in CCM cipher.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Aug 21 09:37:18 2026\nMerged-from: https://github.com/openssl/openssl/pull/32427", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "29b52d2b47fb8fa892695bc941d3e035164481cf", "date": "2026-08-19T11:32:58+09:00", "subject": "Fix relcache reference leak when decoding TRUNCATE", "body": "ReorderBufferProcessTXN() opens every relation referenced by a\nTRUNCATE change. When RelationIsLogicallyLogged() returns false,\nit skips the relation without releasing the reference acquired\nby RelationIdGetRelation().\n\nLooking at the in-core code paths building XLOG_HEAP_TRUNCATE records,\nno relation OIDs would be included if they do not satisfy\nRelationIsLogicallyLogged(). One pattern that could go through is if a\ntable is switched to SET UNLOGGED, but that would not be reachable in\npractice as the decoding happens after a historical snapshot is taken,\nso the relation should still be valid.\n\nThis is a defense-in-depth measure in practice, and we tend to be\ncareful about how Relations are handled when sending changes to output\nplugins, so backpatch all the way down.\n\nAuthor: Chao Li \nReviewed-by: Xuneng Zhou \nDiscussion: https://postgr.es/m/7DD65D03-3B5A-43B2-99AD-8E6AF5372BAB@gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "5c2805dbaf2ece13736ae3fbe2bac849c814c304", "date": "2026-08-18T19:54:05-07:00", "subject": "fix(runtime-vapor): restore rendering suspense boundary in fragment ctx", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "e6c200f9a39a75da1bb829e87e99133ea5fea237", "date": "2026-08-19T12:31:00+09:00", "subject": "psql: Avoid returning oom_buffer from psql slash command scanner", "body": "psql_scan_slash_command() builds the command name in a local\nPQExpBufferData and returns the buffer's data pointer to its caller. If\neither the initial allocation or a later enlargement failed, that data\npointer could be the static PQExpBuffer OOM buffer rather than malloc-owned\nstorage. HandleSlashCmds() could then eventually pass it to free(), causing\nundefined behavior.\n\nDetect a broken command-name buffer before returning it, report OOM, and\nreturn NULL instead. Teach HandleSlashCmds() to treat a NULL command name\nas a command error before trying to compare or dispatch it.\n\nBackpatch to all supported versions.\n\nReported-by: Junwang Zhao \nAuthor: Fujii Masao \nReviewed-by: Chao Li \nReviewed-by: Junwang Zhao \nDiscussion: https://postgr.es/m/CAHGQGwEh3R3=1tx_a5=fTDJ+ycuwxWMEn6bG_Yt4B5P+hE7AVw@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "de38e72a199117ef69d7653f91f849626b4dcf3b", "date": "2026-08-19T12:34:28+09:00", "subject": "psql: Avoid returning oom_buffer from psql slash option scanner", "body": "psql_scan_slash_option() builds option text in a local PQExpBufferData\nand returns the buffer's data pointer to its caller. If either the initial\nallocation or a later enlargement failed, that data pointer could be\nthe static PQExpBuffer OOM buffer rather than malloc-owned storage.\nThe callers could then eventually pass it to free(), causing undefined\nbehavior.\n\nDetect a broken option buffer before returning it, report OOM, and return\nNULL instead. Also avoid evaluating a backtick substitution when the option\nbuffer is already broken, since doing so could otherwise touch the static\nOOM buffer.\n\nThis keeps the existing NULL-return convention for slash options. Callers\nare not generally changed to distinguish OOM from no option.\n\nBackpatch to all supported versions.\n\nReported-by: Junwang Zhao \nAuthor: Fujii Masao \nReviewed-by: Chao Li \nReviewed-by: Junwang Zhao \nDiscussion: https://postgr.es/m/CAHGQGwEh3R3=1tx_a5=fTDJ+ycuwxWMEn6bG_Yt4B5P+hE7AVw@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "e8bc40d290673a74c094c289b00c01c997d36652", "date": "2026-08-19T12:38:16+09:00", "subject": "psql: Fix psql slash option leaks", "body": "psql_scan_slash_option() returns a malloc'd string, but \\getresults,\n\\gset in pipeline mode, \\restrict, and \\unrestrict did not free it\nafter consuming or copying the value.\n\nFree these option strings after use.\n\nBackpatch to all supported versions. In v17 and older, only \\restrict\nand \\unrestrict are affected, so those branches need only that part of\nthe fix.\n\nAuthor: Fujii Masao \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/CAHGQGwEh3R3=1tx_a5=fTDJ+ycuwxWMEn6bG_Yt4B5P+hE7AVw@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "ba32f106323c7ae990e6a1e4232dca9c26b4bec2", "date": "2026-08-19T00:07:57-04:00", "subject": "build(vim-patch): vim9 parse_argument_types is n/a", "body": "Vim9script has incompatible lambda syntax to support argument types.\nTarget v9.1.0270.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "neovim", "sha": "855e973bf2d5873a43657adbe58a89e98ad54ba6", "date": "2026-08-19T00:16:44-04:00", "subject": "build(vim-patch): v9.1.0050 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vite", "sha": "4a261f242831bef92afd2f1aacfb81eab9dec371", "date": "2026-08-19T13:46:15+09:00", "subject": "test(module-runner): add TLA circular import case", "body": "", "style": "conventional", "type": "test", "scope": "module-runner", "breaking": false} {"repo": "vite", "sha": "599b44b6600ec426e10cd556908d53b027b0c4fb", "date": "2026-08-19T13:46:36+09:00", "subject": "test(module-runner): simplify server-hmr tests", "body": "", "style": "conventional", "type": "test", "scope": "module-runner", "breaking": false} {"repo": "neovim", "sha": "c8da9f3df29741a2c7d099fb4d68539d50b1d82e", "date": "2026-08-19T00:49:08-04:00", "subject": "build(vim-patch): n/a patterns for popupwin. v9.2.0636", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "neovim", "sha": "3e487dd1456f661bc7c2f6a5e38083aabf647618", "date": "2026-08-19T01:24:03-04:00", "subject": "build(vim-patch): n/a tabpanel docs from v9.1.1391", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vite", "sha": "2804636ff608d105928009d274ffba7cfbe55340", "date": "2026-08-19T10:39:43+05:00", "subject": "fix(css): don't pass empty targets to lightningcss", "body": "", "style": "conventional", "type": "fix", "scope": "css", "breaking": false} {"repo": "angular", "sha": "17b7c1f643f17340113a5db86566fc66b37068ae", "date": "2026-08-19T08:41:09+03:00", "subject": "docs: add NG0991 error page and document rxResource's completion contract", "body": "RESOURCE_COMPLETED_BEFORE_PRODUCING_VALUE had no guide, no JSDoc on\nRxResourceOptions.stream, and — since the code was positive rather\nthan negative — could never get an auto-linked docs page even if one\nexisted. Flip it to -991, add the NG0991 reference page, and document\nthe \"stream must emit a value or an error before completing\"\nrequirement on stream's JSDoc and in the RxJS interop guide.\n\nAlso documents and tests that an unguarded template read of an\nerrored resource's .value() propagates to the global ErrorHandler,\nand recommends guarding with .hasValue() as defense in depth.\n\nhttpResource can throw the same error, but for a different reason:\nits internal request Observable isn't something app code writes\ndirectly, so an empty completion there is almost always an\nHttpInterceptor swallowing the response (catchError(() => EMPTY))\nrather than a stream authored in the resource() call itself. The page\ncovers both APIs with guidance matched to what's actually going on\nfor each.\n\n(cherry picked from commit 8975b4346d5d3a4b9ed5a7c34026fe0a824ed0ab)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "8975b4346d5d3a4b9ed5a7c34026fe0a824ed0ab", "date": "2026-08-19T08:41:09+03:00", "subject": "docs: add NG0991 error page and document rxResource's completion contract", "body": "RESOURCE_COMPLETED_BEFORE_PRODUCING_VALUE had no guide, no JSDoc on\nRxResourceOptions.stream, and — since the code was positive rather\nthan negative — could never get an auto-linked docs page even if one\nexisted. Flip it to -991, add the NG0991 reference page, and document\nthe \"stream must emit a value or an error before completing\"\nrequirement on stream's JSDoc and in the RxJS interop guide.\n\nAlso documents and tests that an unguarded template read of an\nerrored resource's .value() propagates to the global ErrorHandler,\nand recommends guarding with .hasValue() as defense in depth.\n\nhttpResource can throw the same error, but for a different reason:\nits internal request Observable isn't something app code writes\ndirectly, so an empty completion there is almost always an\nHttpInterceptor swallowing the response (catchError(() => EMPTY))\nrather than a stream authored in the resource() call itself. The page\ncovers both APIs with guidance matched to what's actually going on\nfor each.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "1d9fa392a43229241f80630236f8552ce8f7cd0f", "date": "2026-08-19T13:47:55+08:00", "subject": "refactor: use JSON import attributes instead of readFileSync in constants", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "curl", "sha": "3db791981869bd9ff149a7ae64e973d72d6bd5d3", "date": "2026-08-19T08:18:27+02:00", "subject": "CVE-2026-6276: stale custom cookie host causes cookie leak", "body": "Reimplemented 3a19987a87f393d93", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "6119ec35cee8dcf4528ac7e5154276cd89052177", "date": "2026-08-19T08:25:40+02:00", "subject": "CVE-2026-6429: netrc credential leak with reused proxy connection", "body": "Based on b4024bf808bd558026fdc6\n\nImported Curl_url_same_origin() from 32a513e180ce83d5e9b708 (and edited)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "95fd8d9327ea707f234fbd8bfd4565759cc7fe1b", "date": "2026-08-19T08:39:45+02:00", "subject": "CVE-2026-7168: cross-proxy Digest auth state leak", "body": "Based on c1cfdf59acbaf9504c45 (edited)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "7546b7a80549c9620d52692cb0f5cab4c69b546c", "date": "2026-08-19T09:40:48+03:00", "subject": "fix(core): throw coded RuntimeErrors instead of crashing when hydration/rendering can't find an expected DOM node", "body": "This started from a real production crash trace:\n\n TypeError: Cannot read properties of null (reading 'nextSibling')\n\nTraced through a minified bundle back to siblingAfter() in\npackages/core/src/hydration/node_lookup_utils.ts. It walks forward a\nfixed number of DOM siblings during hydration, based on how many the\nserver rendered. Its only guard, validateSiblingNodeExists(), is\ngated behind ngDevMode and stripped from production builds. If the\nclient DOM has fewer real siblings than the server-serialized data\nexpects — client/server render diverged, or something outside\nAngular (a browser extension, an injected script) removed a node —\nthe loop ran fully unguarded in production: currentNode went null\npartway through, and the next iteration's currentNode.nextSibling\nthrew a raw, uncoded TypeError instead of a coded, debuggable one.\n\nReproduced first, not just theorized: added a test that server-\nrenders a @for loop with 3 items, removes 2 of them before hydration\nruns, and confirmed it actually throws that exact TypeError against\nthe original code before touching anything.\n\nFix: add a null check at the top of the loop, matching the pattern\nalready used for the existing NG0500/NG0502 hydration checks in\nelement.ts — the dev-mode check runs first as before (a no-op in\nproduction), and the new check is a pure safety net that only fires\nonce that dev-mode check has already been compiled away. In dev mode\nthis is provably dead code, so nothing about dev behavior changes.\nCoded as NG0501 (HYDRATION_MISSING_SIBLINGS, already existed).\n\nWhile investigating, found and fixed two more call sites with the\nexact same shape of bug — a dev-only check that leaves production\ncompletely unguarded:\n\n- navigateToNode() (same file): walks an explicit \"firstChild /\n nextSibling\" navigation path recorded for nodes that can't rely on\n simple sibling-walking (e.g. content right after an \n slot). Its only guard was also ngDevMode-only. Added the same\n production safety net, ", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "vitest", "sha": "9f710d304579f8af08c2ba29018ab492f06aff91", "date": "2026-08-19T09:43:59+03:00", "subject": "fix(browser): don't hang when `ui: true` in projects", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "openssl", "sha": "4f15f84b7716b49002f85556efe5a5937801dd47", "date": "2026-08-19T08:44:35+02:00", "subject": "test: fix TAP stream corruption in three test recipes", "body": "'local $/ = undef' leaks into OpenSSL::Test::run(), which then\nemits a bare \"# \" that glues onto the following 'ok' line.\nScope the slurping properly.\n\n'pkeyutl -verifyrecover' outputs binary on stdout; redirect it.\n\nFixes: https://github.com/openssl/openssl/issues/32428\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Mounir Idrassi \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 14:20:32 2026\nMerged-from: https://github.com/openssl/openssl/pull/32429", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0f31f694cf61a5154023688c5dcc86ca09b277d2", "date": "2026-08-19T08:44:35+02:00", "subject": "test: fix TAP stream corruption in three test recipes", "body": "'local $/ = undef' leaks into OpenSSL::Test::run(), which then\nemits a bare \"# \" that glues onto the following 'ok' line.\nScope the slurping properly.\n\n'pkeyutl -verifyrecover' outputs binary on stdout; redirect it.\n\nFixes: https://github.com/openssl/openssl/issues/32428\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Mounir Idrassi \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 14:20:32 2026\nMerged-from: https://github.com/openssl/openssl/pull/32429\n(cherry picked from commit 936250eef1c319c905af85639dcc2e3f67712f94)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vitest", "sha": "294b139a3df178a4bdda7a2b9d38bc8142cfba0a", "date": "2026-08-19T12:15:53+05:30", "subject": "docs: document expect.fail method", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b6b1f89847f7bece7fab2ef392d3c84a7ee6f88d", "date": "2026-08-19T08:46:54+02:00", "subject": "Message style fixes", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "c031e5a1e8fa91309a96a2e1f897530e060499a2", "date": "2026-08-19T15:48:09+09:00", "subject": "docs: add cjs support status in common errors", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "9a90124c42f01faca3be33e74ea61e036d14b987", "date": "2026-08-19T15:48:28+09:00", "subject": "ci: upload vitest results after failures", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vitest", "sha": "ff9dcf4c92d60a1dffcdfe03415996354028d4bb", "date": "2026-08-19T15:48:48+09:00", "subject": "chore(ui): fix html report path for dev ui", "body": "", "style": "conventional", "type": "chore", "scope": "ui", "breaking": false} {"repo": "vitest", "sha": "1f4fcd36354532f16cee8e36c8e573f8a1a0af76", "date": "2026-08-19T15:49:07+09:00", "subject": "fix(ui): reveal explorer virtual scroller on resize", "body": "", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "vitest", "sha": "bce2d3c2f5b63763ec6c1552b85213b1d4f80011", "date": "2026-08-19T15:51:47+09:00", "subject": "feat(ui): highlight editor source locations", "body": "", "style": "conventional", "type": "feat", "scope": "ui", "breaking": false} {"repo": "vitest", "sha": "e5d38fd3c7247c9bcc0980638192bb679277b30c", "date": "2026-08-19T15:51:48+09:00", "subject": "refactor(ui): consolidate CodeMirror options prop", "body": "", "style": "conventional", "type": "refactor", "scope": "ui", "breaking": false} {"repo": "vitest", "sha": "6816f3b17ca8368578bc3d223f872f20fcfff421", "date": "2026-08-19T15:51:48+09:00", "subject": "feat(ui): navigate trace steps with keyboard and remove indirect editor focus", "body": "", "style": "conventional", "type": "feat", "scope": "ui", "breaking": false} {"repo": "postgres", "sha": "ca86b45adddc5385c50e974a0905b7938eb10487", "date": "2026-08-19T08:56:56+02:00", "subject": "Message wording fix", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "fc6c67ce8613ee89956e8397a2c0c3e5169b3913", "date": "2026-08-19T09:03:41+02:00", "subject": "docs: clarify that cookies need domain set to match", "body": "Especially important if reading cookies from HTTP headers, which is a\nstyle we discourage.\n\nReported-by: accl on hackerone\nCloses #22621", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "170c9344defaefb351a9dbd0134f4c37a9847358", "date": "2026-08-19T16:06:00+09:00", "subject": "Don't free fast-path FK metadata from the inval callback", "body": "Commit e484b0eea6 made InvalidateConstraintCacheCallBack() pfree an\nentry's FastPathMeta to plug a leak, but that breaks the rule stated\natop the callback: entries may have active references at invalidation\ntime, so we mark them invalid rather than removing them.\n\nThe metadata is subject to the same rule. ri_FastPathCheck() and\nri_FastPathBatchFlush() copy riinfo->fpmeta into a local, and\nri_FastPathFlushArray() additionally takes FmgrInfo pointers into it\nbefore its \"walk all matches\" loop. That loop runs\nindex_getnext_slot(), ri_LockPKTuple(), and user-supplied cast and\nequality functions, any of which can accept invalidation messages, and\na user function that performs DDL triggers one deliberately, no\nconcurrency required. The callback then freed the object still in use,\nso the loop read freed memory and called through FmgrInfos in it.\n\nFix by unlinking the metadata from the entry, so the next check\nrebuilds it as before, but deferring the actual free to AtEOXact_RI(),\nwhich runs from CommitTransaction() / PrepareTransaction() /\nAbortTransaction() with no RI check on the stack. Detached objects\nare chained through a new next_dead field and released there.\n\nThe queue holds a few kB per detached object until the transaction\nends, which only affects transactions that interleave DDL with\nFK-checking DML. That seems clearly preferable to a use-after-free.\n\nUnlinking alone is not enough for the multi-column path.\nri_FastPathFlushLoop() calls build_index_scankeys() once per buffered\nrow, and that function re-read riinfo->fpmeta each time. A cast\ninvoked for one row can accept an invalidation that clears the field,\nso the next row found NULL there. Pass the metadata down from\nri_FastPathBatchFlush() instead, as the array path already did, so one\nreference covers the whole batch. That is only safe because of the\ndeferred release above; latching without it would turn the NULL\ndereference into a use-after-free.\n\nReported-by: Jacob Brazeal (o", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "abca12838fefabbf146307447c23176ec7eb67ed", "date": "2026-08-19T16:10:18+09:00", "subject": "Give RI fast-path cached FmgrInfos their own memory context", "body": "ri_populate_fastpath_metadata() copies the cast and equality FmgrInfos\ninto the cached FastPathMeta with fn_mcxt set to TopMemoryContext, the\ncontext active at the copy. fn_mcxt is scratch space for the called\nfunction: record_eq(), and the record I/O functions generally, allocate\ntheir per-call cache there and keep a pointer to it in fn_extra.\n\nBecause that scratch is in TopMemoryContext, it outlives the metadata.\nWhen the metadata is discarded on invalidation, whatever the cast and\nequality functions cached is left behind, with nothing pointing at it.\nEach subsequent repopulation allocates afresh, so a session that\nrepeatedly invalidates a foreign key constraint grows TopMemoryContext\nwithout bound.\n\nGive the metadata its own context for the FmgrInfos' fn_mcxt and delete\nit along with the metadata, so the cached scratch is freed with the\nFmgrInfos that point at it. Since the preceding commit defers release\nof the metadata to AtEOXact_RI(), the context is deleted there rather\nthan in InvalidateConstraintCacheCallBack().\n\nThe context deliberately is not reset while the metadata is in use.\nfn_extra points into fn_mcxt, so resetting it would leave those\npointers dangling; the next call would find fn_extra non-NULL and read\nfreed memory. fmgr_info_copy() zeroing fn_extra in the copy is the\nsame invariant seen from the other side. Nothing accumulates in the\ncontext during use in any case: record_eq() and friends allocate only\nwhen fn_extra is NULL and reuse the cache afterwards.\n\nReported-by: Noah Misch \nReviewed-by: Ayush Tiwari \nDiscussion: https://postgr.es/m/20260705210533.ee.noahmisch@microsoft.com\nDiscussion: https://postgr.es/m/CA+HiwqFFB6vzx8v3t2=rbNYyxMistLf5kkJfqzJ81nadFyLrxA@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "4417cbfc9e8f4bf42b3c785ca32d05dbbbbfb9b5", "date": "2026-08-19T16:11:19+09:00", "subject": "docs: fix formatting inconsistencies for better clarity", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "16048483f6fb98c237767d26f7bd6028fc6c53b0", "date": "2026-08-19T16:13:42+09:00", "subject": "docs: clarify about `server.proxy` WebSocket origin check", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "9b29495863c967e2768ea1c517f3ad28a50a9a1b", "date": "2026-08-19T09:32:16+02:00", "subject": "http: custom Authorization: header overrides Negotiate", "body": "As documented and as it does for other methods.\n\nDifficult to test since it needs a successful kerberos ticket.\n\nReported-by: sdgh179 on github\nFixes #22610\nCloses #22622", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "38876e7ea14b74364181b7bb4c12053d5c8ebdaa", "date": "2026-08-19T09:44:10+02:00", "subject": "test framework: always read child output line by line in run()", "body": "If a recipe played with $/ (slurp or paragraph mode), the pipe read\nwould misbehave and emit a bare \"# \" prefix, or prefix only the\nfirst line of a multi-line read, risking TAP stream corruption.\n\nFixes: https://github.com/openssl/openssl/issues/32428\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Mounir Idrassi \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 14:20:36 2026\nMerged-from: https://github.com/openssl/openssl/pull/32429", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9f919186c2d06afa8a185dfaa6f833656fdf3c03", "date": "2026-08-19T09:44:10+02:00", "subject": "test framework: always read child output line by line in run()", "body": "If a recipe played with $/ (slurp or paragraph mode), the pipe read\nwould misbehave and emit a bare \"# \" prefix, or prefix only the\nfirst line of a multi-line read, risking TAP stream corruption.\n\nFixes: https://github.com/openssl/openssl/issues/32428\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Mounir Idrassi \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 14:20:36 2026\nMerged-from: https://github.com/openssl/openssl/pull/32429\n(cherry picked from commit 4fb28c19f5e2e9b9139c603b224a7142996f7ada)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "bbee82b00c77e2593e9f1127b87a636c45a6f754", "date": "2026-08-19T15:49:32+08:00", "subject": "docs: correct Tabs tabPosition API type", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "d3fde917f036b9473d2441b8862c571d396784da", "date": "2026-08-19T00:52:15-07:00", "subject": "fix(runtime-vapor): re-resolve transition hooks on prop change", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "36dd18616f664f88a262fb20b609a6071dbf3c20", "date": "2026-08-19T00:55:33-07:00", "subject": "fix(runtime-vapor): reflow the TransitionGroup's own document for moves", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "f2fa54df3e29511111be92263e154442e4f89293", "date": "2026-08-19T00:58:22-07:00", "subject": "fix(runtime-vapor): flush pending transition cbs on the resolved element", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "2c914ef8a2b9cda8af0a029fac2083be321d268d", "date": "2026-08-19T01:03:05-07:00", "subject": "perf(runtime-vapor): skip TransitionGroup bookkeeping on ForBlock wrappers", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "e0c246ce6abdb72df49b1e94cd5a362bceb18aaa", "date": "2026-08-19T16:06:57+08:00", "subject": "docs: correct Table rowSelection selections default value", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "744d64dbac14c806003970a1214b69f84451eeae", "date": "2026-08-19T01:11:22-07:00", "subject": "perf(runtime-vapor): collect-only child snapshot in TransitionGroup beforeUpdate", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "2854ac505201ac48e31618ab93a1c81a31a39acf", "date": "2026-08-19T16:13:42+08:00", "subject": "docs: add explicit anchors to Chinese headings", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "495d9ff5a7d843ca876a9e49799947a5deb704c7", "date": "2026-08-19T17:20:06+09:00", "subject": "feat(deps): widen `@vitejs/devtools` peer range to v0.5.0", "body": "", "style": "conventional", "type": "feat", "scope": "deps", "breaking": false} {"repo": "ant-design", "sha": "7bdc4a450473fe5f09403345547bbcf2e3708121", "date": "2026-08-19T16:21:23+08:00", "subject": "fix(Cascader): respect null notFoundContent", "body": "", "style": "conventional", "type": "fix", "scope": "Cascader", "breaking": false} {"repo": "ant-design", "sha": "761839eb1588305fcc30d16c4710978462c4533c", "date": "2026-08-19T16:22:34+08:00", "subject": "fix(Tree): handle zero key in DirectoryTree range selection", "body": "", "style": "conventional", "type": "fix", "scope": "Tree", "breaking": false} {"repo": "vue-core", "sha": "c526d45227770a533105bf95045f13b5d83c9cfb", "date": "2026-08-19T01:27:56-07:00", "subject": "perf(runtime-vapor): trim TransitionGroup per-child props tracking", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "9ed5abce3844899c02f97b48b5fb59235efeb9d9", "date": "2026-08-19T16:29:11+08:00", "subject": "fix(Menu): upgrade dependencies for unique popup ids", "body": "* fix(Menu): upgrade dependencies for unique popup ids\n\n* test(Layout): update Menu id snapshots\n\n* test: update dependent Menu id snapshots\n\n* test(Dropdown): update semantic Menu id snapshot", "style": "conventional", "type": "fix", "scope": "Menu", "breaking": false} {"repo": "tmux", "sha": "5047e4eb1a2d0ecdde2148a8b7a82c88d3fd500b", "date": "2026-08-19T09:34:39+01:00", "subject": "Add a way to make a modal pane capture all key presses and use for display-popup compatibility.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "8f666e5dcb97ba33345ed9999b51c29903bc0180", "date": "2026-08-19T12:16:11+03:00", "subject": "refactor(docs-infra): add Angie to the roadmap illustration", "body": "Backport of #70224, which landed in main as 1c2b0575ed with a feat\ncommit and so is only eligible for minor.\n\nadev deploys main to next.angular.dev and the stable release branch to\nangular.dev, so the mascot is live on next but not on the site itself.\nTyped as a refactor here so it can ride a patch. The file is byte\nidentical to the one that landed in main.", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "vite", "sha": "05a003e6a17a84d75f907ea0f1598bc39b8dce6c", "date": "2026-08-19T14:19:25+05:00", "subject": "fix(config): resolve sourcemap paths against sourcemap location", "body": "", "style": "conventional", "type": "fix", "scope": "config", "breaking": false} {"repo": "angular", "sha": "6421832b6deaa3748d8c653450ff0e95ff31e6c6", "date": "2026-08-19T12:27:33+03:00", "subject": "fix(docs-infra): remove stray tokens from adev illustrations", "body": "`routing.svg` and `directives.svg` each carry leftover text after a\n`stroke-miterlimit` attribute, which has made them invalid XML since the\nangular.dev import. They render only because decorative headers inline\nthem into the page, where the HTML parser turns the leftovers into empty\nattributes, but the files fail anywhere they are parsed as XML,\nincluding when served directly as `image/svg+xml`.\n\nNeither drops anything from the drawing. The `directives.svg` leftover\nrepeats a path the file already declares.\n\n(cherry picked from commit f35a3961c422e8d26892af21672841597e00486f)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "f35a3961c422e8d26892af21672841597e00486f", "date": "2026-08-19T12:27:33+03:00", "subject": "fix(docs-infra): remove stray tokens from adev illustrations", "body": "`routing.svg` and `directives.svg` each carry leftover text after a\n`stroke-miterlimit` attribute, which has made them invalid XML since the\nangular.dev import. They render only because decorative headers inline\nthem into the page, where the HTML parser turns the leftovers into empty\nattributes, but the files fail anywhere they are parsed as XML,\nincluding when served directly as `image/svg+xml`.\n\nNeither drops anything from the drawing. The `directives.svg` leftover\nrepeats a path the file already declares.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "023a63df8c112946b57666c55826983e699f41eb", "date": "2026-08-19T12:33:38+03:00", "subject": "refactor(devtools): unsubs from profiler when the perf track is disabled", "body": "Avoid the leak by unsubscribing from the profiler when the performance track is disabled.\nAdditionally, drop the duplicated performance track state flag.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "openssl", "sha": "461c7e8b2818b9dac8b4307b9c69751f8ea3b792", "date": "2026-08-19T11:59:45+02:00", "subject": "Documentation for X509V3_EXT_nconf and X509V3_EXT_nconf_int", "body": "Reviewed-by: Mounir Idrassi \nReviewed-by: Simo Sorce \nMergeDate: Thu Sep 10 19:05:26 2026\n(Merged from https://github.com/openssl/openssl/pull/31636)\n\n(cherry picked from commit 2564382c0bfd5c8d5cfb28671457878ee61ed6df)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2564382c0bfd5c8d5cfb28671457878ee61ed6df", "date": "2026-08-19T11:59:45+02:00", "subject": "Documentation for X509V3_EXT_nconf and X509V3_EXT_nconf_int", "body": "Reviewed-by: Mounir Idrassi \nReviewed-by: Simo Sorce \nMergeDate: Thu Sep 10 19:05:26 2026\n(Merged from https://github.com/openssl/openssl/pull/31636)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "7ec81e23008a8134a2ae326fe8f240375a799221", "date": "2026-08-19T18:18:29+08:00", "subject": "ci: use `$/` syntax to reference local action and remove actionlint", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9bb8e16bd53e2c8a822bf13832c4dcb3905a34e5", "date": "2026-08-19T12:25:06+02:00", "subject": "Tighten ACL check in repack_is_permitted_for_relation()", "body": "repack_is_permitted_for_relation() uses pg_class_aclcheck_ext()\nto silently skip a concurrently-dropped relation. That's wrong\nfor a caller that may already hold a lock on the relation whose\nACL is checked, where missing a relation is not fine, and it\nmakes the single-relation REPACK and CLUSTER cases more brittle.\nSo only detect a missing relation where that's expected,\nfollowing the fix for vacuum_is_permitted_for_relation() in\ncommit 824d5f6241ea.\n\nThe new already_locked behavior is limited to get_tables_to_repack()\nand get_tables_to_repack_partitioned(). All other callers of\nrepack_is_permitted_for_relation() hold a lock on the relation\nthat prevents it from being concurrently dropped, so this commit\nalso adds an assertion to that effect.\n\nWhile at it, update the comment in RangeVarCallbackMaintainsTable to\nalso mention REPACK.\n\nAuthor: Bharath Rupireddy \nBackpatch-through: 19\nDiscussion: https://www.postgresql.org/message-id/CALj2ACX3pyuRS8%2B%2B6L20cJUMRTf_qbbVp69J1btJ3y6%3D77e5gw%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0ee83dd4a9940883b5241fc9c80fd27f63f2e47f", "date": "2026-08-19T13:29:47+03:00", "subject": "postgres_fdw: push down FUNCTION RTE into foreign joins", "body": "A foreign join planning hook now considers a (foreign-table x function-RTE)\nINNER join as a push-down candidate when the function expression is\nIMMUTABLE and otherwise shippable. The remote query absorbs the function\ncall as a FROM-list item (e.g. unnest(...) AS f(c1, c2, ...)), so the\nforeign side returns only rows that match the function-produced set and the\njoin executes entirely on the remote.\n\nAn IMMUTABLE function gives the same result on any server, so the same\nfunction RTE can be a push-down candidate for several distinct foreign\nservers without semantic risk. To keep the planner state consistent\nacross those independent attempts, the per-call stub fpinfo for the\nfunction side lives on the joinrel's PgFdwRelationInfo (new\nouter_func_fpinfo / inner_func_fpinfo), never on the function rel itself,\nand the function side is detected via rtekind rather than fdw_private.\n\nset_foreign_rel_properties() propagates fdwroutine onto a joinrel that\npairs a foreign rel with an RTE_FUNCTION rel so GetForeignJoinPaths gets\ncalled; the FDW retains full control over whether to actually generate a\npath. deparseRangeTblRef and deparseColumnRef gain a FUNCTION-RTE branch\nthat emits the function expression and resolves Vars to the generated\ncolumn aliases.\n\nDiscussion: https://postgr.es/m/e3af56f9b2f1bd9fdf12aff6ca25b18d%40postgrespro.ru\nAuthor: Alexander Pyhalov ", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "e216c46b1e435efb6ed5e2bea6a2cea350292961", "date": "2026-08-19T18:53:43+08:00", "subject": "docs(Slider): document range object type", "body": "", "style": "conventional", "type": "docs", "scope": "Slider", "breaking": false} {"repo": "tmux", "sha": "b28b1b7b9a628c6c1dfa435a51f28e990d1cd8e1", "date": "2026-08-19T10:56:10Z", "subject": "Add a -K flag to new-pane to make a modal pane capture all keys like popups used to.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "8499d8007e90fcc4e7c0f3b3e35490a025b09662", "date": "2026-08-19T11:56:27+01:00", "subject": "new-pane -K test bits.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "e1ae7b7f63f1f57de29b1ac538673f0891306270", "date": "2026-08-19T13:01:51+02:00", "subject": "CVE-2026-8458: wrong reuse for different services", "body": "Cherry-picked from 308adae518b9f3f7057 (which is the fix from the 8.20.X\nbranch) and edited. Fixed differently than official patch due to lack of\nCurl_creds", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c0d796d61674e84b1aad7b30adc6a188060766cb", "date": "2026-08-19T13:01:51+02:00", "subject": "CVE-2026-8458: wrong reuse for different services", "body": "Fixed differently than official patch due to lack of Curl_creds", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "5b1c21f4b89f9afdabf524e55887a9c1c0052069", "date": "2026-08-19T13:03:01+02:00", "subject": "fix(lifecycle): late signal kills Nvim mid-teardown", "body": "Problem:\nA deadly signal arriving during teardown can kill Nvim while it is\npreserving swapfiles. `os_exit()` ignores deadly signals via\n`signal_reject_deadly()`, but `signal_teardown()` => `uv_signal_stop()`\nresets them to the default behavior, so SIGHUP arriving after that\nkills the process:\n\n [Process exited 129] // 128 + SIGHUP\n\nThis is a race when closing a pty: kernel sends SIGHUP to foreground\nprocess group *and* the reads return EOF, so `chanclose()` on a TUI job\nprepares to exit twice. This means it is unpredictable whether Nvim\nexits 1 or is terminated.\n\nSolution:\nIgnore deadly signals once the watchers are closed. Only SIGKILL\ninterrupts it now.", "style": "conventional", "type": "fix", "scope": "lifecycle", "breaking": false} {"repo": "curl", "sha": "775fe3b89923302125ea956ce8af4391f77555c8", "date": "2026-08-19T13:33:14+02:00", "subject": "CVE-2026-8926: password leak with netrc and user in URL", "body": "Fix implemented differently due to lack of Curl_creds - and the original\nfix is a larger netrc retake.\n\nTest 1304 extended in similar spirit as in the original patch.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "743bca9202e96869b102d9883c40cf778dbd0df3", "date": "2026-08-19T21:35:26+10:00", "subject": "fix(zip): lost password if sent before unzip disables echo #41377", "body": "", "style": "conventional", "type": "fix", "scope": "zip", "breaking": false} {"repo": "vite", "sha": "8413052731836d4aaf3eb94a0f25788dd35d2888", "date": "2026-08-19T16:50:18+05:00", "subject": "fix: respect `resolve.preserveSymlinks` when resolving root (fix #23197)", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "openssl", "sha": "de821dc99741904abb402642de0f807f062ea37d", "date": "2026-08-19T13:53:37+02:00", "subject": "file_store_any2obj.c: validate blob length before allocation", "body": "Both msblob2obj_decode() and pvk2obj_decode() computed an\nattacker-influenced buffer size (from `bitlen`, or from\n`saltlen + keylen`) and passed it to BUF_MEM_grow() without an\nupper bound.\n\nResolves #31684\n\nFixes: f40c5f2c53b9 \"PROV & STORE: Make the 'file:' store loader understand more binary formats\"\nFixes: 0f5493e60b9e \"storemgmt: Validate msblob length before buffer allocation\"\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Neil Horman \nReviewed-by: Matt Caswell \nMerge-date: Fri Aug 21 10:11:10 2026\nMerged-from: https://github.com/openssl/openssl/pull/31704", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "be7ac194ea65bc68f6797a703f916cd9503bec3c", "date": "2026-08-19T13:53:37+02:00", "subject": "file_store_any2obj.c: validate blob length before allocation", "body": "Both msblob2obj_decode() and pvk2obj_decode() computed an\nattacker-influenced buffer size (from `bitlen`, or from\n`saltlen + keylen`) and passed it to BUF_MEM_grow() without an\nupper bound.\n\nResolves #31684\n\nFixes: f40c5f2c53b9 \"PROV & STORE: Make the 'file:' store loader understand more binary formats\"\nFixes: 0f5493e60b9e \"storemgmt: Validate msblob length before buffer allocation\"\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Neil Horman \nReviewed-by: Matt Caswell \nMerge-date: Fri Aug 21 10:10:52 2026\nMerged-from: https://github.com/openssl/openssl/pull/31704", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b584ce6ec9693d40d063052e8aca976b9b0eff24", "date": "2026-08-19T13:53:37+02:00", "subject": "file_store_any2obj.c: validate blob length before allocation", "body": "Both msblob2obj_decode() and pvk2obj_decode() computed an\nattacker-influenced buffer size (from `bitlen`, or from\n`saltlen + keylen`) and passed it to BUF_MEM_grow() without an\nupper bound.\n\nResolves #31684\n\nFixes: f40c5f2c53b9 \"PROV & STORE: Make the 'file:' store loader understand more binary formats\"\nFixes: 0f5493e60b9e \"storemgmt: Validate msblob length before buffer allocation\"\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Neil Horman \nReviewed-by: Matt Caswell \nMerge-date: Fri Aug 21 10:10:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/31704", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7fae8613488714c124b83d21de1d163f73a81f7b", "date": "2026-08-19T13:53:37+02:00", "subject": "file_store_any2obj.c: validate blob length before allocation", "body": "Both msblob2obj_decode() and pvk2obj_decode() computed an\nattacker-influenced buffer size (from `bitlen`, or from\n`saltlen + keylen`) and passed it to BUF_MEM_grow() without an\nupper bound.\n\nResolves #31684\n\nFixes: f40c5f2c53b9 \"PROV & STORE: Make the 'file:' store loader understand more binary formats\"\nFixes: 0f5493e60b9e \"storemgmt: Validate msblob length before buffer allocation\"\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Neil Horman \nReviewed-by: Matt Caswell \nMerge-date: Fri Aug 21 10:10:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/31704", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "46970ba0191afd6271c1d6c162db891ae9b592bb", "date": "2026-08-19T13:53:37+02:00", "subject": "file_store_any2obj.c: validate blob length before allocation", "body": "Both msblob2obj_decode() and pvk2obj_decode() computed an\nattacker-influenced buffer size (from `bitlen`, or from\n`saltlen + keylen`) and passed it to BUF_MEM_grow() without an\nupper bound.\n\nResolves #31684\n\nFixes: f40c5f2c53b9 \"PROV & STORE: Make the 'file:' store loader understand more binary formats\"\nFixes: 0f5493e60b9e \"storemgmt: Validate msblob length before buffer allocation\"\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Neil Horman \nReviewed-by: Matt Caswell \nMerge-date: Fri Aug 21 10:09:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/31704", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d078067a26422aa1a1ce5a4cdba0d0da69f8ff1e", "date": "2026-08-19T13:53:37+02:00", "subject": "file_store_any2obj.c: validate blob length before allocation", "body": "Both msblob2obj_decode() and pvk2obj_decode() computed an\nattacker-influenced buffer size (from `bitlen`, or from\n`saltlen + keylen`) and passed it to BUF_MEM_grow() without an\nupper bound.\n\nResolves #31684\n\nFixes: f40c5f2c53b9 \"PROV & STORE: Make the 'file:' store loader understand more binary formats\"\nFixes: 0f5493e60b9e \"storemgmt: Validate msblob length before buffer allocation\"\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Neil Horman \nReviewed-by: Matt Caswell \nMerge-date: Fri Aug 21 10:09:11 2026\nMerged-from: https://github.com/openssl/openssl/pull/31704", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2924476b5591e691e904c4baf57894c526c4b8de", "date": "2026-08-19T13:54:30+02:00", "subject": "test: forward real stdio calls in bss_file unit test mocks", "body": "The --wrap link option rewrites the stdio symbols of everything linked\ninto the test binary. In --coverage builds libgcov reads and writes the\n.gcda files at process exit through fread/fwrite/fseek/ftell/fclose,\nwhich then hit the cmocka mocks with no expectations queued and make\nthe binary exit with 255 after all tests have passed. Forward calls on\nany stream other than the fake test pointers to the real functions.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Thu Aug 20 07:50:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32431", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vite", "sha": "bdfadefb207039bdb06d547da67dc81a92fe8750", "date": "2026-08-19T19:55:56+08:00", "subject": "test(bundled-dev): run hmr-ssr spec in bundled dev mode under test-serve-bundled", "body": "", "style": "conventional", "type": "test", "scope": "bundled-dev", "breaking": false} {"repo": "curl", "sha": "24e012595965b58f096752e57b702ef9a08dd929", "date": "2026-08-19T13:57:25+02:00", "subject": "CVE-2026-8932: incomplete mTLS config matching in conn reuse", "body": "Cherry-picked 7541ae569d82fb308a5e2d then edited", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "211a195acb79f87b320d7567e76249b62ce3c244", "date": "2026-08-19T20:13:08+08:00", "subject": "fix(window): crash when closing new curwin during CTRL-W_x", "body": "Problem: Crash when closing the new current window during CTRL-W_x.\nSolution: Remove a duplicate redraw of a window that should normally be\n already mark for redraw in the previous call.\n(cherry picked from commit 2cd4229d01df4176527e8e8a04eb82abcfa7e2a1)", "style": "conventional", "type": "fix", "scope": "window", "breaking": false} {"repo": "neovim", "sha": "2cd4229d01df4176527e8e8a04eb82abcfa7e2a1", "date": "2026-08-19T20:13:08+08:00", "subject": "fix(window): crash when closing new curwin during CTRL-W_x", "body": "Problem: Crash when closing the new current window during CTRL-W_x.\nSolution: Remove a duplicate redraw of a window that should normally be\n already mark for redraw in the previous call.", "style": "conventional", "type": "fix", "scope": "window", "breaking": false} {"repo": "neovim", "sha": "5a71131282010225745d94b068a95fc7eb4db2a9", "date": "2026-08-19T14:18:26+02:00", "subject": "fix(lifecycle): on Windows, CTRL_CLOSE kills Nvim mid-teardown", "body": "Problem:\n`signal_ignore_deadly` doesn't work for Windows, where the console still\nmay terminate Nvim during teardown (after `signal_teardown`), while it\nis already trying to exit. Besides interrupting any housekeeping we are\ndoing, it results in an unpredictable exit code (flaky tests).\n\n [Process exited -1073741510] // 0xC000013A STATUS_CONTROL_C_EXIT\n\nSolution:\nRegister our own CTRL_CLOSE_EVENT handler which \"blocks\" the signal.\nNote: if exit takes longer than 5s, Windows will consider the process\n\"hung\" and kill it anyway.", "style": "conventional", "type": "fix", "scope": "lifecycle", "breaking": false} {"repo": "openssl", "sha": "8c85b6bb88f52108b3a45a2ef0fbb4cd8de9aadc", "date": "2026-08-19T15:13:46+02:00", "subject": "ci: use long calls for HPPA cross-compilation", "body": "The QUIC radix test translation unit is now large enough to exceed the\nPA-RISC 1.x short-call range. This causes the HPPA test binary to branch\ninto data and terminate with SIGILL under QEMU.\n\nBuild the HPPA CI target with -mlong-calls so calls can always reach\nlinker-generated stubs.\n\nAssisted-by: Codex:gpt-5\nSigned-off-by: Nikola Pajkovsky \nReviewed-by: Matt Caswell \nReviewed-by: Bob Beck \nMerge-date: Wed Aug 19 13:47:32 2026\nMerged-from: https://github.com/openssl/openssl/pull/32432", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9716f88ddf6b468d191a93e5f9fea1c6f98cc1d1", "date": "2026-08-19T17:30:48+03:00", "subject": "Fix compilation warning in 0ee83dd4a994", "body": "Discussion: https://postgr.es/m/CAO6_XqrGH2NRe0oFx_00nLXra1UZezd3jbS9pgoRhZtgzHD9gg%40mail.gmail.com\nAuthor: Anthonin Bonnefoy ", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "588dc547539787bf1d719a9bed97133fe6802e6c", "date": "2026-08-19T15:46:09+01:00", "subject": "WIP: preserve image redraw experiment and recovery plan", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "93d7f718d2e057e37deda42d197e46672fe7f7f5", "date": "2026-08-19T16:54:03+02:00", "subject": "fix(language-service): honor quote style preference when generating imports", "body": "The quick fix and completion auto-import always generated the module specifier with double quotes, ignoring the user's quote preference and the style used in the file.\n\nThe generated import now respects the `quotePreference` from the TypeScript user preferences. When the preference is `auto` (or absent), the style is inferred from the first existing import in the file, mirroring TypeScript's own behavior. Files with no imports\nkeep the previous double-quote default.\n\nFixes #67108\n\n(cherry picked from commit 0f0d52e9c1154f8db9a4a8d9d0ff73f0afdc7031)", "style": "conventional", "type": "fix", "scope": "language-service", "breaking": false} {"repo": "angular", "sha": "0f0d52e9c1154f8db9a4a8d9d0ff73f0afdc7031", "date": "2026-08-19T16:54:03+02:00", "subject": "fix(language-service): honor quote style preference when generating imports", "body": "The quick fix and completion auto-import always generated the module specifier with double quotes, ignoring the user's quote preference and the style used in the file.\n\nThe generated import now respects the `quotePreference` from the TypeScript user preferences. When the preference is `auto` (or absent), the style is inferred from the first existing import in the file, mirroring TypeScript's own behavior. Files with no imports\nkeep the previous double-quote default.\n\nFixes #67108", "style": "conventional", "type": "fix", "scope": "language-service", "breaking": false} {"repo": "angular", "sha": "7c752d4815db41ee04e43d38cc826dd63214432b", "date": "2026-08-19T10:08:06-05:00", "subject": "fix(core): prevent TransferState prototype pollution", "body": "Store transfer state and serialization callbacks in null-prototype dictionaries, and only read values that belong to the store. This keeps special state keys from changing lookup behavior or exposing inherited cache entries.\n\nFixes #70265\n\n(cherry picked from commit 168a324cce0f133ed92953374c2510a75868b790)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "168a324cce0f133ed92953374c2510a75868b790", "date": "2026-08-19T10:08:06-05:00", "subject": "fix(core): prevent TransferState prototype pollution", "body": "Store transfer state and serialization callbacks in null-prototype dictionaries, and only read values that belong to the store. This keeps special state keys from changing lookup behavior or exposing inherited cache entries.\n\nFixes #70265", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "openssl", "sha": "03f200cabbf064ddda03160056fdb82bfc0c2612", "date": "2026-08-19T16:32:56+01:00", "subject": "Make the TLS 1.3 resumption PSK check aware of DTLS 1.3", "body": "tls13_check_resumption_psk() tells tls_construct_ctos_early_data()\nwhether the resumption session can key 0-RTT, but compared the session\nversion against TLS1_3_VERSION only. A DTLS 1.3 session records\nDTLS1_3_VERSION, so the check never matched for DTLS.\n\nThe caller then fell back to the external PSK session, which is NULL\nwhen resuming, and took the suppression path: max_early_data was set to\n0, early_data_suppressed was set, and the early_data extension was not\nsent. SSL_connect() ran an ordinary handshake and returned WANT_READ\nwith an empty error queue, so SSL_write_early_data() returned 0 with\nnothing to indicate why.\n\nCompare against DTLS1_3_VERSION for DTLS, using the version1_3 idiom\nthe calling function already uses.\n\nThe helper is newer than the DTLS 1.3 work: it was factored out of\ntls_construct_ctos_early_data() after that function had been made\nversion aware, so it never picked up the DTLS case.\n\nFixes the DTLS 1.3 iterations of nine sslapitest early data tests.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Tomas Mraz \nMerge-date: Fri Aug 21 18:10:39 2026\nMerged-from: https://github.com/openssl/openssl/pull/32441", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "925cda0d58c143773ba1f54d1fb33e99b1e330b7", "date": "2026-08-19T16:43:21+01:00", "subject": "Use the SSL_CTX cached HMAC for DTLS listener cookies", "body": "dtls_listener_cookie_hmac() fetched an EVP_MAC for \"HMAC\" on every call.\nIt runs once per cookie generation and once per verification, so a DTLS\nlistener paid a provider fetch for every ClientHello it answered.\n\nSSL_CTX already holds a pre-fetched EVP_MAC for exactly this purpose.\nSSL_CTX_new() fetches it up front and fails if it cannot, so ctx->hmac\nis always usable, and ssl_hmac_init() in ssl/t1_lib.c uses it the same\nway. Take the MAC from there instead, which also removes the local\nfetch and free.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Tomas Mraz \nMerge-date: Fri Aug 21 18:10:40 2026\nMerged-from: https://github.com/openssl/openssl/pull/32441", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "f31251b6451c6f5da78457273ff9307c2be0ecb7", "date": "2026-08-19T09:27:38-07:00", "subject": "test_07_upload: skip tests if nghttpx lacks QUIC", "body": "If nghttpx is available but lacks QUIC support, disable the tests that\ndepend on that support.\n\nCo-authored-by: Stefan Eissing\nFixes #22609\nCloses #22624", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a13cdbd3d8b628b49219f70be2772230d36fa565", "date": "2026-08-19T12:31:09-04:00", "subject": "Fix GIN VACUUM posting tree root split bug.", "body": "ginVacuumPostingTreeLeaves swaps a shared buffer lock for an exclusive\none when it encounters a leaf page. It neglected to re-verify whether a\npage that was initially a leaf root page became an internal page due to\na concurrent root page split (during the window when no lock was held).\nIt was therefore possible for GIN VACUUM to spuriously treat an internal\npage as a leaf page, leading to data corruption. VACUUM could miss dead\nTIDs that it was required to remove, leaving behind dangling references\nin the index.\n\nTo fix, re-verify that a leaf page is still a leaf page after an\nexclusive lock is acquired. If it isn't, drop our exclusive lock and\nacquire a shared lock so that the non-leaf root page gets processed in\nthe usual way.\n\nOversight in commit fd83c83d, which fixed a deadlock bug in GIN posting\ntree vacuuming.\n\nAuthor: Peter Geoghegan \nReviewed-by: Andrey Borodin \nDiscussion: https://postgr.es/m/CAH2-Wz=RBpJTQgvOxr6C=J04dExmFSt1E3F-r+cRTQ56hEotkg@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "8aa50462c859053a4f1f780ffad45cc5dda9e267", "date": "2026-08-19T16:36:19Z", "subject": "build: update cross-repo angular dependencies to v22.1.3", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "96f98e0e15bd64eba06528d42d86b78677f53426", "date": "2026-08-19T18:47:42+02:00", "subject": "test: fix TAP stream corruption in 20-test_pkeyutl.t", "body": "'pkeyutl -verifyrecover' outputs binary on stdout; redirect it.\n\nFixes: https://github.com/openssl/openssl/issues/32428\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Mounir Idrassi \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 14:30:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32436", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "3ca22c7b7ccb1f8897ca6d05f4e6c4934c537eeb", "date": "2026-08-19T10:36:44-07:00", "subject": "pg_locale.c, unicode_case.c: use size_t for iteration.", "body": "Already done by e615da8cb21b in master. Backpatch these particular\ncases to make other backpatches in this area safer, and to be safe for\ncallers in extensions that we don't know about.\n\nThis applies to REL_19_STABLE and REL_18_STABLE only.\n\nSuggested-by: Andres Freund \nReviewed-by: Andres Freund \nDiscussion: https://postgr.es/m/v36ssaygf7grb3qzfsjhtdzi7kqd45ds56nyuf7gi5qjml4qbb@ezmfqzmhlrs2", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "8637383f5ff1cb7eefd0f9ea4e3da81a70ca206b", "date": "2026-08-19T13:46:45-04:00", "subject": "Fix GIN multiple-VACUUM-scans pending list bug.", "body": "ginbulkdelete performs pending list cleanup before it searches the entry\ntree (and any posting trees) for dead TIDs. This is necessary to avoid\nleaving behind dangling TID references that index vacuuming is required\nto remove; nothing prevents recently inserted pending list tuples from\ncontaining TIDs that VACUUM already considers dead.\n\nHowever, ginbulkdelete neglected to perform pending list cleanup on\nVACUUM's second or subsequent call. It was therefore possible for a\nVACUUM that requires multiple rounds of index vacuuming to leave behind\ndangling references.\n\nTo fix, teach ginbulkdelete to perform pending list cleanup during every\ncall. In passing, tweak some related comments in the pending list\ncleanup path to make it clear why it's safe for VACUUM to not _fully_\nempty an index's pending list.\n\nThis was arguably an oversight in commit e2c79e14, which fixed a similar\nissue where pending list cleanup by VACUUM could end early, but missed\nthis closely related problem.\n\nAuthor: Peter Geoghegan \nReviewed-by: Andrey Borodin \nDiscussion: https://postgr.es/m/CAH2-Wzmsa-RPA2Ko8A5LaGOnmbpimJ--71xkiBqwgjk3Fq8YEg@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "74b732f63792317a9ca8ad7877729cad445bc985", "date": "2026-08-19T21:53:44+04:00", "subject": "tool_msgs: make notef() respect --silent", "body": "Until 8.17.0 the --silent check lived in voutf(), so it covered every\nmessage type that went through it. 56450ce26f moved that check out to\nthe callers so that errorf() could honour --show-error, and warnf() and\nerrorf() got the check back, but notef() did not. Since then \"Note:\"\nmessages have been printed even with --silent, as long as tracing was\non.\n\nReported-by: Jace Laquerre\nFixes #22623\nCloses #22625", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ccab6d36d0cbe554c380f69ade1598a41c06552a", "date": "2026-08-19T14:06:52-04:00", "subject": "Support tied Perl objects in plperl and associated contrib modules.", "body": "Up to now, if you passed a \"tied\" object to a PL/Perl operation,\nit would most likely get treated as \"undef\", since we failed to\nperform get magic on values we're passed. Tied hashes worked\npartially, in that you could read out their keys, but the values\nalways read as undef.\n\nTo fix, use hv_iterval() not HeVAL() to fetch values from perl\nhashes, and be careful to apply plperl_materialize_sv() before\ntesting SvOK() or SvROK() on any perl SV.\n\nAdd regression test cases exercising some typical usage scenarios\nfor tied objects.\n\nArguably this is a bug fix, but in view of the lack of field\ncomplaints to date, let's treat it as a new feature instead.\nThe risk/reward ratio for back-patching isn't attractive.\n\nAuthor: Andrey Rachitskiy \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/569769.1786901901@sss.pgh.pa.us", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "1fb3b58abe79c4cb1653cb4d4f8b6baac5c28c02", "date": "2026-08-19T18:36:09Z", "subject": "build: update cross-repo angular dependencies to v22.2.0-next.2", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "cb217c4f174f487009bcba3812bbf23b8d8eba4c", "date": "2026-08-19T15:45:56-04:00", "subject": "GiST: Invalidate killed items consistently.", "body": "GiST neglected to invalidate its killedItems[] array on a rescan. As a\nresult, it was just about possible for the wrong tuples from the wrong\nindex page to be LP_DEAD-marked on a rescan. The scan mistakenly\nbelieved that the previous rescan's killedItems[] were for this rescan's\ncurBlkno, causing index corruption.\n\nTo fix, bring GiST in line with nbtree and hash: call gistkillitems from\nboth gistrescan and gistendscan (the existing gistgettuple caller still\nhandles the common case where we need to LP_DEAD-mark before moving on\nto the next page). That way the scan's pending killedItems[] are passed\nto gistkillitems while they still describe items from curBlkno. When\ngistkillitems runs, it'll invalidate the array in passing (and won't\nneedlessly miss out on an opportunity to LP_DEAD-mark eligible index\ntuples). Back branches just get minimal hardening: we invalidate\nkilledItems[] at the places where the master branch gets new calls to\ngistkillitems (and we invalidate curBlkno and curPageLSN on a rescan).\n\nThe test that proved corruption on master didn't result in corruption on\nany stable branch, though only because, without commit 9c9ddf109, we'd\nclobber curPageLSN without also updating curBlkno -- which accidentally\nprevented it. Relying on gistkillitems to not LP_DEAD-mark by passing\nit a curBlkno whose curPageLSN was taken from an entirely different page\nseems like a very bad idea, which is why this issue is being treated as\na bug affecting all stable branches.\n\nAuthor: Peter Geoghegan \nReviewed-By: Andrey Borodin \nDiscussion: https://postgr.es/m/CAH2-WzmwEThnQf17Ju+t0N9_KJLsEQSXzYrFnaS2=s4KnGGrqw@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d064df18af85426f7b152ec77c2834d19e87c840", "date": "2026-08-19T13:54:48-06:00", "subject": "Clamp negative lengths to 0 in ASN1_STRING_set0", "body": "A negative length is never valid here, so store an empty string\ninstead, while taking ownership of the data pointer as usual.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 27 13:52:55 2026\nMerged-from: https://github.com/openssl/openssl/pull/32178", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "50d0348b648258f11d75ec751581220884ff32a3", "date": "2026-08-19T19:55:04Z", "subject": "Stop at the right end line when walking wrapped lines, GitHub issue 5479.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "60a874c3fbedc2ce216bcf783325f4e1291edcb0", "date": "2026-08-19T22:28:04+02:00", "subject": "fix(common): avoid prototype member collisions", "body": "Switches to using an object with a `null` prototype to avoid collisions.\n\n(cherry picked from commit 862a0c8ab3403242eeac3b2fb80f7cf63cf0c892)", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "angular", "sha": "862a0c8ab3403242eeac3b2fb80f7cf63cf0c892", "date": "2026-08-19T22:28:04+02:00", "subject": "fix(common): avoid prototype member collisions", "body": "Switches to using an object with a `null` prototype to avoid collisions.", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "angular", "sha": "3da35cbab94777230871a3e9289455089b80fe7f", "date": "2026-08-19T22:28:11+02:00", "subject": "fix(core): avoid prototype member collisions", "body": "Switches to using an object with a `null` prototype to avoid collisions.\n\n(cherry picked from commit 7ba082e08e59eadec4594ad80661d5c1ec08cb2a)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "7ba082e08e59eadec4594ad80661d5c1ec08cb2a", "date": "2026-08-19T22:28:11+02:00", "subject": "fix(core): avoid prototype member collisions", "body": "Switches to using an object with a `null` prototype to avoid collisions.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "004cf3a27734ae90738a0a745cc0369b52306ca3", "date": "2026-08-19T20:37:39Z", "subject": "build: update cross-repo angular dependencies to v22.1.5", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "34f0e1315bfeea1b8a64909a0039ad7f5a35a0f0", "date": "2026-08-19T20:44:38Z", "subject": "docs: add guidelines for Object.create(null) PRs to PR review skill", "body": "(cherry picked from commit ad3c16562bba0c9895ee770f3d30604495888682)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "ad3c16562bba0c9895ee770f3d30604495888682", "date": "2026-08-19T20:44:38Z", "subject": "docs: add guidelines for Object.create(null) PRs to PR review skill", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "55eeb46418080e2faf64c84eac389c21abea6df8", "date": "2026-08-19T23:50:06+02:00", "subject": "fix(compiler-cli): retain metadata for strict standalone errors", "body": "This fix ensures that metadata is properly retained when processing\nstrict standalone component errors for improved error diagnostics.\n\n(cherry picked from commit 74b294cd51b0f73e3a71e0cd13a176b1d0f3a89f)", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "74b294cd51b0f73e3a71e0cd13a176b1d0f3a89f", "date": "2026-08-19T23:50:06+02:00", "subject": "fix(compiler-cli): retain metadata for strict standalone errors", "body": "This fix ensures that metadata is properly retained when processing\nstrict standalone component errors for improved error diagnostics.", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "curl", "sha": "5d6dc816785358f505ba6922ceb507b4521c5421", "date": "2026-08-20T00:33:42+02:00", "subject": "GHA/checksrc: preprocess workflows to lint more shell code", "body": "`zizmor` keeps being confused by non-\"well-known\" shell designators\n`msys2 {0}`, `\\bash.exe '{0}'`, `cpa.sh {0}`, while `actionlint`\nsilently skips checking such shell code. Though it's all POSIX/bash.\nReplace the unrecognized shell designators with `bash` before running\nthe linters, to remove these blind spots.\n\nzizmor pedantic persona:\nBefore: `No findings to report. Good job! (1 ignored, 61 suppressed)`\nAfter: `No findings to report. Good job! (1 ignored)`\n\nzizmor auditor persona:\nBefore: `No findings to report. Good job! (62 ignored)`\nAfter: `No findings to report. Good job! (1 ignored)`\n\nCloses #22626", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "9a8e4826b9db4359c3f0e7185e9582b7bd08a3ef", "date": "2026-08-20T01:05:06+02:00", "subject": "fix(core): preserve namespace for dynamic component hosts", "body": "Forward the insertion namespace when creating dynamic component hosts inside SVG or MathML.\n\n(cherry picked from commit 2ab5ff56ded8c6246c0035d9df62a0c7fc9e1308)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "2ab5ff56ded8c6246c0035d9df62a0c7fc9e1308", "date": "2026-08-20T01:05:06+02:00", "subject": "fix(core): preserve namespace for dynamic component hosts", "body": "Forward the insertion namespace when creating dynamic component hosts inside SVG or MathML.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "4434340b5769439cc31dbd068a8c7f336eef71ed", "date": "2026-08-20T09:38:36+09:00", "subject": "Reject too many arguments in CREATE TRIGGER", "body": "The number of trigger arguments is stored as a smallint, but there was\nno check that the number of arguments fits with the catalog data type.\nThis could result in an invalid negative value being stored once one\ndefined more than INT16_MAX arguments, with an overflowed value stored\nin the catalogs.\n\nLooking at other catalogs that store a number of arguments, we have\nsimilar protections already in place (aggregates, functions, etc.).\n\nReported-by: Xingwang Xiang \nAuthor: Kyotaro Horiguchi \nDiscussion: https://postgr.es/m/19627-5b72a57e332e2b3f@postgresql.org\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "41c2eb85cf6acc87e9a9ae48d04819fd694c64e7", "date": "2026-08-20T03:34:30+02:00", "subject": "ci: close automated prs and issues after 1 day", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "208d8454e2d386b29e6fdeaec5b651cc0c641d87", "date": "2026-08-20T09:32:53+07:00", "subject": "Fix tick race in radix fault-injection scripts", "body": "Guard any `OP_SET_INJECT_WORD`+`OP_WRITE` with 2 new functions\n`OP_ENGINE_TICK_ENABLE`/`OP_ENGINE_TICK_DISABLE` that enable/disable ticking\nin the QUIC engine. Otherwise, the radix test interpreter can tick between the\n2 ops and cause the peer to close the connection prematurely\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Tomas Mraz \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Aug 26 12:00:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/32406", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "1fb8afe2b3a386f9c879f80841bf5aefc732f807", "date": "2026-08-20T11:46:50+09:00", "subject": "chore(create-vite): mention experimental react compiler support", "body": "", "style": "conventional", "type": "chore", "scope": "create-vite", "breaking": false} {"repo": "vue-core", "sha": "89353dd145650a2583dc39eda7b3df02c228fc1d", "date": "2026-08-19T19:57:02-07:00", "subject": "test(runtime-vapor): split hydration.spec.ts", "body": "", "style": "conventional", "type": "test", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "b7b2841cec0cc10ae734b1ff839d702d959ac8bc", "date": "2026-08-19T20:18:17-07:00", "subject": "refactor(runtime-vapor): move fragment anchor category into __vf flags", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vite", "sha": "9db0b61d4c9c7caad7ea1d9670b637faf2bb6c93", "date": "2026-08-20T12:53:30+09:00", "subject": "fix(ssr): rewrite computed key of destructing parameter", "body": "", "style": "conventional", "type": "fix", "scope": "ssr", "breaking": false} {"repo": "vite", "sha": "cb77e2a93bad2a8ece00b4aa0ef507c092582c45", "date": "2026-08-20T12:53:42+09:00", "subject": "test(ssr): add destructing assignment case for moduleRunnerTransform", "body": "", "style": "conventional", "type": "test", "scope": "ssr", "breaking": false} {"repo": "postgres", "sha": "f3a52a229adeb9c54e5b656b45af609b967874d6", "date": "2026-08-20T14:02:57+09:00", "subject": "Restore after-trigger firing context at subtransaction end", "body": "AfterTriggerEndQuery(), AfterTriggerFireDeferred(), and\nAfterTriggerSetState() bracket their firing loops with\nfiring_depth++/--. The decrement runs after the loop and is not\nprotected by PG_FINALLY, so an error caught by a subtransaction (e.g. a\nPL/pgSQL EXCEPTION block) leaves firing_depth too high. Separately,\nAfterTriggerEndSubXact() unconditionally cleared\nfiring_batch_callbacks, even if the subtransaction began while an outer\nbatch-callback loop was active.\n\nfiring_depth feeds AfterTriggerIsActive(), which the RI fast path uses\nto decide whether an FK check is running inside trigger firing and may\nbatch. A stranded firing_depth makes AfterTriggerIsActive() wrongly\nreport firing as active afterwards. This is reachable and results in\nsilent data corruption: after a caught FK-check error, an ALTER TABLE ...\nADD FOREIGN KEY whose validation runs per-row (RI_Initial_Check() having\nbailed, e.g. because RLS is enabled on the referenced table) calls\nRI_FKey_check() with AfterTriggerIsActive() wrongly true. The check is\nrouted into the batched fast path, but a utility command has no\nAfterTriggerEndQuery() to fire the flush callback. The violating row is\nnot reported, the constraint is marked validated, and the cached PK\nrelation and index leak.\n\nSave firing_depth and firing_batch_callbacks at subtransaction start and\nrestore them in AfterTriggerEndSubXact(), next to the existing query_depth\nhandling. Restoring, rather than zeroing or clearing, is required because\na subtransaction can begin and end while an outer query is firing, where\nfiring_depth is legitimately positive and firing_batch_callbacks may be\nlegitimately set.\n\nReported-by: Noah Misch \nDiscussion: https://postgr.es/m/20260705222115.be.noahmisch@microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "axios", "sha": "4943cc2749393d55f0c8de28a5c57aca403aaeb7", "date": "2026-08-20T07:59:26+02:00", "subject": "docs: add codebase knowledge base", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "f405ae7a75143089e8c9d1b525a1db2d03c13005", "date": "2026-08-20T14:02:04+08:00", "subject": "fix(List): recommend Listy in deprecation warning", "body": "* fix(List): recommend Listy in deprecation warning\n\n* Update components/list/index.tsx\n\nCo-authored-by: lijianan <574980606@qq.com>\nSigned-off-by: 高艳兵 \n\n* fix(List): escape warning text and refresh snapshots\n\n---------", "style": "conventional", "type": "fix", "scope": "List", "breaking": false} {"repo": "openssl", "sha": "6a7aee9dcbabb5a8769df0bace304788c3c8e0b7", "date": "2026-08-20T08:04:26+02:00", "subject": "exporters/cmake/OpenSSLConfig.cmake.in: Added missing crypto libs.", "body": "OPENSSL_LIBRARIES only exported OPENSSL_SSL_LIBRARIES,\nwhich can cause application link failures when libcrypto\ndependencies are not propagated through `target_link_libraries()`.\n\nAppend `OPENSSL_CRYPTO_LIBRARIES` to ensure consumers linking\nagainst `OPENSSL_LIBRARIES` also receive the required crypto\nlibraries.\n\nFixes: https://github.com/openssl/openssl/issues/31318\nReviewed-by: Richard Levitte \nReviewed-by: Andrew Dinh \nReviewed-by: Mounir Idrassi \nMerge-date: Thu Aug 27 09:17:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/31363\n(cherry picked from commit 1b543af5e4fb0dd81a1fe7118bd9735c61fe05b7)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1b543af5e4fb0dd81a1fe7118bd9735c61fe05b7", "date": "2026-08-20T08:04:26+02:00", "subject": "exporters/cmake/OpenSSLConfig.cmake.in: Added missing crypto libs.", "body": "OPENSSL_LIBRARIES only exported OPENSSL_SSL_LIBRARIES,\nwhich can cause application link failures when libcrypto\ndependencies are not propagated through `target_link_libraries()`.\n\nAppend `OPENSSL_CRYPTO_LIBRARIES` to ensure consumers linking\nagainst `OPENSSL_LIBRARIES` also receive the required crypto\nlibraries.\n\nFixes: https://github.com/openssl/openssl/issues/31318\nReviewed-by: Richard Levitte \nReviewed-by: Andrew Dinh \nReviewed-by: Mounir Idrassi \nMerge-date: Thu Aug 27 09:17:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/31363", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "8bec2bd9e80d2a84333cfbdc7727bc473166928a", "date": "2026-08-20T14:04:55+08:00", "subject": "docs: update v4 migration guidance", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "6a1759a02aed282ab1af51207301974fa51bc33f", "date": "2026-08-20T15:17:32+09:00", "subject": "Fix postmaster failing to exit when startup crashes during crash restart", "body": "Commit 9b43e6793b0f removed the PM_STARTUP shortcut that exited the\npostmaster directly when the startup process died, routing the case\nthrough HandleChildCrash() so that children processes running during\nPM_STARTUP are cleaned up rather than orphaned.\n\nHowever, HandleChildCrash() does nothing if FatalError is already set,\nand that is exactly the case while reinitializing after a crash: a\nrelaunched startup process that dies before WAL redo starts would leave\nthe state machine stuck at PM_STARTUP, preventing the postmaster to shut\ndown.\n\nThis issue is fixed by restoring the pre-9b43e6793b0f shortcut behavior:\nif FatalError is set when the startup process crashes, signal the\nremaining children and move to PM_NO_CHILDREN, so as the postmaster can\nproperly exit with nothing orphaned. This is only reachable in v18 and\nnewer versions, the early return of HandleChildCrash() on FatalError\nbeing introduced in f0b7ab725139.\n\nThis issue has been reported on Windows, for a postmaster with its\nconsole gone. A trick to make InitPostmasterChild() fail aggressively\nwas equally able to stuck a postmaster.\n\nReported-by: Kuan-Ting Kuo \nAuthor: Zexin Li \nDiscussion: https://postgr.es/m/19623-f9bd331940be1273@postgresql.org\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "e86e35b1232de6f554b54d1ccf6118362ace94c3", "date": "2026-08-20T14:22:47+08:00", "subject": "fix(runtime-vapor): resolve element namespace at interop boundaries", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "47abd8cb986215ff6ea0e59725ff7ec74b3942ef", "date": "2026-08-20T08:52:58+02:00", "subject": "CVE-2026-11564: Native CA trust persist", "body": "Inspired by d69bfad3fa3daf5e72331, implemented differently", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c36a0df195dca72112c558454bf56916a4842c10", "date": "2026-08-20T08:53:57+02:00", "subject": "Unify error messages", "body": "and some small style improvements", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "3d41d86a3bcbf69fbd41288841b52a9a822cb1e8", "date": "2026-08-20T14:55:03+08:00", "subject": "docs: add Upload data version", "body": "* docs: add Upload data version\n\n* docs: remove Upload RequestOptions.data version\n\nCo-authored-by: Cursor \n\n---------", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "9e116fb52a9f6d438d7ba8e7adea4fb2d27e3a71", "date": "2026-08-20T10:03:24+03:00", "subject": "ci: cache `playwright install-deps` by using custom Docker image", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d2a710c7e9e08e101694fe6c1f6789f18f9d3dcf", "date": "2026-08-20T17:12:53+09:00", "subject": "Track RI fast-path FK-check batches per firing cycle", "body": "Commit 34a30786293 fixed an RI fast-path crash under nested C-level SPI\nby keeping batch-callback lists per after-trigger query depth. That fix\nwas incomplete: the RI fast path still tracked callback registration\nwith one global flag. Once an outer firing cycle had registered its\ncallback, the flag suppressed registration for a nested cycle, leaving\nthe nested batch to be handled by the outer callback, too late and with\nthe wrong snapshot, potentially after the ResourceOwner holding its\nrelations had gone away.\n\nNor is per-depth callback registration sufficient while the cache is\nkeyed only by constraint OID. If nested firing checks the same\nconstraint, it reuses the outer entry, combining rows that must be\nchecked in separate firing cycles.\n\nKey the cache by both constraint OID and query depth. Register a callback\nfor each depth that creates an entry, and make ri_FastPathEndBatch() flush\nand release only entries belonging to the ending depth. Add\nAfterTriggerCurrentQueryDepth() so ri_triggers.c can obtain the current\ndepth; depth -1 represents deferred firing.\n\nAdd regression coverage for nested firing through a cursor portal, whose\nresources must not outlive the nested cycle, nested firing of the same\nconstraint at different query depths, and deferred firing at query depth\n-1.\n\nReported-by: Noah Misch \nReported-by: Peter Geoghegan \nDiscussion: https://postgr.es/m/20260705222115.be.noahmisch@microsoft.com\nDiscussion: https://postgr.es/m/CAH2-Wz=D533JbF_ak_Pc8kP0FKse-ju8DnMxtjvY==yHsP4xgw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "ac23f21a2c222357fbcca1e1b22f38c382060f55", "date": "2026-08-20T09:18:58+01:00", "subject": "More regress tests for zoom.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "df00ed264c79da6a1e435600969502760f66ed80", "date": "2026-08-20T10:43:58+02:00", "subject": "refactor(compiler-cli): add compiler option for enabling source locations", "body": "Adds an internal config options that allows us to enable source locations.\n\n(cherry picked from commit f3c093df24b7c87e4caee51a10a070a2baab2f34)", "style": "conventional", "type": "refactor", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "f3c093df24b7c87e4caee51a10a070a2baab2f34", "date": "2026-08-20T10:43:58+02:00", "subject": "refactor(compiler-cli): add compiler option for enabling source locations", "body": "Adds an internal config options that allows us to enable source locations.", "style": "conventional", "type": "refactor", "scope": "compiler-cli", "breaking": false} {"repo": "postgres", "sha": "539e6a916eef09dc6b7c3abb603722af772b1386", "date": "2026-08-20T11:45:51+03:00", "subject": "Skip bogus find_composite_type_dependencies() call on sequences", "body": "A sequence has no rowtype. We called\nfind_composite_type_dependencies() with InvalidOid, which is harmless\nbut pointless. Skip it.\n\nThis started to happen with commit 344d62fb9a97 in v15, which added\nthe ALTER SEQUENCE ... SET LOGGED/UNLOGGED subcommand. Before that,\nsequences were never rewritten. While this is harmless, backpatch to\nkeep the code the same on all branches, to make backpatching future\npatches a little easier.\n\nBackpatch-through: 15", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "31de0d69fc745b342352f04a03cbcb90211c4edc", "date": "2026-08-20T05:18:36-04:00", "subject": "fix(cmdatom): lhs not always reported in CmdAtom #41386", "body": "Problem:\nAn unreplayable Visual operation does not emit a CmdAtom event. That's\nmaybe not super important, but it hints at a flaw in how `vatom`\n\"voiding\" is plumbed: `vatom.state=kVatomVoid` replaces the \"kind\", so\nthat info is lost to later parts in the lifecycle.\n\nSolution:\n- Define `VatomState` as \"flags\", so `kVatomVoid` can \"poison\"\n `vatom.state` without losing its kind flag.\n- Unify `lhs`: always report the original user input in `CmdAtom.lhs`,\n for all kinds of user actions: visual, \"translated\"/\"stuffed\" cmds,\n and dot-repeat (\".\") itself.\n- Unreplayable Visual atom emits CmdAtom with non-empty `lhs` and empty\n `keys`, like a mapping/macro composite.", "style": "conventional", "type": "fix", "scope": "cmdatom", "breaking": false} {"repo": "tmux", "sha": "554c4bbde418e5e9b8220e4d4ceea5753db53a41", "date": "2026-08-20T10:19:19+01:00", "subject": "Tests for modal panes on top.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "95a71707098b1cf60a626adae4b2fbcaa8da4dee", "date": "2026-08-20T09:19:24Z", "subject": "Add a way for floating panes to stay above zoom (-A) flag. Use by default for modal panes to match popups.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "67ef7a19bd039aa91ce49741b7966f4d508ce361", "date": "2026-08-20T12:46:22+03:00", "subject": "fix(worker): bind stdio's early in case overriden", "body": "", "style": "conventional", "type": "fix", "scope": "worker", "breaking": false} {"repo": "ant-design", "sha": "82da292c90200d92e75fa87ad3670b2077250f93", "date": "2026-08-20T17:46:33+08:00", "subject": "feat(Typography): support shimmer effect", "body": "* feat(Typography): support shimmer effect\n\n* refactor(Typography): use CSS variable helpers\n\n* refactor(Typography): simplify shimmer configuration\n\n* refactor(Typography): streamline shimmer config merging\n\n* fix(Typography): adjust shimmer animation timing\n\n* fix(Typography): refine shimmer animation and demo\n\n* feat(Typography): support separate shimmer motion and wait times\n\n* chore: update size limit\n\n* test(site): update Typography table snapshots\n\n* test(Typography): cover shimmer duration tuple", "style": "conventional", "type": "feat", "scope": "Typography", "breaking": false} {"repo": "openssl", "sha": "26d762a108e532bb0ca8d510e5d56ac9718dd02d", "date": "2026-08-20T17:02:48+07:00", "subject": "Fix tick race in radix fault-injection scripts", "body": "Guard any `OP_SET_INJECT_WORD`+`OP_WRITE` with 2 new functions `OP_ENGINE_TICK_ENABLE`/`OP_ENGINE_TICK_DISABLE` that enable/disable ticking in the QUIC engine. Otherwise, the radix test interpreter can tick between the 2 ops and cause the peer to close the connection prematurely\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Norbert Pocs \nMerge-date: Wed Sep 2 06:32:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/32335", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "73020da8906aadd69c33d4141a62100b4b63334c", "date": "2026-08-20T13:09:58+03:00", "subject": "refactor(docs-infra): share the angular.dev origin constant", "body": "The angular.dev origin was declared twice as a local `ANGULAR_DEV`\nconst and hardcoded inline in two more places. It now lives once in\n`core/constants/links.ts`, next to `ANGULAR_LINKS`.\n\n(cherry picked from commit 593f2d7e2ed548c557ad4aba43803b89f5db42fc)", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "593f2d7e2ed548c557ad4aba43803b89f5db42fc", "date": "2026-08-20T13:09:58+03:00", "subject": "refactor(docs-infra): share the angular.dev origin constant", "body": "The angular.dev origin was declared twice as a local `ANGULAR_DEV`\nconst and hardcoded inline in two more places. It now lives once in\n`core/constants/links.ts`, next to `ANGULAR_LINKS`.", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "openssl", "sha": "a981006cc8ec94d6108639af57e47a435ab83bdb", "date": "2026-08-20T11:29:03+01:00", "subject": "Guard the QUIC poll test hook so no-quic builds compile", "body": "poll_translate() calls ossl_quic_poll_translate_test_step_cb, a test\ninstrumentation hook whose declaration and definition both sit inside\n#ifndef OPENSSL_NO_QUIC. On master that is safe, because the whole of\npoll_translate() is inside the same guard.\n\nThis branch widens that guard to\n\n #if !defined(OPENSSL_NO_QUIC) || !defined(OPENSSL_NO_DTLS)\n\nso that SSL_poll() works for DTLS, so poll_translate() is now also\ncompiled when QUIC is disabled. The hook call came along into the wider\nregion and does not compile there.\n\nGuard the call itself. Configurations that disable QUIC, directly or by\ncascade from no-tls1_3 or no-tls, then build again.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Tomas Mraz \nMerge-date: Fri Aug 21 18:10:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/32441", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a847149bf4dcb5b30a5b112be49d5048ae288c6a", "date": "2026-08-20T11:29:13+01:00", "subject": "Reformat a foreach member call for the new ForEachMacros list", "body": "ossl_dgram_conn_lookup_foreach() calls a struct member named foreach.\nclang-format has a built-in ForEachMacros list which contains foreach,\nand it formats anything on that list like a control statement, with a\nspace before the parenthesis. The space was therefore correct when this\ncode was written, and check-style accepted it.\n\nMaster has since set ForEachMacros explicitly, to name the OSSL_RBT_*\nmacros. Setting the option replaces the built-in list rather than adding\nto it, so foreach is no longer on it and is now formatted as an ordinary\nfunction call, with no space. That turned an already compliant line into\na check-style failure without the line itself changing.\n\ngit rebase does not run the pre-commit hooks over the commits it\nreplays, so nothing reformatted this on the way past.\n\nAssisted-by: Claude Code:claude-opus-5\nReviewed-by: Ryan Hooper \nReviewed-by: Tomas Mraz \nMerge-date: Fri Aug 21 18:10:42 2026\nMerged-from: https://github.com/openssl/openssl/pull/32441", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "ee63c0a8131348dbaba399373e894fbe6a0f089b", "date": "2026-08-20T13:33:15+03:00", "subject": "refactor(docs-infra): hold editor diagnostics in a signal", "body": "`DiagnosticsState` carried a TODO to move off `BehaviorSubject` once\nzoneless was turned on. adev has been zoneless for a while now, with no\nzone.js dependency and an empty `polyfills` array, so the condition is\nmet and the TODO can go.\n\nBoth consumers now derive from the signal instead of subscribing. The\nConsole tab badge becomes a `computed`, and so does the code editor's\nerror list: the diagnostics are produced by a `linter()` configured with\na 400ms delay rather than arriving as a stream, so the rxjs chain was\nonly adding a further second before they were displayed. The errors box\nstays dismissable through a `linkedSignal`, which reverts as soon as the\ndiagnostics change.\n\n(cherry picked from commit 238d8bf98aaeaa9ce4ceddff26eb15aab5d013b6)", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "238d8bf98aaeaa9ce4ceddff26eb15aab5d013b6", "date": "2026-08-20T13:33:15+03:00", "subject": "refactor(docs-infra): hold editor diagnostics in a signal", "body": "`DiagnosticsState` carried a TODO to move off `BehaviorSubject` once\nzoneless was turned on. adev has been zoneless for a while now, with no\nzone.js dependency and an empty `polyfills` array, so the condition is\nmet and the TODO can go.\n\nBoth consumers now derive from the signal instead of subscribing. The\nConsole tab badge becomes a `computed`, and so does the code editor's\nerror list: the diagnostics are produced by a `linter()` configured with\na 400ms delay rather than arriving as a stream, so the rxjs chain was\nonly adding a further second before they were displayed. The errors box\nstays dismissable through a `linkedSignal`, which reverts as soon as the\ndiagnostics change.", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "curl", "sha": "aab0518d04599627368ab78b1f8713683098b435", "date": "2026-08-20T12:39:59+02:00", "subject": "url: fix handling of empty user in ntml matching", "body": "Line should already have been removed in 7103a93, but neither\nhumans nor clankers caught it. tststs...\n\nPrior to this change it appears all reuse of empty cred connection was\nblocked for ntlm instead of the intended behavior of allowing\nreuse in the limited circumstance when it is the same connection used by\nthe transfer before.\n\nCloses https://github.com/curl/curl/pull/22629", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "6b3d0946b41bc306c1d8242c20ee3ace18b541ab", "date": "2026-08-20T12:25:43+01:00", "subject": "docs: preserve established image semantics in redraw plan", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "ea0ae8e1266e312ff28e8482002a14483e4e982b", "date": "2026-08-20T13:32:10+02:00", "subject": "CVE-2026-8926: password leak with netrc and user in URL", "body": "Fixed differently than original fix", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "c13de729b929efd5e7c7a86679f410f52de8b0a9", "date": "2026-08-20T14:16:02+02:00", "subject": "chore(ai): ask for a manual approval", "body": "", "style": "conventional", "type": "chore", "scope": "ai", "breaking": false} {"repo": "angular", "sha": "0739e3a1235404a634ee446fab8de477f9a871fc", "date": "2026-08-20T17:19:56+05:00", "subject": "docs: warn against using in-memory-web-api in production", "body": "(cherry picked from commit 3438614f5e7dd495f13a9f5823694945c18f3fa7)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "3438614f5e7dd495f13a9f5823694945c18f3fa7", "date": "2026-08-20T17:19:56+05:00", "subject": "docs: warn against using in-memory-web-api in production", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a5dffd58f3d22a4bea991ef1bd7f9b8ebaff8c84", "date": "2026-08-20T14:26:03+02:00", "subject": "cipher_overhead_test.c: Init libssl to sort the ciphers", "body": "Otherwise ssl3_get_cipher_by_id() cannot work properly.\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Aug 20 12:59:43 2026\nMerged-from: https://github.com/openssl/openssl/pull/32443", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "5fa2c9ad327d53ce14404d0710ee35dfeb689625", "date": "2026-08-20T14:28:12+02:00", "subject": "CVE-2026-8932: incomplete mTLS config matching in conn reuse", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "08de88e50294c4e01f6cae97eceeb578da55792b", "date": "2026-08-20T18:00:57+05:30", "subject": "feat: handle underflow in no-loss-of-precision", "body": "* fix: handle underflow in no-loss-of-precision\n\n* chore: retrigger CI\n\n* refactor: address review feedback", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "curl", "sha": "7f8f525c376413d6cd527e9dbd2522dd930261d5", "date": "2026-08-20T14:50:46+02:00", "subject": "CVE-2026-9079: stale proxy password leak", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "193ee5f87092a539deb78fc8819466ace5423b3f", "date": "2026-08-20T15:00:02+02:00", "subject": "CVE-2026-9080: UAF after pause in socket callback", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "5da7fa66f5c469e02399a9f89789ed08a628205e", "date": "2026-08-20T15:02:07+02:00", "subject": "docs(core): clarify toObservable synchronization behavior", "body": "(cherry picked from commit 045746c18ce3081c5f26aabc0097dcfa1715730b)", "style": "conventional", "type": "docs", "scope": "core", "breaking": false} {"repo": "angular", "sha": "045746c18ce3081c5f26aabc0097dcfa1715730b", "date": "2026-08-20T15:02:07+02:00", "subject": "docs(core): clarify toObservable synchronization behavior", "body": "", "style": "conventional", "type": "docs", "scope": "core", "breaking": false} {"repo": "angular", "sha": "29d3066505995d814866057545611d6fdf280432", "date": "2026-08-20T16:05:01+03:00", "subject": "refactor(docs-infra): type the API manifest in the nav entries", "body": "`getApiNavigationItems` cast the manifest to `any` behind a TODO waiting\non #66252, which was closed without merging, so the three `any`s it\nguarded were never going away on their own.\n\nThe real `ApiManifestPackage` cannot be imported here: navigation\nentries are built as a standalone `ts_project` so the route generator\ncan consume them at build time, and it has no dependency on the app\ntree. Declaring the handful of fields this file reads gets rid of the\n`any`s without touching the build graph.\n\n(cherry picked from commit 4997521685eec564d684eb23664f0e3a5530ea83)", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "4997521685eec564d684eb23664f0e3a5530ea83", "date": "2026-08-20T16:05:01+03:00", "subject": "refactor(docs-infra): type the API manifest in the nav entries", "body": "`getApiNavigationItems` cast the manifest to `any` behind a TODO waiting\non #66252, which was closed without merging, so the three `any`s it\nguarded were never going away on their own.\n\nThe real `ApiManifestPackage` cannot be imported here: navigation\nentries are built as a standalone `ts_project` so the route generator\ncan consume them at build time, and it has no dependency on the app\ntree. Declaring the handful of fields this file reads gets rid of the\n`any`s without touching the build graph.", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "openssl", "sha": "c4c0b8cc2650b9731c91608a07ee4ee17b432c5c", "date": "2026-08-20T09:09:29-04:00", "subject": "Allow X448MLKEM1024 to be fips approved", "body": "According to lightship, we can allow X448MLKEM1024 to be an approved\nhybrid kem, we just can't allow it for use in TLS1.3 (which isn't\ncurrently possible anyway, given that there is no code point for it).\nSo set FIPS_DEFAULT_PROPERTIES on this alg rather than\nFIPS_UNAPPROVED_PROPERTIES.\n\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nMerge-date: Fri Aug 21 14:34:08 2026\nMerged-from: https://github.com/openssl/openssl/pull/32444", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "42d1e84219e4595edb1ed133c66e749ecb3fe743", "date": "2026-08-20T09:09:29-04:00", "subject": "Allow X448MLKEM1024 to be fips approved", "body": "According to lightship, we can allow X448MLKEM1024 to be an approved\nhybrid kem, we just can't allow it for use in TLS1.3 (which isn't\ncurrently possible anyway, given that there is no code point for it).\nSo set FIPS_DEFAULT_PROPERTIES on this alg rather than\nFIPS_UNAPPROVED_PROPERTIES.\n\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nMerge-date: Fri Aug 21 14:33:14 2026\nMerged-from: https://github.com/openssl/openssl/pull/32444", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0f0aabfcba92f693d8c05e3a2d2226fb67479965", "date": "2026-08-20T09:09:29-04:00", "subject": "Allow X448MLKEM1024 to be fips approved", "body": "According to lightship, we can allow X448MLKEM1024 to be an approved\nhybrid kem, we just can't allow it for use in TLS1.3 (which isn't\ncurrently possible anyway, given that there is no code point for it).\nSo set FIPS_DEFAULT_PROPERTIES on this alg rather than\nFIPS_UNAPPROVED_PROPERTIES.\n\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nMerge-date: Fri Aug 21 14:15:23 2026\nMerged-from: https://github.com/openssl/openssl/pull/32444", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b64f68a94e61fa2363c598c75444482b48056697", "date": "2026-08-20T09:09:29-04:00", "subject": "Allow X448MLKEM1024 to be fips approved", "body": "According to lightship, we can allow X448MLKEM1024 to be an approved\nhybrid kem, we just can't allow it for use in TLS1.3 (which isn't\ncurrently possible anyway, given that there is no code point for it).\nSo set FIPS_DEFAULT_PROPERTIES on this alg rather than\nFIPS_UNAPPROVED_PROPERTIES.\n\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Paul Dale \nMerge-date: Fri Aug 21 14:15:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32444", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "67f58341fb37f311a74fceaa1211da7401896f0a", "date": "2026-08-20T16:20:48+03:00", "subject": "refactor(docs-infra): remove the orphaned home animation component", "body": "The new home page in #63662 stopped rendering `HomeAnimationComponent`\nand its directory was left behind. The animation that runs today lives\nin `features/home/animation`, which replaced it.\n\nNothing references the class, its `adev-home-animation` selector, or any\nfile in the directory, and no stylesheet or build target pulls it in.\n\n(cherry picked from commit eb6570b06dda4252da9f3f83672c4457279f1b29)", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "eb6570b06dda4252da9f3f83672c4457279f1b29", "date": "2026-08-20T16:20:48+03:00", "subject": "refactor(docs-infra): remove the orphaned home animation component", "body": "The new home page in #63662 stopped rendering `HomeAnimationComponent`\nand its directory was left behind. The animation that runs today lives\nin `features/home/animation`, which replaced it.\n\nNothing references the class, its `adev-home-animation` selector, or any\nfile in the directory, and no stylesheet or build target pulls it in.", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "curl", "sha": "d2bac17f3e941cad63097b76acb9b219c2dc1031", "date": "2026-08-20T15:43:17+02:00", "subject": "CVE-2026-9545: exposing HTTP/3 early data", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b12dc923c7e05b7cf1b984b2ff05eebb4082a1f8", "date": "2026-08-20T15:46:49+02:00", "subject": "CVE-2026-19931: Negotiate ambient user conn reuse", "body": "Different than original patch", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "c6174a6cf2847b23075ee35ba5151c2184a70594", "date": "2026-08-20T16:59:13+03:00", "subject": "fix(coverage): v8 to ignore Vite SSR's generated import bindings", "body": "", "style": "conventional", "type": "fix", "scope": "coverage", "breaking": false}