Datasets:

willtheorangeguy
/

2017-Request-for-Commits-Transcripts

+**Break:** \[00:00\]
+**Nadia Eghbal:** I'm Nadia Eghbal.
+**Mikeal Rogers:** And I'm Mikeal Rogers.
+**Nadia Eghbal:** On today's show, Mikeal and I talked with Evan You. Evan is the creator of Vue.js, a JavaScript framework that recently reached two million downloads. Evan works full-time on Vue and currently funds his work through Patreon.
+**Mikeal Rogers:** Our focus with Evan was crowdfunding and community organizing. We talked about what it's like to use Patreon to fully fund yourself, why he decided to do it, and how he balances his own paid work for the growing community of contributors.
+**Nadia Eghbal:** We also talked about running a community project in the midst of other corporate players and where he sees the future of Vue.
+Evan, you started Vue while you were at Google... Is that also where you started getting into open source, or did you have a background in open source before then?
+**Evan You:** Well, it's an interesting question because I had a small project when I was still in school called HTML5 Player. It was kind of like my first ever project that got some attention. It had several hundred stars on GitHub. That was my first taste of people paying attention to your open source work. But it was not particularly an open source project, it was more like just some code that I decided to post online for whatever reason... So I think Vue is technically the first open source project that I put online and was dedicated to maintaining it.
+**Nadia Eghbal:** Pretty awesome first experience.
+**Evan You:** It depends on how we define it, yeah.
+**Nadia Eghbal:** Did you contribute to open source before then, or...?
+**Evan You:** I think I did a little bit... I think I started contributing more as I started working on Vue, because when I was working on Vue I also used other people's projects, and I started to run into bugs in their projects, and I started submitting PRs to fix them so that I could use them for Vue.
+**Nadia Eghbal:** \[04:01\] Nice! So was that like completely terrifying then, that Vue was your first intense open source experience?
+**Evan You:** I would say that, yeah.
+**Mikeal Rogers:** Why did you end up making Vue? What was the impetus behind getting that created?
+**Evan You:** I think in the beginning the motive was very simple. It was something I wish I had when I worked on some of the projects I was working on. At the same time, it seemed like a good opportunity to just flex some technical muscles and sort of like -- the feeling that we see some great ideas or some interesting ideas that you see in other people's code, and you just want to try it out and write it yourself and see if you can actually implement it. It's a combination of the two.
+The motive for the project changed over time. It started more like an experiment, but it gradually evolved into something that I want to open source and maintain, but it was more like just giving it a shot, and it turned out people were actually liking it. The more people used it, the more responsibility I felt that I should keep it well maintained, and the more I felt motivated to make it better.
+The scope of the project grew, and I guess the goal for the project also kind of evolved along the way. The goal today I would say is more like providing a framework that helps make it easier for people to build the applications. It sounds crazy ambitious. When I first started, that was definitely not what I had in mind.
+**Mikeal Rogers:** You said that there were some projects that you were doing at Google at the time that kind of drove you needing it... What were those kinds of projects and what are the kinds of projects that people are using it for now and is there any difference between those?
+**Evan You:** Yeah, definitely a lot of difference. The department that I worked at Google was called Google Creative Lab. It's a very special department where we don't work in production engineering products, but instead we work on a lot of prototypes, experiments... Some of the projects were more like things you saw with Chrome experiments, but some of them were internal and they were a collaboration between the product teams, and it helped them design the next iteration of what their products would look like.
+Some of them would be super experimental, like just imagining what this product could be in five years or ten years. These required a lot of rapid prototyping, where we would maybe come up with 3-4 crazy ideas and we want to see them become tangible in a very short amount of time. Basically, my job was to create these things as a web-based prototype.
+The whole idea was I needed to turn ideas into tangible prototypes as fast as possible... Some of the common elements in building web applications today, for example we want declarative rendering and components and all that would become very helpful in these scenarios too, but at the same time they probably didn't need some of the things that were geared towards larger scale applications, so that is why I felt Angular was a bit too heavy for my use case at that time.
+The initial version of Vue was essentially a version that extracted the parts that I felt useful from Angular and threw away the things I felt I didn't need at that time.
+**Mikeal Rogers:** \[08:10\] So now that the project has grown, presumably people are using it for more than just quick applications, like they actually need to maintain it. Has that shifted some of the goals of the project, or have you really maintained that you don't wanna sacrifice any of that speed?
+**Evan You:** Yeah, that's interesting... Obviously, there are many people using Vue to build large production apps today, and the framework today is also very well-suited for those purposes. But this initial version of Vue that focused on this ease of use and this use case for rapid prototyping is still there today. So as we look at how you can use Vue, you can still use Vue by just dropping it into an HTML page with just a script tag or a link to a CDN file and instantly start writing code.
+This is kind of now a feature that we had when Vue was initially released, and it's still true today. Alternatively, if you want to build the professional way, you can obviously use our CLI to scaffold a full project, with boilerplates and Webpack and all the build tools, testing tools, with all the fancy stuff built in. So we offer different tasks, depending on how much you actually want from the framework.
+This is also why we call it a progressive framework, because it's incrementally adoptable. You can use very small pieces, just the core, for simple use cases, for rapid prototyping, but you can use the full stack for more ambitious apps. So the whole stack is incrementally adoptable; you don't have to use everything all at once.
+**Nadia Eghbal:** I have a question about the early stages of Vue... I read that the Chinese community had kind of helped you find popularity around Vue and get it out there... Can you talk a little bit about your involvement with the Chinese developer community?
+**Evan You:** Sure. I myself am Chinese; I grew up in China and I came to the U.S. after high school. I also am pretty active on the Chinese social networks - basically, the Chinese version of Twitter, the Chinese version of Quora, and I will obviously talk about Vue or answer your questions about Vue on those social networks.
+Interestingly, a lot of people discover Vue first because of -- it's kind of like Vue got popular in the U.S. and then some people in China discovered it and realized "Oh, this is actually written by a Chinese guy" and they got really excited about it. Then they found out I'm actually active on Chinese social networks.
+It was kind of an interesting round trip, but somehow Vue also got really popular in China. I'm not sure how much of my social network stuff contributed to it, but I think me being Chinese definitely played a role in it... But it's also because maybe in Chinese also helped some contributors from China to voluntarily translate the docs into Chinese, which helped a lot in the adoption.
+**Nadia Eghbal:** That's awesome.
+**Evan You:** Yeah.
+**Nadia Eghbal:** The thing is it all happened pretty naturally. I didn't intentionally try to promote it in any way. I think probably the biggest contributing factor in Vue's popularity in China is because of my answers on Zhihu which is sort of like the Chinese equivalent of Quora.
+\[11:59\] A lot of people directly ask questions about Vue and then ask me to answer it. A lot of times, when I have nothing better to do, I just answer those questions. I think that helped people a lot; it made them feel connected to this project more, because they were able to see the direct interaction with the author of the framework and it made them feel like the project is more connected with them.
+An equivalent example would be Dan Abramov interacts a lot with React users, and I think that helps a lot of users connect with React better. So I guess that's the positive part of having someone representing a framework being really active on social networks.
+**Nadia Eghbal:** That's awesome.
+**Mikeal Rogers:** Yeah, I've seen this in the Node community, too. There are certain projects where because the maintainers are in China and available on the social networks, there is just more localized support for it. It's not like those projects aren't used anywhere else, but they certainly have more of a following in China than elsewhere, which is really interesting. I've noticed with the coroutines stuff in Node.js. There's a lot of people that follow that, particularly in China.
+**Evan You:** Yeah, I think one of the core maintainers of Koa is Chinese, and a lot of the active Node contributors are from Alibaba, because they use Node.js pretty heavily in production, and they've open sourced a lot of modules. I think it's a very good thing; they contributed a lot to Node.js's popularity in China because of all the open source work they do.
+**Nadia Eghbal:** From my perspective at least, being newer to open source, a lot of conversation at least seems dominated by the U.S., Europe and Australia... For the both you - have you noticed that developers outside of these areas can feel siloed off? How do you make them feel involved? It sounds like part of what was cool about Vue was that it just made them feel connected to you and to the project.
+**Evan You:** I think primarily it's still the language and partly the cultural barriers - they're kind of inevitable. A lot of the developers in China - they can read and understand English, but a lot of them don't feel comfortable communicating in English. It's a lot of extra effort for them to say... The simplest example is when they want to open an issue. A lot of them have trouble just articulating some of the subtle concepts in programming using a language that is not native to them.
+I think this is just a natural barrier for them to be able to contribute more, which is also why programmers in China often say "English is as important as your programming skills if you want to become a really good programmer." Because so much knowledge and so much open source work and so much resources are written in English, it is almost essential for a programmer to be proficient in English to be able to stay up to date with the latest resources.
+Most of the programmers actually can read English, but when they try to use it to convey some of the more subtle concepts, it becomes a struggle. I think that's the primary reason for them to be less active on the main stage of open source. But because I'm Chinese, I'm able to get a peek at the Chinese open source scene - it's actually very active. It's mostly kind of siloed off because of the language barrier.
+**Mikeal Rogers:** \[15:59\] Yeah, in the Node project we have some metrics, and it's huge. It's like 12% of all of our users, which is like a lot of people.
+Moving along a little bit, I just wanted to know... You've taken a step to kind of go full-time working on Vue.js, or at least attempted to. What was that decision like? What made you wanna start working just on your open source project full-time, rather than also doing a lot of the other work that you've been up to?
+**Evan You:** Yeah, so I was working for Meteor in 2015, I think... Wait - when was that? Yeah, 2016. Early 2016, that was when I started working on Vue full-time. Before that I was working for Meteor. At Meteor I already started feeling the pull of this unsustainability of having a day job and at the same time maintaining an open source project that's growing more and more popular. The amount of issues and maintenance and the growing scope of the project was just -- it just started feeling too much to be something that I work on in my spare time, and I felt it was not sustainable.
+I started seriously thinking about which do I actually want to work on more, and I think the answer was "I want to work on Vue more." So that's when I started to think "Is there any possible way for me to turn this into my full-time job, instead of something I can only attend to in my spare time?" I guess that's kind of the question a lot of other open source maintainers are also asking right now. I didn't really have a good answer at that time. In fact I still don't, because the Patreon campaign was more like just an experiment or an explorative move on my part...
+I thought that there were people sustaining themselves by creating content, and I compare that to Vue.js or I compare that to an open source project. I felt like "I'm working on this project and it's creating value for people. If I'm creating value for people, is there any way for them to somehow give back in a financial form so that I can sustain myself?" and this kind of model seemed probably the most direct model that I see. If someone uses my software and they feel like it has helped them, and they don't want the project to die, then at least they have an incentive to give me some money so that I can keep working on it.
+So I started the campaign. I didn't really think too much about it, and I just threw it out there. It turns out people actually wanted to give me money, and some of the companies were really generous. I had the tiers of like $100/month, $500/month, and there was one tier that was $2,000/month. I put it out there just thinking no one would probably ever do it, but there was this company called Strikingly - it's a startup... They were a Wi-Fi company, but they somehow moved to China because the founders were Chinese and they couldn't get a U.S. visa... And they don't actually use Vue, but they have this fund that they just use to donate to open source projects that they felt were doing a good job. So they signed up for the $2,000 tier and they were basically saying "We liked your project, we want to support you, so we're just giving you $2,000/month."
+That was a really huge help in the beginning. They did it for six months, and that was probably like -- if it wasn't for them, I don't think the campaign would ever grow to what it is today. Today we have like $9,000-ish a month, which is already enough for me to somehow sustain the family and all that.
+\[20:17\] I still consider myself extremely lucky to have pulled this off. Whenever some other open source maintainer asks me for advice, I'm always hesitant to recommend them going this way because I don't feel this is something that's easily repeatable, and it really depends on how much traction you have gained, and it depends on what financial situation you're already in. I have some money saved up, and I was basically planning to do it for free for a few months just to see if it would ever work... But it turns out I got to $4,000/month pretty fast.
+I was almost able to sustain myself, and it just kept growing until it became -- today I'm pretty proud to say I'm fully sustained by open source work.
+**Nadia Eghbal:** That's awesome.
+After the break, Nadia and Mikeal talk with Evan about how he's been able to make open source his full-time job, treating it like a job for better balance in life, better balance with his family. We also talk about an often uncomfortable topic - funding. Who should get paid? Where should money be spent? Should this project have a full-time paid developer and how to sustain that? All this and more after the break.
+**Break:** \[21:47\]
+**Evan You:** I felt that I wanted to work on Vue more as my full-time job... And I generally felt much, much better after I switched to working on Vue full-time, because now I can actually treat it like a job. I set hours for myself, I don't have to stay up late at night anymore, so I can actually spend time with family, having spare time as I actually should have, and I feel like this is something that's much more sustainable now than having a day job and working on it at the same time.
+**Mikeal Rogers:** How many other contributors were there to the project when you started the Patreon campaign, and what does the team look like right now?
+**Evan You:** I can't remember exactly how many when I started the Patreon campaign... I would say probably only a few. Today we have a pretty decently sized team, but we have like 20-ish people in the Vue.js organization on GitHub.
+\[24:08\] The team is pretty loosely organized. Everyone is just contributing on a volunteer basis, and we don't have a very formal structure of assigning people tasks or expecting them to do something by a deadline.
+The team also grows pretty organically. Whenever someone makes quality contributions on a regular basis for a while, I just invite them to the organization and ask them "Do you want to do this on a regular basis? Do you want to become a part of the team?" but overall it's still pretty loose. I don't really feel comfortable -- because they are doing voluntary work, I don't feel comfortable assigning them stuff as if it's their duty.
+**Mikeal Rogers:** It's pretty easy to put yourself in that same position, being that you were just doing that recently. You wouldn't want people assigning things to you when you were working on it in your own time, right?
+**Evan You:** Yeah, definitely. But I have to give a shoutout here, because the team is doing such an amazing job at dealing with the most tedious part of open source work, which is triaging the incoming issues. It think the amount of issues we get across the whole Vue.js organization is totally becoming unmanageable for one single person, and luckily we have a team helping me to filter out these issues that can be relatively easily dealt with, and leave me with the ones that are critical or are really important, so that I can better focus on the things that matter.
+Other than that, obviously some of them also make high-quality contributions to the codebase itself, which are really helpful.
+**Mikeal Rogers:** That's interesting... I wanna hear more about that. You have a growing community, people are taking on more responsibility, a lot of new stuff is rolling in... How did you end up pricing the Patreon campaign? Because you know that you need your time, but like you were saying, you're already doing pretty good even with where it's at... You have a bigger goal. So eventually it sounds like you wanna spend more money on the project beyond just your expenses.
+How did you kind of price the Patreon campaign and what was the goal in terms of supplementing what was already going on on that community side?
+**Evan You:** I'm not sure how you would define pricing the Patreon campaign. It was more like... I just set the numbers based on just jumping up in tiers. I didn't really think that much about how to distribute it between contributors at that time, and I think Open Collective didn't even exist back then. Patreon was kind of like the only obvious solution there... And even today -- I actually had a talk with one of the more frequent contributors, Chris... He's been making a lot of contributions to the docs, and I even discussed like -- I wish I could have allocated some of the incoming Patreon projects to him, to help him spend more time on this stuff, but he didn't want to. It has this very subtle social aspect of it, because he felt if he took some of the money, then it seemed fair for other contributors who also did the work, and it becomes a problem how to properly quantify and measure the work that each contributor does. It also creates this really awkward dynamic between the team members... So this is actually a question we are still thinking about how to best solve.
+\[28:16\] At this moment, we're trying to somewhat -- one of the alternative ways we try to give back to these contributors is, for example, Chris has got a deal for working on educational content for O'Reilly, because he's part of the core team and we connected with O'Reilly. He was able to get that, which offers pretty good compensation on his part. There's also -- I think some of the members got contacted by publishers to work on books and other sort of content... So these are somewhat indirect ways that we can give back to the contributors. Because they are now part of the core team, they get much better recognition for their expertise in Vue.js.
+**Nadia Eghbal:** That's awesome.
+**Evan You:** Yeah, but we just feel that direct allocation of money is somewhat tricky, because it's really hard to properly quantify it and we still don't feel there is a very elegant solution to this problem.
+**Mikeal Rogers:** Yeah, I mean, the Patreon campaign is really set up to fund you, really specifically... In fact, you could extend this to some other -- if you had another idea, you could allocate some work to that, right? People really are backing you and investing in you, even though you probably have enough left over for the project.
+We've actually heard from other people too, it's like when you get money for the project, it's actually hard to figure out what to spend it on...
+**Nadia Eghbal:** Right... What is the project, versus what are you doing...
+**Mikeal Rogers:** Right, right. But when people are funding you, they're literally saying "I want you to work on open source full-time."
+**Evan You:** Yeah, that's a different way to look at it. In some way, this is basically the majority chunk of my income now. I need to use this money to pay the bills, I need to put these into my savings, even pay for retiring and all that stuff... So in some way Patreon is more directly connected to the creator, the person, rather than the project.
+It's a little different from the mindset or the model that Open Collective is pushing forward, where the funding is going to the project. I think it's an interesting dynamic, and it kind of depends on how the project is structured and how the creator or the maintainers want to frame the story.
+I just feel like this is an interesting difference between some of the different funding models that we see in some of the more popular open source projects.
+**Mikeal Rogers:** Yeah, I find it really fascinating. One of the hazards that we keep hearing about is that when you start paying some people, other people don't feel as valued sometimes... And it's interesting - you're not running into that. You actually have a pretty thriving community that's still growing, and even within that community, they're kind of saying "No, we actually don't wanna be paid directly... Forget this stuff!" \[laughs\] It's still a concern, even though you have your Patreon set up to continue to sustain the project.
+**Nadia Eghbal:** And it's really interesting to see how there are like other benefits to being a regular contributor that might not be "I'm getting paid to work on those, but my reputation grows, and then I have this book deal..." That's really interesting to think about. My brain's going in like 100 directions right now. \[laughs\]
+**Evan You:** \[32:01\] Yeah, and another aspect that I wanted to mention is one of the core members actually started to work on a project that's fully dedicated to Vue resources... Kind of like making Vue videos and all that. I think it's somewhat like a way for him to both contribute to Vue and provide more content for the users, and also in this process try to financially sustain himself.
+**Nadia Eghbal:** Interesting. Do you say that was a Patreon campaign?
+**Evan You:** It's not a Patreon campaign, it's more like a subscription service, similar to Laracasts.
+**Nadia Eghbal:** It'd be interesting to think about, like, if different contributors feel motivated to raise money to work on Vue for their own time, I guess, but it's still different from raising money for Vue; it's "Raising for my work on Vue", in this aspect.
+**Evan You:** Yeah. It's something Chris has been thinking of, too. He backed off, because he felt it could be distracting to have multiple campaigns that are in fact for the same project, and I think this increasingly reveals the problem between "How do we support the project? Do we support the people behind the project directly, or do we support the project and then somehow let the maintainers figure out how to allocate the money?"
+**Nadia Eghbal:** Yes, that's exactly it.
+**Evan You:** I think this is something that we haven't really figured out at this point. I hope we can find a good model for this in the near future.
+**Nadia Eghbal:** I'm wondering a little bit about just like the logistics of -- knowing all that now, is there a legal entity that's associated with Vue? Is there a company?
+**Evan You:** Currently no...
+**Nadia Eghbal:** Wow, that's so interesting.
+**Evan You:** Yeah. I mean, I have a single-person LLC, but it really has nothing to do with how the project is organized; it's mostly just for me to send invoices, and stuff. But as far as I know, I also looked if it's possible to start a non-profit organization for it, but it's awkward because Vue is at a scale where it's not big enough to warrant a full foundation-like work, but it's also big enough that it's not sustainable as a spare time project.
+It's stuck in this awkward place where we're basically trying to figure out what to do with it.
+**Nadia Eghbal:** Yeah... That's so interesting, because people -- I refer to it as like "Vue is funded through Patreon", but really it's your work that's funded through Patreon, and Vue is still an open source project without an associated entity, which is actually really cool.
+**Evan You:** Yeah.
+**Nadia Eghbal:** I have to ask the obvious question of like why didn't you decide to start a company and raise venture? I mean, you worked at Meteor - they kind of went that path?
+**Evan You:** Yeah, there were actually people asking me that. There were some VCs in China that actually wanted to give me money, but the biggest obvious problem is I don't have a very clear business model for making money off of Vue, because it's an open source project; it's MIT-licensed. I don't want to sell it. I don't want to license it and then ask people to pay money for it, so I just don't see a business model that would align with what a VC would want to see. With that, it just feels awkward to take VC money.
+I also don't want to feel the pressure of scaling and monetizing beyond what is enough to sustain myself, at least at this stage. That's something that's been put on the table before, but I just don't feel like it's the right choice, at least for now.
+**Nadia Eghbal:** \[36:25\] It makes sense.
+**Mikeal Rogers:** Not having a business model hasn't seemed to stop most startups in San Francisco... \[laughter\]
+**Nadia Eghbal:** Yeah, at least you're honest. \[laughs\]
+4: I think it's interesting, because frameworks like React and Angular are essentially backed by huge business entities, but Vue is completely independent, I would say. It's an independent open source project. For similar projects in the same boat with Vue, independent open source projects that are pretty popular but not backed by a big business entity, I think this is unknown territory where sustainability is really something we haven't found a successful or a replicable model to sustain this kind of project.
+I kind of want to mention a project, which is Babel. Many people kind of have the impression that Babel is backed by Facebook, which in fact it is not... Although Sebastian, who is the original author of Babel, works for Facebook now. He actually no longer maintains the project. It's largely maintained by Henry Zhu, in case you don't know it.
+Henry worked at Behance, but he actually worked on Babel mostly in his spare time. I talk with Henry pretty often, and he's struggling with finding more time to work on Babel. He asked me for advice because he saw me have a pretty successful Patreon campaign. When we talked about how we could potentially fund his work on Babel, we went into all these little problems that we've just mentioned, like how -- because he is not the only maintainer of Babel, so how do we justify him taking the majority of the fund, and would it make sense to start his personal Patreon, or would it make sense to start one for the Babel project... The subtle difference here is I am the creator of Vue and I am still the -- people essentially associate me with Vue, so they feel comfortable giving me money, knowing that it's going to support Vue.
+In Henry's case, a lot of people don't even know he is the main maintainer behind Babel now, so it's really hard for Henry to start a campaign and say "You should give me money for me to work on Babel." We've discussed this a lot behind the scenes, but sadly we haven't really found a really good way for him to be able to dedicate more time on it.
+**Nadia Eghbal:** Absolutely.
+Coming up after the break, we go further in this conversation. Nadia, Mikeal and Evan go into the struggle of corporate and community-funded open source. We asked questions like "Do you have a sense of being David battling Goliath? Do your goals change in open source when you start working on it full-time?" and as Vue grows, does Evan have a plan or know how to fund beyond the current Patreon campaign? Stick around.
+**Break:** \[40:00\]
+**Nadia Eghbal:** Evan, you were starting to touch on Henry's story with Babel, and the perception that Babel is supported by, or is an official Facebook project... I'd love to dig in a little bit more around -- I think a lot of people are perceiving right now that companies are getting involved in open source, and that community-generated and supported open source is harder and harder to support. Do you think that there is a future where corporate and open source worlds overlap and can live harmoniously, or is the future gonna be one or the other?
+**Evan You:** I can't really make a prediction, but I think if we want to have a world where corporate and independent open source live in harmony. I think the enterprise and a lot of the businesses that are using open source need to have their mindsets somewhat transformed or updated.
+Sometimes - I've been thinking about this; maybe it's just speculation, but it just seems the incentives for businesses to give open source projects money is at this moment just too weak. They're getting value out of these open source projects, but they don't feel obliged to do it, even if the open source project is actively asking for donations or pledges.
+One of the things I've encountered in Vue's campaign is in the survey, actually 25% of potential donors backed out because they couldn't get their managers or whatever to approve the charges. And I think this somehow shows that some of the businesses don't even consider donations to open source are a valid type of expense. They feel like this is not something they've done before and they don't ever want to do it, or they just don't feel like this is a justifiable form of expense. I think this is somewhat problematic.
+If we can somehow clear this barrier, where when a developer's saying "Hey, we're often using this open source project and we should support it. We should donate money to it", and if we can somehow make it so that these businesses would gladly agree that they should do it, I think would be in a considerably better situation.
+I know that a lot of developers don't even ask their managers, because they know they would not approve it. This is somewhat sad. I think if we can somehow raise enough awareness that the industry actually considered donating to open source a very regular, very common thing, then we'd have a much better world.
+**Nadia Eghbal:** Absolutely.
+**Mikeal Rogers:** \[44:01\] I think that there are some companies that at least are willing to put in people resources; they're at least willing to open up their people to contribute back. But you're right, there's this whole other section of companies that aren't willing to do that and aren't willing to put in any money for things that they critically depend on, right?
+**Evan You:** Yeah.
+**Nadia Eghbal:** I'm just thinking back to how Patreon is really about funding your work more than Vue as an entity, and whether -- I think in Henry's situation, he got his employer to agree to give him more time to work on Babel during work hours... I wonder if that's sort of like a viable solution, instead of donating?
+**Evan You:** Yeah. In some way, I think Behance can be considered the main sponsor of Babel now, because they allow Henry to work on Babel for some portion of his time. And I think that's cool. I don't want to put Henry in an awkward situation, but I think --
+**Nadia Eghbal:** We'll just have to get him on here...
+**Evan You:** Yeah... This kind of arrangement still has its potential problems, where the maintainer's energy is kind of pulled between what the company would expect of him, versus what the open source project itself calls for. But as far as I know, there are some cases where similar to how some engineers at Google work full-time on V8, some people work full-time on Webkit at Apple. These are more mature and ideal examples of people getting paid full-time to work on open source, but technically they're actually still employers of a bigger enterprise.
+**Nadia Eghbal:** Right.
+**Evan You:** The thing is these projects originated from within those companies were -- as the case for Babel is... Babel is essentially still an independent open source project. No one owns it. The code is maybe MIT (I don't know); it's not technically owned by a company. It's fully driven by the maintainers. Although Henry is now the main maintainer, he was originally a part of the community and he is still part of the community, so we can say Babel is fully community-driven, and that's kind of still very different from the kind of open source projects that are largely controlled by a company and just made open source. There's still some difference in there.
+**Mikeal Rogers:** Let's talk about that a little bit. I think that you've identified one of these interesting cases where corporate sponsorship - for a lack of a better term - or even in this case just the appearance of corporate sponsorship is actually kind of a liability, because if they're not putting in enough people to actually maintain the project, then it's actually harder to get funding for people like Henry for the project, because people are like "Oh, that's a Facebook thing, they'll take care of it." But there's this whole other section of projects where they really do use that corporate backing not just to help out with people resources but also to add like an air of legitimacy to the project.
+I remember when Angular was released... The fact that it was used for Google-scale web apps was a selling point, and the same thing for React from Facebook. You've been on the other side of that, sort of building up a project and not saying, "Oh, we're backed by a mega-company" and you turned that into an advantage, right?
+**Nadia Eghbal:** I wouldn't say that has given us much advantage; I think we've always been fighting an uphill battle as in that we almost every company adopting Vue are asking the question of "Why wouldn't we choose React or Angular, because they're backed by Facebook and Google?" I've been asked that question so many times, and honestly, I still don't have a good answer for it, because people just naturally pick corporate backing as somewhat as a badge of -- they just feel more comfortable using something backed by a big company.
+\[48:17\] To that, I usually ask them "What do you actually think it means for an open source project to be backed by a big company?" and they say, "Oh, it's more stable. You know because they rely on it that this project wouldn't die all of a sudden", and I guess in some cases the buzz factor is smaller... So usually in response to that I would just point them to look at some of the stats. I would say "If you are worried about how well maintained a project is, you can look at the track record" because everything is open, it's on GitHub. You can go and look at their commit history, you can and look at the issues, you can see how many total issues have been filed and how many of them have been handled and how fast they're handled, and how many pull requests they get, how fast the pull requests get merged and reviewed, and how many contributors they get.
+I think it's usually more reliable if you look at these numbers directly, rather than making assumptions just because some project is backed by a big company. In reverse, you make the assumption that if a project is not backed by a huge company, it must be somewhat inferior in certain aspects. That's usually the best answer I can give, but... I don't know. I guess being backed by a huge company is still a pretty big positive factor in terms of adoption for open source projects.
+**Nadia Eghbal:** It's interesting to think about... Because there's also so many examples where a company has shut down an open source project, or kind of just abandoned it or not really paid attention to it, so it's funny that the assumption is that somehow it's more stable or supported.
+**Evan You:** Yeah, I usually try to avoid mentioning that, but it's true... The thing about company-backed open source projects is that in a lot of cases big companies open source their projects sometimes because they want to make it sort of an open standard for a certain industry, or sometimes the simply open source it to serve as some sort of publicity improvement to help with recruiting. But fundamentally, this project needs to be supporting the company's internal work, their production projects.
+If this project no longer serves that purpose, then most companies will probably just cut it, or (in other terms) just give it to the community and let the community drive it. I don't know... I'm not arriving at any conclusions here, I'm just trying to...
+**Mikeal Rogers:** Well, I think you actually covered the contrast earlier a little bit. You talked about earlier how the needs of the community and what people were building continued to kind of drive where Vue.js was going. In these corporate-backed open source projects that's not always the case. They're driven by whatever that particular company is doing with it, because the needs of the community are kind of secondary to what the company wants.
+So their whole notion of "Well, it's always gonna be around..." -- well, it may not be around in a way that's even usable to you. Angular 2 went off and did a giant breaking release that I don't think a lot of their community was asking for. It was sort of driven by the people internally at Google that wanted to make that change, whereas your solution to that is "Look, you can contribute to this project, you can become a member of the community and help drive it... You can also put in money and pay for it and continue to drive it", right?
+**Evan You:** \[52:16\] Yeah.
+**Nadia Eghbal:** Mikeal and I have talked about this - my experiences talking to especially older generations in open source... They see what's happening now and they're like "This is the thing that we were trying to prevent." This is why you have like -- I don't wanna speak for Apache, but you have stuff like Apache where that can sort of like diversify risk, where people aren't participating as companies, but they're participating as contributors and they're trying to avoid the situation where everything is pinned on one sponsor who might change their mind.
+That's why I also - especially earlier on - would hear a lot of resistance to people bringing money into open source... Because they said, "The amazing thing about open source is there is no money involved" or "The project is not being directly funded, and that means no one can take it away. People are just contributing because they want to" - I don't think that that's sustainable anymore, but there are some interesting lessons to be learned from those older generations of open source, where community-funded is really great because they can also diversify risk.
+**Evan You:** It's true, yeah.
+**Mikeal Rogers:** I also don't wanna lose another thing that you said, Henry, that was really, really good, which is that...
+**Nadia Eghbal:** Evan...
+**Mikeal Rogers:** Sorry, yeah... We were talking about Henry for a while... \[laughs\]
+**Nadia Eghbal:** We talked so much about Henry... \[laughter\]
+**Mikeal Rogers:** We'll interview Henry eventually, don't worry... \[laughter\] You mentioned that a lot of times these companies open source these projects so that they can get a standard within a particular industry, and I find that kind of fascinating because what they used to do for that was write standards... Companies used to control a lot of the standards bodies and drive a lot of standards, and there was like a decade of driving standards in this really weird direction that nobody cared about.
+Now it seems like they're just skipping the standards phase and going straight to open source projects.
+**Evan You:** Yeah, I think the problem with directly drafting standards is you have this big risk of just having something written as specs and it turns out people just don't like it, or it's really hard to work with. Or it turns out after you have published the specs you realize it hasn't really gone through enough field testing, so it's actually not good enough, but then you're stuck with it.
+You generalize from something that's already working in the field, so that you spec based on top of that. I think the way ES is spec-ed right now, somehow we have these different stage approvals, and Babel kind of serves the role of letting people play with it before it's actually become part of the spec. Now we have a much better feedback loop, because you actually get real feedback by people using these features before they are even part of the spec.
+I digress, but I think it's actually a good thing that we try to -- I'm not sure it's essentially a good thing, but from an implementation perspective I think having some open source project being successful kind of paves the way of having a solid spec on top of it. It becoming successful means it's working for people and it solves a real problem.
+**Nadia Eghbal:** I'm actually wondering about that as you kind of move into a more mature phase of Vue... I'm thinking about what are all the different stages of an open source project, and I imagine early on it's sort of like spreading the word, getting downloads, and at some point you start to double down on users... I mean, how do you know who's using the project and how do you make sure that people are actually using it because your funding depends on it, right?
+\[56:07\] If people think it's cool and then kind of get bored with it, they're just gonna stop funding you.
+**Evan You:** Yeah, I think we get some sort of very general sentiment based on tweets that I get, and I somewhat keep very loose track of, say, the Google search trends or GitHub stars, or npm download counts... None of these alone are very good metrics to determine the growth of the project, but when they are combined together, you would get this pretty good combined growth curve. I've basically been using that as an indicator.
+Actually, a better indicator is probably the weekly active users of the official DevTools plugin... That's kind of like -- the real developers, the actual developers that are actually using Vue are using DevTools. It's kind of a good metrics that I look at pretty often. Aside from that, we also have -- I also added the feature in our DevTools plugin which allows us to... Basically, the logo just brightens up whenever you visit a website that's using Vue, so sometimes community members would sent tweets at me saying "Hey, I just discovered this website is using Vue."
+**Nadia Eghbal:** That's cool!
+**Evan You:** Yeah... So that's kind of one way of doing it. I've done a few surveys in the past, asking what companies are using Vue in production. The last one I did I think was almost a year ago maybe... But overall, I'm constantly surprised. Sometimes for example I'm on Twitter and someone mentions me, and you see this pretty cool website that's built with Vue...
+Another channel -- as we are preparing VueConf, I get a lot of -- so we are reviewing the talk submissions, and there you start to see all these real people who are excited enough about Vue to submit talks to your conference and you see what products they work on, what company they work for, and it's pretty eye-opening.
+I think I'm not even fully aware of all the people that are using Vue across the world, because by the statistics there are more than 100,000 active Vue developers as of now... Then how many applications they must be working on?
+Honestly, I don't think I myself have a good enough grasp of the scale it's now at. I think it's maybe time for another survey.
+**Mikeal Rogers:** \[59:14\] I think on that note -- I think that's a great way to take us out, with the burgeoning, amazing user community. Hopefully we'll have you back on in a year or so, once you figure that all out, and then maybe we'll have something more to say as well.
+This has been great... This has been a really great chat. We really got into a lot of really great stuff, so thank you.
+**Evan You:** Great!
+**Nadia Eghbal:** Thanks for coming on, Evan!
+**Evan You:** Yeah, thank you for having me! I think that's it...
+**Nadia Eghbal:** Go to bed! \[laughs\]
+**Evan You:** Yeah, I should go to bed... It's like almost 3 AM now. \[laughter\] Alright...

Documentation and Quitting Open Source_transcript.txt ADDED Viewed

	@@ -0,0 +1,345 @@

+**Nadia Eghbal:** On today's show Mikeal and I talk with Ryan Bigg. Ryan is known for his work on Rails documentation, including the Rails guides and Rails 4 in Action. In 2015 he announced he is quitting all open source work, as he wanted to spend his free time elsewhere.
+**Mikeal Rogers:** Our focus with Ryan was burnout. We talked about his time as a community manager for Spree, writing documentation for Rails, and what lead him to quit open source. Twice.
+**Nadia Eghbal:** We also talked about his occasional contributions since then, getting paid to work in open source, and whether there's a sustainable, happy medium.
+**Mikeal Rogers:** So why don't we start off by just telling me a little bit about how you got into open source and how you got into Ruby?
+**Ryan Bigg:** Sure. I was interested in Ruby on Rails from about 2007 when a friend of mine showed me the 15-minute blog by DHH, where he used Whoops a lot, and he's like "Look at all the things I'm not doing", and from that point on I was really interested in building a forum system in Rail. I started building that, and I open sourced it and started doing work on that, and people were contributing to it through the SVN repository at the time.
+Then a few years later GitHub came along and I moved the project over to GitHub and people were contributing through that. So that's really where I got my start, with my own little forum project.
+**Nadia Eghbal:** You have a background as a developer, but then you made a lot of contributions to Rails through documentation, right?
+**Ryan Bigg:** Yeah, that's right. In 2011 I ran this little Pledgy campaign; I raised $2,500 to document Rails. I think it was around that time that Rails 3 was coming out, and the router had been rewritten by Sam Stephenson, but while the code had been rewritten, the documentation hadn't. So I went through all the code and documented all the methods that you could use in the Rails router. Out of that, then I think it was Mike Gunderloy who wrote the Rails routing from the outside in guide, and that's what I refer everyone to if they wanna understand what the Rails router is. From that documentation came better documentation. It's nice to see that evolution.
+I've also rewritten the Getting Started guide, which is probably the one that people know the most about. My first contribution though was the Active Record Query Guide, which was previously called the Active Record Finders Guide; because it does more than finders, we renamed it to querying. The Configuring Rails Applications Guide was mine, the Asset Pipeline Guide started because I was baiting DHH on Twitter, I was like "Hey, this asset pipeline is pretty good, but where's the documentation?" and he's like "Why don't you put on your big boy pants and contribute to Rails? What have you ever contributed to Rails?"
+\[03:59\] This was about after my 400th commit to Rails... So I replied to him and I said "Here's the list. Here's why I am on the contributors' list." I think he was busy that day, because he didn't end up replying back.
+Anyway, after baiting him on Twitter, I ended up writing the beginning of the Asset Pipeline Guide. That's how that came about. Then the last guide that I wrote that I can remember is the engines guide, which was actually a chapter of Rails 3 in Action and I extracted the content from that and built that into the official engines guide.
+**Mikeal Rogers:** What was it like being a documentation contributor on Rails? Was it something that was really encouraged, was there a lot of infrastructure to support for it, or was it something that you really had to drive yourself?
+**Ryan Bigg:** There was a lot of infrastructure support for it. Xavier Noria ([fxn](https://github.com/fxn) on GitHub) started that project, I think with Pratik Naik ([lifo](https://github.com/lifo) on GitHub). They started the doc Rails project, so looking for contributors, and they were really encouraging people from the community to contribute; I was one of those people, and I reckon we had a team of around ten people who were running the documentation at one point. That's why we've got the Rails guides that we have today, because of the work of Xavier and Pratik That's how I got my start.
+**Nadia Eghbal:** So you mentioned - you kind of glossed over this, but you mentioned the first documentation project you did you crowdfunded?
+**Ryan Bigg:** Yeah, that's right.
+**Nadia Eghbal:** I was curious, why did you decide to crow fund that?
+**Ryan Bigg:** I was in between freelancing jobs at the time, and I was like "Rails needs documentation and I need money, so let's combine the two and we'll start this crowdfunding project, and hopefully I'll raise enough money to survive a month", and I did. It was really nice to be able to spend an entire month working on -- I worked primarily on my book, Rails 3 in Action, and also... I reckon it's more of a 50/50 split actually between Rails 3 in Action and Rails documentation, and just helping out in the community, answering questions on Stack Overflow and that sort of thing. It's really good to be paid to do that kind of work, I really enjoyed it.
+**Mikeal Rogers:** You mentioned that there was some support inside of the Rails project for people doing documentation, but you eventually kind of transitioned into more of a community manager role for an open source project with Spree, so what was that transition like and what are your thoughts on how to do community management?
+**Ryan Bigg:** That transition was completely by accident. I was working for a company - you probably haven't heard of it, because it's an Australian-based company... It's called the ABC, and it's the Australian equivalent of the BBC. Everyone's heard of the BBC, no one's heard of the ABC. They are a big government-funded broadcaster, they do a lot of interesting programs, and from these programs they sell DVDs and merchandise and whatnot, and to sell that online, they had an old ASP store, and then they were switching it over to Rails and they wanted the Spree.
+The company I was working for at the time, ReInteractive was hired as the consultant on that project. That project persisted for eight or nine months or so, and I got involved in the Spree forums (the Google groups, I should say) and a friend of mine, Phil Arndt (who does the Refinery CMS), pointed out this post in the Spree list asking "We wanna make Spree a Rails engine, so how can we do that?" I was like, "Well, you know, I've got some time to reply; I'm gonna just write a couple of paragraphs and then they can take it from there." It ended up being a page and a half of my thoughts, and out of that Sean Schofield, the creator of Spree, contacted me and he's like "By the way, we've just received this funding and we're looking for a community manager, and you're really passionate about this because of all your answers on the forum. Would you like to come and join us?"
+I was like, "I get to work on open source, and I get paid not a pittance, so I can actually do things." And the other offer on the table was that I get to go around the world, speaking at conferences, and I was like "Yes, yes, yes. Absolutely. I will do this, because this is an amazing offer." So I said goodbye to my consulting gig and moved over to Spree. From then on it was really just about -- it sounds glamorous that you have to travel the world and speak at conferences and maintain open source and get paid for it, but then eventually after two and a half years I ended up burning out from it, and that wasn't very good.
+**Nadia Eghbal:** \[08:14\] It's like a totally different function from documentation as well, right?
+**Ryan Bigg:** Yeah, totally. I'm always interested in helping out communities and building communities and helping people understand code. Documentation is one of the ways that I do that, and the community manager role was definitely another way I could do that... Because I got really involved with the spree project and I could understand the code better than most people could, and that feeling when you can transmit an idea from your brain into somebody else's brain and they go "A-ha, I get it now!" - that's where I get my biggest thrill from, it's when people understand something because I was able to explain it clearly enough to them that they could understand it.
+**Mikeal Rogers:** Did you bring any kind of particular focus on non-code contributions, since that's kind of the world that you came from when you started community management?
+**Ryan Bigg:** I welcomed all kinds of contributions; we did eventually have Spree guides, so we did end up doing that, but a lot of the contributions I would say on Spree were code contributions, people improving the framework and that sort of thing. Documentation wasn't kind of our focus on that project.
+**Nadia Eghbal:** Maybe this is a dumb question, but because Spree is a user-facing application, which is kind of different from the types of open source project we've had on here before, was Spree being open source - did that change how you thought about the community manager role? Was it like being a community manager at any other company? Was anything about that particular to open source?
+**Ryan Bigg:** I've never been a community manager at any other company so far; Spree is my only experience with that. What we were doing at Spree was mainly answering emails, maintaining the GitHub issues, doing new releases and that sort of thing. I don't know if it's any different to any other community manager role. I didn't really get to talk to many other community managers and I really wish I did, because I think I would have got a lot better -- I would have built up a support network, and that may have prevented the burnout.
+**Mikeal Rogers:** You mentioned that you were traveling a lot and going to conferences and stuff like that - what was your schedule like? How much were you actually traveling?
+**Ryan Bigg:** Well, I think it was 2013 that I traveled six and a half times around the world, the equivalent of... So about 180,000 kilometers. That was a lot of flying. My longest flight was 42 hours I think it was, from point to point. It was crazy.
+There was this time where I was consulting in the U.K., then I flew back to Melbourne and then had to fly back to New York for a conference the week after, so I spent three or four days of that week in airports. That's the less glamorous side of going around the world and traveling and speaking at conferences.
+The glamorous side is that you get to meet all these people and they're like "Oh my god, it's Ryan Bigg! He wrote the book and he maintains Spree! He's actually real, he exists! He's been really helpful, and I love him!" That kind of fanboyism is hilarious. My wife got approached at a conference that we both attended; she came along with me to DC, and this guy kind of cornered her and he's like "I have this idea I wanna tell Ryan, but he looks busy at the moment so I'm gonna tell you and hope that you pass it on." She doesn't remember what the idea was... \[laughter\] She was just like "He's right over there, you can just go up and speak to him", and he didn't... So I don't even know who the guy was, but if he's listening, I'm keen to hear what his idea was, still.
+**Nadia Eghbal:** Do you think some of that contributed to the feeling of burnout in that type of a role?
+**Ryan Bigg:** Yes, absolutely. It's this -- what's the word I'm looking for? Not Messiah complex... Fanboyism, hero worship probably is the better term for it. People think you're the most amazing thing and you do all this amazing work, and therefore my feelings for that were "People think I'm amazing, so I have to continually be amazing, I have to continually push the envelope and continue working and reply to all these issues and reply to all these emails and just keep going, keep going, keep going." I think that's definitely a contributing factor to the burnout; I felt like I had to be the hero.
+**Nadia Eghbal:** \[12:26\] Do you feel the same way about documentation as well? That sort of hero worship?
+**Ryan Bigg:** Yeah, in a different kind of way. With documentation typically the projects are a much slower burn, if that makes sense; the open source projects are -- you know, if an issue or a pull request sits there for a couple of weeks, people are like "Why is it taking so long? He doesn't love me anymore." But if you're writing a book or a guide, people don't care that it's taking you weeks to write, because it's this slow process of getting the content right, getting it ordered, reviewing it, editing it, making sure it's perfect before releasing it out into the world.
+With the Rails guides there was never any pressure to get them done. With the books I've written, there was pressure from the publisher, but as Douglas Adams says, "I love the sound that deadlines make when the whoosh past."
+**Nadia Eghbal:** One of the reasons I was interested in having you on this show was because you've contributed in a variety of different ways to projects, which I don't know that a lot of people can say they've written code and done documentation and done community management, so you've seen sort of like the whole view... A topic that's come up before has been code versus non-code contributions, and kind of reading about your background - I'm just sort of thinking out loud here, but sometimes people portray that as like you have to find the non-code people to make those non-code contributions, or that's the implication, and there's this sort of like false dichotomy between developers code and non-developers do non-code things, but you're a good example of someone who does all of that, right?
+**Ryan Bigg:** I don't see any point in writing code if you can't explain what it does in the form of documentation. If you're developing an open source project and you want people to use it, yes they can look at the code, but having documentation where it's like "Step one, step two, step three, step four, and now you're winning" - that is extremely helpful. And even documentation explaining the thought process is sometimes useful as well. And you're right, there is that dichotomy of people who -- developers are like "No, I don't write documentation, I write code", and they're all hoity-toity about it.
+It's a very strange dichotomy and I'm trying to crack it, but I can't convince people to write documentation. They think they suck at writing documentation, but typically they don't, they're okay. Documentation is sometimes sparse, and that's why they think that documentation sucks, because the documentation they write is sparse. I don't know if you know the meme of like "Draw two circles, and then draw the rest of the owl." So it's like "Step one, step 16", but there's like 15 more steps in between they're missing out.
+My job then, if I'm working with somebody like that, is to encourage the documentation to be written, however they wanna write it, and then I work with them in reviewing the documentation, like "Hey, by the way, did you think about adding this in? We could explain the wording here. How about having an image describing the flow of this state machine, for instance?"
+Like Spree's payment gateway - you've got to authorize the capture, the refund, all of that; you can explain it with words, but you can also explain it with a pretty picture, which helps people understand the flow much better.
+**Nadia Eghbal:** It's almost just like having those methods I guess out there... It helps people help themselves, right?
+**Ryan Bigg:** Yeah, that's right.
+**Nadia Eghbal:** Just thinking about it, I feel like there's almost -- the explanation that developers don't do that kind of stuff is almost just trying to pass the buck to some mythical person who's gonna come in and write all this stuff, but there's value in just having people learning how to do at least a little bit of it themselves.
+**Ryan Bigg:** \[16:02\] Yeah, try it. I mean, I didn't just instantly become good at writing documentation; I don't even think that I am good at writing documentation still. It's a practice, it's an art, it's a craft; any skill that you learn - you get good at if from practice. You learn the piano by playing the piano a lot. You get better at writing code because you write a lot of code. You get better at writing documentation because you write a lot of documentation, and that's what I have done. Again, I don't' think I'm any good at it, and if people feel like they aren't good at writing documentation, I'd still encourage them to try it. Trying is at least practicing, and you will get better through practice.
+**Nadia Eghbal:** It's better than zero documentation.
+**Ryan Bigg:** Exactly.
+**Nadia Eghbal:** Do you think that that changes for a big project versus smaller projects? I know you've maintained smaller projects on your own and then you've contributed to bigger ones like Rails - do you think that responsibility changes when it's a small project versus a big one?
+**Ryan Bigg:** Yeah, on a small project you probably -- if it's a really small team, it's up to the developers of that project to do that documentation, whereas on Rails it's not necessarily up to the Rails core team to do that documentation, although I wish it were, because then they would have had routing documentation, they would have had the asset pipeline documented, and probably engines as well.
+On the larger projects you've got people who add new features, and you probably have people who want to write the documentation, and getting them to work together is the key there.
+**Mikeal Rogers:** Also, the smaller projects probably don't have anything as complicated as the asset pipeline to document, right?
+**Ryan Bigg:** No, absolutely not. I still don't understand the asset pipeline.
+**Adam Stacoviak:** Coming up, Nadia and Mikeal talk with Ryan about his departure from open source. In November 2015 he wrote a post announcing his departure, so we asked why he wrote it, what were the events leading up to it and what was the response from the community. We also talk about how he stepped down and handed off his projects, and which strategies he might suggest for those wanting to do the same.
+The interesting thing is that Ryan didn't actually quit open source; he still occasionally contributes, but the big thing we figured out is if quitting was the right decision for him... And you might be surprised what his response is. Stay tuned.
+**Break:** \[18:30\]
+**Mikeal Rogers:** I wanna get into this whole time when you burned out for a minute. So before you burned out, before you wrote this blog post, what were your days and your weeks like? What were you doing on the daily?
+**Ryan Bigg:** On the daily I'd get up in the morning and I'd check my inbox and see what open source stuff there was; I'd reply to issues that people had left comments on overnight, and then I would go to work and do my full day of work. On the train home I would check my inbox, and then before going to bed I'd check it again and do the same thing. And that was every single day, even on weekends. Especially on weekends I had free time, so I was able to do some actual coding work on these open source projects. That just meant that I was working all the time.
+The same thing happened at the end of my Spree term. That was a similar kind of feeling where I had all this work to do and it felt endless, and I kind of just collapsed in me. With Spree for instance, there were several days, I reckon, several months of this where I would wake up and there would be 200 new messages in my inbox... Over 200. This is the day before completely clean, empty, and then the next day there's be 200 messages from the user group from GitHub, from the open source projects I was maintaining, and it just felt overwhelming to have to jump through that every single day.
+When I quit open source in 2015 I felt the same way. I had all this work to do, I was trying to write Multitenancy with Rail, the second edition of that, and I had my wife -- I was trying to spend time with my wife and not be feeling guilty that I'm having a good time while all these open source projects are sitting there stagnating and nobody is working on them, and people are actually using them, but nobody cares about them enough to contribute.
+It was a really difficult time, and I just felt like something has to give; either I have to spend less time with my wife, I have to spend less time at work, or I have to spend less time on open source. The wife situation isn't going away, absolutely not; that would be terrible. Absolutely terrible if that went away. Work is enjoyable and it gives me money that I can use to spend on actually having a roof over my head and food on the table, and open source usually brings me joy. At that point in time it wasn't bringing me joy, so I decided that's the thing I was gonna cut out of my life, it's the open source work. I would spend my time that I was spending on open source focusing on Multitenancy with Rails.
+What happened after that was from November through to (I think it was) June or July 2016, I was able to complete Multitenancy with Rails, and I got a big thrill out of that when I finally did complete it.
+**Mikeal Rogers:** You mentioned that you were a community manager at Spree. A lot of people, when they go for these kings of roles, they go for them because part of that role is usually working on open source. So I think the assumption when you said community manager was that a lot of that open source work was part of your job now. So was it that it actually wasn't part of your job and that it was entirely separate, or just that there was so much extra open source stuff that you couldn't get it done in your regular workday?
+**Ryan Bigg:** Yeah, there was just so much extra stuff I just couldn't get done in the regular work day. Because the work day wasn't just all about the open source issues and code and pull requests and documentation, it was about replying to users on IRC, there were long discussions there; big issue threads on GitHub and also on the mailing list. It was just all of these conversations that were ongoing, kind of like "touch this one, touch that one, touch that out. Okay, good, now I can get back to open source." And it felt like this massive pile of work that I had to do. It was very tough.
+**Nadia Eghbal:** Had you tried anything prior to straight up quitting? Like "Maybe I just need to manage my email better" or whatever, and then did you just hit a point where it was like "Nope. No, this is more than that."
+**Ryan Bigg:** Yeah, I kept trying to manage the email, but it's hard. "This news stuff that wasn't important, bounce it to the next day" sort of thing; I was trying to ask for support from Spree (the organization), I was just asking for somebody else to help maintain the open source issues, because that was just me doing that.
+\[24:04\] There was interest from this company called FreeRunning Technology who is now called Stembolt, and they were doing some open source maintenance, but they weren't as interested as they are now. They actually run the fork of Spree called Solidus and they're maintaining that, because after I quit, Spree kind of fell down; it had to lie down. So these Stembolt guys picked it up as Solidus... I wish they would have done that earlier; I would probably be continuing my work with them, but you know... Things happen.
+So yeah, it was just hard. I did try to manage it better, I did try to get support, and I guess I just didn't reach out to enough people or the right people to get that kind of support. I felt like that was the only thing I could do, it was quit.
+**Nadia Eghbal:** What was the response like when you wrote that post? Were people understanding? Yeah, people were very understanding of the quitting open source part; more understanding than I thought they would be. I thought people would be like "But, but hold on, I've still got this issue... Can you just look at this issue?" People were actually not like that. They were really respectful, and they were like "Okay, that's fine. I can fork your project and merge my pull request, that's cool... And I'll just run off this fork for the time being."
+Other people in the community were actually volunteering their time to help out. I've got this guy called [Johnny Shields](https://github.com/johnnyshields) who's maintaining [by_star](https://github.com/radar/by_star/), John Hawthorn ([jhawthorn](https://github.com/jhawthorn) on GitHub) is maintaining [Paranoia](https://github.com/rubysherpas/paranoia). The forum - I wish there was a maintainer for that, but there isn't. I can't think of any other projects now.
+Redis - I was maintaining the Redis suite of gems But still, these projects will have a single maintainer. I wish there was a community around that was interested in maintaining these, not just people who are using it passively, but people interested in actually the ongoing support of these projects.
+**Nadia Eghbal:** It's funny, this makes me think about Mikeal's post actually... When Nolan Lawson recently wrote the post about starting to feel burned out by open source and Mikeal wrote a response just being like "You need to step away in those situations, and someone will always take over" - Mikeal, I'm putting words in your mouth, but...
+**Mikeal Rogers:** Well, I wouldn't say always, but stepping away is an opportunity for other people to step up, right?
+**Nadia Eghbal:** Right. That was the thing that stuck out in my head about your post. It sounds like there's often this fear that if I step away people might be mad, or everything will disappear or fall down or whatever, but often times people are okay, and if you just say you're not doing it, then you're just not doing it, right?
+**Ryan Bigg:** Yeah. I should have just asked for support. It's absolutely what should have happened. It shouldn't just come down to quitting, it should be like "Hey, by the way, I'm having a hard time here. Here's what I need help with." But it's the anxiety about it; I couldn't phrase it that way before. Of course, with hindsight I can, but at that time I just couldn't. It just felt overwhelming; it was like "By the way, that's it. I'm pulling the plug."
+**Mikeal Rogers:** Well, I think when it's that overwhelming, the right thing probably is to just say "You know what, I'm stepping away entirely." Maybe six months before that it might have been good to be like "These are the things that I need help with", but once you're in a certain point, your mental health is more important than whatever you feel like you owe to the project.
+**Ryan Bigg:** Yeah. And I do feel a lot more mentally healthy (is that a thing?), I feel happier now that I've got less responsibilities in terms of open source work.
+**Nadia Eghbal:** I was wondering about that, because you still make occasional contributions, and it seems like you're still paying attention to things in open source. Was that post more just about having to draw this line in the sand and say "I'm not maintaining projects, and I can contribute when I feel like it", or did you have a point where you were like "Oh, I kind of miss this and I wanna come back"?
+**Ryan Bigg:** It's true, I do continue to maintain some open source projects. I've got this open source project called Elastic which talks to ElasticSearch servers using Elixir. I actually use it at work, so when I'm working on that, I maintain it at work, and I never do any maintenance on it outside of work, because it's a work project, technically speaking. And they're completely cool with that; Culture Amp is great for that sort of thing.
+\[28:10\] And then I also have another open source project that I'm maintaining called Twist, which is my book review software, which I've been tinkering on for about six years. It's open source, but nobody else contributes to it, pretty much it's just me.
+**Nadia Eghbal:** I'm curious just to hear more about how you handed off your projects and facilitated that process of stepping down. You said you just put out a call for volunteers and you found some maintainers; was there anything that you officially had to do on your repos? Just for other people that might be interested.
+**Ryan Bigg:** Yes. So when I quit, I left a big note in the readme saying "This project is no longer maintained. If you want to maintain it, please contact me at ryanbigg.com." I also put that in the blog post, too. I've got several emails from people who were interested in maintaining it, and I was like "Well, these people I don't really know and they've never really contributed anything", but then there's these other emails from people who I do know and that have contributed, so I made those people maintainers, because I felt like they would have more longevity with the project than the ones who didn't, who were just like "Oh, it's an open source project I can probably maintain."
+**Nadia Eghbal:** That I wanna hear more about. I noticed you had something on -- I think it was translation-gem maybe... You were helping to triage issues, and then expressed a little bit of frustration that whoever brought it up wasn't helping out.
+**Ryan Bigg:** Oh, you saw that too...
+**Nadia Eghbal:** I was just wondering, how do you know who's gonna be a "good maintainer" or committed maintainer? How do you actually pick someone to hand off to?
+**Ryan Bigg:** So there was this guy in the community who I've known since I was a pup... There was this guy called Jason King, and he has a lot of opinions, and some of them are right. This particular opinion was that the i18n gem was unmaintained and abandoned and nobody loved it, and Sven is a horrible person for doing that. That was his opinion, and that opinion happened to be horribly, horribly wrong, and I disagreed with it so much that I came out of my open source hiatus and triaged every single issue on i18n that was still open, every single pull request, and assigned them into milestones, and just on the train to work and train home I was able to "maintain" the i18n gem, and actually did a new release of the i18n gem and fixed a lot of long-standing issues. I think the oldest one was from 2011, so that's a six-year-old issue now that I've been fixing on i18n gem.
+I'm probably going to step away from the i18n project now that it's done and maintained in a relatively healthy state. So if anyone else wants to step up and maintain... I'm sure Sven and I can find somebody else who can do that.
+**Nadia Eghbal:** Do you feel like because you sort of made a conscious decision to step in, did that make it easier to contribute where maybe others didn't? Is it just you cared more than other people? It sounds like it was an easy problem to solve, or you kind of just came in, did it and then you left.
+**Ryan Bigg:** That's a tough question. I think it needed to have two things - it needed to have somebody who cared somebody who had time; well, that person needed to have both of those things. They needed to have the care and they needed to have the time. I happened to have both at the time, because I'd finished writing Multitenancy With Rails and was kind of between books, and I was looking for stuff to do on the train, and I found i18n needed help, and that's what I did.
+If anyone else feels the same way, I'm sure Sven and I could find somebody else to maintain it as well over the long haul.
+**Nadia Eghbal:** Would you recommend handing -- for people that are newer to open source and looking to build their reputation, would you trust someone like that to take over a project that has been handed off, versus a project that they started themselves? Does that make sense? Is that the right kind of role?
+**Ryan Bigg:** \[32:02\] I wouldn't recommend, like, for Sven and I to hand off i18n to somebody who's new to open source. I reckon what we could do is have this handover period of probably a year where we monitor the issues and just give guidance, that kind of thing, so if people are interested in open source, they can be like "Hey, what kind of issues are there?" and they could try and triage the issues themselves and they can give them guidance on like "You know, maybe don't call users idiots because they don't understand how the software works. Maybe don't disparage them using a lot of horrible languages..." It happens on a lot of open source projects. "Be kind, be generous with your time, but don't overdo it and don't burn out..." Kind of give them that guidance of like "It's not critically important that you work on this and if you feel like it's too much, then it's okay to step away at any point. But while you're here, we'll give you guidance, we'll give you love and we'll support you 100% of the way."
+So yeah, I think if new open source contributors wanna come in, by all means...
+**Nadia Eghbal:** Do you feel like you had that mentorship for yourself as a maintainer?
+**Ryan Bigg:** In the early days no, but with practice and Spree and a lot of the community discussions around that I believe I did get a lot of mentorship there. And there was also a lot of people who kind of grate on you a little bit. There's this one guy - I'm not gonna name him, because you know, name is withheld to protect the innocent - one day he just really got on my nerves; I was so mad at his petulance, his entitlement, and I was like "Screw this, I'm going for a ride on my bike."
+So I get on my bike and I ride two kilometers, a mile from my house, and I have an accident. I come off my bike and break my left arm. So I'm mad from this guy, and then this car cut me off and I broke my arm and I just cracked it really hard. I pick my bike up off the road and threw it over to the curb and I just went to the driver and stuck my arm out towards her, finger extended, I'm like "You!!!" and she's like "I'm sorry, I'm sorry... Do I need to take you to the hospital?" and just that sentence, like "Do I need to take you to the hospital?" completely simmered me off. I just had the crappiest day, and that was not a good way to end it, but then they gave me some really good painkillers, and that made my day so much better.
+**Nadia Eghbal:** \[laughs\] Worth it!
+**Ryan Bigg:** Yeah, it's that kind of thing -- in open source people can be discouraged because of the hatred and vitriol and entitlement; every open source maintainer comes across it. I know Aaron Patterson talks about it a lot on Rails, and there's other Rails contributors who talk about it a lot... That sense of entitlement, of like "Why aren't you paying attention to what I'm saying? I'm right, you're wrong." It's hard to deal with as a new open source maintainer, and having that guidance of like "It's okay, just be nice."
+One person who I think does this really ridiculously well, who always seems cheery and happy and chirpy is Jose Valim on the Elixir project. He consistently is happy, treats people with respect even if they're being a bit blunt or a bit obtuse, and he's just an inspiration absolutely. I wanna be more like Jose.
+**Mikeal Rogers:** It seems like you still pick up new projects here and there... Do you think that to some extent you enjoy picking up a new project and getting involved in new projects more than the inevitable maintenance of them? Is that kind of part of why this continues to happen as well?
+**Ryan Bigg:** I'm a sucker for it. I'm such a sucker for it. With i18n I think I am not going to maintain that any longer with anything else, but the Elastic package - I'm probably going to maintain that as long as I work at Culture Amp or as long as I need it. But I try and realize that I'm slipping back into that lifestyle I guess you could call it, of like "Oh, I'm just gonna check these issues. I'll just reply here and I'll help out here", and that sort of thing.
+\[35:59\] I try and catch myself doing that, and I go "No, that wasn't healthy before, and it's probably not gonna be healthy now, so let's just not do that. Let's find something else to distract our brain with", so that's what I'm doing with my writing at the moment; that's my distraction from open source, if that makes sense.
+**Mikeal Rogers:** Have you had any projects that you picked up like that or sort of working on that then did grow the community and that you were able to just kind of step away from without people noticing? Are there any practices that you might have learned along the way?
+**Ryan Bigg:** Paranoia is probably the only project that did that. It's in accessparanoia.alternative/equivalent for active record, and then when I did quit this open source work there was a substantial community around - I guess substantial is like 20 people - it who were using it actively. One guy stepped up and that was really good, and he maintains it now. But I wish there was more of a community ethos around maintaining the open source projects, not like "Oh, Ryan will merge this pull request..." Or maybe it should have just been that I should grant people permission to do that, I don't know.
+**Nadia Eghbal:** So sort of wrapping up this segment - so people use the term burnout a lot to talk about when they leave or step down or whatever... Do you think that people need to call it burnout in order to leave? Is it okay just to leave because you're kind of done with something and you wanna focus on something else? Do we always have to call it this failure?
+**Ryan Bigg:** Yeah, burnout is used as this convenient excuse; it's like "My mental health is being affected by this project and now I feel like leaving. I'm going to leave now." I think while that is mostly true, in some cases it may not be true; it may just be used as a convenient excuse.
+Also, the reason I didn't quit or hand over the projects earlier is because I had the anxiety around "If I quit, then people will hate me, they'll not use the project and this will have all been a waste of time. It will all just crumble into dust", and that's not true at all. That didn't happen. People continually use the projects now, and there's people who still maintain it and they didn't crumble into dust. It's the anxiety that I had about maintaining it that kept driving me to that point of burning out.
+I think what we need to have is more of a discussion in the open source community like "If you are no longer interested in the project, if your heart is no longer interested, just like a regular job, you're allowed to leave. You're not chained to your desk, you're not forced into the building every single day." You can choose to leave at any point in time, but it's polite and encouraged that you tell people you're thinking about leaving and you commence hand over procedures if possible.
+**Nadia Eghbal:** That's what made me start thinking about it, because it is like with a job, where... I imagine for people - if you're really passionate about your job and you might be afraid to leave because you're like "Well, who's gonna everything that I was doing?" or "The next person may not be as good" or whatever and it's hard to leave a job, but I mean... In the end it's possible to just quit and move on, not because you hated the company or it didn't work out, but you're just sort of like "I'm done and I've done the things I wanted to do." I feel like there's something to be learned there for the open source as well, or any kind of -- I mean, whether it's volunteer or a paid hobby/activity.
+**Ryan Bigg:** Yeah, absolutely. I agree.
+**Nadia Eghbal:** Something that Adam was musing about that I also wanted to ask you about is do you think that we're hearing more about burnout now, just because open source as a practice or community has started to mature, and there's sort of this question now if it's not just about creating stuff in open source, but now we're kind of getting to the point where people are starting to wanna age out of a project or move on? Do you think that's contributed to why we talk more about burnout now?
+**Ryan Bigg:** \[39:51\] I think that's definitely the case. I think what we're seeing is the first generation of open source contributors - or at least on GitHub projects... I mean, I'm only new to this game; open source has been around longer than I've been alive. I think we're seeing these first-generation open source contributors go "I'm no longer passionate about this project. I'm thinking on moving on, but my work is going to crumble into dust and it will be meaningless if I do so", and then they burn out from it and move on.
+So what we should be encouraging is the second generation or next generation of open source maintainers to come along. Encourage the first generation to say "By the way, I'm looking for help", exactly like what I should have done. "I need help with these projects. I'm trying to maintain them, but I'm struggling." It's okay to admit defeat in that way.
+So when they do that admission, then they should look around the community, like "Who's been contributing to this project? Who can I tap on the shoulder to be the next maintainer or the next group of maintainers for this project?" That's probably why we're hearing more about burnout, because that first generation is getting tired.
+**Adam Stacoviak:** Coming up we talk with Ryan about getting paid for open source work and whether or not money actually helps. We talk about his thoughts on compensating open source developers, funding an individual contributor versus funding a project, a happy path to maintainership if there is one, and so much more. We'll be right back.
+**Break:** \[41:25\]
+**Nadia Eghbal:** I wanted to touch on what it's like to get paid for open source and how money and open source mix. When you were at Spree, you were being paid to work on open source, and it sounds like even some of your open source work now is on work hours. Do you think that the presence of money helps keep you in open source?
+**Ryan Bigg:** Certainly. It absolutely does. If for instance I was not paid to work on Spree, I probably wouldn't have done as much work as I did. It allowed me to live my life, have my roof over my head, have food on my table sort of thing, for two and a half years. Within that two and a half years I did a substantial amount of contributions to Spree; I think it's over 4,000 commits. Without that pay I wouldn't have contributed 4,000 commits; maybe 40, maybe 400, who knows? But it allowed me to do a lot of good work on open source, and not worry about where my next -- how I'm going to pay rent or how I'm going to get food. It was really beneficial.
+With regards to culture and paying me to do open source contributions, they benefit indirectly from that because we're using the Elastic package at work, and if that Elastic package didn't exist, then we would probably use another package or somebody else would have to contribute to another open source project to do the things with Elastic that we wanted to do. And even then we would be contributing to open source, so either way we're going to have to spend time, and time is money, so we're gonna have to be spending money on open source contributions, too.
+**Nadia Eghbal:** Are there tradeoffs to it? Are there things where being paid to work on open source makes things harder, or just compromises you in some way?
+**Ryan Bigg:** Yeah, your ethics are kind of compromised in the way that "We're paying you to do open source, so these are the things we think are important and these are the things you should work on." At Spree, because I was paid to do Spree work, you've got this large group of work that you can do, and there's a large boundary around the work that you can do; there's this huge area of the work that you can do. There's a never-ending stream of work there, and you could contribute to the Spree main project, or you could contribute to extensions, you could contribute to the Google group, you could contribute to the IRC discussions... And they didn't mind as well that I was contributing to other things at the same time.
+If I was paid to work on a particular open source project -- let's say I worked at Facebook and worked on the Babel project for instance, even though I'm not a JavaScript developer; a large open source project like that, used by thousands of people, because I'm paid by Facebook, Facebook gets their say of what I contribute to Babel, for instance. I think that compromises the ethics in that regard, because they get the features they want, rather than the project getting the features the community wants/needs.
+**Nadia Eghbal:** Yeah, it's a different kind of project, right? Which I guess again goes to show that open source is not all the same across all projects.
+**Mikeal Rogers:** I mean, I think it's rarely that direct... It's usually pretty indirect, right?
+**Ryan Bigg:** Yeah, it is... It's not like "By the way, we're gonna fire you if you don't add these features in." It's never that direct.
+**Mikeal Rogers:** Yeah. I think maybe a better example is the Go project, where the majority of the committers are at Google, and Google's certainly not telling these people what to do, because they are the kind of developers that are not told what to do... But they are sort of constantly inundated with Google's problems, and Google's scale problems, and the types of engineers that Google has, and that really paints who they view as their audience, right?
+**Ryan Bigg:** That's exactly right, yes.
+**Nadia Eghbal:** \[47:39\] I guess part of me wonders what makes open source not something that is fun as a volunteer hobby...? That's sort of going through my head right now, but then I think "Well, some people do enjoy contributing to open source in their spare time... And I guess I was just kind of thinking about, like, are there different stages of open source either in your own career as a contributor, or different stages of a project where doing it for free or doing it as a volunteer makes sense, and then there's a point where it's kind of like "I need to be paid for this, or it doesn't work"? Does that seem coherent?
+**Ryan Bigg:** Yes, it does. it does absolutely seem coherent. So when I was starting my contributions to Spree, it was a lot of times of doing Google group answering, and a couple of issues here and there, but nothing substantial. And then working on the project itself is enjoyable, so when I was contributing to other open source projects, you start out like "Oh, I'm just gonna make this contribution here" and then "It feels great, because people get to use my code; that's mainly why I contribute to open source. There's going to be somebody else out there - there's a market, in a sense, of people that wanna use my code", and that's a good feeling. "I wrote something that people are using and benefitting from", and that is why you contribute to open source in the first place.
+But then open source is not free as in beer, it's free as in puppy, and you have to maintain it, and people come up with weird and wonderful feature requests, and bug reports that are literally a single line of like "It didn't work." I'm like, "What didn't work?" That grinds on you, that just gets to you over (I don't know) years and years of maintaining the same open source project.
+And you know, your interests change over time as well. That's another issue. You get started with an open source project and you love it and you work on it every day because it's something that you're passionate about. Then if the passion kind of fades with time, you're like "Well, that's no longer interesting to me, so I'm not gonna contribute to that anymore."
+The money helped a lot with Spree. While I wasn't in love with the code as much as I was at the start, after two and a half years, the money kept me going. I was like, "I get paid to do this; I'm just gonna keep thinking about these problems and trying to solve them, even though the code is no longer enjoyable to work with", like it wasn't really my passion at that time, at the end of it.
+**Nadia Eghbal:** It's almost like there has to be some reward at some point, whether it's building your reputation that's a suitable reward, or at some point then it just kind of becomes money.
+**Ryan Bigg:** Yeah, the reputation definitely came out of that, and I still get approached at conferences, like "Oh my god, it's Ryan Bigg. Oh, it's him!" "Oh, hello." "You wrote the book, you contributed to Spree", I'm like "That was years ago... I've done a lot more now." I don't know, it's very strange having this reputation... This reputation is a niche of a niche; the general layman view of a celebrity is somebody who's famous everywhere - their photos are in magazines and whatnot. And so when I tell people that I'm a celebrity - jokingly, mostly - they're like "Oh, what do you do?" and I say "Well, I write books and documentation and I contribute to open source", and they're like "Oh... What, what do you do?" \[laughter\]
+But in the niche of a niche of a niche of a niche where we exist, in this Rails world, people are like "Oh my god, it's Ryan Bigg!" It's so weird to see that change of people who are like "Oh, I don't care about you", and other people who are like "Oh my god, it's you!"
+**Nadia Eghbal:** It's kind of the best of both world, right? Because then you can go to something else outside of open source and no one knows your name.
+**Ryan Bigg:** Yeah, that's right; lots of anonymity there.
+**Mikeal Rogers:** It reminds me of like actual celebrities that are on a sitcom, and then they move past that sitcom and do like major independent films, and everybody's like calling them the name from their character 20 years ago... \[laughter\]
+**Nadia Eghbal:** Yeah, totally.
+**Mikeal Rogers:** You're like, "I've done more than Rails, come on!"
+**Ryan Bigg:** \[51:54\] Oh no, that's okay... Any contribution is fine; they can name anything I contributed to. I don't care if it's Rails 3 in Action or Multitenancy with Rails. Anything. It's just good to be known, and I try to get to know them as well. I just don't want to be like hero worship, but I wanna know who is doing the worship, if that makes sense. Because by talking to these people and not just going "Oh, that's good that you like what I do." I'm like "What do you do?" and I find out a lot of interesting stuff in the community through just that. So the fame comes with a lot of finding out information from people and finding out these cool things they're working on, or projects they're looking into. It's really good.
+**Nadia Eghbal:** Does any of it surprise you, that people from totally random walks of life that take interest in it?
+**Ryan Bigg:** Yeah, it does. Absolutely. I had all kinds of people - old, young; all sorts of things - approach me, and it's just... And when they come up to you and they're like "I read your book, and now..." So, that's right. I went to a Rails meetup in Sydney, and this lady called Metanya introduced herself to me and she said "Hi, I'm a real estate agent. Or at least I was until I read your book. Now I'm a Rails developer, so thank you." That was awesome, and I keep thinking of that when I'm writing and I'm getting down in the dumps about that, kind of that writer's block, like "Oh, I'm never gonna finish this project; no one loves me anymore and this project is no longer useful to anyone." I think well, maybe this project will be useful to another Metanya down the line. Maybe somebody will read this book and go 'I understand this concept now. I'm gonna become the most amazing Rails developer', or the most amazing whatever developer. It's a really good feeling.
+**Nadia Eghbal:** That's awesome.
+**Mikeal Rogers:** There's been a lot of self-reflection -- I'm just trying to think, like, is there anything else that you've learned in the last few years about managing your workload, and just sort of being better at managing your work and not burning out, and just kind of feeling happy in your daily life? What would you tell yourself two years from now, and by extension, to a lot of developers out there, before they hit burnout as well?
+**Ryan Bigg:** So I'm good at really thinking about how terrible I am at managing my workload, and my wife is also really good at thinking about how terrible I am at managing my workload; not only that she's good at telling me how terrible I am at managing my workload, and...
+**Mikeal Rogers:** Your advice is to get married? \[laughter\]
+**Ryan Bigg:** My advice is not necessarily to get married, to have somebody like a close friend for instance - even that; you don't have to go the whole way of getting married, but somebody who can say "You're being a ridiculous idiot. Stop working so hard. How can I help you? How can we stop this burnout? You're grumpy, you're really easily agitated, you're not paying attention to these things... Why are these things happening?" and having somebody there to talk through it is great.
+For a new developer - two years from now, I hope that he would know better practices of maintaining workload. If I was able to speak to Ryan from two years ago, I would say it's not that critical that you continue doing all this work. None of it matters that much that you have to kill yourself to do it, that you have to spend all this time doing it. You're allowed to go out and have fun, you're allowed to go out to a movie and watch the movie and not think about open source. You're allowed to go for a walk in nature and not think about open source. You're allowed to go out for dinner with your wife or an anniversary and not think about open source. It's okay, nobody is going to die if you think about open source. It's not important.
+**Nadia Eghbal:** It's a good advice for open source communities too, right? Even reinforcing that to each other, and not expecting that the other person is always gonna be there all the time.
+**Ryan Bigg:** Yeah.
+**Mikeal Rogers:** Yeah, I have a similar story to this, except I've been with my wife for like 12 years now. She would be like "You're being ridiculous now", but now I can actually recognize it before she needs to tell me. This is exactly what happens, actually - there's certain things that she's just never gonna do and I'm never gonna do and we'll slightly annoy each other, and when I start to get really mad about one of them, I know that I'm actually not mad about that, I'm getting sick of my job. \[laughs\] Literally, I'll start to be like "Man, she did not refill the water thing again that goes in the fridge", and I'll start to get really mad about it and I'll be like "Ohhh, I need a new job..." \[laughter\]
+**Nadia Eghbal:** \[56:28\] That's a good degree of self-reflection there. \[laughter\]
+**Mikeal Rogers:** It's taken a long time to get there, but yeah.
+**Ryan Bigg:** I realized that when I come home grumpy it's probably not a good thing -- if it's a single day, like, everyone has bad days. People ask me like "When does programming become easy?" It doesn't. You just have more easier days than you have hard days. You still have those hard days. And when you have those hard days you come home grumpy. But if you have a series of hard days in a row or hard weeks in a row and you're constantly grumpy, that wears you out and it wears the people around you out. Personally, I need to realize when I'm getting grumpy and how long I've been grumpy for, and go "What's making me grumpy and what can I eliminate that..." -- eliminate sounds bad, doesn't it? It's like I'm asking for an assassin. "What can I get rid of out of my life that will make it easier for me to be happier? Can I stop contributing to open source? Do I need to put down the writing projects that I'm on? Do I need to pay more attention to my wife and my daughter? What do I need to do to make me happy?" It's about setting aside that time to have that retrospective thought of "What's been making me grumpy and how can I fix it?"
+**Nadia Eghbal:** Hypothetically, is there anything in open source that would need to change or that could change that would make you come back and start contributing more regularly?
+**Ryan Bigg:** You could pay me to do it. You could pay me what I get paid to work full-time on proprietary software to work on open source software, just like I did on Spree. That definitely helps a lot. To come back to open source projects -- I just simply don't have time anymore. I've got my wife, I've got my daughter, I've got work, I've got stuff I do on the weekends, I've got stuff I do at night... There's just no room in the schedule for it.
+When I do have free time is on the train to work - it's half an hour in and half an hour out. I spend that time writing usually, so there's really no time for open source.
+If you gave me a full-time job where my purview was literally "contribute to open source however you see fit" and you paid me a livable wage, by all means, I'm open to job offers.
+**Nadia Eghbal:** Good answer. When I've seen you advocate for people getting paid to work on open source, do you think that just has to come from companies, versus raising money for yourself, or whatever?
+**Ryan Bigg:** Yeah, I'm a hard left-wing kind of guy and socialism is high on my agenda, and it's not high on people who have the money, it's high on people who don't have the money. With regards to that, while I would love if Culture Amp would hire me purely just to contribute to open source projects that they use, or funnel money into open source projects they use... Like, we use Webpack extensively, we use React extensively, Babel, Rails, Ruby... We're not contributing anything back to these projects directly, and while we do encourage open source contributions that are relevant to the work we're doing now, we don't encourage developers necessarily to go trolling through the open source projects and lending a hand.
+\[59:55\] It's not like we have a day where it's like "Go and work on open source, it's fine...", because it doesn't make sense financially for the company to do that; you're wasting the company's -- that's why they say it's wasting the company's time contributing to these open source projects. But the way I personally see it, you're not wasting the time -- you're spending money now to save money later. You're spending money now to maintain these open source projects, and because they are maintained in a good state, those projects will be around longer and you'll be able to benefit from them longer.
+For instance, let's say Spree, which uses the Active Merchant gem extensively - if that was maintained by one person... It's maintained by Shopify, but if that was maintained by one person, and that one person decided to quit, and that project then fell into disrepair, Spree would have to pick up the maintenance of Active Merchant, because it seriously depends on it, or the community would need to exist around Active Merchant. But because Active Merchant belongs to Shopify, Shopify pays its developers to work on Active Merchant because it's important to Shopify's business.
+With a project that isn't directly important to a business, the business doesn't see any value in contributing to that open source project. Culture Amp doesn't see any value in, well, I'm speaking for another organization. My view of Culture Amp, my experience of Culture Amp is that there isn't a view of "We should contribute more to Rails and spend time helping out Rails." Even though we usually use it extensively, we shouldn't spend time contributing to React and Webpack, because that isn't relevant to the work we're doing right now, and that isn't earning the company dollars. But in the future, if those projects were unmaintained and fell into disrepair, then we would need to do work and probably switching to a different project, and that would be more money that we would spend doing that work, if that makes sense.
+**Nadia Eghbal:** So wait, are you advocating for only in a situation when a company directly benefits should they encourage those contributions? Do you think that's right, or do you think it should be different? I can't tell which side you are.
+**Ryan Bigg:** I'm sorry, yes, I did speak for a while...
+**Nadia Eghbal:** It's good, you spoke both sides really well.
+**Ryan Bigg:** Thank you. It's like that quote from Lord Of The Rings, if you're gonna ask the elves, they're gonna say "Yes and no." I'm on the side personally of companies should probably allocate at least one, two, three developers to work on open source, and not necessarily full-time, but perhaps a day every two weeks would be good... Just to say, like, "Go and work on whatever open source project you feel valuable", and if there's no open source work, then they can come work on regular work.
+I think because these open source projects are vital to the work that the companies are doing - it's like you report, the roads and bridges, right? It's the underlying infrastructure to these businesses; if these open source projects didn't exist, these businesses would struggle. If Rails didn't exist, Culture Amp would probably be on Asp.net, and I wouldn't wish that on my worst enemy. If Active Merchant didn't exist, there'd be all this buggy payment process, and someone would have gotten hacked, or credit card details would have gotten stolen, or something. If Webpack didn't exist, we'd all be stuck using Gulp, or whatever else there is.
+\[01:03:14.11\] These are vital parts of the infrastructure, and we do need to spend company's money - and we do need to spend time convincing the people with the money at the companies to contribute developer time to these projects, because they are our infrastructure and it is vital that they are maintained. So that's what side I'm on.
+**Nadia Eghbal:** Awesome. Well, I can't think of a better note to end on.
+**Mikeal Rogers:** Yeah, me neither.
+**Nadia Eghbal:** Thanks for coming on, Ryan.
+**Ryan Bigg:** Thank you!

Maintaining a Popular Project and Managing Burnout_transcript.txt ADDED Viewed

	@@ -0,0 +1,201 @@

+**Nadia Eghbal:** Tell us how you got started as a contributor, and then eventually become a maintainer of Mocha?
+**Christopher Hiller:** Well, I actually started as a maintainer. In 2014, when TJ left Mocha and all these other Node projects, he put a call out that said "Hey, I need to have somebody take over my projects. They're up for grabs." I at the time was a user of Mocha and said "Hey, I would like to help. I enjoy using this software and I don't wanna see it die." Basically, he just gave me the commit bit and didn't say anything, and that was it.
+**Mikeal Rogers:** Yeah, TJ was sort of unique... Most of his projects really didn't have any other contributors to them. A pull request here or there, but he was really pushing the solo maintainer thing. Express did have a few other people working on it, but most of his projects were literally just him being the only maintainer.
+**Christopher Hiller:** Yeah, Mocha had at least one other person, maybe another couple people at the periphery at the time I joined, but they still weren't that active. So yeah, that's kind of the story, I just joined up.
+I didn't really know what I was getting into. I had contributed to my own projects, I contributed some to Angular UI, which was an early kind of widget toolkit around Angular; it started pre-1.0, so I had done some open source contribution there, but I hadn't really been that involved in open source until Mocha.
+**Nadia Eghbal:** It looks like a pretty crazy ramp-up, from being a user to being a maintainer. Did anything get done before contributing to other projects prepare you for what that was gonna be like?
+**Christopher Hiller:** \[03:58\] Yes and no. There are things in my career as a software engineer or what have you... I have learned, for example, how to do a code review and how to receive feedback. I don't view the code I write as an extension of myself; code is code. I wasn't really bashful about sharing it, and I think that was something I had prepared for.
+But as far as dealing with -- I had never dealt with... Again, because of being a software engineer, a lot of companies will kind of insulate you from your user base directly. So if a user of the software you're building has a problem with it or whatever, they contact support and support deals with them. Maybe if the problem is bad enough, support is gonna contact you, and in some cases - this is not true, obviously, across all companies - eventually an engineer will get to talk to an end user. But I didn't have a lot of that, so I wasn't really well versed in how to deal with users or what they expected.
+There were many issues, and I like to think that I learned fairly quickly how to not be a jerk, and be polite about pull requests and stuff. But I certainly made quite a bit of mistakes there; I didn't really know too much about semantic versioning, and that caused some problems. I remember a big one was at some point I decided - and there were a couple other maintainers - to change the default reporters... So Mocha is a testing tool, and the reporter is how you receive your test results.
+I changed the default one from one which is very sparse to one which is more verbose, and we did this in a minor, and a lot of people were really upset about that. I caught a lot of flack for changing the output, not the API, and that was kind of surprising. Since then, I've really tried to drive this home with other contributors that we need to play it safe, given the size of the user base.
+**Mikeal Rogers:** So looking back -- it's interesting, you're talking about a lot of the biggest challenges have been kind of community-oriented stuff... When you first took over the project, did you view the biggest challenges as being technical, or being community-related? Did you see a long list of features or bugs or technical issues, or did you really see it as like "Okay, we have to get contributors and make people happy"? Because it sounds like in retrospect the hard part was actually the community side.
+**Christopher Hiller:** When I came on we had a lot of open bugs. I looked at those bugs and I said "Oh my gosh, look at this thing; look at the terrible state Mocha is in", and so I wanted to go in and fix them, I felt like that was the highest priority. Well, you go and you start fixing bugs, and you can do that to an extent, but the coverage for Mocha wasn't so good at the time of itself. Mocha would have tests to test itself, and we would kind of -- I would get into trouble when I was trying to fix a bug, because I would break something else.
+So at the time I felt like "Well, there are all these things that need to get fixed and all these features that need to be added", but pretty quickly I realized "Well, I can't do this all myself. I need help", so I started to reach out to basically get more help.
+\[08:14\] Some people that were with the project originally maybe had learned something that I didn't at the time; they had said that Mocha is in maintenance mode. And I was like "Meh... Neah. We've got some features we can add. I think it's a good idea to add these things, because there's a lot of people that want them", but I think what I maybe didn't realize was when a project isn't adding features, it's not necessarily for technical reasons; it's because there's just not a whole lot of people interested in contributing to them.
+So at the time no, I didn't think it was a community problem, but I came around and started to see "Well, I just can't do this all myself, as much as I'd like to, and I need help." That kind of started me down the path of looking for ways to keep the project sustainable.
+**Nadia Eghbal:** How long did it take you to figure that out? So you took it over in like 2014-ish...
+**Christopher Hiller:** Yeah, at some point I decided that it would be good, because there were a handful of features that were basically impossible to deliver with the current architecture. I felt like doing a refactoring of the core of the system would be a good idea, and I got some way on that. I was doing it by myself. It didn't seem like anyone else was really interested in helping, though I wasn't very vocal about it.
+I worked on that refactor for about six months and it didn't really go anywhere. Meanwhile, Mocha was just kind of languishing during that time, so it took me -- I don't know... I wanna say, I was working on the codebase for at least six months, and then this rewrite for another six months after that, and then once I kind of came up for air, I realized that wasn't getting anywhere. That's when I said "Okay, we've gotta figure out this community problem."
+**Nadia Eghbal:** Do you feel like you've learned those types of challenges about being a maintainer on the fly? Was it sort of through experience and realizing when things weren't working, and sort of guessing it would happen next? Did you have any mentorship, or did you read stuff from anyone else to figure that out? How did you learn how to be a maintainer?
+**Christopher Hiller:** Well, a lot of it was by the seat of my pants. Different projects demand different types of maintenance, I suppose. Mocha in particular is a project with a very large user base. The code itself is pretty touchy, so I learned to be a very cautious type of maintainer. I didn't have a mentor. I had become interested in sustainability about when I came for air after the rewrite and decided I need help. "How do I do this by myself? I wanna learn how to do this better", and I started reading articles and such, but I didn't really have anybody to tell me how to do it. I would have loved that; I think there are resources now that I didn't have back then that I would have loved to have.
+**Mikeal Rogers:** \[11:55\] One thing that we've touched on on this podcast before, but we haven't really explored enough is how hard it is competing for people's attention, especially contributors' attention, and I'm just recalling back to 2008 when John Resig first wrote QUnit, one of the things that he kept talking about was that he wasn't saying "Use my test framework", he was literally talking about how fun it is to write a test framework and how much you learn as an exercise.
+Still to this day, there are just more test frameworks than maybe any other type of library, and there are more people that just decide randomly to write their own test framework, and I'm wondering if that has been a particular challenge with maintaining a test framework, where people just are too willing to go "You know what? I think it would actually just be fun to go write one of these myself..." Have you experienced that?
+**Christopher Hiller:** I've definitely experienced some disgruntled users that did not like our decisions, and "I'm gonna write my own blah-blah-blah..."
+**Mikeal Rogers:** We can say Aaron Hammer, it's okay... \[laughter\] He's a good friend, I can call him out.
+**Christopher Hiller:** Actually, I think Lab happened long before I came on, so I was not privy to that discussion. But no, that hasn't really been a thing I've been too worried/concerned about; certainly, there have been some newer frameworks that are gaining steam or have got a lot of interest... I'm not worried that somebody who could contribute to Mocha is instead gonna go contribute to AVA or Jest or something that's the new hotness. I'm not worried about that.
+What's a little bit frustrating is simply the allure of the new thing, and how "Oh, this is a new project. We need maintainers, we need contributions, it's really exciting! Hey, everybody, let's join up! Give me a bunch of GitHub stars and let's do this!" That's been a little frustrating, because I feel like there's a lot more attention on those other frameworks while there's still many, many more users of Mocha, and it gets a lot of mindshare, but... That's one of those things where it's -- I don't know, it's kind of a bad analogy, but I was gonna say the loudest voices in politics tend to be focused on one particular issue, and you have this feeling that there are actually more of them than there really are... That's not really the case.
+I would love to find a way to make a project that's already existing -- and maybe this is what I was thinking about with the rewrite, at least in part, was make it new, make it exciting. Some sort of marketing push, a new website, social media - all that stuff. Word of mouth, like "Hey, what's old is new again. Let's contribute to Mocha! Look how awesome it is!" That would be cool, but I have no idea; I have absolutely no idea how to get that attention for the project.
+It wasn't until very recently, I think sometime late last year that Mocha actually crossed 10,000 stars, and there are projects that cross 10,000 stars -- this is just like the little silly popularity contest type thing, and there are projects that will cross that within a week if they get on the right website. It's kind of just -- the old standby; it's kind of boring and dull, it does what it does, it usually does it well, and I think that's why it's so popular. It's just simple and easy and it works.
+**Mikeal Rogers:** \[16:22\] I think in the npm rankings it's the number one depended upon, right?
+**Christopher Hiller:** Yes, and the number two is Chai, which is the assertion library that a lot of people use with it. But yeah, by a long way Mocha is used by a ton of projects on npm, and even more on just GitHub projects that aren't necessarily published. I never knew any of that before library style, of course.
+**Mikeal Rogers:** Yeah... I mean, just for reference, there may only be 10,000 stars, but there are over 200,000 downloads in the last day, so it's definitely being used and depended on at a rate much higher than people are sort of saving it on GitHub.
+**Christopher Hiller:** Yeah, and I think that's kind of a criticism that I've heard before, that GitHub stars don't really have anything to do with the actual real-life popularity of a project or the maturity of a project. They're just kind of -- I don't even know what they are. It was cool when we did pass 10,000, but I just wish there was a good way to get more attention, and not do it in a way which is "Help, we're drowning!"
+**Nadia Eghbal:** I'm just thinking out loud a little bit, it's probably a known problem with not just maintaining open source projects - I think it's exacerbated there - but any kind of maintainery duty or infrastructure or anything, even outside of software, that it's really hard to keep people's attention, and I'm trying to think about what does keep people's attention, and I think that often it's community... It's what I see in older software projects, at least. There's sort of like a cult of brand or personality or something where people will feel this love and devotion for a project; even if it's not the new hotness, it's the people that will keep it there. That's the only thing I can think of that really competes against shiny new technology.
+**Christopher Hiller:** What about instead of just a project that you consume, what if that project was also a platform and people could build upon it and make new things all the time?
+**Nadia Eghbal:** That's another way of thinking about it. I don't know if it's possible for every project to be able to do that.
+**Mikeal Rogers:** I think even when you get into that space -- Node is sort of in that space; there are other libraries, even smaller ones like Browserify, that you would think that at some point they can maybe be done and that everything would just be built on top of them, but it just never really happens. The world on top of them is changing enough that the demands and the environments that they end up in continue to evolve, as well.
+Adam Stacoviak: Up next we're talking about funding experiments. Mocha was asked early on to participate in funding their project on Open Collective, and something unexpected happened. They started getting donations. More donations than they know what to do with. It became a challenge to know what to do with the money being donated; what the project really needed was time, attention and community, and they started asking questions like "How do we raise the profile of this project? How do we turn the money being donated into what the project and the community really needs?"
+Money tends to complicate things and create conflict, so is money the real solution to sustain open source? If you have money, where does it go? How do you spend it efficiently? We ask these question and more after the break.
+**Break:** \[20:24\]
+**Mikeal Rogers:** Let's get into some funding experiments that you've gotten involved in. First, why don't you tell me a little bit about Open Collective and what made you wanna try that out as an experiment for getting the funding injected into Mocha?
+**Christopher Hiller:** I hadn't really considered going for funding before Open Collective. I can't recall, but I feel like that team actually approached me, and I can't recall why... But at the time, they hadn't had too much interest, I believe; there may have been a few projects that had what they call "collectives." I think Mocha was the first really kind of popular one.
+They talked to me, and I said "Free money? Sure, what could it hurt?" I also thought "Well, if this takes off, maybe it's a really good way to get some more attention around the project." Now, I went into it with no illusions that we were gonna get so many donations that I would be able to quit my job and go work on Mocha full-time and live off donations. I didn't think that was gonna happen. That said, that would be cool if that happened, but I didn't expect it.
+I felt that sort of thing was more of a means to an end. The end was community, and it wasn't really money per se that the project needed, it was time and attention, and I was hoping that by getting some funds injected into the project, we could use those funds to raise the profile of the project, or that sort of thing.
+\[23:58\] Now, I absolutely don't have any sort of marketing background, I don't know anything about social media; I don't know if I had the money, what would I do with it, and it turned out that not only did I not know what to do, but I didn't even have the time to spend the donations that we had. So it became kind of a challenge to figure out "Well, what do we do with this money and how do we turn this money into what the project really needs?"
+Our monthly recurring expenses were $14/month before npm decided to give organizations to open source or public projects. They made that free, so now we have no expenses, but we have some money, and I don't really know what to do with it. I would love to figure out a way to maybe send T-shirts, or something, to a conference, or who knows what? But as I said, I don't know what's the best use of that money to get more contributors.
+**Nadia Eghbal:** Where did that money come from? Is it from companies, individuals? How did you end up having $12,000/year?
+**Christopher Hiller:** We had like a one-time donation from Yahoo! of $500, I think. We had a one-time donation from [Auth0](https://auth0.com/) and the other thing was SoftLabs has been donating $500/month for quite a while now, so they're kind of Mocha's big sponsor in terms of corporate sponsorship. There aren't any other corporate sponsors of Mocha other than SoftLabs, and the rest is individual backers. It's usually $2/month, $5/month, $10/month, $1/month, and these are just simply individuals.
+**Nadia Eghbal:** That's a lot of people donating $2 or $5/month that adds up to the budget that you have now; it says something about people that love it enough to donate, I guess, right?
+**Christopher Hiller:** Yeah, and there are quite a few of those people. What I wanted to do, and I was so excited when we got our first backer, and I sent out a tweet, I was like "Hey, thanks for backing!", and then the next five people that decided to back us and donate, I sent out tweets. Then there was like 100 people, or whatever... And I get several emails a day - "So-and-so has donated $2 to Mocha." I want to shake everybody's hand and say thank you individually, but it doesn't scale very well.
+I don't know if Sean Larkin and Webpack have figured out how to do that. He spends a lot of time on Twitter, but you know... That doesn't scale. I would love to be able to do that, but I haven't found out a good way to do that yet that works for me.
+**Mikeal Rogers:** Yeah, there's definitely a scale issue there. \[laughter\] I raised a very tiny amount of money in Git Bounty, or Bounty Source, or one of those back in the day, and I think the money is just still sitting there, because the Request project just could never figure out what to do with it; we've really had the same problem. You obviously have a lot more money, so you could figure something else out... But what's keeping you from paying somebody - or even yourself - for a few weeks to build out a new feature, or something like that? Are there some incentives that you're worried about, or messaging that you're worried about around that?
+**Christopher Hiller:** \[28:04\] Yeah, we tried that, and there were a couple things that needed to happen, to get done. There was myself and a couple other maintainers; we worked on these tasks and build for it, and we had kind of decided, "Okay, we're gonna work on these and build for these. You do this, and I do this" etc. That worked out, and then what happened was I wanted to work on something else, and said "Okay, I'm gonna build for this, too", and it was part of this rewrite. One of the other maintainers disagreed and said "We shouldn't build for that." I asked him why, and he says "Well, it's fun." \[laughs\]
+That was kind of strange, and I thought "If I'm not having fun, why am I here? And if I can't make money doing thing that are fun, what's the point?" But the point is that people are gonna have different ideas about what is worth actual money. Money tends to complicate things. Who is to say that if I do some task and build for it and get paid, what if somebody came along and said "Well, I would have done that for free"? It just complicates things; I think it creates conflict, where it doesn't seem like the way we tried to do it worked out at all.
+I'm not sure how -- I know Webpack has an idea of how to spend their funds; I haven't look too deeply into it, but I know it didn't work out well for us, and we had actually received advice from some other projects that says "You just wanna keep money out of the development, and use it for things like recurring expenses, services and stuff", because it's... What's the word I'm looking for...? Perverse incentives, or something. It can cause issues, and I think a lot of people in open source software are kind of afraid of money, and I understand, just from personal experience, that it can cause problems.
+**Mikeal Rogers:** Yeah, I spoke with Sean a little while ago and one of the things that he said was that he focuses on things that nobody else wants to do because of this particular problem, but at the same time Sean has a full-time job and he's doing like 20 hours a week of work that nobody else wants to do on Webpack. That's not sustainable, he's gonna burn out. I don't see a world in which he doesn't burn out doing that.
+**Nadia Eghbal:** Something I think has been a recurring theme this season too, thinking about if you have money, where does the money go and how do you spend it efficiently, and for both of your stories -- you know, a small amount of money can be really awkward amounts of money, because what do you do with that? A big distinction that's been coming up for me is "Do you fund a project? Do you fund Mocha, or do you fund your time?"
+The examples I've seen that have been successful are where the project itself can be sustainable in the sense of money is not coming in and out of the project - beyond, like you said, recurring expenses - but it's more about if someone already internally wants to work on the project and is able to find time, whether it's from their employer, or whether they raise money and get people to do it, or whatever... That works. So I'm wondering, have you ever asked an employer to work full-time on Mocha, or have you thought about finding ways just to sponsor it yourself?
+**Christopher Hiller:** \[32:26\] Well, I have a full-time job as well, so working on Mocha in any regular capacity is just kind of beyond me right now; I'm not able to do it, and of course, I tried, but I burned out. So if I was going to do that, it would need to be through an employer, but I have yet to have an employer where working on Mocha kind of made much business sense for them.
+It's a tough one, because Mocha is part of the test stack, so number one, your product manager or what have you is not going to care too much about what that test stack looks like, and is only going to really care if the software is quality. So basically, we're gonna leave the testing up to the developers, and that's their thing.
+It's really hard for developers to get visibility for these types of tools that they use, and justify to employers that "Yes, this tool we use is very important." There are other tools, like maybe a Webpack, that help build production websites, and that's big. That's something that they can kind of wrap their head around, but a project like Mocha, or maybe a project like Babel - these are behind the scenes things that aid development, they don't have a direct impact on your download speeds or whatever of your site, or what have you.
+So it's kind of tough to just -- I can't even pitch it to my employer that I should spend time working on this thing, because I can't justify it to myself... Like, "Yeah, this project needs help, but it really has nothing to do with the business goals." It's just all about "Yeah, we want quality, so write the tests", but does that test framework really need me to hack on it to make that quality happen? The answer is no.
+**Nadia Eghbal:** If you were starting all over again and you had to think about -- or let's just say even if you were TJ and you were trying to figure out where this project goes long-term, is there anything that could have been done differently, or is this just sort of the inevitable way that projects unfold?
+**Christopher Hiller:** I don't wanna blame TJ, because it's not really his fault... If we as the open source community had known in 2011 (or whenever he wrote this) what we know now about how to grow and sustain a project, sure he would have done things differently. Why wouldn't he?
+If I was going to start over and do it again, sure, I would do it differently. I would make things much more friendly to contributors; I would work a little harder on documentation. We would maybe have chat rooms, forums, all sorts of other things that weren't really part of the idea.
+\[36:16\] I think Mocha was another project where a developer decided to scratch their own itch and make a tool that was useful for them. I don't know if he envisioned that it would get as large as it has - large in terms of the userbase - but yeah, definitely things would have been done differently if it was starting today, like new projects do. I see lots of new projects with contribution guidelines and very clear bullet points on how to contribute, how to become a maintainer and codes of conduct etc.
+Actually, just today I merged a code of conduct for Mocha, and that was a long time coming.
+**Nadia Eghbal:** Nice!
+**Christopher Hiller:** I can't remember who it was, but somebody sent a PR, and yeah.
+**Nadia Eghbal:** That's awesome.
+**Mikeal Rogers:** Awesome.
+**Nadia Eghbal:** You also joined the JavaScript Foundation recently, right? Did you pursue any other form of project-focused funding or support, any of those types of grants?
+**Christopher Hiller:** Yeah, so Mozilla has an open source funding program called MOSS. What the program does is it awards grants to various open source projects. I thought "Well, maybe Mocha would be a good candidate for that." The reason I thought that was because Mozilla publishes a list of open source projects they most widely use internally, and the top project on that list was Mocha, so I thought "Mozilla is invested, in Mocha... Why not?" So I reached out to some people at Mozilla. The feedback was kind of mixed. I heard that somebody would be willing to sponsor me...
+So part of that grant process is that you need an internal Mozilla employee to sponsor or vouch for the open source project, and I found somebody to do that, but I had also talked with a few people and I had listened to this podcast, the Request For Commits podcast, and sometime back there was an interview with Max Ogden, I wanna say...
+**Nadia Eghbal:** Yes!
+**Christopher Hiller:** And the whole episode was about grants and funding, and how to get a grant for your open source project. So I listened to that and I was kind of like "Eh, I don't think so..." \[laughter\] If you're doing that, if that's your thing and you have time to dedicate to that - sure. It sounds like there's a lot of networking involved and hobnobbing. I believe one of those things that Max said was "Don't apply for a grant unless somebody at the Foundation approaches you and asks you to apply for a grant", and I thought "Well, nobody asked me, so maybe that's not a good idea." But furthermore, it certainly sounded like there was a lot of time and energy involved in securing grant funding that I simply didn't have.
+\[39:55\] Another issue with Mozilla in particular was if they are to award a grant, they are -- I don't know if the text says they will not or that they're unlikely to, but it's something like they are unlikely to award a grant to an individual. So if I was to get a grant from MOSS, I would need some other entity, a legal entity to accept the funds and disburse it to me. I don't have a legal entity, I am just me. Mocha does not have a -- there's Open Collective sponsors, sure, but there's no company behind Mocha. It's a pure bootstrap open source community project. It came from an individual.
+So I went to the JS Foundation and asked "What if you went ahead and wear my legal entity if I was to apply for this MOSS grant?" The reaction was mixed again, because I don't think they've had great success with paying individual or disbursing funds to individuals for code, so I wasn't sure what the story was there, but he said that "We would consider it, but you would have to apply and you'd have to write up the grant proposal."
+At that point I was like "I don't know... I don't have time to write up this grant proposal. I'm not even gonna embark on this if I'm not pretty sure I'm gonna get it", so I didn't. That's kind of the story of the MOSS grant.
+**Nadia Eghbal:** I'm glad we played some small part in that decision. \[laughter\] I mean, one way to read it is it's a not ideal outcome to not have applied for a grant, but I think... I mean, this is my experience with adventure as well of people -- it's better than the belief that some people have of "This is just free money, it's super easy to get; I'm just gonna try it", and not knowing all the hard work... It's like a full-time dedication kind of thing to be able to get that kind of money, so in a way I think it's actually a really good thing to know that about yourself and be like "You know what? I'm not gonna do that."
+**Christopher Hiller:** Well, it's not even about me, it's just like I wouldn't have known that. I wouldn't have know that "Oh, it's not just free money; I can't just apply and wait for a track", you know? I wouldn't have known the difficulty because I don't know anything about grants - I've never applied for a grant - without the show, or just talking with a couple of people, but I knew that I was simply not gonna have the time to dedicate to pursuing grant funding. Again, full-time job, so I chose not to do it.
+Adam Stacoviak: In this last segment we ask questions like can a project do all it needs to do and be done? Can a project be done? When can a maintainer walk away and say "That's as good as it gets." Specifically, Nadia asks Christopher when does he feel like he can step away from Mocha and what keeps him from doing so, if he hasn't. Stick around.
+**Break:** \[43:29\]
+**Nadia Eghbal:** We've talked to other folks before who have talked about whether a project can just kind of be done at some point, and what that would mean to be a maintainer on a project that could just sort of be done. And you've talked a little bit about this, of trying to figure out the scope of Mocha moving forward, or whether it needs new features or not... Do you feel like Mocha could ever just be done and you could walk away and say "That's as good as it is"?
+**Christopher Hiller:** Yeah, I think we can get there. It could be done now if we wanted it to be done. We could decide not to add any new features. There are certain things that we do have to keep adding, for example as the Node project adds new flags to its coming online client. Mocha kind of supports those and passes them through, so those would need to be added. But what has kind of stopped me from just -- I guess that would feel to me like I failed in some way... It would feel like giving up, because I know there are a lot of people that could use a couple new features in particular. I know that I would like some of those new features. I know that what I'd really like to see for the project is a really solid API, a really solid foundation on which people can build other tools for testing, and really great integration with IDE's and text editors, and kind of like stuff that TypeScript has done.
+Mocha simply is just not there, it's not where at least personally I am kind of happy with it. I don't really want it to be done, because it's -- I guess it's kind of personal; I feel like we could make not just myself happy by adding some things, or refactoring some things, but quite a lot of people, too. There's still plenty of work to be done on making the development experience better. For example, Mocha still has a makefile, so you have to have that whole toolchain installed; it's not just pure JavaScript. You can't just run npm, run whatever, and have your scripts run, and everything be JavaScript. If you're on Windows. You've gotta install Visual Studio, or what is that thing - [MinGW](http://www.mingw.org/) - to even contribute to Mocha, to even build.
+So there's plenty of things that need to be fixed there. I don't really want Mocha to be done, but it works, and it works really well for a lot of people, and from what I can tell, it's pretty solid. Certainly, there's some issues that people come across from time to time. There are quite a few issues that have never been resolved, but they're not things that we're getting just hammered on. If we break something major, we hear about it; that's gonna happen simply because of the number of people that are using the software. So I don't want it to be done.
+**Mikeal Rogers:** \[49:54\] I'm curious what you think the scope of Mocha is... Because I can see this kind of alternate world where the scope has been completed. You define the test format. A lot of these other competitors that we were talking about earlier actually use your test formats, and you can use the same tests between the different runners and stuff like that. So there's obviously a lot of extra value that Mocha does on top of that, but there's this alternate world where you can say "Well, that's our scope, and if you want TypeScript support, there's this library, and if you want Visual Studio integration, there's this library."
+It sounds like what you really wanna see from the project, what your scope is is a much more useful, full-on test runner... So I'm just curious what you feel the full-on scope of Mocha is.
+**Christopher Hiller:** I don't think it's much different from what it had always been declared to be, which... Mocha does two things - it provides a convention for writing tests; actually, it provides several different ones, so you can choose. But it provides this convention for writing tests, and then it provides a way to execute the tests, whether that's in the browser or in the command line, and... I guess this is three things - so then it provides a reporter that reports the test results. So it's not like a "batteries included" thing, and this is a criticism of it; some people like it, some people don't... It's just one of those things.
+It doesn't have assertions in it, so you need to pull in another library. Mocha won't do much just by itself. I don't think the scope is much more than that. I think what -- I'm talking about in the differences, it's that it does those things that it does, but it does them better, and it does them in a way that allows people who want Mocha to do more to build on top of it and make it do more. That's what I really wanna see out of it, and right now it's just not there. It doesn't have a -- the API is roughly documented; for all intents and purposes it's undocumented, and there are quite a few packages that are built on top of Mocha anyway, and consuming output that was never intended for machines, and stuff like that. That makes it difficult to make certain changes, knowing that if I changed this human-readable output just, just like -- I draw a parallel between and Node, where if you change the text of an error message, you're gonna have to put that in the next major.
+**Nadia Eghbal:** So what about Mocha being done, but then there's you and then there's the projects. At what point do you feel like you could step away from Mocha? And if you haven't done it yet, then what keeps you from doing so?
+**Christopher Hiller:** Well, late last year I started feeling a bit burnt out on the project after feeling very frustrated about what I saw as kind of a failure to get the attention that I wanted for the project, and get the contributions and maintainers that the project needed... So I decided to take a break from it for a few months.
+I did that, and it was okay. There are a couple other maintainers right now who have picked up their contributions. I had kind of raised an alarm on Twitter, I had written a message in our project's readme that we need help. About that time, when I did those things, is when I decided to just kind of take a break... So in the meantime, while I was away, we got some decent progress on the project's needs, and one of those was better test coverage... \[54:05\] But there have been quite a few people who have approached me and asked how they can help. Of those people, the number of people who followed through is less, and of those people, the number who want to become maintainers is even fewer. We have yet to get somebody on board who wants to be dedicated to it. That's tough.
+There's a few options here. One, if I feel that I've done what I can and there's no way this project is gonna get to where I want it to go, I'm just gonna quit. That's just life. It's kind of out of my hands. Obviously, I can't do it by myself; I need help from other people to get it sustainable, and I need help from other people to take the project where I think it should go. And if I feel that's not gonna happen, then maybe I can walk away if I feel like there are people on the project who are dedicated enough to pick it up and run with it. And maybe they have their own ideas, and those ideas are good - then great, let them do it. Again, there are other projects that I'm happy to work on.
+So those are kind of the two ways in which I might feel like my time with Mocha is done. If I can kind of limit my involvement and make sure I don't burn out and not let myself get too frustrated, then I can keep with the project indefinitely. We can make baby steps to where it needs to be, and that's kind of where it's at right now.
+I started contributing again just a couple weeks ago. It felt good, and I'm just gonna keep at it. I was really happy to see that some people answered the call; this project needs a lot of help. And yeah, we'll just play it by ear.
+So that's how I might end up leaving the project. I don't wanna quit Mocha if I don't have to. Or if I stop using Mocha for whatever reason, I might. But I'd like to stay with it.
+**Mikeal Rogers:** I think that's a really good note to take us out on. Thanks for coming on, this has been a really great conversation, actually.
+**Nadia Eghbal:** Yeah, thanks Chris.
+**Mikeal Rogers:** We really appreciate it. We got some incredible insights here, so thank you very much.
+**Christopher Hiller:** Thank you, I had a lot of fun!

Open Source History, Foundations, Sustainability_transcript.txt ADDED Viewed

	@@ -0,0 +1,331 @@

+**Mikeal Rogers:** We talk about open source sustainability on this show, it's a really hot topic now... But you've been working on open source sustainability I think maybe before it was called open source, so why don't you give us some of that kind of history and back-story through the lens of sustainability?
+**Danese Cooper:** Sure. So when I got into open source, it was just kind of the beginning of companies caring about open source, and there was an early vanguard of companies that either they already kind of had open source as part of their history, although maybe it was called 'free software' when that started, or they had some vision about why it was gonna be important to them... So the big players were Sun and IBM, who has been in it since very early. And there were some companies that were really against us, and we had the classical light and dark, if you will... So Microsoft was pitted directly against us, and there were a couple of other big companies that it really was gonna threaten the way that they did business, so they were pretty clearly against us.
+And then there were up and coming companies like Red Hat, that were just starting out, and were tiny, but were so deeply based on open source that they cared a great deal, and they were building alliances all the time... So that's sort of the point at which I came in.
+It's important to say that there was a good 15-18 years of people developing software in this way, that would be called the modern era... Because we all know that in the early, early, early days of computing everybody shared everything, because it was just a hobby, and they had to share everything because it was the only way you could reasonably get anything done.
+\[03:50\] The point at which that all changed has been classically tied to the Homebrew Computer Club and Bill Gates' realization that he wanted to charge for essentially access to the sources, although that's not how he framed it. But he wanted to start selling non-source-accessible executables, and he made a very successful business out of that. The modern era of open source wasn't exactly a reaction to that, although it's been characterized that way. I think some of it was just people who it was more convenient for them to work in the old-fashioned way, because their problem space was new enough that they needed that kind of sharing, and I think the BSD community falls pretty evenly into that... And maybe also `sendmail` and `bind` and all that stuff but there were other things that were definitely starting to head in the direction of money, and money changes everything... So that's about when I came in the door.
+In those days, the corporations wanted in on the marketing lift of the coining of the term 'open source'. Tim had that famous meeting and they came up with that cool term, and then he used his media arm to make that term interesting to people, and almost immediately there were people that just wanted to take advantage of the lift the term was getting.
+Some existing projects that had been around a long time and were pretty successful all of a sudden had to deal with that, the hype factor. People were code dumping, for instance, at Apache. Apache had to change its license and also some of its practices to make sure that people got it, that they didn't have access to the Apache brand just because they dumped some code at Apache.
+**Nadia Eghbal:** Can we talk a little bit about just the media lift part of it of how he made sure that people were using the term open source? Because I think people don't necessarily know about that.
+**Danese Cooper:** Well, I don't think he enforced anything. It was sort of the anti-free software movement, if you will, and I don't mean that in the sense that they were pitting themselves against free software; I mean they were using different tactics. Tim and others like him saw amazing software coming out of the free software movement that was being hampered by the political attachment that the authors of the GPL had. Basically, in GPL v.2 the first page and a half of the license is a political manifesto, before you ever get into any legal terms. That made a lot of people uncomfortable in the straight corporate world, and also governments and places like that.
+Tim saw an opportunity to reframe the whole thing under a new term that, first of all, didn't have the inherent ambiguity of 'free', because in English 'free' means both liberty and gratis, and that was a problem, because people got them confused. Free as in gratis doesn't necessarily lend value to the endeavor, free as in liberty does. Free as in liberty is very definitely the message that the Free Software Foundation was pushing, free as in gratis was what was driving people's adoption, so there was an ambiguity... Not all people, but certainly corporations were picking things up for that reason. So open source was a little more honest; it wasn't ambiguous, I guess, would be the answer.
+So he came up with the term, and then there was a Perl conference that was pretty successful, that he'd been running for a while, and he started it around the time that there was a lot of contention in the Perl community because of the work of ActiveState to creative a Windows-native version of Perl, which felt like crossing the streams, if you will, of loyalty, since Microsoft was a pretty vocal critic of open source (or free software) generally.
+\[08:06\] He changed that to be the first open source conference, OSCON, and invited everybody, and used the fact that the publishing industry gave him disposable income to push memes to get more people thinking about open source; he wrote extensively about it. His blog was well read -- actually, pre-blog, his writings were well-read, and he had a venue to write. Before blogs, all of that was rare, right?
+And almost exactly at the same time we started to see the rise of peer-to-peer computing. It was only maybe a year later that the first peer-to-peer conference happened, which was also one of his... That was both attempting to address some social ills, which is interesting, given the separation or apparent separation between open source and free software. I actually think they're the same thing, just trying to express themselves in a different way, but free software led with social responsibility, but I think most open source people had that bent as well at that point... So in the peer-to-peer movement there was a whole lot of helping political activists, and using the lightweight, importable nature of peer-to-peer networks to do interesting things, starting with sharing music more widely.
+Famously, there were early music sharing platforms that the music industry really disliked, and one of the really successful ones was Gnutella. Gnutella was written by a bunch of people who saw the central architecture of the first experiments in music sharing and decided that a massively decentralized architecture was gonna be harder to quash, because there would not be a single point of failure, or a single throat to choke, so they created this endlessly replicating "everybody is a server" or "everybody is a client" architecture for music sharing.
+Almost immediately, Larry Lessig showed up and started talking about copyright problems, and the confluence of the change in the music industry and the attempted change in copyright law and open source - they all converged at the same moment. So for a really long time there, every open source conference had a huge component about music sharing and media sharing generally, and that gave Tim a much bigger springboard for his message about open source. So some of it was his cleverness, and some of it was just a lucky confluence of time and place.
+**Mikeal Rogers:** Interesting. So this is happening around the same time that Linux was also on the rise, so before we kind of leave this era, one question that I have is that -- one of the things that I have to talk about a lot right now is how there are a million reasons why people contribute to open source, and legal obligation is usually not one of them... The kind of legal obligation you have with copyleft. But often, people point to "Well, in this previous era, the only way to get people to contribute at the same time was this legal obligation." How important do you think that was, being that there were these other forces that they got people into open source that weren't tied specifically to the legal obligations?
+**Danese Cooper:** \[11:48\] You know, I don't remember that that's why Linus chose the license. I think Linus chose the GNU license because it was the only one he knew. Richard Stallman did a lot of running around, talking to universities about why freedom was important, and his style is very sympathetic to people of that age and that ilk and a lot of projects from that era are licensed under a GNU license because they had heard Richard speak and they knew it better than any of the other options; they didn't really understand what the other options were.
+But the BSD project, which was the permissive licensing that Bill Joy invented, happened almost -- well, now I'm in trouble, because MIT was working at almost the same thing at almost the same time... But permissive licensing dates from almost the same era as the free software licensing, and the projects pre-date the projects that Richard Stallman launched under that new license... So I think there's always been both sides. I have to say that -- I mean, I have a bias, because I've been an Apache member for a long time... I think that in my experiments for Sun with pretty much every licensing model you could possibly imagine, including all the possible hybrids, permissive licensing definitely drives adoption better, if you happen to be a deep pocket, trying to get non-members of your community to join in, right? Which is a bunch of ifs. But there is that situation.
+In the early days of corporate involvement in open source, the very first corporate project was Mozilla, and Netscape at the time was considered a big company; not as big as IBM, but they had a pocket and they were worried when they constructed their project. They did it as a Hail Mary, because Microsoft had just destroyed their market by distributing Internet Explorer with every copy of Windows for free, and Windows was 98% of the install base at that point... So this is why that whole lawsuit happened about anti-competitive practices that Microsoft was doing. And they'd been doing various other things, it wasn't just open source were picking on. They were anti-competitive all over town, but this was a really obvious example.
+So Mozilla, among other things, did a Hail Mary and put their client software that was competitive with IE as open source, thinking that that would drive adoption and keep them alive. And it worked, right? But they were so worried that people would not feel comfortable as individual contributors, because at that time most open source people were hobbyists and individual contributors. They basically waded everything in their license to reassure the individual contributor that they weren't going to harm them, while still being fairly legalistic about things like copyright jurisdiction, patents and a bunch of other stuff. They were really bending over backwards. As it turned out, they needn't have worried because their initial code push was kind of a mess, it wasn't really set up for anybody to use it.
+The next thing that happened was they got a new owner, and the new owner didn't like the way that open source worked - this was AOL - so they were gonna pull the plug, because they couldn't get their own stuff in fast enough, and because the community vetting process was frustrating to them.
+We were in the middle of that lawsuit, and Sun decided that it couldn't afford to have that browser go away, and so another Hail Mary thing happened where IBM and Sun's lawyers went and talked to AOL and convinced them to release the IP into a foundation, and then to give a little bit of funding; Sun and IBM each also gave a little funding, and that's how the team was provisioned to write Firefox.
+**Mikeal Rogers:** Interesting.
+**Nadia Eghbal:** You were talking about copyleft versus permissive licensing, that it started around the same time... How deliberate - when you're saying things like "permissive licensing was useful for driving adoption, and still is today", how deliberate were things like the idea that copyleft could prevent anybody from taking your stuff and running with it?
+**Danese Cooper:** \[16:13\] You know that those licenses, even today, all but the AGPL, are still trigger on distribution of the code. People don't really distribute code so much anymore, because now we have software as a service, so as the pendulum swings back and forth between client server and peer-to-peer - we're kind of in a peer-to-peer place now... And there isn't distribution happening, there's no separation, so most of the requirements that you give back are gone now, except in rare cases, and they're in places where the software is going to meatspace like the automotive industry. There's still lots of GPL violations going on in places like that, because the code is distributed physically in a product; it's not a server-oriented thing.
+Back in the day, the free software people really genuinely felt that they had to compel anyone who touched the code to give back changes. When we were explaining licensing to corporations, we used to say "For a free software person, the worst thing that could happen is a piece of code gets used in a non-free way, and for the permissive people, the worst thing that can happen is a piece of code doesn't get used." \[laughter\] It was really that simple.
+The Apache people were like "Look, we would much rather have gifts of code that people wanna give us - they're gonna be higher quality than gifts of code that people have to give us, and they're gonna do the least they must in order to fulfill that obligation", right?
+**Nadia Eghbal:** Yeah.
+**Danese Cooper:** But I think that the free software people were absolutely coming from a place of trying to push the right behaviors, they were just using the stick instead of the carrot.
+**Nadia Eghbal:** Yeah. I hear it uses this example now of "Well, if only everybody had been GPL, then you wouldn't have this problem of people using and not contributing back", because that was the point, but nobody wanted to use that license.
+**Danese Cooper:** But I don't think that's true...
+**Nadia Eghbal:** I agree, yes.
+**Danese Cooper:** ...because now we have software as a service and they don't have to anymore. On that model, we would have almost no contributions now, because nobody would be required to do it.
+**Nadia Eghbal:** ...or adoption.
+**Mikeal Rogers:** Right, and if adoption of the software is one of the ways that you attract contributors, then anything that harms the adoption of the software is going to lead to a lack of contributors, especially if this other legal mechanism isn't even there.
+**Danese Cooper:** Right. I mean, when WebSphere, which was just Apache's web server wrapped by IBM and turned into a product which is legal under permissive licensing - when that happened, the whole free software movement was getting out their popcorn, because they thought that it was gonna fail, or it was gonna show up the permissive model as insufficient. But what actually happened was IBM for a long time had their own version and maintained their own version and did all the extra work things that proprietary companies think they have to do, and then at some point it got to be too painful to have to keep backporting their own version, as the open source version was being improved by people that weren't them. So they finally got it that it made sense to send their own stuff upstream so that they could stop with the backporting and also with the dealing with conflict when the open source community solved it in a different way that they solved it. They got it that the point was the solution at the end of the day, and not necessarily who wrote it, and both of those things were pretty well proven out.
+**Mikeal Rogers:** It seems like there is a concern on the free software side of people using the software and not contributing back, and there's definitely the view now, too -- I think you've described it before as like people worried about freeloaders, or whatever... \[laughs\] How has that changed over time, because the language that people use around it now is very different than I've heard in the past.
+**Danese Cooper:** \[20:12\] Right. Well, I think that a lot of the enforcement activities that the Free Software Foundation engages in are around people who use the software and extend it, and don't give back their interesting extensions. That's because they have curiosity about how people are using it, and they want everybody to be able to take advantage of those changes.
+I don't know if you guys are old enough to remember the Nordstrom chocolate chip cookie, but...
+**Nadia Eghbal:** Yes!
+**Mikeal Rogers:** Yeah...
+**Danese Cooper:** There was a recipe that running around on the web for a while, and there was a story that went with it, which I think got debunked... But the story basically said "Somebody at Nordstrom started to charge me $14 for a cookie - and it was a really good cookie - and I got the recipe and I'm gonna share it with all of you because $14 is too much for a cookie" was basically how the story went, right? That's an interesting story because Richard Stallman used to use a cooking metaphor when he was explaining free software. He would say, "You know, that time you added pecans to the cookie recipe and they came out a lot better, and all your friends really liked them, and you were able to give them that recipe because there isn't a copyright protection on the cookie recipe - that's what we're trying to do with software, right?"
+So what they were trying to make sure was that if anybody came up with a clever hack like pecans, everybody got to have the pecans and not just the -- people couldn't make money off the pecans, and adding them to a recipe that everybody else contributed to... So it was a sharing thing, and kind of a fairness thing, and it probably still is, in their world.
+But the freeloader problem is where people aren't contributing both code and support, right? There's different ways to get involved in open source. One of them is if you've done something clever with the code, you give that back, but that's not even always all that desirable, because what you did with it might be esoteric and not all that applicable to the general public. What you did with it might be your secret sauce; Google has never given back the bulk of their special changes that they've made to Linux to make it work for their scale, because it's tied to the way they do their business. They think that it would be giving up a competitive advantage. They're not required to, because their services are offered as a service, so they've gotten around the distribution problem... But they came up with the Summer of Code, and they open sourced a lot of other good stuff, and they try to mitigate the fact that they're unlikely to give up those changes to Linux; they do give some of them up, for the same upstream reason, because it's a pain in the ass to backport, but they don't give up the ones that comprise their secret sauce.
+So are they a freeloader? They're not giving up those changes. Maybe those changes would be super useful to everybody else and maybe if they would give them up, the commoditization of search engine technology would mean that we didn't have to deal with that AdSense quite so much. On the other hand, they're trying to balance the scales through some of their other activities, so are they a freeloader or aren't they?
+I hear from younger open source people a concern about the flow of money, and I think that's a really interesting question. Some of the modern foundations are really designed to attract a lot of money, and that attraction of money is then gonna be spent on things that are last mile projects, or things that the group of people that were doing it before the money came in weren't finding time or resources to do.
+\[24:11\] That problem of "well, the group that have attracted this can't get through all the work that we'd like to do" problem isn't a new problem, right? But some of older foundations are designed to run on almost no money, and I think that that was a conscious decision. The Apache Software Foundation runs on almost no money, and the reason that they do that is conscious, because they first of all didn't wanna be a deep pocket that was gonna attract patent trolls and things like that.
+Originally, when Brian Behlendorf first talked to me about the desire to create a foundation - it was obviously before the foundation happened - as he's explaining it to me, one of the design concerns was a concern about patent trolling, and they had looked at Mitchell Baker's work on the patent piece license part of the Mozilla license and they wanted to add that to the Apache license, and they wanted to do it in a way that made it a nice idea for Apache to become a patent pool, because they thought that they needed to have a patent pool in order to protect the different entities that were coming into Apache, and as it turned out, that didn't really happen... It was sort of an idea that never got off the ground. Even more recently, the Open Innovation Network is trying to do that now, and it's been kind of a hard sell, and that has to do with the way patent law works.
+But the idea that patents could make it in as part of a give was an interesting idea for a while. IBM opened 500 patents at one point. Scott McNealy famously told everybody that the patent situation with the Solaris was -- basically, he was indemnifying all the members of the community against any kind of patent action because he owned all the patents on Solaris, and they were gonna be unenforceable after he open sourced the code.
+**Nadia Eghbal:** I think this is a really good lead-in to foundations, and...
+**Danese Cooper:** Well, especially that Mozilla story is probably the best lead-in I gave you to that topic.
+**Nadia Eghbal:** That's a pretty good story.
+**Danese Cooper:** Yeah. There's a piece of that, by the way, Nadia, that I always wanted to tell you.
+**Nadia Eghbal:** Oh, great.
+**Danese Cooper:** So the piece I left out of that story has to do with Mitchell Baker... Have you ever met her?
+**Nadia Eghbal:** No.
+**Danese Cooper:** You know who she is though, right?
+**Nadia Eghbal:** Yes.
+**Danese Cooper:** She runs Mozilla. She has run Mozilla since the beginning of time... So she started as a lawyer and she wrote the license, and then she got intrigued by the project. She had actually worked at Sun before she went to work for Netscape, so they knew her as a lawyer. And she got intrigued... She got a job as the general manager of that division, and her title was made Lizard Wrangler almost immediately. So she was doing her job, and then one day AOL buys the company, and they wanna get their changes into the browser, so they're trying to push her to get them in faster, and she's sticking with the process that the community has agreed to, and vetting everything, and they decide that she's the problem, so they fire her.
+She had a pretty good severance package, she had a golden parachute of some kind, so she wasn't worried about money. She went home, she dusted herself off and she logged on and she kept running the project, and all of the engineers kept deferring to her, more importantly, right...? \[laughter\] And to be fair, they didn't all work for AOL by then; a lot of them worked in other people's companies, and they were just contributing. But everybody in that community was still deferring to her, so AOL said "Oh, I see. Yeah, okay, we get it. We didn't understand, but we still don't like this, so we're gonna pull the plug."
+So she called me up one day and she was just panicked. She was like, "They're gonna pull the plug", and that's when we got it together to strong-arm them to create a foundation, basically. But think of that chutzpah, man, right? That's open source.
+\[28:15\] When I give the talk now to people about standing in their power as open source developers -- because honestly, the influx of new people that have come into open source, the most troubling this for me hearing them talk is a lack of understanding about where they can push back, and that they're gonna lose the whole game if they don't continue to push back on the things that will kill open source, so that's why I do that talk now... My brief sustainability talk that's got the seven points that I think you really have to focus on when you're thinking about standing up and saying "You know, this isn't right", which people have to be willing to do, just like Mitchell did.
+**Break:** \[29:00\]
+**Nadia Eghbal:** Having that history is so important because what I've seen is a lot of people feeling isolated and not having that sort of long historical lens of seeing examples of where the other people have been able to push back and really understanding that, and I think almost some of that comes from open source being so default right now that people don't think of it as "Oh, I have this unusual power." They think "This is just sort of what everyone is doing, and I'm just gonna sort of accept things the way they are and be miserable", which is not great... But it's stories like that that make you think "Oh wait, there's something really powerful about the way we work now, and we have a way to leverage that."
+**Danese Cooper:** Yeah, if you were an engineer before we did all of this work to change things -- I mean, it would be a lot like being an engineer now in a lot of places like Accenture and those places. Think of yourself as a cog in a wheel for a minute; the problem is that engineers are artists. A good engineer is attached to the creation of the thing that they're building just exactly like an artist is. Paul Graham wrote about this a long time ago, but we all kind of knew it before that. And not all engineers are that, and not all need to be.
+I mean, one of the reasons that Microsoft was so opposed to open source I think was their developer base -- when Steve Balmer ran around, screaming about developers with the sweaty armpits and all that, he wasn't talking about people that were gonna get to invent Microsoft products, he was talking about consumers of Microsoft products, and Visual Studio and Visual Basic were basically designed to make the job of programming easier, so that non-artists could still do it. It was kind of like paint by the numbers, in a way. Visual Basic is very much just stringing together other people's work, like beads, right?
+\[32:26\] So there's two kinds of programmers - there's people who invent stuff, and there's people who work with Microsoft's tools, and that's why it was so dangerous to them. And that was a gross generalization; of course, there are people who invent stuff with Microsoft's tools, but there's a really large percentage of the kind of programmers that just show up every day, and they're not bad people... They're working, but they're not motivated or trained up enough to really be inventors, and open source was founded by inventors.
+**Nadia Eghbal:** I think that's a big distinction that's growing now, especially as programming is becoming easier for a lot of people... And that's sort of the point, right? You have these tools that you don't have to go deep into the weeds and figure out how to build things.
+**Danese Cooper:** Well, that's what abstraction does, right? We're working on five and six GL languages now, and those abstractions save you from having to pay attention to pointer math, or any of the things that you used to have to really have got, like, powers to learn how to use the language well. But along with that comes the free style of working that we have; there's a lot of other things that come from us being the one irreducible quotient, right...? We are the thing you have to have in order to make software.
+**Nadia Eghbal:** That's how I felt coming into it as an outsider... Like, "Wait a minute, all these people are forming the foundation of software for everyone else, but they seem to not have the leverage that they should", and I think that that's the part where even for an open source developer today who's pushing things forward and innovating, I think even they still don't recognize their own influence and power sometimes.
+**Danese Cooper:** Yeah, which is why I do that talk... \[laughs\]
+**Nadia Eghbal:** Yeah, which is really important.
+**Danese Cooper:** ...which we're gonna call the superpower talk. It has a superhero in it, and everything; he happens to be a little boy, but... Yeah, I think it's really important for people to realize that you've gotta be uppity, because you know, it's kind of like democracy, right? If you wanna talk about the real freeloader problem, the real freeloader problem is people not showing up at all, and just expecting somebody else to do whatever it is.
+When we started out, people were just so excited to have an opportunity to work together and get stuff done so much faster and so much more efficiently than it was happening in their day job. People used to say things like, "Look, if my employer ever tells me I can't do this, I will quit, because this is the only thing that keeps me going." And they didn't mean their day job, they meant the evenings and weekends they were spending on their open source projects.
+**Mikeal Rogers:** Wow... That's such a difference from now, where people just flow in and out of contributing to open source in their jobs, a lot of the time... Whereas back then it was like, "Oh no, this is what I do to stay sane on my \[laughs\] nights \[unintelligible 00:35:29.24\]"
+**Danese Cooper:** Yeah, very much so. It very much was that way. And part of why I've been pushing inner source so hard is because I think we'll see a time - as with democracy - where it's so much an accepted fact that we start to lose ground on what we want for ourselves... Because we want a lot of autonomy and a lot of choice, and a lot of... I mean, we're seeing it now - IBM is making everybody come back to work in the office, right? One of the things we kind of fought for was the right to work wherever the hell we happened to be, and also WHENever the hell we happened to be, right? Because as you know, Mikeal, programmers work better at night, they do. They just do. And not all, but a really large percentage. So dragging everybody in the office from 9 to 5 and asking them to code is a ridiculous request...
+**Mikeal Rogers:** \[36:21\] Well, also a lot of projects are global; it's always night somewhere, so...
+**Danese Cooper:** Exactly. Before the internet, we didn't have good enough communications to do global projects. That was baked into open source, that assumption that somebody was always awake.
+**Mikeal Rogers:** Yeah. So moving on through this a little bit...
+**Danese Cooper:** Yeah, you're trying to get to foundations... I apologize.
+**Mikeal Rogers:** Well, it's okay. This is all great, so it doesn't matter. You've been involved in more foundations than I can count... I mean, just for the audience to be aware - you mentioned Mozilla and Apache, the Node.js Foundation you've been a part of since we started...
+**Danese Cooper:** Yeah, the Open Hardware Association - I helped them get off the ground, the Drupal Association... I've had conversations with Ushahidi, I've spent a fair amount of time -- I've done a lot of "Let me just help you with this one little problem" with lots and lots of different foundations.
+Before there was the Linux Foundation, there was something called The Open Source Developer Labs (I think that's what it was called), and I helped them out, although they were really set up so that Linus would always have an employer, because Linus took a job with a chip manufacturer at one point, and announced that he was gonna put Linux on the chip, and Intel lost their mind... So when that folded, they never ever, ever wanted to go through that nightmare again, so they and IBM funded OSDL so that Linus would always have a job...
+**Mikeal Rogers:** Yeah, that's a word of warning for future BDFL projects. When you have a BDFL and a bunch of companies depend on your project, you need to be terrified that one of your competitors will eventually hire that person. It's one of the old things that LF (Linux Foundation) still does - they make sure that Linus is at an independent entity and not at one of the competitors.
+**Danese Cooper:** Well, and to circle back on companies, I always say that there are no open source companies, there are only companies that are good to their open source employees, and if you look at Guido van Rossum, for instance, who invented Python, or Rasmus Lerdorf, who invented PHP - both of them had really, really good relationships with their employers for years and years, where they were allowed to still run their project for the best outcome for the project, and they got air cover from the company that employed them, and the company employed them because they used that project, but no one could say "Wow, Google has really had an undue influence on Python..."
+**Mikeal Rogers:** Yeah... So getting back to what I actually wanted to do... So a lot of these foundations have very different contexts that they came out of. When does a project need to start a foundation or they need to go into a foundation? What are the constraints that it's under where it needs that kind of institutional support, versus the majority, the 99% of other projects that are happening in open source that just aren't at the point where they've had to get that?
+**Danese Cooper:** They're just on GitHub somewhere, yeah. So in the early days, the foundations were another way to convince people that you were serious about open sourcing, and you weren't gonna try to control it over much. There were a lot of attempts in the early days to leverage open source by big companies that failed, like Apple, when they came out with OS 10, that was based on the Mach kernel, which is a variant of BSD, and they made a commitment to the Mach community that they were going to continue to run an open source project that took advantage of all of the R&D that they were doing to make the Apple OS work better and better... And they did hire Jordan Hubbard and set up a really nice little playground for them outside the firewall of Apple, but they never made the step of actually doing their development, which would have been unimaginable for them, over that wall.
+\[40:25\] They kept throwing tarballs over the wall everytime they did something clever, but unfortunately that completely ticked off the people that were working on the public project and trying to solve some of the same problems... Because it'd be like, you know, I'm building a snowman - I've done the body, I've done the chest, I just did the head and I'm about to put the facial features on it... Oh, crap! Here comes in a bigger snowman from over the wall, that's gonna kill my snowman, because it's another way to solve that same problem... And Apple's tarballs are always gonna win, right? So there is a way to not do it, right?
+If Apple had created a foundation and made a commitment to working within that foundation to build their product in addition to everything else, that would have been unimaginable for them, but there were a lot of projects that went "Okay, let's make it a foundation, that way anybody that wants to can come in and work hard and gain reputation in this thing and become a leader", and several of the umbrella foundations, like Apache, were set up explicitly to be transparent scrum (scrummage) grounds for new implementations of emerging standards.
+After the web server was a done thing and Apache was wondering about how to be relevant, Brian was really big on "Well, let's do Java. Java is the newest thing since sliced bread; they're gonna need to do some open source, let's make it the home for Java." Sun ended up not wanting to use that home, but at the same time XML was happening, and IBM and Sun were kind of fighting over who was gonna lead the XML implementation. There was SOAP and there was something called WSDL. You don't know about WSDL anymore, because nobody uses it; everybody used SOAP. Not that SOAP was great, but at the time it was better than what there was before.
+But that argument, that battle happened transparently at Apache. Anybody that cared about that cared about that could watch it happen, which is different than standards bodies, where everything happens behind closed doors by design, because none of the companies want anybody to see the political machinations they're going through to get what they want. So it was a democratization, if you will, of the de facto standards process. That could only happen at a foundation. You had to create a neutral ground where the dinosaurs could come for water, as one of my slides generally shows. But these days it's a lot different. The reasons that you might come under a foundation are different, as well.
+Well, let's see... We might as well talk about Node, since both Mikeal and I work on Node. Node ended up at a foundation because we were trying to heal a fork. The community was rebelling against the trademark holder, because it was set up as a BDFL organization, and the thing that people don't understand about BDFLs is they're called "For Life" for a reason - you can't pass that baton; there's a lot of attempts to try to pass those batons, but functionally, people are at least partially organized around the personality of that individual, or at least to his creation, and they can't make the transference... So you don't get to just nominate another BDFL, and then another one when that one doesn't work out, and that had kind of happened at Node.
+\[44:13\] There were a lot of disagreements, there was a lot of desire on the part of the community to move forward, and the trademark holder couldn't keep up with the pace of interest in change, because they hadn't resourced it to deal with the growth that was happening... Because it wasn't part of their direct line of business, it was just something they happened to own. So getting into a foundation felt like the only way that the fork could be healed and the trademark holder could feel comfortable that anarchy wasn't going to reign fundamentally, because they were wedded to a certain amount of control, so giving up the control - they felt like they needed some structure, and that's how it ended up there.
+**Nadia Eghbal:** It's interesting, I feel like I'm hearing a lot about... Foundations are - and especially historically - meant to be a governance mechanism, right? And I think there's sort of a misinterpretation of foundations sometimes as the way to raise money in itself or pay a maintainer to work on a project, but it sounds like historically open source foundations as you had said earlier, are often designed to run without money, or are really meant to bring transparency or the governance aspect to a project.
+**Danese Cooper:** Yeah, the deal about the money... I have to say, I've spoken out against pay-to-play boards, which upsets Jim Zemlin. I think everytime I say it, Jim Zemlin thinks that another kitten has died in heaven, right? Because he is genuinely trying to do the right thing; he's not a bad guy, and his conceptions of what's wrong in open source are just as valid as anybody else's.
+One of the things he really saw was the inability to attract top talent because they just couldn't get paid at the same scale to work on foundation work, because the foundations didn't have that kind of budget. Now, Apache really kind of only spends money on sysadmins, and I think they do pay them pretty close to scale.
+Mozilla had the same problem, actually, that Jim identified. They were losing their top talent to, say, fellow travelers - not so much competitors, just fellow travelers - like Google, because Google was the new hotness and they were sucking up all the talent in town... But Mozilla really felt like they needed to make a change in order to be able to support industry competitive salaries, and it wasn't because they didn't have enough money - because of their deal with Google they had lots of money, by most standards, and even today they have a pretty solid endowment. But nevertheless, they went through all of the work to create a tax-baring entity, Mozilla.com, that was completely owner by Mozilla.org.
+This was kind of new thinking on their part, because it was challenging the way that the tax authorities like to see things done. But they were trying to do the right thing; they were saying "No, we'll pay taxes on our profits, and we'll generate profits and that will allow us to hire people and pay them what will keep them involved with us." And it was challenged eventually by the service, and they won in that examination, so I guess that's a viable model now, although there's a whole lot of moving parts if you wanted to replicate it.
+But Jim's idea was "Let me streamline fundraising" -- and he inherited OSDL, which the buy-in cost to be part of OSDL was pretty high, and it wasn't clear what you were getting for that money, so a lot of his funders were really balking at the high price. So he reimagined all of that, and part of it was "I'm gonna hire the very best people that I can to share services, and I'm going to charge discounted rates to projects that come in and need those services, but the people are gonna get paid industry standard rates because I'm gonna raise money on each of these little foundations that I spin up."
+\[48:34\] He was aided in this quest by the fact that the whole idea of an open source foundation became controversial to the government... Because when they started out, we looked like raggle-taggle bands of hippies, right? And 10-15 years on, all of a sudden there's huge value being created in these foundations, and it's not clear who's getting taxed on that, and they're wondering if there should be some taxation on that. They're trying to figure out if they made a mistake allowing open source foundations to happen, so they stopped making new ones. And there weren't any fewer projects that wanted to get there, so all of a sudden it became very difficult... Where before it was like a two-week process (maybe four at the most) to become a non-profit in open source, all of a sudden it was taking years... And Jim was able to streamline that for people, because he and his lawyers figured out how to spin up new foundations under their umbrella.
+Apache went the other way - there's only one Apache Foundation, but you can be a project under that foundation. And there are a couple of others we should probably mention. There's one that came out of the free software movement called The Software Conservancy, and there are a lot of projects that are there, and... God, Mikeal, I've just lost the fourth one. Do you remember?
+**Mikeal Rogers:** There's a bunch of other ones... There's the FSF, there's the Software Freedom Law Center, there's the Eclipse maybe...
+**Danese Cooper:** Yeah, exactly... I clearly forgot about them. Well, yes, but no... Eclipse was built for a different reason. It was originally organized around a project that IBM created, a development environment for Java, and it was an anti-competitive move, or a competitive move against Sun, because Sun had an open source project that was an IDE, and it was important to own the IDE market, I guess... So IBM came up with this other idea, and that was what Eclipse Foundation was about.
+Now, you're right that Mike Milinkovich has moved it into a space where there are a lot more than just that there, and not all of it is Java anymore either, so yeah, I guess maybe they are... But it's almost like neighborhoods - a certain kind of person is a tenant at the Eclipse Foundation, versus the Software Conservancy.
+Software Freedom Law Center is not actually a home for software, it's a home for legal advice...
+**Mikeal Rogers:** Right, right.
+**Danese Cooper:** ...but the FSF has been, and maybe was the first foundation that accepted other people's projects. Their way or running things is very particular. You're basically trusting the Free Software Foundation to run your project forever forward. In fact, they made an assumption early on that if you put your code there, you're gonna accept whatever changes to their licensing scheme that they created, and that's a big ask, right? You're really giving up control completely. You're giving up your copyrights to the Free Software Foundation. Most of the corporates-started foundations had to split hairs there, and came up with some exotic ways to say "We want to act like the copyright holder, but you wrote this stuff, so you're also gonna have some copyrights."
+**Break:** \[52:20\]
+**Mikeal Rogers:** So I'm trying to pull this back a little bit from the history and get it more into what these foundations actually offer to projects, so why you would wanna spin one up or join one. I think there are some huge disparities in just what these foundations do, for instance. You mentioned that Apache runs on basically no money, which is great if your project doesn't need money, but one of the reasons why people like these member organizations is because what they do with the money is they don't hire developers - you still want individual contributors to help the project, but they offer a bunch of other support, like marketing and PR that you get when you're trying to go up against other proprietary entities. That's just something that Apache doesn't offer, and if some of the projects inside of Apache have interest behind them that are sort of competing over those spaces, right?
+So with the member organization, you definitely need to have a wall between the project and the board, because the board is the corporate interest, right? Like the pay-to-play board, and you can't have them just sticking their thumb in the project...
+**Danese Cooper:** That's how Jim has it set up. There are other possible setups, right?
+**Mikeal Rogers:** Yeah, well there are infinite possible setups.
+**Danese Cooper:** He's using board membership to drive fundraising.
+**Mikeal Rogers:** Right, but also there is not a connection between board membership and, for lack of a better term, project ownership. So you have this wall between the project governance and the board governance, and the board governance is mainly dealing with the institutional aspect... Whereas I think in Apache because developers come up through the process, because they're not doing all this other stuff, the board is staffed essentially by developers, and they do make certain executive decisions about the projects, right?
+**Danese Cooper:** Well, only in extreme cases, actually. 99% of the time as appeals come to them to deal with issues, they turn it back to the project committee.
+**Mikeal Rogers:** Yes, but the projects also operate under a process that is pretty strict, and the interpretation and how that process is written is owned by the board, right?
+**Danese Cooper:** Yes, that's true, but getting them to take the kind of action that you're talking about... Like, troublesome member of my project - this person on my project is really creating a lot of problems for the whole project, and everybody's mad at them and we can't figure out how to get rid of them; the board answer would be "That's your problem."
+They only get involved in really the most extreme cases, or if that person is actually on the board, they would deal with it because they are the project, right? But yeah, it can be vexing actually, the extent to which they wanna turn it back to the project, and a lot of the -- it's been very successful for projects that are focused on technology and have few political vectors, but the minute that politics enters the picture, it becomes kind of ugly.
+They have a couple of times stepped in and restructured a project that they thought was otherwise healthy (had a lot of contributors), but there was some kind of poisonous aspect to the project brewing. They have put in almost like a special master, who is nominated by the board to get the project back on track... Less than 1% of their projects have had that problem, so it's definitely a last resort thing; they really don't wanna get in and do that, partly because they're all volunteers and they all have day jobs, too.
+**Mikeal Rogers:** Right, right. I wanna get into the future of open source a little bit, and I think that I know how to bring us there. In the io.js days, in the fork, when we were considering where we could go, like "Could we put this into a different foundation? Could we go into Apache?" were we going to need a new, unique foundation for the healed fork, or just for io.js before the offer was kind of on the table to come back together? When we started evaluating all these places, one of the troubles that we had was that -- we were a very modern project, we were really built in this kind of GitHub era. We had all these new people coming in, and we had a very new governance process.
+It was less than six months old, and we had been iterating on it a lot, and we were wildly successful, but we were not confident enough that what we had written down at that time was not going to need to evolve and change, and it definitely has had to evolve and change over the last few years... And one of the constraints that basically every other umbrella that we would consider had was that we wouldn't really own that process; we would have to change a lot of the parts of that process and the governance rules around that for the entire institution. So we weren't really able to take on the kinds of experiments that ended up kind of making this very new governance model that we have at Node.js now.
+\[01:00:06.25\] So if we're looking at like where new projects go -- I think just in the last few months Apache said that you can even have your project on GitHub, so... If you wanna use new tools, if you wanna use new governance structures, there are constraints with some of these other foundations. And then moving away from there, if you have all of this tooling happening in the ecosystem, if you have all this new governance going on in the ecosystem, do you need to join a foundation anymore, or do we need umbrella foundations that are even less involved in the project than they are today?
+**Danese Cooper:** Well, it's an interesting question. The Drupal Association was having some problems last year - not the problem that recently happened, but more funding problems the association was having, because they make their money from a conference, so they're subject to the vagaries of the conference business... So Jim was helping us think through how to fix that, and one of the things he was saying was "Why are you hosting your own development environment when there's GitHub?" So for him it's a decision of cost, right?
+**Mikeal Rogers:** It's a liability, it's not an asset, right?
+**Danese Cooper:** But he was unwilling to take a project on that didn't host itself on GitHub, so he's made a tooling trace as well.
+**Mikeal Rogers:** There have been projects that were brought on in that time, that use non-GitHub tooling. I think that the distinction though is that they don't have tooling that they entirely have to own the maintenance and hosting of, right?
+**Danese Cooper:** Right. Well, everybody had to do that back in the day, because SourceForge, although it existed, you couldn't really run a project on it. I think going forward, there's a general consensus that spinning up a thousand foundations is probably not gonna keep scaling, partly because it's so hard to do now, and even though that has eased a little bit, it still comes up as an issue, because it looks like value that's not being taxed; it's not the last time we're gonna have that conversation, probably.
+It depends on how you define umbrella. Apache thinks it's an umbrella, the Linux Foundation also thinks it's an umbrella... Each of the Linux Foundation foundations are technically separate - they have separate by-laws and separate articles of incorporation - but they're under the Linux Foundation family, if you will.
+**Mikeal Rogers:** Also, some of those individual foundations are their own umbrellas, as well. It's complicated...
+**Danese Cooper:** Right. Jim is point out that those foundations can leave the nest anytime; they lose some things, but they might also gain some things by doing that. I think if too many of them do that and he becomes known as a foundation mill, then he may have trouble, if it turns out that the tax authorities are not wanting more foundations.
+And then there's the whole question of the locus of the foundation. The vast majority of open source foundations exist in the U.S., and that's because most of the really deep pocket funders also exist in the U.S. and they get tax breaks if it's a non-profit. And also, it's relatively easy to start a non-profit in Europe, but there are different rules, like in some countries you have to be very careful with your accounting and basically spend most of the money that you earn every year.
+**Nadia Eghbal:** So there's the question of whether to spin up your own foundation or not, and I think we've talked about that. Then there's also the question of "Why join an umbrella foundation now at all, especially (I think) if you're not, let's say, a Node-sized project?" What does Apache or Linux or any of the existing umbrella foundations have to offer smallish scope projects on GitHub with maybe one or two maintainers that just kind of wants to find a way to keep it going?
+**Danese Cooper:** \[01:04:09.12\] Right. Well, a smaller scope project probably wouldn't do well in Apache, because there's an assumption of ongoing contributorship that isn't just you. If you are looking for additional contributors and it was sufficiently interesting, then it might make sense. The tiny projects that Jim has been raising funds to help, like the one that caused Heartbleed a couple years ago - that's an example of one and two and four-person projects that were wildly successful and became the underpinnings of the internet, but never grew to additional contributorship, and the people that started it are aging or otherwise unable to continue, and it's problematic now because the internet rests on some of these things. So should they have joined in a foundation? Maybe so...
+Now there's a foundation trying to retroactively help them without them actually being part of the foundation, and it's a little bit "cat herd-ey".
+I think that a middling sized project can totally just live on GitHub. I think that companies need to not create foundations to hold their IP. Microsoft did that at one point about five years ago, maybe four years ago, and there were some other companies that did it because it looks attractive, because you can create a safe harbor and put all of your liability in that foundation... But you're not fooling anybody about what you're trying to do there. Basically, what you're trying to do is have your cake and eat it too, and nobody's gonna love you for that.
+When PayPal asked me if they should create a holding foundation for their open source assets, I said "Oh, hell no!" I think that's been proven as a bad answer.
+But for an individual project, I think you start on GitHub, you see how much traction you can gain... If you get to the size where big, deep pockets are starting to come calling, you're probably gonna find yourself pushed into a foundation, if only because they're more comfortable that the legal procedures are somewhat regularized that way.
+A lot of the way that we do things in open source comes down to Sun legal and IBM legal coming up with things that made them comfortable. Most of these licenses rest on copyright law, and copyright law is a good choice because they're almost immediate remedies if you can find infringement; you can stop software from shipping if you find that, where a patent-based complaint can go on for months or years, and the remedy might just be "I'm gonna make you show your patents to these people." It's just not the same as "You can't sell it anymore."
+So copyright law is what they rest on, but U.S. copyright law, if you have to go to court, you have to assemble all of the people who have any claim to copyright and get them to - at least the majority of them - agree to your line of defense, and that's not gonna be easy if you haven't already aggregated the copyrights. That's why Apache does copyright aggregation, because IBM wanted to play there; they paid for all the legal work to set up Apache, and of course, they set it up to meet their needs.
+Modern projects are asking whether things like a contributor agreement, which creates the copyright aggregation, is too much of a barrier to entry, because they're optimizing for contribution instead of optimizing for long-term legal viability. But you've gotta remember that IBM famously stuck through a lawsuit that they really probably were the only house at the time that had the resources to get through it... There was a challenge to Linux - and actually kind of to UNIX, but mostly to Linux - called the SCO lawsuit, where a single company was claiming that they owned a bunch of stuff and they were gonna sue everybody else for using their ideas.
+\[01:08:21.07\] If you know about the UNIX wars, establishing who owned UNIX is a pretty tricky thing to do, but IBM, instead of settling that suit, instead of doing anything else, actually saw it all the way to its bitter end - at least we think it's done - and it was proven that SCO didn't have a claim, and that Linux is safe to use. That kind of legal work is only possible if you can actually pay to have the provenance checked, or you've already aggregated the copyrights.
+So they actually spent an enormous amount of money on figuring out exactly who owned all the parts of UNIX and Linux and all the other IX's and making it clear that SCO's claim was spurious, right?
+**Mikeal Rogers:** To bring us back in push towards wrapping up a little bit - we talked a bit about why it's now just more important than ever to increase adoption and then get contributors, right? And I think one of the things that you're seeing in these news communities as pushback is that what they are feeling and what they really understand is that they need to retain contributors, and any barrier to that really has to meet a pretty high bar of scrutiny.
+**Danese Cooper:** Well, we don't know yet what the legal challenges to all of that are gonna be; it's gonna be interesting to see.
+**Mikeal Rogers:** Right, but that's such a hundredth-order problem from the problems that they're dealing with, right? And in particular if you're talking about establishing who wrote the work with these utilities, who would need them from day one. And in reality, we have them with zero of day one software, right? Nobody starts with the CLA anymore. So it's really hard for a lot of these communities to get their head around that this is such an -- not that this isn't a real legal argument or that this isn't going to be a real challenge, but that it's important enough to sacrifice their short-term sustainability... For if they succeed and they become a legal problem for IBM, then why do they care, right?
+**Danese Cooper:** Well, there are a couple of really famous projects that are not aggregated, and one of them was a challenge to J2EE (Java Enterprise Edition), and that software -- the guy (Marc Fleury was his name) was an ex-Sun employee, and he was very concerned that he was gonna get sued personally, and his lawyer told him not to aggregate copyrights, because it was the only strategy that would confound a lawsuit because there was no target.
+**Mikeal Rogers:** This was the Linux model for a long time too, right?
+**Danese Cooper:** Well, they didn't do it intentionally; they did it because they thought it was gonna be a barrier to contribution, I think... Or they just didn't think of it. I think initially it wasn't a high-order issue. The whole contribution question didn't really come into being until about the time that Mozilla pushed out their project or Apache was started. They were started within a couple of years of each other and it was very much on everybody's mind at that point. And it might have just been as simple as IBM, the largest IP holder in the world for software going "Wow, this is a whole new model. We've gotta get our hands on this..." - I don't know. But I do know that it has been the case that it's been useful to have some of that information on projects as the legal challenges come in.
+\[01:12:00.12\] I mean, I'm immensely encouraged because there's now gonna be (it seems) a formal legal challenge to the GPL. We just found out that GPL was a good enough contract that they were willing to let it be tried in court in the U.S., which is -- every other time in the U.S. that that's happened, it's been settled out of court, so that's kind of exciting.
+**Mikeal Rogers:** So we're thinking about this from the point of view of these open source projects on a somewhat individual basis, and I think that we have an assumption that over time it gets more adopted and it becomes more of a target and so on, but if you actually take a modern application and work through it and look at all of the different software that created this project and what its legal standing is, the idea that you would be able to enforce any of these mechanisms for even half of the stack is impossible, right? We moved to this model of all of these different small projects making such a huge small network of dependencies that it's very hard for a lot of individual project maintainers to see what use is it actually going to be to a person with an application being threatened with a legal challenge that "I did this thing" that it's going to put off a lot of my contributors, right?
+**Danese Cooper:** Well, but then you see people who are getting sued for stealing trade secrets and stuff between industries, right? That stuff expresses itself in open source, too, so... It's a thornier problem than you think, but I understand the reordering of the issues based on what feels like the most important thing is probably appropriate, but it may be a bit painful down the line, that's all I'm saying. We just don't know yet. Is it gonna be a fair tradeoff? Well, there's so many projects; certainly they're not all going to simultaneously come under any kind of legal action, at least hopefully not... So it probably is worth the risk, but it is a calculated risk, that's all I'm saying.
+I think that contributors to projects like Node realize that, but again, when I said at the beginning "Money changes everything" - most of this stuff came up around the money, not around the contribution, right? So I think as with everything, if you follow the money, you can understand the motivations for things, so...
+**Mikeal Rogers:** Yeah. I think that people tend to think of money as being directly flown into the project, but money is flying around all the time, and if you don't think about how it's influencing stuff, it'll just sneak up on you, right?
+**Danese Cooper:** Yeah, and it takes a lot of vigilance to keep that from happening. Keeping the same people involved -- so the BDFL model I think is pretty difficult to sustain for any length of time, because people get old, and you can't pass on that right... I mean, you really can't; we've seen it tried many times and it just doesn't work.
+**Mikeal Rogers:** They get bored quicker than they get old, but yeah...
+**Danese Cooper:** Right. Well, there's been some hacks on it. The Debian people vote in a leader every year, which is kind of an interesting thought... Because there was a BDFL, but then there's also a Debian leader. The BDFL is gone, but he's also not been involved for a really long time.
+So the consensus model seems like it's winning just because there's no way to live long enough to keep the BDFL thing going... Although some of the most powerful software is still BDFL software.
+All of these -- you notice that there's two ways to go always in open source, or in this whole conversation? There's free software and open, there's permissive and inherited, there's BDFL and consensus, there's foundation and not foundation... It's really interesting how there's always two choices, and just like seeds in a forest, both ways work for somebody some percentage of the time. Growth happens out of all of those possible avenues.
+\[01:16:08.10\] And we've gotta remember - 20 years seems like a long time; it seems like a long time to me, but really, it's just a blip. There's gonna be a lot more time that people try on these different structures.
+There's a bunch of people who think that open source is doomed because software will change in a way that means that it isn't created this way anymore and it doesn't matter, but I still think that the moment in history that we took advantage of to create it and bring it this far - it ws worth doing, because... I think it describes the ideal conditions for creating the best possible software. It's pretty clear that the ideal conditions -- Apple is like the exception that proves the rule. What it costs them to do things the way they do them is kind of crazy, and I think the reason that they still exist... I mean, Microsoft is having to change the way they do things now to become more open. Apple continues to buck that trend, but I think the only reason that they can do that is that they've got the monopoly of the hardware and the software going for them, and a fandom following that will eventually erode. It has to. That's what happens.
+I'm so happy to have spent the time that I've spent supporting this work, and trying to connect the dots where it looked like a dot needed connecting, and I've really had a good time talking to you guys about it, too. I don't know if it's unintelligible still, or if it helps to hear these stories...
+**Mikeal Rogers:** No, it's been perfect. And that was kind of a perfect wrap-up as well.
+**Danese Cooper:** Okay, I do what I can.
+**Mikeal Rogers:** This has been great, Danese. Thanks for coming on.
+**Nadia Eghbal:** Yeah, thanks for chatting with us.
+**Danese Cooper:** Of course, of course.

Open source and supercomputers (Spack)_transcript.txt ADDED Viewed

	@@ -0,0 +1,300 @@

+**Nadia Eghbal:** Todd, you work for the government... Tell us what your actual job is.
+**Todd Gamblin:** Okay, so I work for a specific part of government... I work for Lawrence Livermore National Laboratory - it's part of the department of energy. I'm a computer scientist, that's my job title (we don't really have official job titles here). Effectively, I'm a researcher, I'm a project lead, and I'm also a developer. I work on an open source project Spack, which is one of the things we're gonna talk about, and I also lead a bunch of other projects - I lead the DevOps project here at the Lab for simulation teams, and I do research in machine learning for parallel performance.
+**Nadia Eghbal:** And for those unfamiliar, what is Lawrence Livermore?
+**Todd Gamblin:** Lawrence Livermore is one of three labs in the National Nuclear Security Agency. It's part of the DOE, and we're responsible for a whole lot of different national security missions. That includes things like nuclear non-proliferation, making sure that people don't sell nuclear materials... We're also responsible for the U.S. stockpile, making sure that it remains viable, and then we do work on a lot of just basic research projects. We do some climate simulation, we have people working on that, we're part of the -- I don't know if you remember Obama's Cancer Moonshot, we have people working on that initiative - there's a lot of interesting data analysis and machine learning there - and just general computation.
+\[03:55\] I think all sorts of missions at the Laboratory are based on computing and our Livermore Computing Center, which is where all the different clusters are that we run. I work both with the research organization and with the computing center, and also with the code teams, who are from all different parts of the Lab.
+**Nadia Eghbal:** So you play with nuclear weapons all day?
+**Todd Gamblin:** I do not play with nuclear weapons all day. \[laughter\] What I do is mostly in a support role. We're working with the simulation team... One of the things that we simulate is nuclear weapons, that's true. We also have a whole lot of open science codes; we work with all those different teams to help them get their simulation up and running on our machines... My role is mostly taking what they do, looking at it and trying to make it go fast, and also just making sure that they run effectively on the machines. That's why I run this DevOps team... We call it the HPC Developer Ecosystem team. That's about collaboration tools, we're trying to deploy secure CI for our center, and work with the teams to help them package and share all the different parts of their codes.
+In the research role, what I do is I work with a lot of students in academia, I have post-docs who work for me, and we'll usually talk to the code teams, try to figure out what their problems are, what kind of things are they having issues with in terms of making sure that their simulations run fast, and then we try to come up with ways to speed them up. So my background is really performance tools. We'll go look at a parallel simulation and try to measure "What's it doing at runtime? Where is it spending its time?", we'll use profilers, trace tools, "Is it bound on memory? Is it bound on CPU? Could we do vectorization better?" Stuff like that. Basically, we support the code teams in lots of different ways.
+**Nadia Eghbal:** Cool.
+**Mikeal Rogers:** \[05:52\] So in terms of like your team and everything... Is this a bunch of people in the same physical location? I think when people hear of a lab, they're thinking like some place with a giant hadron collider, or something... Is this institution slightly virtualized, as well? You mentioned you're working with different students, and stuff like that - are they all on-site in the same place, or is the institution a little bit more virtualized?
+**Todd Gamblin:** I would say it's both. Livermore itself is a one square mile laboratory in Livermore, California; it's like an hour from San Francisco. We do have a giant laser here... We have the National Ignition Facility, which is the world's largest and highest energy laser; basically, that's like 192 beams that focus on something smaller than a pea and try to implode it. That's a fusion experiment. Some of the simulations are simulating that, so we sort of iterate with them.
+So that's more like on-site stuff... We would go and visit the code team, or they would come to our office, we have meetings, like a normal company... I sit in a building with the big computers, so basically I have an office building and there's essentially like a 48,000 square foot data center attached to it. But yeah, there's a lot of collaboration that goes on here.
+We also collaborate across the DOE and with a whole lot of universities. I have telecons all the time, I have collaborators at the University of Delaware, University of Arizona, University of Illinois and other places, with students, and then we also collaborate with people across the DOE. The Exascale project that I think I mentioned earlier - that's a collaboration of six different laboratories; it's all three of the NNSA labs - that's our agency, the National Nuclear Security Agency; that's one part of DOE, and then there's also the Office of Science. You may have heard of Argonne or Oak Ridge National Laboratory, or Berkeley Lab - those are Office of Science laboratories and we have telecons with people there all the time.
+To some extent they're similar, because they have large physics experiments on their sites too, and all those big labs have big computing centers, but... We travel around the different labs and also to conferences and do things like present papers... So it's a fairly diverse job, working for the DOE.
+\[08:10\] I like it a lot because I'm not always working with programmers or computer scientists, I also get to talk to people who are doing physical sciences and other stuff. It's a pretty cool environment from that perspective.
+**Nadia Eghbal:** Can you talk a little bit about your lab's history with open source? I think you've mentioned in previous conversations that your lab has open sourced a bunch of other types of projects before, and I was wondering also how did you personally get into open source in your current role?
+**Todd Gamblin:** If you look at Livermore's history, I think from the Lab's founding in 1952, we've deployed fast computers; building software for them has been a long part of the lab's history. We built this thing called the Livermore Time Sharing System - I say "we", but this was way before my time... And that was one of the first time sharing operating systems to run on like a supercomputer, and that was so that the physicists could swap out time for their simulations.
+We also used to build compilers here. Apparently, we had a compiler team, there was something called "the Pastel Compiler." I think Richard Stallman actually wanted to base the original version of GCC on that, but I believe the memory requirements for it were way too high for ordinary computers... It wasn't gonna run on something that wasn't like one of our supercomputers, so he went and built GCC on his own.
+Then as the machines evolved, we've deployed more and more fast machines, and in the '90s they started looking more like clusters. We were one of the first labs to look at deploying Linux on our machines and maintaining our own Linux distribution for HPC (high performance computing) machines. So we did that...
+We have a developer here who ported ZFS to Linux from Solaris, and he maintains that port, that's pretty popular... And I think in general in the research community people have done open source a lot. I don't think that necessarily means that they've taken steps to popularize or necessarily build really large communities around their tools, because it's kind of a niche area, but we do have lots of open source projects.
+ZFS I think is one of the major ones now, and then Slurm is used all over the place, and that was invented at Livermore. That's a resource manager... It's basically like a batch system for submitting jobs to clusters and managing the nodes and the memory allocating time to users.
+**Mikeal Rogers:** These are pretty high-end use cases... Who are the other users of these open source projects, other than Lawrence Livermore?
+**Todd Gamblin:** Other labs deploy Slurm, and actually a lot of university computing centers will run that. Our Linux clusters all run Slurm; even our IBM Blue Gene machine - which is like a million and a half cores - runs Slurm, and other national labs also run Slurm on their systems.
+**Nadia Eghbal:** \[11:09\] The name makes me laugh every time.
+**Todd Gamblin:** Yeah, I think it's actually from Futurama... So yeah, we have lots of interestingly named tools. And then ZFS I think is used in the industry. I'm not a file systems expert, but I know that a lot of companies have started using ZFS on Linux; it's fairly widely used all over the place.
+We ported ZFS because our interest is in -- so we also have developers who work on a parallel file system called Lustre, and that's what we run on our clusters. Lustre is a parallel file system where there's a local file system that it's based on. Lustre runs on top of ZFS in our current configuration, and we're pretty psyched about that, because we get some good performance out of it.
+**Nadia Eghbal:** Was Spack the first project that you had open sourced at Lawrence Livermore yourself?
+**Todd Gamblin:** \[12:03\] It wasn't, actually. In addition to Spack, I've had a bunch of different research projects. For my PhD I worked on a scalable clustering algorithm. That's open source, it's called Muster and it's on GitHub, too. That was sort of for finding nodes in a parallel application that very similar performance characteristics.
+I don't think that really caught on... It was sort of a research project; it wasn't generally useful, like a package manager is. We also had a project here called -- well, originally I wanted to call it ClownCar, but that was deemed not serious enough, so we renamed it to CRAM, which stands for ClownCar Renamed to Appease Management. \[laughter\]
+**Nadia Eghbal:** Oh, government...
+**Todd Gamblin:** So that dealt with a real problem that we had here. When we deployed the Sequoia machine, which is the big IBM Blue Gene/Q system, originally people anticipated running maybe like 200, maybe 400 jobs at a time on that thing, and they were thinking that each job would be maybe 10,000 cores, or something like that, on the order of 10,000.
+Now the mission has sort of changed, and one of the things that we're very interested in is uncertainty quantification, so trying to figure out what inputs is the simulation sensitive to? We decided that we wanted to run lots and lots of small jobs on the machine; we had people who really wanted to run a million and a half one-core jobs on the system, and the resource manager couldn't handle it, because it imposed some OS requirements, sockets, memory on the front-end of the system, so we would have to have a process per job that was running in the system, and they couldn't handle that many...
+CRAM basically takes one job and splits it into lots of them, and manages that sort of on the cluster, as opposed to on the front-end. So it's kind of a stopgap until we get a more scalable resource manager.
+Those are just some examples, but I would say in the research world there's lots of open source software. That's kind of the default for people who are doing research, publishing papers about it... We have a compiler project here called ROSE - it's a source-to-source compiler; that's been open source for a very long time, and actually they've gotten a lot of traction before my project. They've been really well funded by different parts of the DOE historically.
+**Mikeal Rogers:** Do you have a sense of how long it's been the default? Because it hasn't always been the default to do everything open source. Do you know when that shift happened? Or just like as long as you've been in it, you've been able to do everything open source?
+**Todd Gamblin:** As long as I've been in it, yes. I started here at the lab in 2008, and I know that it goes back further than that. Our early efforts with TOSS, which our Linux distribution (based on Red Hat) - that was in '99 or the early 2000s. Actually, there's a policy document from the DOE from 2004, and it basically says that for all the software developed under the Advanced Simulation Computing Initiative, which is what funds most of our machines - we can talk about that as well - it should be open source, unless there's a reason not to make it open source.
+An open source nuclear weapon simulation sounds like a really bad idea, but some of the other software that we develop to make our systems run - that's pure computer science; we can put it out there, and other people can use it and benefit from it... And they do.
+So yeah, there's a DOE document that says we should make this software open source from like 2004, and it's interesting because -- you know, that's something we're struggling with now... We do have an IP organization as part of the Laboratory. Livermore has had some successful commercial software come out of some of our projects. Dyna is a physics simulation package that a lot of auto manufacturers use to model things like crashes, and that's been very popular in the industry, and it was originally a Livermore product and I think we've made a significant number of royalties out of it.
+\[16:20\] But other projects, like Spack or a lot of this infrastructure stuff - I don't think we're gonna be able to sell that, and I don't necessarily think that we should, because we're in sort of a niche here. There aren't that many sites that do large-scale high-performance computing, and I think we have the same motivation that a lot of these larger industry open source projects do. We wanna share resources among the different computing sites, and try to develop a lot of the infrastructure together, instead of doing something siloed at different laboratories and developing it independently, because I think it's just redundant effort at that point.
+**Nadia Eghbal:** So what's it like when you think you have something that you might wanna open source? What does that process look like for your Lab? Do you have to talk to the IP Office?
+**Todd Gamblin:** Well, yeah... So the IP Office wants to make sure that the thing that you're putting out there as open source is not something that we could potentially get royalties off of. That's kind of interesting, because the government in general, technically -- I don't think we're actually allowed to have intellectual property if we were actually a part of the government, but we're a contractor. So because we're a contractor and because we're run by this LLC, we can own things and potentially license them out under commercial licenses, I believe. I'm not an expert on that.
+But yeah, so we have to go through the IP organization when we put things out there. For most of my stuff, getting the approval isn't actually that hard; it's the actual process that's kind of tedious. Livermore's software release process involves burning two CDs and carrying them to three buildings throughout the Laboratory, one of which is in like a classified area... So I can't have some of my post-docs go do that, because they can't go into classified areas. That part is kind of tedious, and there's forms that you have to fill out, and they're actual paper forms, so that can be frustrating. But in general, the approval is pretty good, and I think the Lab is okay at a high level with open source software. They actually are starting to recognize the benefits.
+The thing that I would really like to do here is start thinking about open source a little more carefully in terms of like "What is our open source strategy with this thing?" Look at open source as a potential software sustainability strategy for DOE projects. That's kind of like what we're doing in Spack; we've actively tried to build a community around that.
+Depending on what it is, maybe that's a good strategy, maybe it's not. There's some things that we wanna develop internally that we wanna keep a well-staffed team on, and there's other things that we can probably share effort with other laboratories on and that we would wanna build like a larger community maybe, even out to Academia or industry around. I think it really depends on what the software is, which one of those things you wanna do.
+I was on a working group here that was trying to come up with -- I guess it started out as like a software engineering working group, and I think by the end of it we had changed the name of the working group to Software Sustainability. I think everyone thought we were gonna come back and say "Okay, everyone should do waterfall, or everyone should do Agile, and here are the tools that you should use", and what it ended up being was "We think we need more than that to sustain our software here. We need an actual plan for how the project is gonna develop, who we're gonna include in the community, who we think is our target audience for this thing and how we're gonna sustain it, who's gonna contribute to it...", so we sort of came up with a plan that centered around that, instead of trying to be prescriptive about how people can develop their software.
+**Adam Stacoviak:** After the break, Nadia and Mikeal talk with Todd about the back-story of Spack and where it's going. We talked about Spack's growing community, how they're finding contributors, especially in a government context. We also talked through the details of Spack becoming a NumFOCUS affiliated project and what that means for it. All this and more after the break.
+**Break:** \[20:15\]
+**Nadia Eghbal:** Todd, you've mentioned Spack a couple of times as an open source project that you've created while at Lawrence Livermore... Could you explain a little bit about what Spack is for people that are not high performance computing people, and what inspired you to create it.
+**Todd Gamblin:**\[21:47\] I have an easier time explaining Spack to people who are not high-performance computing people than to people who are high-performance computing people, in the sense that like I think people who are not high-performance computing people are familiar with package managers... Spack is a package manager; it's not a binary package manager, it builds things from source, and it's for running on these large machines that we care about. That's anything from like a Cray machine or like an IBM Blue Gene, or just a Linux cluster. We run lots of different things like that.
+I think the reason that I built my own package manager is because the way that we distribute software in HPC is kind of different from what you would expect from most software products. People actually build their code from source on the supercomputers, typically because they wanna optimize it for hardware, but also because that's kind of how scientists put their code out there. If you download a simulation code, it comes as a tarball and it probably has dependencies on maybe some math libraries or a numerical solver, or something... And these can get really complicated.
+If you look at the climate community, they have just a whole bunch of Fortran models that are tied together in different ways. Our codes tend to be like C++, some C and maybe some Fortran down in the numerical libraries, but primarily they're C and C++. We have Python drivers for some of those things; I think recently we're getting into Lua.
+But yeah, so one simulation could be 47, 70, maybe 100 libraries in C, C++, Fortran and Python. So building that and linking it all together and getting fast native libraries there is kind of hard.
+And the other thing that we care about is exotic compilers. So we don't just build with GCC, we build with GCC, Clang, Intel compiler, Portland Group compiler, and then maybe the Cray compiler on the Cray machines... Lots of different ways to optimize your code. And there really wasn't a good package manager for experimenting with all that and swapping compilers and compiler flags, and for building lots of different versions of things.
+\[24:05\] Another problem with the software ecosystem here is that people will distribute their code, and maybe it relies on a very specific version of some research library, because maybe that research library is published like by a university, maybe they don't have the greatest release cycle set up, or maybe they don't even do releases. Maybe they just publish to the head of their repo. It's hard to rely on some of those things, and so we have to be specific about a specific commit, maybe a specific revision, or even like a specific version where something isn't broken. And you may get two packages that depend on two different versions of something. So we needed something to support all that.
+Essentially, Spack is a tool to build software the way that I was building it, where I would have lots and lots of different versions of things, and essentially any dependency graph that you make is a new version, so we assign a hash to that. And that's a lot like -- if people are familiar with Nix or Guix or some of these, they call them "functional" package managers... It's a lot like that, but we have the added requirement that we wanna rely on the vendor libraries. So when you get a super computer, it comes with an optimized version of MPI (message passing interface); that's what people use to communicate among the processes in their application. So we wanna be able to build packages that actually rely on vendor products, as well as things that we can build from open source. Putting all that together led to Spack.
+The idea is that you would be able to make a package that could build across all those different platforms and that people would be able to install it and say "spec install foo" and have foo, which is not generally the experience on supercomputers.
+**Nadia Eghbal:** Interesting just hearing some of the things that you have to be particularly concerned about, and it feels almost like you're doing open source but for a smaller but more engaged community of users, beyond your Lab. It sounds like a lot of it is for other labs, or that type of audience, more than a general public, which might be how we normally think about open source.
+**Todd Gamblin:** Yeah, I mean... I think for a lot of the simulation codes, the usage model is very different from what you see in web development. Depending on the team -- some of the teams are their own users, so it's a research project, they're developing this thing and they're testing out different versions of it and they're doing experiments with the latest version of it. But they might rely on some math libraries, or something.
+Other teams do publish software that can be used by lots of people and they'll actually bother to package it themselves, but I think the build process on these machines is so complicated that your typical computational scientist doesn't wanna get into all those details; they wanna focus on the science.
+We actually talked to different teams in the community and try to figure out "What's your deploy process?" Some of them had good practices, and others said "Well, we have one user on every machine, because we build the code once, and there's this guy who runs it for everyone." Then everyone talks to him, they send him their inputs, and he goes and runs it and he sends them the outputs. That doesn't seem scalable to me, but that is how some of the teams operate, and it has worked for them historically.
+Now we're starting to see more sharing in the community, and also this sort of push for exascale computing has caused people to really think about how portable their code is... So things like this have started to matter to them.
+We have people who are really concerned with performance portability, which is another good reason to package your software and to use libraries that work across different architectures... For a long time - since like the '90s - if you could run on a Linux cluster you were probably okay and the processor architecture on a Cray maybe didn't look so different from what was on a Linux machine. But now, if you wanna get a really powerful system, you might need to use GPUs, you might need to use like a Xeon Phi, which is like a many-core from Intel... Or maybe you wanna run on ARM, or some other exotic type of architecture, and you have to take the same parallel code and get it to run into those models.
+\[28:14\] A GPU is very different from a Xeon Phi in terms of how you would parallelize code for it, and it's very different from your multicore processor. So I think people have started to realize that they have to rely on libraries for that. They really wanna separate that concern out and give it to someone else who works on maybe like the math library, or some abstraction layer for their loops, whereas historically the tuning was not so different across the different architectures that you had to outsource the task to someone else; the teams could get by by writing parallel code for their own system.
+**Nadia Eghbal:** So how does that play into thinking about getting contributors for an open source project like Spack? Because I imagine -- I noticed that Spack has a lot of contributors, and there are only so many people that are using Spack in the first place, or your total audience is smaller... Do you feel like you kind of know everyone who could contribute to it, or are there people out there that you think should be contributing but aren't aware of it?
+**Todd Gamblin:** I think that one really intriguing place where we could get more contributors would be like industry... I think definitely we could grow the audience for it right now. I think there's a lot of people who still just build things by hand.
+I think when you look for contributors you only have to think about what's the structure of the community. For HPC actually I think there's maybe more roles than people are used to in the software community. I think you're used to thinking like developers and users, but actually we have -- so there's users, there's like scientists who want to get on a machine and they wanna run some application and do some kind of simulation. There's also the developers who actually make that simulation (they might wanna distribute their tool) and there's people who run HPC centers. So actually, in this field, your typical HPC center has a user support team, and they deploy a lot of the common libraries and things that people need on their system.
+Actually, one of the motivations for Spack was that that particular task was getting to be very overwhelming at Livermore computing, where I am.
+In my case - I was a researcher and I wanted to deploy things on the machines for students and post-docs who were working for me who don't necessarily know how to build all these things, but they want to run them. Then we also have a users support team that deals with the application teams, so they deploy software for them. Then some of the application teams actually deploy their own software.
+The original contributors to Spack were people who were at these HPC centers who were just sick of building things by hand day in and day out, and wanted to deploy things on these machines. Our deployment model is a little different from what you might be used to from the cloud... Most of these machines have like a shared file system across all the nodes, and will go and build a copy of an application and put it out there, and the users can use these things called modules. So they can say "module load Hypre", or PETSc, or some library, and that gets the tool into their environment, so that they can use it.
+So yeah, we targeted the HPC centers to try to get them to work together initially... But it turned out that we started getting more contributions from actual application developers and people kind of hacking on HPC code on their own, and I think that's actually served to really grow the community, but that wasn't something that was necessarily planned.
+**Mikeal Rogers:** The other HPC centers - are they as adept at open source as Lawrence Livermore is? Do they have similar policies around it, or are you a little bit ahead in that regard?
+**Todd Gamblin:** I don't know that we're necessarily better or worse than the other labs... I think there's strengths and weaknesses in terms of the process. There is no standard process for releasing open source software; it kind of varies from lab to lab. In terms of actually having popular research software, I think -- Argonne has worked on MPICH for a really long time. MPICH is like the most popular -- well, I don't know if it's the most popular, because there's Open MPI now, but it's a major MPI implementation... So they've maintained that for years and they've actually got lots of funding to do that (or at least some funding). And then they also have developed math libraries like PETSc, Oak Ridge has developed IO libraries...
+\[32:19\] I think in large part -- I mean, all the different laboratories have had some large open source project that they've put out there. I think DOE has had an open source software ecosystem for a while. I don't know that they've always had licenses or thought about the licensing aspect of these things, or thought about building communities around them. I think a lot of them are sort of maybe confined to a team at a particular laboratory who maintains the tool.
+Accepting pull requests wasn't really easy before GitHub; even setting up infrastructure outside your laboratory could be difficult. A lot of Livermore teams have had trouble setting up all the hosting infrastructure that you need to host an actual project with collaboration tools, Subversion, things like that. So we've had people actually host their projects outside the laboratory to make it easier to get the hosting done for the open source communities. Not all the teams were willing to do that...
+Al Livermore recently we've consolidated our GitHub presence, we've gotten more and more people to join the GitHub organization for LLNL; I think it's become easier, so more and more people are getting into that and really starting to think about how they put their software out there and how they do build communities around it.
+**Nadia Eghbal:** I wonder how much - I know government is not a monolithic thing, so maybe just for you guys - of open source practices are being influenced by industry, versus they're doing it on their own... Do they care what companies are doing now with open source and looking at that and saying "Oh, we should do it like that?" Like getting on GitHub, stuff like that, or is it in their own bubble?
+**Todd Gamblin:** I think different people are pushing that at different laboratories; it just really depends on the part of the laboratory. You could have two groups on the same hallway who feel very differently about this. We have some teams who are reticent to put their development version out there, they don't wanna release things before they're done, they don't know about doing development out on GitHub, and then some of them have actual sensitivity concerns around what they're doing that maybe prevent them from doing that. Or at least the lab doesn't provide 100% clear guidelines on when you can do development out on GitHub and when you can do development inside. Like, you can release your software, but it was for that version; what do you do about incremental changes to the thing? Can you host development outside?
+Yeah, so we have people here who have really pushed to get -- Ian Lee, the guy who presented at GitHub Universe on Livermore's open source software, he really pushed to get people to consolidate there and to use open tools, to use sites like Readthedocs and stuff like that, and other labs have people like that as well.
+But you know, like I said, I think that labs have for a long time -- software projects, they don't think about building a community around it, and they don't think about it in the same way that the industry does now, in the way that I saw... Like, when I was at the Open Source Leadership Summit (the Linux Foundation event), it seemed like a whole lot of companies had bought into this notion that there were just basic infrastructures that they didn't wanna pay for themselves and that they wanted to build communities around so that they could focus on differentiation and not doing the same infrastructure stuff over and over again. I'm pushing for that here, for us to start thinking that way about a lot of our infrastructure projects, and I think other labs are starting to do that, too.
+I wouldn't say though that historically the labs have been against that, I just don't know that they've thought about collaborating among the different labs as much.
+**Nadia Eghbal:** In terms of where you decide to start going with Spack, I noticed that Spack joined NumFOCUS as an affiliated project... For anyone who's listening, NumFOCUS is an umbrella organization for a lot of scientific and academic related open source projects, so I was wondering why did you decide to join NumFOCUS, what does that mean to you and what was the process like?
+**Todd Gamblin:** \[36:25\] I don't know, it was kind kind of on a whim. I went to the NumFOCUS web page and I looked at their supported projects and their affiliated projects. I mean, I like all the stuff that NumFOCUS is doing; they're doing all kinds of awesome things, especially for the Python community, and R...
+I think that was right after Fernando Perez, the Jupiter guy (he's at Berkeley Lab); he had come out and given a talk here. So I don't know, I was inspired to go and do it. They had a list of requirements and they were like "You can be a NumFOCUS affiliated project if you this, this and this", so I wrote them an email and said "Okay, here's why Spack fits all these criteria." I think one of them was you had to have a community of contributors, you had to have more than three core developers, or something else, and you had to make significant use of their supported projects... So I just wrote an email that says "We do that. We package like 12 out of 18 of them, and we have a community... So what do you think? Could Spack be an affiliated project?"
+And what does it mean to me? Well, I like what those guys are doing, and they said that if your project on the NumFOCUS web page, they can encourage people to contribute to your project. So I think for me it was that I'd like to be associated with this community so that potentially these other scientific developers could look at Spack as a potential package manager that they could use. So it's a way to get contributors and maybe users.
+**Mikeal Rogers:** You've said something interesting earlier, which is that you've gotten more contributors from the user side of things than you have from the people maintaining the cluster side. Do you think that that's because you've democratized the whole role a little bit more, so you've made it easier and gotten it more accessible to them, or do you think it's just a difference in how those people think about the project and open source and contributions? Is it a cultural difference or is it actually just like a skillset difference?
+**Todd Gamblin:** I think it might be a cultural difference. One thing I've found with the HPC centers is they don't adopt things easily; they have processes in place that they use to deploy software; that's been more of a socialization effort, like talking to them and saying "We're really behind this. This is a solid thing that you can rely on, and we have developers here at Livermore who are working on Spack, and we will work with you and help you to port it to new platforms."
+For example, one thing we did to get NERSC (which is Lawrence Berkeley Labs supercomputing center) on board is we worked with them to actually port Spack to work in the Cray environment. That was a fair amount of work, but they actually put in some developer effort and so did we, and we thought that was valuable because we worked on CRE, too.
+I think one of the things that we did with Spack that was really helpful for getting more casual users to contribute was -- I mean, we looked at Homebrew and some of these other projects... I mean, Spack's package format is based on Homebrew, it's just Python and not Ruby. We looked at what we could do to make it really easy to download Spack and install a package.
+All you really had to do with Spack is clone it. It doesn't require you to be able to run Pip or some other Python package manager; you can just clone it and then you can run the Spack executable out of the directory there.
+I think that has helped get the regular hackers to start using it. That, and we specifically chose Python because Python is a popular language for scientific computing. I should say that Spack is not the first attempt to build an HPC package manager, there have been others. Oak Ridge had an internal package manager that was based on Homebrew, but it was written in Ruby, and they had a lot of trouble getting people in this community to write packages, because nobody wanted to learn Ruby.
+\[40:07\] Actually, there's another popular package manager that's written in Python from HPC at the University of Ghent (that's in Belgium). It's called EasyBuild. They've done really good things for HPC packaging, but their tool is focused mostly on (I'd say) administrators of clusters, and it doesn't make it easy to say "Do things in your home directory." You can, but I think it's not the model that people are used to, where they just say "Install this thing." They're mostly focused on being like an installation and build tool, and they don't have a lot of the package management features, like uninstalling software and managing dependencies after the packages are installed, that we tried to put in to make things easy.
+I was kind of modeling the Spack contribution model on Homebrew, because that seemed like a successful thing, and it's actually something that people in this community use on their Macs, and it seemed like they did a really good job of making it easy to contribute. So to a large extent, I think it was because we made it easy for people to run it in their home directory that they've got us the casual contributors.
+**Mikeal Rogers:** Right, yeah. It sounds like the big differentiator for you has been focusing on the users rather than on the cluster maintainers. And, aligning with NumFOCUS is part of that, because most of what they work on is there; being in Python is definitely a part of that, too.
+**Todd Gamblin:** Yeah.
+**Mikeal Rogers:** Coming up, we get into funding for Spack and how Todd keeps this thing alive, what it's like working on a project from grant to grant versus ongoing programmatic support, and the challenges for open source in government, especially the Department of Energy. Stay tuned.
+**Break:** \[41:54\]
+**Mikeal Rogers:** So let's dig into it a little bit... Tell us a little bit about how you get funding for your particular projects. You mentioned a couple projects already that get you a fair amount of funding... How does the funding flow into your project specifically?
+**Todd Gamblin:** So for Spack, right now we are programmatically funded. What that means is we're funded by a program. I guess in DOE (and in government in general) a program is a giant source of funding that's been allocated potentially by Congress; that's kind of the best kind of funding, because it doesn't end, unless something drastic happens (fingers crossed). You're basically doing production work at that point, or you're doing some important research project and you have milestones... I think that's fairly stable funding.
+There's lots of research funding in government and in DOE. We have an internal funding source, so for all the grants that we get at Livermore, we tax them and we have this thing called LDRD - I think a lot of the other labs do, too; it's called Lab-Directed Research and Development... So if you are doing research, you can propose a project there - that's not necessarily for development though, so I think that's one of the complications.
+\[44:08\] There's competitive funding grants from the Office of Science, so that's another part of DOE; it's outside the NNSA, but we can apply for funding from there. In fact, a lot of our basic science research is funded by that.
+Then internally we have some funding pots that are discretionary... Basically, the management of the computational organization LLNL has discretionary money that you can go after. I think it's a more informal proposal process, but that's for things like hardening and maybe porting something to a new platform... Things like that. Sort of more mundane things, not way out there research. Those are the main ways that people get funded.
+Other things that people have been known to do - there's a whole SBIR program in the federal government... Are you familiar with that? It's Small Business Innovation Research. That is funding for small businesses to get started. A lot of things that started out as maybe programmatic projects and then sort of outgrew them, they will spin off a small company and then apply for SBIR funding. That's a fairly common thing to happen.
+Academic teams will do that, too. We've got some collaborators who built parallel programming models who've gone and spun off companies for things like that.
+There are a whole lot of ways to get funded. Navigating all that I think is fairly difficult, and then one observation that I guess I would say I have about the whole process is that there is a lot of research funding in our area for cutting edge things, but as far as actually maintaining the software, that is not considered nearly as much. The value is placed on the cutting edge research and publishing papers and building prototypes, but then actually turning that into a product - there's a gap between the research funding and the production funding that you have to sort of cross.
+I think doing that is often a lot of work... You need to socialize your thing, you need to convince someone who controls programmatic funding that they need this thing. Often times, that's getting it into one of our simulation codes, making them rely on it; that's something that would catch people's interest.
+Other things would be making it a critical product for the compute center, like if we're actually using it to deploy clusters and we need it, and it makes us more efficient - that's another way that you can get programmatic funding. But in general, it depends on what the software is and what mission it supports how you get funding like that.
+There's also a distinction between hard money and soft money. Are you guys familiar with that from Academic research?
+**Nadia Eghbal:** No.
+**Todd Gamblin:** Hard money would be like the programmatic money, where it's ongoing, it doesn't necessarily go away... It's either overhead for the organization, or it's part of how Livermore is supported over time. Soft money is stuff that you have to apply for, and it tends to have a short lifetime. If you get a research grant from our internal LDRD process or from the office of science, then it might end after like 3 or 5 years... So you need to have an exit plan or something that you wanna do, some way to productize the thing after that, or some other way to get funding.
+Usually, the exit plan for a lot of projects here is get programmatic funding for the thing, but that doesn't always happen. It's not an easy task to get a project funded, I would say, and to get it to grown. I think that's pretty similar to elsewhere.
+**Nadia Eghbal:** Nobody cares about maintenance, huh?
+**Todd Gamblin:** Right, and I think that's actually -- I don't think people think about maintenance as much as they should. I think people think that once you do the research, that it's magic, or something, and the software continues to work because you've already done the development for it... But there's a ton of maintenance that we have to do, so we're pretty conscious of the maintenance cost for things.
+\[48:07\] Depending on the part of the organization it is, the compute center here is very conscious of the maintenance costs. I don't necessarily think that the people who are running research programs are always aware of that, or thinking about how that thing might take off after the research program. I mean, that's not 100% true because there are programs...
+Tech Transfer is something that we care a lot about. We're supposed to make products viable for industry, and people like it when you do that... But on the software side, I think that -- I think one thing is that the labs... Like, we're fundamentally a science organization, right? We have a science mission, and we have national security missions that we support, so those are the main focuses of the lab. Software is something that supports those missions, but I don't necessarily think that we think about software maintenance in the same way that a software company or a dotcom would. I don't think there's as much consciousness of the ongoing maintenance costs and of what it takes to support software long-term.
+For example, for our simulation codes I think there was this study done, and we found that the number of engineers per lines of code that they had to support and maintain on our teams was a lot -- we had a lot more lines of code per engineers than industry teams do. That was interesting to me, just from a "what do we think about maintenance" perspective.
+**Mikeal Rogers:** I feel like we're covering some of the instability in the funding and we're kind of focused on that, and what it doesn't do... But I'm just trying to compare this to other funding sources that we've talked about. If you start a startup, you're gonna have to go out and beg for money every two years...
+**Todd Gamblin:** Yes.
+**Mikeal Rogers:** If you do grant funding, you're gonna spend a year to get a year of funding, and then have to beg for money again, year on year on year for grants. I think compared to those funding sources, it's actually relatively stable.
+**Todd Gamblin:** It is. Programmatic funding is definitely stable compared to those, and I think that's a good thing, right? I think you've talked on this show in the past about software as a public good, right? Maybe the infrastructure is something that should be funded by some part of government for maintenance. I think the thing that you have to figure out there is which software products have something to do with the mission that the government cares about? Because that's the stuff they're gonna pay for.
+On our side, the things that I'm pushing for are -- we've gotta think about things that could have a broader impact, we've gotta think about things that we could build communities around, because then we could get both contributors for our projects, and also the investment would pay off for us because we'd be supporting something and getting more out than we put in.
+That's actually something that I've tried to point out with Spack - I have this chart that shows "Here's the amount of programmatic funding that we have for Spack", it's like two engineers, or something like that, and "Here's all the contributions that we've gotten." Over the course of a year, the number of packages developed by Livermore went from almost like a 100% of the packages in Spack to less than 25%, so we're getting most of our packages now from external contributors, and that's stuff that we wouldn't have ever been able to sustain ourselves. So I think that we have to find places like that and apply the stable funding there for the ongoing maintenance investment and use it to try to build communities that can help to sustain the software... But that's a hard thing to think about. That's not something that most software engineers think about and it's not necessarily something that most researchers think about, how to actually sustain the product like that.
+**Nadia Eghbal:** I'm wondering, for getting contributions specifically... That must be influenced by academic cycles as well, right? So if you have people that are contributing while they're doing their post-doc, and afterwards they stop - is that a thing?
+**Todd Gamblin:** \[52:04\] Yeah, that's a really good point. I think that might be another aspect of the research funding structure that maybe doesn't work so well as the long-term sustainability strategy. In research you're really encouraged to be a PI (principle investigator) and to start new things, so to advance your career in research, you need to publish, you need to run things, you need to get grants, you need to start new projects, but maintaining the old thing doesn't get you additional research kudos necessarily.
+I think some projects have managed to make it do that, projects that are widely used... I mentioned the ROSE compiler here. They've published tons of papers, and they always have new collaboration going on some piece of their compiler infrastructure. But there's only certain products that can do that.
+We have a paper on Spack, we submitted it to the State-of-the-Practice Track at Supercomputing, which is the big conference in my area, and I think that got us a lot of publicity, but how we would publish more papers, or continue to publish papers about it is up in the air, so we have to rely on the programmatic funding there, and on actual software developers, not just researchers.
+**Nadia Eghbal:** Can you talk a little bit about Exascale? Because I know you mentioned that that's a project where the focus has been on actually developing the software and not just on writing papers or getting funding that way, so how did that happen?
+**Todd Gamblin:** The Exascale computing project is a major collaboration between the six national laboratories. So it's Livermore, Los Alamos, Sandia, Argonne, Oak Ridge and Lawrence Berkeley Lab. PNNL is in there as well - that's Pacific Northwest National Lab... And I think some other labs are collaborating. But it's a project to build a software stack for an Exascale computer. What that means is we like to measure things in FLOPs in HPC, where a FLOP is a floating point operation. I think your typical web application doesn't deal so much with FLOPs. An Exascale machine would be capable of a billion billion FLOPs - that's floating point operations per second.
+So there are plans to build large Exascale machines, and this project is to design the software stack. That's at exascaleproject.org. That was sort of motivated by an immediate need to go and actually build the software stack so that when the Exascale machine gets here - which is something that we're planning to do - that we actually have scientific code that can run on it. Another focus of the project is to make it so that industry can actually take advantage of some of the things that we built.
+There's 15 or so simulation codes in the Exascale project, everything from like nuclear fusion to the climate community, to molecular dynamics... There's lots of useful science there that could be used and could have an impact on industries, like for making cars better, making more clean energy in different areas... But we'd like the industry folks to be able to use that. So the project is definitely focused on delivering that software stack.
+I'm involved in it in two ways - one is with Spack. I would like Spack to be the package manager that's used for the Exascale project, so that that's how we deploy things on supercomputers, that's how we built things and that's how we make it easy for, say, someone in the industry to pick up one of these codes and use it.
+I'm also the lead on a software productivity project - I'm the Livermore lead on a software productivity project (Mike Heroux and Lois Curfman-McInnes from Sandia and Argonne are the main leads on this). But there's actually like an effort within this project, because it has to deliver software, to make the developers more productive.
+That involves things like putting training out there, familiarizing scientists and some of these computational scientists who develop the applications with how to build communities and how to write documentation, how to use source control well, how to have a release cycle, things like that.
+\[56:12\] I think it's kind of new territory in a lot of ways, because I don't think we've had a coordinated effort to build a software stack quite as large, so one of the things we've been talking about lately is "How do we coordinate releases of 15 applications in 80 different software projects?" That's not easy. And "Do we need a coordinated release cycle, or should we teach teams to release things on their own pace?" Those are things we're debating here.
+**Nadia Eghbal:** And just to clarify, that is an open source software stack...
+**Todd Gamblin:** So not all of it is open source. For example, some of our weapons simulations are included in the Exascale project, because they need to be able to run on Exascale machines, but those are not open source. Sorry.
+**Nadia Eghbal:** Phew!
+**Todd Gamblin:** Yeah, but in large part -- a lot of the science parts of it and a lot of the math libraries and the computer science infrastructure are open source; this is stuff that people could build on. I think it would be really awesome if we could build simulation frameworks and things that someone could come along and build an open source code on top of, because that helps both Livermore and other organizations out.
+Our code teams are sort of starting to think about that. We recently had an effort internally to look at how our simulations are structured, and I think they found that something like 40%-50% of the code is pure computer science - no physics, nothing sensitive - and that part could be factored out as like a general tool kit that we could use, and open sourcing that seems like a really good idea, because we could get contributors and we could help open science with the stuff that we're building.
+**Nadia Eghbal:** So how do you get industry contributors involved in situations like that? Do they come to the same sort of conferences that you do? Do you have to reach out to them individually?
+**Todd Gamblin:** Yeah, that's a problem. They don't always come to the same sorts of conferences. Industry HPC has always been -- it's something that people shoot for, but not a lot of companies really get into it. There are companies like Exxon that have HPC clusters and they kind of do their own specific simulations for things like oil wells, how the ground is structured, what's gonna happen when we drill here... And those are really big users of HPC. Aerospace also, the Boeings of the world...
+But one of the goals of this project would be to sort of expand the middle tier of HPC. It's always been sort of elusive. We'd like to have smaller companies able to use these kinds of resources, but I think the complexity of getting into supercomputing has been so high that they haven't necessarily jumped on board. But there are really good success stories for using HPC in industry...
+I think Procter & Gamble - I'm blanking on the guy's name, but he came out here and gave a talk about all the different ways that they use HPC. You've heard of Procter & Gamble I guess, but they have lots of companies under them that they run, so he talked about how they'd used simulations for everything from making the diaper production line more efficient, to making better scentless detergents, and things like this, or improving the way that detergents are mixed. So there's all these applications out there that people could take advantage of this stuff, but before we get industry contributors, we're gonna have to make it a lot easier.
+The other problem with getting industry contributors I think is that in many cases for something as complex as like a piece of simulation software, the industry folks really want someone to support it, they want someone to call and to say "We're having problems deploying this. What do we do?" I think for that we would have to look into starting small companies to support parts of this infrastructure, or to have maybe a support contract for like an Exxon or a P&G.
+**Nadia Eghbal:** \[01:00:06.16\] Well, we're gonna close out on that note... Do you have any final thoughts to share about lessons learned from your experiences open sourcing in a fairly (I'd say) comparably difficult context?
+**Todd Gamblin:** One of the things that I've learned from Spack I guess is that you really have to think about the broader context for the thing that you're building, and that means giving up some degree of control. I think there are a lot of projects where they've grown up in one lab and they've served that one team, and they could be more broadly useful, but making it easy for people to either jump in and contribute or to pick up the thing and use it for maybe something that the users didn't intend - it wasn't always one of the team's goals.
+For Spack we've tried to be really open about contributions; maybe not necessarily implementing things for people, but helping them to implement things and thinking outside of our own use case. That's one of the ways that I think we've gotten contributors and that's how I think we've been able to grow our community.
+**Mikeal Rogers:** I think that's a general open source problem... A lot of people who start projects don't wanna give up control, but they want people to work on it.
+**Todd Gamblin:** Yeah, I totally agree. And teaching people to do that... Actually, that's one thing I could say - a lot of our teams are scared of what will happen if they put something out there as open source and they start to get a lot of pull requests. They're like, "What do I do? I'm gonna have to support all these things? How do I deal with this?", so one thing we're trying to do here is teach people to deal with those contributions and to say "I don't have time to implement that, but we can help you, and we can maybe improve the documentation in that area and then you can contribute, and we'll have to figure out how to maintain that feature, because maybe we're not gonna use it." So just being open to things like that I think can go a long way.
+I don't have a silver bullet for "How do we maintain this feature you contributed after you contributed it?" I think that's the harder problem and you have to get creative.
+**Mikeal Rogers:** Definitely. Thanks for coming on.
+**Todd Gamblin:** Sure, thanks for having me. It's been great!