text stringlengths 0 1.28k |
|---|
**Coby Chapple:** Yeah. The crazy thing about London is coming to terms with the fact that even if you had a full lifetime, you could never see it all. It's just so big and there's so much history, and it's constantly changing; you never can see it all. Once you do accept though and you relax a little bit, it can kind ... |
**Jerod Santo:** Cool. So you're a product designer at GitHub - tell me about that. |
**Coby Chapple:** Sure, so a product designer at GitHub basically means I'm focused a lot on the interactions and the workflows within the products; we have a fairly well-established visual style, so most of my work isn't visual design as such. There's some of that, but most of it is looking at the workflows that we ha... |
**Jerod Santo:** What's the biggest single change that you've made, that has had the most impact across the product? Even if it's just a refinement. |
**Coby Chapple:** Probably the thing that I've worked on directly that I think has had the biggest impact was a lot of the workflows around doing web-based edits to files on GitHub. From the very early days at GitHub, from very early on in the product, it was possible to edit files in the web, but you couldn't create t... |
**Jerod Santo:** Right, it was very limited. |
**Coby Chapple:** You could only edit them. If you wanted to create a file, you had to create it in your editor locally, commit it and push it, then you could edit it online. So we had Ace just embedded in the website for a long time. So one of the things that I worked on a while ago was making it so that you could cre... |
The other thing following on from that workflow-wise was the quick pull request feature. So it's one thing to be able to be on a branch and quickly create a file, but that's a direct commit, and if you do that just directly on master, there's no review, there's no cycle. But if you wanna make a quick change and also ha... |
That's the kind of stuff that I like trying to identify in a product - a shortcut to an existing process that otherwise would be difficult to do or not obvious that you should have this reviewed, but actually making it possible for people to opt in easily to a process that should happen. |
**Jerod Santo:** \[04:12\] I love when a product that you use or a website that you use a lot slowly gets better -- maybe even quickly gets better, it's better than slowly, but you just notice over time... You're talking about shortcuts, and as a user, because you have such a technical product and people use it all day... |
**Coby Chapple:** It is satisfying, and that's one of the reasons I enjoy product design. Before GitHub, I worked for a long time as a freelancer, so I was still doing design work. Sometimes it was on products, but at the end of the day it was client-focused work, and I got sick of that. I kind of realized that what I ... |
**Jerod Santo:** A single thing. |
**Coby Chapple:** ...not necessarily a complete product, something as giant as GitHub, because there's so many different parts of it, but what I like is the responsibility for the things that I build. If I build something, I like being able to see how that's used over time, use it myself, and if something isn't as good... |
**Jerod Santo:** I'm not sure if you mentioned on air or not, but you mentioned that the aesthetic, the feature set, the general interface had been established for a long time, and that you're working on specific workflows, specific aspects, refinements... As a product person, do you strongly differentiate between inte... |
**Coby Chapple:** That's a good question. I think that I don't get too caught up in the labels; I think a better question to ask is what do you actually do on a day-to-day basis. |
**Jerod Santo:** Okay, let me ask you a question... What do you actually do on a day-to-day basis? |
**Coby Chapple:** So there are some people who class themselves as UX designers that do a lot of coding; there are some people that class themselves as UI designers that don't touch code at all. And I think it doesn't matter. If you care about someone's experience, you might design UI. If you care about interfaces, you... |
**Jerod Santo:** Why do we give each other these labels? |
**Coby Chapple:** I don't care as much about the labels, I just look at "Okay, here's a product, here is something that could be better about the product. What physically needs to happen on it...? Who needs to sit down at the computer and talk to who else? How much brainstorming do they need to do? Do they need to just... |
**Jerod Santo:** Ultimately, isn't it all user experience? |
**Coby Chapple:** Yeah, at the end of the day you're creating a product. It's probably a business, it's probably for achieving a certain set of tasks... If they could do that better, then cool, make the change. |
**Jerod Santo:** As Steve Jobs said, "The design IS how it works", it's not like one versus the other. Form follows function... |
**Coby Chapple:** It doesn't matter what the job title is of the person who made the change, what matters is that the change needs to happen. |
**Jerod Santo:** \[07:53\] Yeah. So you guys have been doing refinements for a long time - those small shortcuts, those tweaks... I love all that, but recently at GitHub Universe you had some major new features. One that I love is the code review feature... Mostly because now I can create a bunch of comments and send a... |
**Coby Chapple:** You're drafting... |
**Jerod Santo:** Yeah, grammar, and we're helping and we're giving advice and thoughts... So I may review a 3,000-word document for somebody, and I may send them 17 emails to do so. And I actually apologize afterwards. Like, "I apologize that you just got 17 emails, but I have line notes", you know? So that ability is ... |
The other big one is the Projects feature. Tell me about both of those in summary first. |
**Coby Chapple:** Sure. With the code review thing, we have the exact same problem ourselves. For a lot of the blog posts you write or press releases or internal documentation, we use pull requests ourselves and we have the exact same problem of like, okay, not only do we just get a lot of emails, but sometimes the fir... |
**Jerod Santo:** Right. |
**Coby Chapple:** That's kind of one of the goals we had, because we had this problem and that's what we wanted to solve. |
**Jerod Santo:** One thing about that problem is like, even though it annoyed me, I was empathetic to it because I knew that it was hard to solve. There's a lot that goes into actually batching that up and providing it as a single thing... The user interface itself has a lot of concerns, there's a lot of ways you could... |
**Coby Chapple:** The reason it's complicated is as soon as you go from, okay, "Here is a pull request that you can comment on", to "Here is a process that you can go into", it creates this state where something can be unfinished, but still saved in the product, and that creates a whole set of interface problems of lik... |
**Jerod Santo:** Do you remember what happens in that case? This hasn't happened to me, so I don't know. |
**Coby Chapple:** Right, so I haven't used that part of the product enough and had things go wrong yet to kind of be able to say... |
**Jerod Santo:** I haven't, either... There's so many things to think about. |
**Coby Chapple:** Yeah, there's a lot to think about, and it's a technical product to start with. That's one of the challenges of being a product designer: "How do you take something that is deliberately technical, deliberately complex - because it's a technical product for technical people - and make it approachable?"... |
**Jerod Santo:** Why not? |
**Coby Chapple:** We don't wanna dumb it down, because the people that use our product are highly intelligent people, doing complex work that requires... |
**Jerod Santo:** Nuance... |
**Coby Chapple:** Nuance, and a lot of... |
**Jerod Santo:** Faculties...? |
**Coby Chapple:** They might wanna use lots of different tools at different times, so you don't wanna remove it and say, "Just use this hammer for everything." There should be a lot of options to use, and it's a case of making them approachable and... |
**Jerod Santo:** Usable? \[laughs\] |
**Coby Chapple:** Approachable is probably the best word for it, because it can still be very complex and it can still be very technical... |
**Jerod Santo:** What about "discoverable"? I think that's maybe a little bit different angle, but... |
**Coby Chapple:** Yeah, discoverability is a huge thing, and that's a big pet peeve of mine I have... |
**Jerod Santo:** Let's hear it. |
**Coby Chapple:** A pet peeve of mine that a lot of people do in product design or in interaction design is things that show up on hover. There's a bunch of things that I personally get -- |
**Jerod Santo:** \[12:01\] Because people will never know that that exists. |
**Coby Chapple:** The analogy that I use to explain how frustrating this can be is imagine you're in a room where the doors only show up when you're close enough to them. \[laughter\] How do you get out? |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.