Forgot your password?
typodupeerror
Open Source Programming

State of the GitHub: Chris Kelly Does the Numbers 34

Posted by timothy
from the git-along-little-hub dept.
I talked with Chris Kelly of GitHub at last week's LinuxCon about GitHub. He's got interesting things to say about the demographics and language choices on what has become in short order (just six years) one of the largest repositories of code in the world, and one with an increasingly sophisticated front-end, and several million users. Not all of the code on GitHub is open source, but the majority is -- handy, when that means an account is free as in beer, too. (And if you're reading on the beta or otherwise can't view the video below, here's the alternative video link.)

Timothy Lord: Chris, can you explain your role with GitHub?

Chris Kelly: Yes, so you know, I do all of our outreach work. So, it’s us going out into the world, talking to people and getting them involved in their communities and really trying to amplify the work that’s already being done in the OpenSource world and how GitHub can support and promote that.

Timothy Lord: GitHub started obviously with Git, talk a little bit about the progression of going from a particular organization tool to having how many repos right now?

Chris Kelly: We have about 18 million public repos currently and we’ve got 6 million active users with accounts, so it’s gotten pretty big. So Git is what, 10 years old next year and so we started with that. It was just – the founders were really are getting involved with Git and needed a solution for solving like sending patches around via email problem, which was kind of a nightmare and so started building GitHub for themselves and that’s how the company has built itself all the way up for the last six years is building the tools to solve the problems that we have and we are all software developers and so it’s grown that way.

Timothy Lord: And it’s not like there is a lot of advertising that goes around saying to use GitHub; GitHub has grown pretty organically?

Chris Kelly: Yeah, so a lot of it has to do with our outreach. Being at conferences, meeting people, and really understanding what their problems are and really just being supportive of the community and then I think that’s the best way to really engage with any kind of audience, any company that’s interested in engaging with the developers is go be developers and go be part of the community. You can’t sell to that audience, you can just be a part of that audience and that really works.

Timothy Lord: Talk about some of the big projects that are on GitHub right now and what sort of scope, we’ve got one aspect of the scope of the entire site is number of users, but talk about some projects and how many commits you see happening?

Chris Kelly: Oh, I’d have to look at some graphs for number of commits. But we have really large projects coming on to GitHub all the time. Microsoft just put Typesafe on to GitHub recently. We’ve got the Rails project, is one of a classic example that the number of commits spiked the second I think Rail moved onto GitHub and started doing the development cycle. So we’re finding that a lot of open source projects, very large ones are moving under GitHub and seeing a lot of adoption and engagement from the community because we think we make it so much easier to both communicate about the project as well as contribute back to it.

Timothy Lord: Talking about the actual development because obviously people use GitHub as their own tool for developing software of their own. What’s the development work like for the software that runs GitHub?

Chris Kelly: So we used GitHub to do everything basically. So it manages all components of the way we work, so between GitHub itself and our open source chat system Hubot that sits on top of Campfire and Slack and HipChat. We can do all things you have in the very open and so that’s what we built GitHub for, is to try to change companies to start working more and more open and so we do it in the open, every project and every product feature gets talked about really publicly and everyone can contribute to that conversation. Everything that happens in our office, the events we put on, all happen in GitHub repos. The planning of this conference and showing up here is done in the GitHub repo. We do literally everything, GitHub repos with issues and port requests, because we think it’s the right way to work and so we’re really kind of stressing the way GitHub can function by doing that and then going on to the world and talking to other companies and that want to change the way their development cycles go and so that’s what we really promote.

Timothy Lord: Now can people build their own GitHub, if somebody calls from a large company and says we want to have our own install?

Chris Kelly: Yeah, so GitHub has two products. We have GitHub.com which most people are familiar with and we have both open source projects and then a type of repos, large organizations use GitHub.com. But if you have any sort of compliance requirements that you want me to go through or need some security concerns, you can have GitHub itself inside your own internal using GitHub Enterprise. So that’s one of our major products. Really excited about like it’s advancement, it’s basically everything at GitHub.com is wrapped up in a VM and put into your own firewall, so developers can have that familiarity but also with the security of putting it behind a firewall.

Timothy Lord: It’s interesting that the ratio of users to projects is pretty extreme, talk about that, with some kinds of development you would actually think there would be more users than projects and now it’s the other way.

Chris Kelly: Yeah. So it depends on any number of scenarios, like Node is a very repo heavy community. They have lots of very small modules and so they all have individual repos, so one person could have 50 different repos all for their just Node ecosystem. But other communities, say the .NET and Java communities are working one single repo with many people contributing to those things. So, over the balance of things, you end up with what, three repos per active user, and people spin up repos all the time for just random projects, they want to experiment, they do forks, it just depends on whatever they want to work on. So I think the experimentation and the play that we allow with the open source repos and the fact that we give them away, really helps people kind of experiment and try it out long before they have to like

Timothy Lord: Anyone can clone and start one?

Chris Kelly: Yep, anyone can clone, mess with it, break it and see how it goes, and really get involved in the open source community very early, very easily.

Timothy Lord: What are the languages used by users, how does it break down, what are the most popular languages that we see being used by developers?

Chris Kelly: Yeah, on GitHub it marries with what the broader analyst industry will say, so JavaScript is the number one language on GitHub, a lot of that has to do with Node plus JavaScript is in many web applications, but beyond that we have – the Ruby community is very strong, Java is very strong, .NET is very strong and then you have everything all the way down to things in C and Go, that’s up and comer, so we see the broad trends of the industry playing out basically exactly the same with almost no change on GitHub. So people that are building things for their companies in these languages are building the same things in open source and we can see those as well, so.

Timothy Lord: Now, one thing, there are sites, SourceForge being one, but others too that are about more of the distribution of software, distinguish a little bit if you’re making a piece of software in GitHub, what things you might use in other’s company for service wise and what things can GitHub do that weren’t available in that same way five, ten years ago.

Chris Kelly: Yeah, so GitHub is really about we focus on the collaborational code and that’s about writing the software and getting it done. Beyond that things that we don’t do are like your Ci systems. So you might be using something like Travis or Circle Ci. So that when you push the new code, it will automatically run on a build system and see if tests are passing and then allow commits beyond that. We don’t focus as much on distribution, like you can use GitHub as a distribution mechanism for your libraries but we really want to encourage people to do the actual work and the actual software development on GitHub because of the openness and that promotes versus using it for ticketing systems and bug tracking systems GitHub issues is a great flexible way to solve a lot of that problem, so, yeah.

Timothy Lord: And if someone is using another development system but they want to move to GitHub, is that a simple process?

Chris Kelly: Yeah, so for SVN projects, for example, we have a one button like importer for getting SVN projects into GitHub, in fact every GitHub repo is a Git repo and an SVN repo, so if you’re trying to transition a team over to using Git, they can start by using GitHub and using SVN commands and we have a bridge that maps all of that over to Git, and then you can slowly get them over into the Git workflow. So it’s simple as it can possibly be. And we work on importing lots from all kinds of other source control systems with all our teams. We have a training team and a services team that really helps and our support team really helps and get anybody what wants to get on GitHub as smooth as possible.

Timothy Lord: And before we started the camera, I mentioned I kind of wanted to use this as a state of the GitHub.

Chris Kelly: Yeah.

Timothy Lord: Can you talk about any other statistics that people might find interesting about the growth that you’re experiencing, I mean you are up to getting close to 20 million projects right now?

Chris Kelly: Yeah, so we’re really excited about what’s happening in the large open source projects. GitHub started as a very small community driven projects where anybody like myself can just spin up a little gem. I am a Rubyist by trade, I can spin up a gem and get it on there and now we’re seeing large open source projects moving over to GitHub and really excited about what that means and that we’re getting more and more developers adopted into it. They are getting – if something, let’s say, the Linux Kernel ever happens to move over to GitHub if we get lucky one day to have that, like the number of contributors to the Linux Kernel, the number of contributors to any of those large projects, can really expand that into other companies and that’s our goal. It’s not just to change the open source world, but to make businesses and software developers inside of companies function like the open source world where the flexibility and like the tools that you are used to using are available inside your company and that’s what we’re really excited about. So companies like Google, Oracle, Twitter, all of them have kernel developers. Imagine if we got the Linux Kernel on to GitHub, how many more people would be using GitHub inside of their companies, other development teams will be using GitHub, and that’s really exciting.

This discussion has been archived. No new comments can be posted.

State of the GitHub: Chris Kelly Does the Numbers

Comments Filter:
  • by Anonymous Coward

    can we get a non video recap?

    • by i kan reed (749298) on Thursday August 28, 2014 @03:51PM (#47778041) Homepage Journal

      Hide/Show Transcript

      It does this:
      $("#sdtranscript").toggle();

      • by Ethan Bernard (2954293) on Thursday August 28, 2014 @09:02PM (#47780797)

        Timothy Lord: Chris, can you explain your role with GitHub?

        Chris Kelly: Yes, so you know, I do all of our outreach work. So, it’s us going out into the world, talking to people and getting them involved in their communities and really trying to amplify the work that’s already being done in the OpenSource world and how GitHub can support and promote that.

        Timothy Lord: GitHub started obviously with Git, talk a little bit about the progression of going from a particular organization tool to having how many repos right now?

        Chris Kelly: We have about 18 million public repos currently and we’ve got 6 million active users with accounts, so it’s gotten pretty big. So Git is what, 10 years old next year and so we started with that. It was just – the founders were really are getting involved with Git and needed a solution for solving like sending patches around via email problem, which was kind of a nightmare and so started building GitHub for themselves and that’s how the company has built itself all the way up for the last six years is building the tools to solve the problems that we have and we are all software developers and so it’s grown that way.

        Timothy Lord: And it’s not like there is a lot of advertising that goes around saying to use GitHub; GitHub has grown pretty organically?

        Chris Kelly: Yeah, so a lot of it has to do with our outreach. Being at conferences, meeting people, and really understanding what their problems are and really just being supportive of the community and then I think that’s the best way to really engage with any kind of audience, any company that’s interested in engaging with the developers is go be developers and go be part of the community. You can’t sell to that audience, you can just be a part of that audience and that really works.

        Timothy Lord: Talk about some of the big projects that are on GitHub right now and what sort of scope, we’ve got one aspect of the scope of the entire site is number of users, but talk about some projects and how many commits you see happening?

        Chris Kelly: Oh, I’d have to look at some graphs for number of commits. But we have really large projects coming on to GitHub all the time. Microsoft just put Typesafe on to GitHub recently. We’ve got the Rails project, is one of a classic example that the number of commits spiked the second I think Rail moved onto GitHub and started doing the development cycle. So we’re finding that a lot of open source projects, very large ones are moving under GitHub and seeing a lot of adoption and engagement from the community because we think we make it so much easier to both communicate about the project as well as contribute back to it.

        Timothy Lord: Talking about the actual development because obviously people use GitHub as their own tool for developing software of their own. What’s the development work like for the software that runs GitHub?

        Chris Kelly: So we used GitHub to do everything basically. So it manages all components of the way we work, so between GitHub itself and our open source chat system Hubot that sits on top of Campfire and Slack and HipChat. We can do all things you have in the very open and so that’s what we built GitHub for, is to try to change companies to start working more and more open and so we do it in the open, every project and every product feature gets talked about really publicly and everyone can contribute to that conversation. Everything that happens in our office, the events we put on, all happen in GitHub repos. The planning of this conference and showing up here is done in the GitHub repo. We do literally everything, GitHub repos with issues and port requests, because we think it’s the right way to work and so we’re really kind of stressing the way GitHub can function by doing that and then going on to the world and talking to other companies and that want to change the way their development cycles go and so that’s wha

  • Thankfully... (Score:2, Informative)

    by dnebin (594347)
    I'm not using beta, cuz I don't want to see live videos in the news feed, in case anyone @slashdot/@dice even cares...
  • by Eric Smith (4379) <<moc.ahahuorb> <ta> <cire>> on Thursday August 28, 2014 @04:57PM (#47778803) Homepage Journal

    Not all of the code on GitHub is open source, but the majority is -- handy, when that means an account is free as in beer, too.

    I'm not privy to any details of GitHub's finances or business model, but most likely it's a good thing that there are non-open-source projects using GitHub, because that's probably what's paying for the free open source use. I've recommended to several clients developing proprietary software the use of GitHub rather than running their own in-house repositories, because the interface is easier for them to use and they don't need as much in-house expertise to manage things. Because Git is distributed, they could of course do both, or easily transition away from GitHub later, and that's a selling point.

    • by m00sh (2538182)

      Not all of the code on GitHub is open source, but the majority is -- handy, when that means an account is free as in beer, too.

      I'm not privy to any details of GitHub's finances or business model, but most likely it's a good thing that there are non-open-source projects using GitHub, because that's probably what's paying for the free open source use. I've recommended to several clients developing proprietary software the use of GitHub rather than running their own in-house repositories, because the interface is easier for them to use and they don't need as much in-house expertise to manage things. Because Git is distributed, they could of course do both, or easily transition away from GitHub later, and that's a selling point.

      Wasn't the whole point of git to not have central servers and such? That you could use a directory or any other source as a repo instead of centralized server repos.

      • No, it was that any of the clones would have the full version history of the entire source tree. You still want a primary repo from which you make your releases, but all the developers can have a copy of the whole thing. They can also make changes to their local clones when offline, then merge all the changes later. Anyone can branch from any point, and the whole thing works on snapshots of the repo instead of sets of incremental changes to each file.
      • by Anonymous Coward

        A central "master server" is stilll useful from an organizational point of view.

        The point behind git is that it's not special for any reason other than organization. In case of a problem (technical, political, maintainer disappears, whatever), there's no technical factor preventing you from failing over to another one very easily.

        In all technical respects, the repository on my laptop is just as good as the central server. I don't have to be on line to examine history or make commits.

        This also facili

  • by turp182 (1020263) on Thursday August 28, 2014 @05:52PM (#47779303) Journal

    Really? What are the stats on Classic vs. Beta pageview counts?

  • by Anonymous Coward

    And if you're reading on the beta or otherwise can't view the video below

    No video clutter - finally, now I can say something good about Beta.

  • Yeah, so (Score:2, Funny)

    by dbc (135354)

    Yeah, so, I start almost everything I say with "Yeah, so,..."

  • GitHub has become a facebook.com alternative for the 1337 haxxor set and alot of people use it for free hosting to put up a personal site

    that's my experience anyway...the idea is great, a website that hosts code for coding projects...but the whole abstraction layer of calling it a 'Git' still irks me...it's not a 'git' it's a computer file that contains code...

    any frequent uses of GitHub care to comment? what does /. think?

    • by m00sh (2538182)

      GitHub has become a facebook.com alternative for the 1337 haxxor set and alot of people use it for free hosting to put up a personal site

      that's my experience anyway...the idea is great, a website that hosts code for coding projects...but the whole abstraction layer of calling it a 'Git' still irks me...it's not a 'git' it's a computer file that contains code...

      any frequent uses of GitHub care to comment? what does /. think?

      No the git comes from the fact that Linus is a git.

    • i have no idea what you're talking about, i download code off of it and use it, it's a hub of git repositories... am i missing something?
      • ok...thanks for the reply...

        download code off of it and use it, it's a hub of git repositories

        now we can converse...see...what is a 'git'...as in GitHub...it's a "hub" for "gits"...

        this is a 'git': http://en.wikipedia.org/wiki/G... [wikipedia.org]

        it's is a software program that is used as **version control for making other software programs**

        the version control system is not the program/code you are making...they are two separtate things completely...theoretically a text editor, discipline, and good filenames is all you ne

        • what's software? is it soft? does it "ware" out. what's a program? am i watching TV? you're just bitching because you don't like the name, it's just as clear as the naming process for a billion other concepts. i hate the idea that you're down on people for hosting their person projects on there, what's your problem with people having a good place to host their stuff? i use projects hosted on there pretty much every day, do you work at sourceforge or something?
          • ...but that's not the main point...my main question is about if others think GitHub is the facebook for haxxorz

            i put my point in the headline so you wouldn't miss it...and copied above the text from my previous comment (which you responded to)

            my question...this whole time...has been if others think GitHub is becoming a defacto social network

Everyone can be taught to sculpt: Michelangelo would have had to be taught how not to. So it is with the great programmers.

Working...