Interview: Chris DiBona, Google
Trevan McGee talks to Google's open source and public sector programs manager, Chris DiBona.
Chris Dibona has worked within the open source community years before joining Google. As writer and editor for Slashdot and the coeditor of essay collections "Open Sources" and "Open Sources 2.0." Now, as the open source and public sector manager at Google, he oversees the company's open source endeavors, supervises the distribution of resources to open source projects and generally loves what he does. Trevan McGee had an opportunity to sit down with Mr. DiBona during OSCON 2010 where they discussed Go, Android, and the future of open source technologies.
Trevan McGee: What are you up to? What is Google up to?
Chris DiBona: Well my role at Google is looking after open source and that's sort of broadly defined in a lot of ways as making sure that open source remains healthy, making sure that it remains as vital resource for us, looking after open source infrastructure, and we do that by both just providing it on code.google.com, where we host hundreds of thousands of projects and also by funding groups like the Oregon Open Source Labs, which hosts Mozilla and so many other projects. We also help pay for things like kernel.org and their machinery and admin staff, so that projects that matter to us, projects in general remain quite healthy. That's the infrastructure mission and then there's releasing code from Google because we feel that it's nice to be a resourceful friend, it's better to give code, though. We have lots of Googlers patching all the time into hundreds of projects. We probably patch two-three hundred projects a month now. We release 2-5 projects a week and that's not counting large projects like Chrome and Android. We've released a little over 20 million lines of code into the outside world and it's pretty exciting.
The third mission is creating more open source developers, whether by enabling Google engineers to become those people or by creating new ones from scratch with Summer of Code and projects like it. And in that, we also maintain compliance for all of the open source licenses used in the company. Every build, every product. We just make sure that every product is within and respecting those licenses that are so important to us. If you screw that up, everything else is hard to do.
I have a team. Folks like Danny Berlin and others who are very, very sharp about these licenses and their implications and we help out Google.
TM: When you mentioned allocating resources whether it's time or money to projects that you think matter, how does Google determine which projects matter? Like the Oregon Open Source Lab for example.
CD: That's just me and my group, we make those decisions. We look at what needs help and who can invest in it. It's actually really, really hard to convert money into actual code and actual useful resource within the open source world, so when you find them, like the Oregon Open Source Lab and like The Linux Foundation and Kernel.org, it matters. You try to help them out because they're actually doing important work and they need help to do it. That's strictly financial help and we do that.
As far as like, “How do we encourage Google to open source things,” it's part of the product direction. In the case of Android, Chrome, Wave, Chrome OS –– those need to be open source in order to be successful. The smaller projects: The engineers want to do it, we make it easy for them, so they do it.
As an engineer at Google, you can in about three days get authorization from my group and thus from legal, trademark, patents and all the rest and have a piece of software out in three days. If it's ready to go, it's ready to go.
It usually take the engineers a little bit longer to clean it up to where they want it to be, so that it's buildable outside of Google and all the rest. We have a philosophy of getting out of people's way and keeping people out of your way and we sort of act as your linemen in some ways to keep people out of your way, so that you can release.
The same thing goes with patches. We look at the first couple of patches and what it might to a project and then we tell them we trust them, so we have people patching into projects all the time, they let know about it, but we stop being gatekeepers after a little while and thus, code flows. And it comes very fast, so you know, we've released more than 20 million lines of code, that's more than anyone's done with these kinds of programs, so that's pretty exciting.
I see myself as being actually a very effective bureaucrat, I know that sounds like a strange thing to brag about, but I'm really good at keeping everything moving very smoothly, well oiled machinery for getting code out.
TM: You mentioned licensing as well and there was an issue with the WebM license when it was released. Did you and your crew have a hand in that?
CD: Oh sure. We felt that it would be incompatible with the GPL, it would need some tweaking, but that it was fine to launch and then do some tweaking. That's what happened.
We wanted to articulate to the world that patent policies matter and so we did that. Now we have something that's a little more effective, I think, because it's broadly combinable with GPLV2, V3, and other projects. Especially Apache projects.
TM: Are you concerned about claims from people like Steve Jobs or members of H.264 that WebM is within patent violation?
CD: I honestly don't care about Steve Jobs' opinion on codecs. I do care about H.264 and the rest, but they have actually not been so bad. Here's the thing: H.264 is a cartel, it's a cartel of a lot of companies, so they have to get through that cartel. They collect usage fees from their partners. There's no reason to think that those are going to stop, first of all. If they feel threatened by WebM, they would have to go to the cartel and say, “What do you think?”.
We feel that one, we're very confident about the state of WebM, but second of all, we're also very confident that the members of that cartel will see that WebM is a net good for everybody. I actually don't think it's going to hurt them in the slightest, so I'm not worried about it.
TM: What are some other projects you have coming up or initiatives that you're excited about?
CD: Too many. You mean actually at Google or in general? [laughs]
TM: At Google.
CD: Well, actually the WebM thing is the most important thing we're doing, long term. In the short term, Chrome and Android. Android just blows my mind. If you'd told me we'd be shipping 160,000 phones a day when we started this five years ago, I would just be like, “Yeah, whatever.” It's pretty shocking, how successful it's been. It's Android, Android, Android, and then some Chrome. And we have Chrome OS officially launching in October and then GoogleTV launching shortly after that, so it's a pretty exciting time for open source at Google. It's part of everything we're doing in our client strategy. It's cool.
TM: When you joined Google did you imagine it would phones and TV set-tops?
CD: No. I didn't. There's no way. I would have thought they were crazy, I still do.
I had been at Google for about a year when we bought Android. It was a very small company of about five people when we bought it and I was like, “We're buying a cell phone operating system company? That's crazy. But they want to be open source, so I'm happy.”
And so I got to meet the team and they were just brilliant. And Andy Rubin and Brian and those guys grew that team huge so that they could ship a really good phone and it's been awesome to watch. Also, if you look at the different spaces where we operate in, whether it's browsers or phone operating systems and tablet operating systems and Web operating systems like Chrome OS it like, “wow.” It's pretty surprising to do all the things that we're doing. As an open source guy it's pretty great to see it all come out as open source and not locked up and I don't think they would have had nearly the reaction if they weren't released as open source.
TM: You've been involved in open source for a long time now. What's next? What are some of the challenges you see on the horizon?
CD: When I joined Google I decided, “I've been doing open source for awhile, let's see what else is out there.” And so I sort of helped with the Blogger team and how they interacted with the outside world and I found out that I really like the open source world. Blogging was a much bigger representative sample of humanity, so you had all kinds of people. Well one thing I noticed about open source was that the people generally have these extremely admirable ideas around moving computer science forward, moving each other forward, and sort of having a shared purpose. And that's something I feel is kind of missing normally in the computer business and it's something I see at Google and it's something I see in open source.
I still see that. I still see people who believe in technology as a way of moving society forward and I think that's pretty exciting and I think that open source is sort of the purest substantiation of that. There's always ebbs and flows of different technologies that are always exciting to talk about whether its CMSes or Kernels or database systems or whatever. I think we're going to be in a lull for kernels for a little while before new and interesting things start happening there.
I think we're in a fertile time for databases, but it won't be noticed by the outside world for another couple of years. I think it's a fertile time for planting database seeds. And you're seeing that with Drizzle and the rest and I think you're going to see the fruits of that probably in a about a year.
CMSes are having a crisis of conscience trying to decide who they want to be. Part of what happens when CMSes get really popular is they build up these huge stables of plugins that are married to specific versions and then they have trouble moving forward, so what do they do? You're seeing that now with Drupal and Joomla's been sort of dealing with that too, but they're not as popular, so it's not that hard for them.
WordPress, they're seeing the kind of pressures, but in a different way. A lot of their customers are very marquee customers, so they're seeing more security attacks, they're seeing that kind of thing. And so they're having to reassess how they do security and plugins and that sort of thing and so you're seeing some anxiety there, but it's really good because we'll come out of it healthier.
It's a very fertile time for languages. I think that Go is great. Being able to help launch that, it's just a huge treat.
TM: Have you played around with it much?
CD: A little bit. I've done “hello world”-type stuff, some simple server stuff, but nothing I'd consider awesome work, worthy of discussion. But it's pretty interesting. I've built some play Android programs too. The mobile space is extremely interesting for open source and a place where we can have a foothold that will never go away, which I think will be really good for consumers and really good for the world. Period.
Interview: Chris DiBona, GoogleTrevan McGee talks to Google's open source and public sector programs manager, Chris DiBona. Those that are convinced that the headline of this content page is informative, probabilities are you are thinking about using Linux as your number 1 or just an optional Operating-system if you find yourself basically done with your old OS. Even though the majority assume that Linux is too challenging that it is of interest only to geeks, Linux-magazine.com will be able to confirm that this incredible os is in reality for most people. It is actually attention-grabbing to learn about Interview: Chris DiBona, Google.
H.264It is written correctly as H.264. Not H2.64.
Linux Foundation's big event celebrates the 25th anniversary of Linux
Linux has evolved from a “won’t be a professional” project to one of the most professional software projects in the history of computers.
Competitors get in the game with RHEL without Red Hat
Security researchers have already notified Microsoft; some fixes are available
The company is collaborating with Google and Intel to use Kubernetes as an engine for Fuel
Customers can take a free test drive of SLES for HPC on the Azure Cloud
San Francisco-based chip company announces their first fully open source chip platform.
The whole distro gets rebuilt on glibc 2.3
Ubuntu Vendor tries to solve app packaging and distribution problem across distributions.