Do Read/Write Web
A few weeks ago, I had the opportunity to hear Chris DiBona, the Open Source Program Manager at Google, speak at a TiE Pittsburgh event.
At the event, Chris provided an extremely insightful presentation on the state of the open source market. After the event, Chris and I exchanged emails and he agreed to do a interview for Read/WriteWeb.
Before joining Google, Chris was the editor of Slashdot. You can also read more of his thoughts on open source and life at Google on his personal blog and listen to his podcast with Leo Laporte, FLOSS Weekly.
I tried to cover three major themes in my questions to Chris:
How Google Leverages Open Source
How Google Contributes to Open Source
General Questions about the Open Source Software Marketplace
I decided to provide the entire interview in this post rather than only highlight specific responses. Part of this is due to the fact that I found it all very interesting (even the questions where Chris wasn't able/willing to respond.) I do end the post with a few of my thoughts in the conclusion, but even more I'd love to leverage the Read/WriteWeb audience's collective intelligence - please leave your thoughts and observations in the comments below .
How Google products leverage Open Source
Q: What is the review process before an open source component is included in a Google product?
A: We have a variety of tools that we have written to ensure compliance with the licenses that are used by open source projects, and we code review incoming programs and libraries.
Q: How many open source projects does Google currently leverage across all Google applications?
A: Sorry, we don't talk numbers. Many.
Q: What open source projects are leveraged most deeply?
A: There are two projects that we'd miss more than any others. The first is the Linux kernel and the second is MySQL. Both are in heavy use at the company.
How Google contributes to Open Source
Q: What is the process Google goes through to decide what projects are released as open source?
A: The primary issue is if we have the engineering staff to work on the problem. We have a variety of engineers donating their 20% time to open sourcing software that they'd like to see released from the company. There are some other short legal reviews around trademarks and patents, as well.
Q: (Since it is that time of year) What ‘summer of code’ project last year were you most proud of?
A: I can't pick just one! I was most proud that we were able to serve so many students in so
many countries. It was quite a feat to run a truly global student program.
Q: What other initiatives (beyond summer of code) do you have to support open source projects?
A: We donate a fair amount of money and time to projects including places like the OSU Open Source Labs, the Apache Projects, the EFF, creative commons and smaller efforts like MusicBrainz. We also employ a variety of developers who work on external projects as their job. For instance we employ Linux second in command Andrew Morton, Samba co-developer Jeremy Allison, Guido Van Rossum of python and so many more.
Miscellaneous "Open Source Marketplace" questions
Q: In your opinion, where in the software stack will there be the most new Open Source projects
started in the next year?
A: I don't have a considered opinion for this question. I think the voip stuff will continue to
develop and be quite exciting.
Q: Are their any commercial applications, which you believe are especially vulnerable to an open source competitor emerging?
A: I think that vulnerable is the wrong word. Here's how I think of it: if a commercial entity finds that they can no longer sell their software because a viable open source project has risen to displace them, they really have to decide if they want to start selling services around the open source offerings or get out of that particular line of business. IBM has done a terrific job of the former.
Q: What is Google’s view of the different Open Source Licenses?
A: We think there are too many of them, and that is why we support only a subset on our project hosting system on code.google.com. This subset includes what we think are the most important licenses, including the BSD, Apache, GPL and other licenses. That mirrors how we feel about them internally. I like to think that we have a very nuanced view of how licenses should be accommodated internally.
Q: What percentage of current GPL projects do you think will move to GPL v3? What do you
anticipate the effect of this being?
A: Well, all of the FSF projects will switch, as will Samba. Others will switch over time if the GPLv3 drafts continue in the direction they seem to be heading and thus the effect will be pretty minor. That said, anytime a license revs, there will be a period of confusion while people expand their thinking to include the new license.
Q: Seeing as most of Google’s products are browser-based, and given Firefox 3 will be an ‘information broker’ and will support offline web apps, Google’s relationship to Mozilla seems crucial over the next few years. So how closely is Google working with Mozilla on these developments?
A: I don't handle the Mozilla relationship, so I'm going to pass on this one. We are happy to have them involved with the Summer of Code, for sure!
In John Battelle's interview with Eric Schmidt at last week's Web 2.0 Expo, Eric talked about their being four 'major thrusts' around Google (see interview on YouTube - four major thrusts reference starts @ 7:28) The first of these being "building the world's most interesting super computers that are running new data services, application services, search services and so forth." Note: the other three 'thrusts' were end user services, advertising, and embracing their unique culture.
While Google has not open sourced the page rank algorithm and other pieces of their systems, it was interesting for me to look at Google's commitment to Open Source as it relates to building these supercomputers. There is little doubt that if any company could attempt to recreate applications like MySQL and Linux it is Google. However, they have instead chosen to embrace and support the various open source communities and leverage their existing application tools.
Additionally, I find Chris' observations about IBM "selling services around the open source offerings" particularly insightful. It is an interesting extension of some of the topics discussed at the Web 2.0 Expo on open source business models.
Finally, I'd like to thank Chris for taking the time to answer my questions. I know he is quite busy and I hope you found the interview as interesting as I did.