Google and Open Source

In May, Google held its annual IO Conference for developers  in San Francisco. During the conference, the Austrian technology site, derStandard.at, had an interesting interview with Chris di Bona, Google’s manager of Open Source.  (Most of the site is in German, but the interview itself is in English.)  I have mentioned some aspects of Google’s involvement with open source here before, but the interview gives some additional insight into how pervasive open source really is at the Googleplex.

The Chrome web browser, and the Android and Chrome operating systems (both derived, in part, from Linux), are probably the best-known of Google’s open-source projects, but there are many others, as di Bona points out:

We have released something like 1,300 open source projects to the outside world in the last five years. That amounts to 24-25 million lines of code, using a variety of licenses.

Asked specifically about where Linux is used within Google, di Bona said:

Everywhere. Every production machine / server inside of Google is running Linux, Android of course, lots of desktops.

He goes on to say that engineering desktop machines overwhelmingly run Linux (Google engineers can in most cases use what they want).  Mobile devices are perhaps 70% Mac OS X (itself a UNIX derivative), with most of the rest Linux.  There is a very small population of Windows users.  (Google, as a software developer, of course needs some Windows machines for testing.)   He also described the way the internal networks for engineering are set up [emphasis added]

We have our own Ubuntu derivative called “Goobuntu” internally for that, integrating with our network – we run all our the home directories from a file server – and with some extra tools already built-in for developers.

I was struck by this, because the idea of having all home directories (user files) on a file server is one that we used with Sun UNIX workstations for securities trading 20 years ago.  (I mentioned this in an early post on Chrome OS.)   Doing it this way — we referred to it as having “dataless” workstations, with only the OS,  X Window System binaries, and the swap space, on the local disk — had several advantages:

• The only files that needed regular backups resided on a file server, which was under IT Operations’ control
• The only files with internal, possibly sensitive data, were on a file server, with physical and network security
• A faulty workstation could be replaced very rapidly with a pre-built spare, getting the user back in business quickly
• All user machines were built with a standard configuration, making the setup of a new machine a routine exercise.

Sun was also a proponent of this approach.

Mr. di Bona also discusses some of the differences in the way that the releases of the Chrome OS and Android are handled.  Chrome OS releases, including source code, are public as soon as the code changes are officially accepted, or committed.  Android has a schedule of periodic releases, which di Bona explains is due to the differences in the mobile device market.

If you look at Android we have lots of partners. We have chipset partners, we have handset partners, we have carrier partners. They all want to use Android and they all want to have something special about themselves.

Coordinating all these players takes more time.

Finally, the interview touches on some interesting questions about the future of the Chrome OS project and Android, and their market acceptance.

The really big question here is, will people accept the Linux desktop that looks like a ChromeOS machine, will they accept a Linux desktop that looks like Android? And if the answer is yes – and I think it is actually – then the Linux desktop will grow to be quite popular. But I don’t think the “classic” Linux desktop will ever be as popular as Mac OS X or Windows.

Working in technology for years you realize quickly how insecure most peoples machines are, how compromised they are, how compromised servers are. And I know when I use a ChromeOS machine that I don’t have to worry about this anymore, because it’s actually very very difficult for it to get compromised.

I think it’s quite possible that some security-conscious organizations will find the Chrome OS or Android model quite attractive, for at least some of their users, and especially for mobile devices.  The average user is not really able to be a competent systems administrator, and I don’t expect that to change; the user’s job, after all, is to do his or her job, not to be an amateur IT person.