Roundtable Transcript, Part 2 (Continued)

Collaboration in Open Source Communities
[14:18] Phipps: But that's why open source communities are typically not very good at doing collaborative innovation between corporations. That's why you tend to discover that corporations go into back rooms and do work and then bring out into the open later is because a fundamental difference in the modus operandi of a group of individuals in the open source community and a group of corporate developers is that corporate developers typically are not working for themselves and the people they are working for tell them that there are competitive environments to be had. It seems to be inevitable that we need to have a space that works that way and, again, what we've seen Apache doing is Apache has gritted its teeth and learned to work within that environment. They're not desperately happy about having to do it, but they do it because they realize that if they don't do it, the corporate participants aren't going to come to the table.

[15:09] Mike Bechauf: But let's be clear, I mean, you make it sound like as if open source and corporate developers are mutually exclusive, and that's not the case. I think what is really important is to come up with an economic model where the calibration models of open source don't exclude later commercial success. Eclipse is, I think, a very good example of it, because there are rules, there are clear committees in place, in terms of how the releases are planned, it's predictable; there's a clear architecture in place, and projects can be planned in a way that the corporate developers have come together and basically shared a common vision but nevertheless, of course, add to the vision in a way that they can later, of course, yield commercial success. That's not mutually exclusive.

[16:03] Farrell: In all fairness, Eclipse wouldn't be what it is today without IBM. I'm not saying that's a bad thing; it's just that is a corporation that has a strategy backing an open source. Whereas, if you have a project in Apache, you're not going to get that same kind of [commitment].

[16:16] Milinkovich: I'd point out from the very central perspective of Eclipse, it's kind of unfortunate that IBM bought Rational, because Rational was in up to its eyeballs before IBM bought them. So we wouldn't be having this "it's an IBM thing" conversation if they hadn't gone and bought Rational.

[16:33] Elloy: Made us very happy, by the way.

[16:35] Milinkovich: Pardon me?

[16:35] Elloy: Made us very happy, by the way.

[16:38] Milinkovich: But there are lots of different corporations cooperating within Eclipse, and it's working very successfully.

[16:48] Bechauf: But it does require planning. You can't just send your guys there and just like, "Go, contribute." You have to be very clear about the scope of the project, what you do with the open source, and the intellectual property agreements that we have signed. There's planning involved. But as a collaboration model, it's nothing but fantastic.