The hardest part about open source isn't the code -- it's the community. Examples of this come up all the time, with Sun being one that has come up a good deal lately -- not just because of its acquisition of MySQL (which I'm still fairly positive about), but the way perceptions of its behavior can affect its acceptance. Even if you do the right thing, it needs to also look like you're doing the right thing.
The other day, longtime Linux kernel developer Theodore Ts'o wrote a post in his blog, "What Sun was trying to do with Open Solaris," in which he chided Sun for making the Open Solaris development and contribution process so difficult. In his purview, this wasn't a reflection that only the best code could make it into OpSol after careful peer review, but a sign of how obstinate Sun's corporate culture was.
I think it is fair to say that from a marketing point of view, the tactic [of taking Solaris open source] has been at least partially successful -- although as John [Plocher] has admitted, the goal of creating a full community with application developers, university students, and so on, hasn't materialized for Open Solaris. Sun has the dream; the Linux community is living it.
... I wonder if Sun will really be able to sustain their Solaris engineering team if they will really be doing all of the work themselves, and outside contributions continue at the rate of 0.6 patches per day. After all, the margins when you are selling low-cost AMD servers are much lower than when you are selling Über-expensive SPARC servers.
Ted believed that Sun was more interested in simply keeping Solaris relevant in the marketplace than in making it into a valuable open source project, and to him this was revealed in all sorts of ways -- like the glacial pace at which patches have been submitted and processed. After feedback from readers (some of whom were from Sun and who took exception to his characterization of things), he refined his argument to talk about open source development that was community-centric rather than company-centric, where the decisions were made by the needs of the community as a whole rather than a "pointy-haired boss". (To me this is that much more proof that Dilbert has done more to raise awareness of stupid corporate behavior than just about any other consciousness-raising efforts.)
One counterargument is that Sun is simply "doing things differently", which is not in itself a bad thing. Perhaps it is in Sun's interest to do things its way -- it's a hardware vendor, and as far as I can tell it sees its open source work as an extension of that business. But you can't build hardware and software the same way. What worked for Sun internally as a hardware company may just prove stifling and alienating to its open source participants in the long run. I'm just glad the MySQL team is, from what I can tell, mostly being left alone to do its thing.