A curious insight has come from all the recent talk about MySQL / Sun / Oracle. People talk about a community around a given open source product, but there's at least as much talk about a team within it. Let's not neglect the importance of either of those things.

Some of this thinking arose from talk about some of the gloomier prospects for MySQL, speculation which goes as far back as Sun's purchase of the company. What good is the software without the team of people who built it? It's difficult to replace the synergy you get from such people (and I hate that word, but it's really the only one that fits) with folks who have been largely on the outside looking in.

Likewise, you also can't replace the outside community of users and contributors with an inside team. The perspective you get from the community often isn't anything like what you get from the core developers -- and that's precisely why it's so valuable. Each sees things the other doesn't.

And yet ... I get the impression there are people who would just as soon find a way to live without one or the other. Unrealistic. "Delusional" might be just as good a word. Few programs of scale can be written effectively in a wholly decentralized fashion. Look at all the major open source success stories, from the Linux kernel on out: all of them needed a strongly organized core development team to get where they are now. That sort of discipline is no accident.

It's also rarely possible for such work to get the street cred it needs without people on the outside, contributing back. I'll venture further and say that such contributions don't always have to be code, although that's the default form of contribution. In Firefox's case, you have the whole culture of add-ons and plugins, the vast majority of which were authored by other people.

That's why whenever I hear people dismissing the fate of MySQL with a shrug and lines like "It's open source, they can do what they want with it," I get antsy. Mere access to the source isn't a substitute for people who have made that project into their mission. And downplaying dev teams in favor of code-in-the-abstract is a great way to kill a perfectly good project.

InformationWeek has published an in-depth report on new software models. Download the report here (registration required).

Twitter: Me | InformationWeek
Facebook: InformationWeek