Lots of people have written in to ask my opinion of the Oracle/InnoDB acquisition – and one of our contacts has written in with some very interesting insights that unfortunately must be kept private, but frankly there are better sources to turn to for information on that subject. Start with Jeremy for some of the more worrisome implications for MySQL, and then head over to Zack’s entry on the topic to get the MySQL side of the story. For good measure, here’s an entry from the Orablogs feed. The only thing I’d add to those bits is that I’ve heard from a couple of LAMP types that found the news rather depressing, which is obviously not what MySQL had anticipated with the release of 5.0.
Overall, I think the actual implications of this deal have yet to play out. Short term, I’d say it’s mostly bad for MySQL, as the FUD is spreading fast and furious at the moment. The longer term implications are less clear, because a.) InnoDB isn’t the only storage manager out there, nor even the fastest (though I’m told its the most functional), and b.) how the MySQL/Oracle relationship will play out is not as predetermined as some have concluded. So my take is: let’s wait and see. MySQL’s got ubiquity on its side – you know, the whole LAMP thing? – and therefore is unlikely to be going anywhere for the moment. As a result, reports of their death are, to paraphrase one Samuel Clemens, a bit of an exaggeration. IMHO, of course.
Anyhow, the logistics of this deal are useful for exploring an important issue not specific to the three companies involved, however, which is the question of value: is it the people or the source code? The answer, in my mind, is both. But it seems that not everyone shares that opinion; perhaps the number one suggested response to to the InnoDB/MySQL question was this – “Just fork it.” Besides the fact that a forked codebase couldn’t be dual licensed as before, given the fact that original IP wouldn’t be owned by a single entity, this argument to me ignores the value in the people who originally developed that source code.
The source code is hugely important, obviously. And the protections that the GPL affords that code are, as MySQL has hastened to point out, a guarantee that the rug won’t be pulled out from under you overnight. But the people, to me, are still important. When people talk about the potential acquisition of 37Signals, for example, do you think it’s for the code to Basecamp or Writeboard? Or to get access to the minds of people like David Heinemeier Hansson? My money’s on the latter, meaning no disrespect to those products.
The fact is that the value of any given codebase is tied directly to the people that support it. Like the betamax, the best technology is no guarantee of success – the best community is. You could see this at work in the recent Mambo/Joomla split. While I think it’s too early to call the winner there, given that the forked Joomla has, in Optaros’ Seth Gottlieb’s words, “the buzz and the passion, and (I would say), the leadership,” I’d be reluctant to invest a whole lot in Mambo.
Does this mean that InnoDB can’t be forked successfully if it comes to that? Nope. It just means that the three guys that Oracle picked up are not irrelevant to the conversation, and we shouldn’t pretend that they are.