Most of my nightmares tend to be mundane nonsense about being late for school. Folks in the open source community have nightmares about open source products becoming closed source properties. That's nightmarish, to be sure, but I have to ask how much of the nightmare is not wholly real.
Here's how this nightmare works. You discover one day that your favorite open-source product -- like, say, to choose an example totally at random, MySQL -- has been bought by a commercial outfit. And then you watch in horror as something that you depended on because it was open and came with little or no cost attached to it devolves into something ... proprietary.
In a piece entitled "Open Season On Open-Source Software", NewWest.Net founder and editor-in-chief Jonathan Weber put the implications of the nightmare this way:
... Companies that make open source being acquired by more commercial concerns "could result in better products that will still be quite cheap, if not free, and that would be a good thing. And completely free open-source versions will presumably continue to exist: the core code, after all, is still in the public domain. But there is some risk that these products will, over time, evolve into expensive proprietary products that are quite different from the original open-source versions, and that could present problems for companies like ours. If we couldn't afford the commercial version, the alterative might be a buggy open-source version that was less compatible with the Web software ecosystem.
Is he justified in his fear? Maybe, but not to the point that all open source acquisitions should be suspect.
One, it may be possible for an open source product to turn into a closed source one, but only in the sense that the closed-source item will be as future iterations of that product by a given development team. As long as any open-source version exists, someone else can pick it up and run with it if they feel the closed-source version has languished or strayed from its original intentions. That's not something you can do with a proprietary product at all -- not without licensing it directly from the manufacturer (and that's assuming they ever bother to allow such a thing).
Two, to my mind, an open source product probably has a better chance in the long run of staying on the cutting edge of Web technologies, precisely because it's open and anyone can step up and contribute code to make it work with those standards. That's not a guarantee such a thing will happen, but it's also no more or less a guarantee than the possibility that your favorite open-source app is going to get bought and turned into something proprietary.
I'll run the risk of sounding like a fool, and restate the obvious. If MySQL the company fell apart tomorrow, MySQL the product as it stands right now could be taken and continued by whoever had the inclination to do so -- and believe me, looking around the world right now, there are plenty of people who would have the inclination. No, that fate wouldn't be as convenient or easy as having the original development team to offer support and continue the work unburdened by restrictive licensing, but it's a heck of a lot better than nothing.
So -- no, there are no guarantees here, much as there are no guarantees in many walks of life. But there are things that, to me, make up a great deal for the lack of a guarantee. They make the nightmare a lot easier to wake up from and see for what it is: bad dreams, but not necessarily real.