Sun's GlassFish stack is, in a way, Sun at its best: creating infrastructures around and powered by its star product, Java. It's also its newest attempt at monetizing several pieces of its portfolio at once -- including MySQL, which many people outside the company are now biting their nails over.

There's a lot to like about GlassFish from a pure specs standpoint. For one, it's cross-platform -- it isn't a Solaris-only, or even Linux-only, stack, so you can stick with whatever platform you're currently running on (as long as it runs Java -- and who doesn't?). It also supports two extremely popular application languages for the Web, PHP and Ruby (plus Java, natch) -- and also contains MySQL as part of its stack, something I imagine anyone with two minutes' experience with Sun could see coming. And of course Sun plans to monetize the whole thing by selling sliding-scale support depending on the size of your implementation.

To be honest, GlassFish makes more sense as a platform strategy of sorts than OpenSolaris. If there's a platform that people associate most commonly with Sun, it's Java -- certainly a much more broadly useful platform than OpSol, since it's already proven itself in a staggering variety of contexts. And it doesn't require migrating to Solaris to be useful.

I worry, however, about Sun's ability to draw people in, even when it has a technically superior product or three. Its quandary now reminds me of the same situation IBM was in with OS/2: in its heyday, it also was superior to DOS and Windows, but not enough people picked up on OS/2 to make it matter except in niche applications. Java is in nowhere nearly the same quandary, thank goodness, which is why I have decently high hopes for GlassFish -- higher than I currently do for Solaris, actually.

Learn more about Web 2.0 at TechWeb's Web 2.0 Expo, March 31-April 3. Join us. (registration required).

Follow me and the rest of InformationWeek on Twitter.