Version 11 of Red Hat's Fedora Linux has hit the streets. I'm downloading it as I write this, although rather than simply picking through its feature set, I find myself thinking more of what each successive major-distro release means for Linux as a whole.

Whether or not we've tried to do so, Linux has begin to consolidate itself. This hasn't happened from the top down, though -- in the form of a single "standard Linux" that everyone starts from and then adds their own deviations onto. Instead, it's happening from the bottom up -- where three or four basic distributions (Debian, Ubuntu, Red Hat, SuSE) become the target platforms and everything else either branches out from them or just becomes an afterthought.

The bad news, again, is it means a lot of duplicated work -- that there's often no releasing a particular program "for Linux", but rather for a specific Linux distribution -- or, sometimes, a specific revision of a specific distribution. It's akin to a wisecrack about academia I once heard: Nobody just studies literature in college. You study Shakespeare -- the tragedies -- the late tragedies. (The only way to get around this right now is to write a platform-neutral application, in which case you're not writing for Linux specifically anymore in the first place.)

Perhaps, then, it's best to not think or talk about the various Linuxes at endpoints in the first place. Even if you can pop in a live DVD and run them as-is and get a fairly decent set of end-user applications in the process, that's not the real product. The product is what you make of it, not just something you get out of a box as a drop-in replacement for Windows or the Mac. Most of the talk about this sort of thing is confined to specific distributions. Gentoo, for instance, which makes no bones about being raw material. Linux's strength is in its flexibility; this has been covered before many times.

What I see happening with Linux on the desktop, then, is not that a given distribution will be the Big One. Instead, it'll be a starting point for someone else to build the Big One -- with other things on top of it that may well be proprietary and closed-source, but also centrally managed, engineered to a tight fit and designed to Just Work. I've mentioned this before, and the more I see the way the different distributions evolve -- especially Fedora, which starts from as free-and-open as possible as a zero point -- the more it's seeming like the way forward.

Plus, this approach won't stop people from sticking with the original, unmolested and free-to-do-with-as-you-wish version, though -- how could it?

InformationWeek Analytics has published an independent analysis of the current state of open source adoption. Download the report here (registration required).

Follow me and the rest of InformationWeek on Twitter.