I've been a long time fan of Solaris the technology, and I've been running OpenSolaris for personal use for several years now. I've tried hard to explain to people the merits of what I view as an exceptional but lesser known OS, with minimal success. The OpenSolaris community never really took off the way I (or Sun, probably) had hoped; Sun never really treated community members as first class citizens, and the "real" work on OpenSolaris all came from inside of Sun.
Well, I guess Oracle had enough of all this mess, and they've shot the OpenSolaris project right in the head. OpenSolaris as a product is completely dead, and the nightly builds and code repositories are closing up shop. The CDDL will be used (to some extent) for bits of Solaris code, but that code won't be released until *after* Solaris releases ship, which is a stark contrast to the more open approach Sun took with OpenSolaris.
As a pre-emptive strike against this potentiality, key community members have already spooled up the Illumos project, a fork of OpenSolaris with the proprietary bits replaced by open source software.
Oracle's decision to keep the repositories closed until after release forces Illumos's hand a bit; they no longer have the luxury of simply rebuilding a free Solaris derivative and keeping it up to date with the latest efforts from inside Oracle. Now, they have to choose: do they accept a huge lag and spool up a post-release open source derivative, or do they fork proper and kiss Oracle goodbye forever?
I have to say, the latter option is much more attractive, and it's not entirely out of the realm of possibility either; the Illumos main page now shows a brief response which seems to imply they're moving in this direction. Illumos is a real community project; if OpenSolaris's woes were really, as many of us felt, due mainly to Sun's handling of the community, this represents a tremendous opportunity to address those issues and build a healthier and more vibrant community than OpenSolaris ever had. Oracle is even helping out by scattering former Sun engineers to the wind; how many of those engineers will want to keep working on the OpenSolaris code base as a labor of love? Are there enough of them interested in continuing the work that they started at Sun as a community project? Will they ultimately make Illumos even *better* than Solaris?
Solaris proper is rapidly dying to me. I'm not the customer Oracle wants, and they've made it abundantly clear that they have no interest in having me. They want big enterprise customers with big Oracle database deployments, and screw you open source and startup hippies, your pockets aren't deep enough to even think of playing that game. Well, so be it. But Illumos is providing an opportunity for all of us hippies and startups to rally around a new project, with a new community, that we may finally be a real part of. I think the possibility really is there, and Illumos could become the open source success that OpenSolaris only dreamed of becoming.
Am I sad to see things go this way? A bit, yes, but it's also a huge burden lifted: now we know the real score. There's no more begging Oracle for table scraps that may never come, it's either do or die by the efforts of the community alone. Whichever way that goes, at least there's nobody left to blame.
I won't lie, I've mostly lurked in the shadows of the OpenSolaris community, justifying my lack of participation with an observation that participation was pointless anyway. Why would I even bother when the community was really just an afterthought to Sun? Well, I (and people like me) are being called on that bluff; if we don't step up and contribute now, it's all over for real.