I guess I'd like to see just what the Sun folks, and others here, are looking at as the main target class: embedded, desktop, or server. Seems to me there's current applications possible for embedded and server, but not so much (aside from game consoles, which I'd tend to view as more like embedded than like traditional desktop) for desktop (keeping in mind _current_, i.e. something that's not only available but still being made and likely to continue being made for longer than it takes to get the port functional on that platform). Also, while one can undoubtedly get the Solaris kernel to run in <= 256MB, I doubt a functional userland would look particularly familiar (and I doubt very much that it could be running a bloated desktop like GNOME); so an embedded "distro" might deviate quite a bit from the traditional Solaris environment.
Whatever the ultimate objective, is there (a) useful work towards it that can still be done on the ODW, and (b) are there really more people that can and want to do that work than have access to an ODW? Until there's a common understanding on those points, I don't see how there can be an understanding of what possible alternative development platforms would be enough of an asset to development to be worth the trouble of supporting at the very least additional chipsets, etc. That is, how can you know what is a viable alternate route for a stage of a journey unless you know what the destination is, how far away the alternate route is, and whether the route you're currently on is in fact carrying more traffic than it can handle efficiently? This message posted from opensolaris.org
