I’ve been playing around with the Ubuntu 8.04 pre-release versions in a VM. So far, things are looking good. Of course, I won’t know if things are really good until I run the thing on real hardware.
Firefox 3 seems like an overall improvement, except for the new autocompletion window (too busy). The Gnome desktop seems mostly the same, which I consider a good thing. Metacity’s new composition manager option seems pretty usable (and it works in a VM!), though I’d like to be able to turn off the minimize animation, and go back to the old Alt+Tab widget without window thumbnails (anyone else find those silly?)
I hope the Gnome people will continue to build on top of the solid foundation they have right now. People talk about lack of vision for the desktop — that people have no idea of what 3.0 is going to be or whatever. I say, who cares? I don’t really think anyone (in FOSS or commercial companies) has that vision. Gnome is no longer significantly different as a desktop from OSX or Windows Vista. No, I’m not talking about hardware support or availability of 3rd party applications. I’m talking about the core desktop UI experience. I’d say Gnome even has the advantage in some respects (configurable panels and metacity’s kick-ass Alt+Tab implementation). Gnome should continue to incrementally and organically evolve. That is the best for the users. People shouldn’t be thinking about what 3.0 should be until it’s blatantly obvious. Besides, there are still so many problems with the Linux desktop that need to be solved. While it’s fun to think of grand, pie in the sky ideas, those are hardly the ones that are going to benefit the platform.
Commercial software has the perennial problem of needing to have revolutionary features every release so that customers will upgrade. Open source doesn’t have this problem. This difference should be exploited. Improve things slowly and incrementally, and build a platform that can’t be beat. In fact, because OSS has much fewer developer resources, it should be even more wary about changing things without justification, replacing working code without tangible, obvious benefits. Vista tried to do a lot of this, and look where that got them.