The current state of GNOME, after the 2.10 launch in March, is that we have a nice development environment, still rough around the edges, which is reasonably usable. For the userish types HAL has been improved & major-suite integration has been applied liberally. Evolution now integrates with the desktop clock/calendar, which is fantastic.
For GNOME 2.12 we look forward to the new power-management features which should be coming out of Project Utopia. Hibernating a Linux desktop through GNOME is pretty high up there on my wish-list. With any sort of luck whatsoever, we'll have Cairo (so fancy graphics which should ooze Clearlooks-themed style) & can start integrating apps like Tomboy (a wiki-ish note taking system, which I thought was pretty innovative), F-Spot & Beagle. These apps all use Mono, so don't stand much chance of getting integrated in 2.12, but they desperately SHOULD be.
So while Usability has been abuzz with talk of moving to a 'user-interesting-object model', stuffing drag'n'drop into every nook & cranny, automatic versioning & even migrating to an OSX-like global menubar, it seems to me that we should reach some kind of consensus as to which programming languages should be properly supported by the GNOME project. If Mono just isn't right politically, then can't we implement a GNOME framework to build apps within? Using VFS, GConf & even GTK automatically when you begin developing HAS to be good for the project, right?
I'm not saying copy