I am really liking Gnome 3 and gnome-shell. But, I admit that the Gnome 3 has been quite unstable, and 3.6 seems to be about ready for use.
But I can see the desire for old interfaces, and this sounds it might have some future.
I'm worried that since the desktop has become the center of attention some distributions have started their own desktop projects shifting focus from distribution to software developement. This is taking work force from distribution tasks like improving and fixing upstream software and its distribution methods. I've steered away from Ubuntu for that reason, they're no longer the bug-free, user-centric and improved distribution that they seemed to be at first. Trying to be too much will kill them or make them no better than Microsoft or Apple.
Since the desktop has become the center of attention some distributions have started their own desktop projects shifting focus from distribution to software developement. This is taking work force from distribution tasks like improving and fixing upstream software and its distribution methods. I've steered away from Ubuntu for that reason, they're no longer the bug-free, user-centric and improved distribution that they seemed to be at first. Trying to be too much will kill them or make them no better than Microsoft or Apple.
People knock MATE, but it has been the only thing that has allowed you to maintain the stable well tested GNOME2 experience in modern desktops for the past year and a bit. For some people that has been very valuable. It has always been installable along side GNOME3 packages, so you can install it without messing up anything else. I dont think there is any free desktop environment that has had the long term testing that late versions on GNOME2 had.
Building a GNOME2 style desktop with more modern technologies is not something that can be done overnight. And getting it up to the same level of stability will take several years. whether the best way to do that is to start from scratch building things with GTK3, or to port MATE to GTK3 (on roadmap), or to modify GNOME Shell to work like GNOME2, or to port XFCE to GTK3 and add features. who knows. its to early to tell. and there is nothing like a bit of friendly competition to spur them on.
It looks like consort is a mix of approaches. for example they forked metacity 2.34 just like MATE did.
Personally I didn't find Gnome 2 as stable as a desktop like it should be after so many years of testing .. But compared to MATE ..
For example in Mint 14, every time I opened appearance properties to change theme/background the process would stay after closing the window consuming 100% CPU. Pretty cool for those who want a stable tested environment .. :)
(Yes, the bugs have been reported)
Also I seemed to be completely unable to change the panel theme. Like it was hardcoded or something(it wasn't, but it behaved kinda like it was).
Of course, I tried putting a solid colour or a backround image, cause I couldn't stand all the white-ish colour.(Changing the rest of the desktop would make it kinda not fit in.
But wherever there were applets, windows in the taskbat etc it would insist adding the Mint theme's panel background, making it even worse than before. Yes, that was regardless of my theme settings .. xD
And the usual applets crashing and stuff ..
Those users who want Gnome 2 because of stability etc, look more like they don't want anything else cause it's different. Even if it's actually not. 0_o
Because Gnome Fallback is basically Gnome 2 with GTK3 and even some bug fixes. I mean seriously, why fork Gnome 2 and not that?(or just maintain it?)
As much as I disagree in some aspects with Gnome 3 devs, I agreed when I read that if there is such a high demand for a classic desktop, why doesn't anyone step up and maintain Gnome Fallback/Classic?
I think consort desktop is the best approach personally out of the ones taken, by far ..
In fact think about this. If MATE is to be ported to GTK3 won't they have the same hurdles that Gnome 3 had? So where will the claimed stability go?
By then Gnome 3 and Fallback could already be much much more stable. And putting the interface aside, we're talking pretty much about the SAME programs that do the SAME things, but we have them two times, except the one of them will have to be ported to GTK3 AGAIN.
Of course they're free to do so. But I myself find the reasons for doing this exact fork and how it's done null.
for example this https://bugzilla.gnome.org/show_bug.cgi?id=663690 its been an issue in GNOME3 since 2011. switching to MATE allowed me to work with my windows where i wanted them. MATE was quick on the scene (because they took a path that got quick results for little work). Its great that there might be more modern GNOME2 style desktops in the future, but none of them provided what i needed in 2011 and 2012.
Am I the only one who noticed the TLA collision?
(Consort Desktop Environment vs Common Desktop Environment)
Also, it strikes me that the longterm goals are similar to MATE; it's just that they started with GNOME3 stuff when possible.
I would be surprised if there isn't some sort of merge down the road (probably once Consort gets closer to the goal).
MATE actually works. Like GNOME2 did.
Those other efforts are nice but will take *years* to get to the level of useability of GNOME2/MATE. (KDE4 did the same mistake is _almost_ useable now. At least its UI never really sucked, it was just slow, buggy and featureless ;)
So yeah, I don't care if MATE uses "old gnome stuff". It works, everything freaking works. I don't need to get new buttons every 6 months. I really don't. I just need multiscreen to work, stuff to load fast, etc.
Sorry, couldn't resist.
Maybe it has something to do with the computers I've used, I first encountered gnome2 with Dapper on a Thinkpad 600x/550MHz PIII/128 MB.
But gnome has never felt fast, while icewm is lightning.
On the rest, I'd agree.
But with RHEL6, you probably wouldn't ever need to touch gnome3. Just go till 2017 (or is it 2020 now?), then see what the current abominations are.