Just got around to trying out a recent XFCE. Yeah, it's nice. A good, inoffensive desktop environment with all the usual metaphors. It's become rather less spartan than I remember XFCE being, as well. I can see the comparison to Cinnamon, given Cinnamon's stated goals. My only gripe is that I don't see any way to configure the positioning of monitors in the monitor configuration--a minor one,since I'm used to handling that myself with xrandr in my .xsession.
Tried out Awesome again. If you took dwm, completely ruined it, added on that 6-point font, right-click application menu from E16, then added extensibility in the form of a Lua API, you'd have Awesome. I have a completely unnatural hatred for it that does not belong on this Earth.
I had plans to use it's Lua scripting and API to essentially bring it back to dwm as a starting point, but it seems that this will be non-trivial and full of rough edges--I got bored and quit before my study of the documentation really bore fruit. I may come back to this later (the way it handles window rules still remains enticingly not-compiled-in).
After ranting to a friend about about it (who unsuccesfully tried to sell me on E17 again--he took my 'well, it seems okay, but I'm not sure it's for me' as 'I can be convinced that this is good' instead of the 'I hate this, but I don't understand why' I should have said), he recommended
CLFSWM to me, touting it's Lisp-programmability. I've been meaning to get (more) into Lisp, but it's that Emacs in the way. I just haven't been able to get past its 'finger feel'. And if I did? I'd probably be writing a ton of code to essentially transform CLFSWM into a WM I actually wanted to use.
Writing code to turn an extensible WM into
exactly what I want is beginning to look like more work than it's really worth.
Also: heard that Linux 3.12 has significant gains in FPS/performance for games on the free AMD driver (and possibly all games on all drivers, due to a change in the performance governor code). Not for me, apparently! Though that was on this RC 3.12 kernel that showed up in Experimental--I'll be trying it again when 3.12-proper makes its way into Debian.