[00:48] <bryce> Laney: https://edge.launchpad.net/~bryceharrington/+archive/ppa - git snapshot of what will be going into 6.12.0
[00:48] <Laney> I'll give it a test
[00:53]  * Laney restarts x
[00:56] <Laney> (did -ati used to support compiz? I think I remember it doing so but maybe I'm wrong)
[00:57] <bryce> on 5xx and earlier it does
[00:57] <Laney> hm
[00:57] <Laney> anyway I think this driver is a bit better
[00:57] <bryce> I've got it on my ati system right now
[00:58] <Laney> http://www.nibbledish.com is very choppy when scrolling in FF though
[00:58] <Laney> and metacity's compositor still brings it to a crawl
[00:58] <bryce> works fine for me, with compiz
[00:59] <Laney> I cannot enable compiz
[00:59] <bryce> oh I see, you still have fglrx installed
[00:59] <Laney> do I?
[00:59] <bryce> look at your bug report - https://bugs.edge.launchpad.net/ubuntu/+source/xorg/+bug/334066
[00:59] <bryce> NonfreeKernelModules: fglrx
[00:59] <bryce> that's your problem
[01:00] <Laney> I killed it all since then
[01:00] <Laney> tjaalton already told me this
[01:00] <bryce> https://wiki.ubuntu.com/X/Troubleshooting/FglrxInteferesWithRadeonDriver
[01:00] <Laney> yeah
[01:02] <bryce> ok hmm, not sure.  try the ppa driver I posted once it's finished building.
[01:03] <Laney> I already built it myself and installed it :)
[01:06] <bryce> ah gotcha
[01:06]  * bryce kicks ppa
[01:07] <Laney> woah, wait a second
[01:07] <Laney> I don't know what I did, but it's alive now
[01:08] <Laney> I did a full restart (had done one after killing fglrx but not after installing the PPA version), that must have done the trick
[01:09] <Laney> thanks for your help bryce \o/
[01:09] <bryce> Laney: no prob, thanks for testing that
[09:51] <mvo> bryce: re focus/grab problems I described yesterday. I reverted to the jaunty version of -evdev - the problem has not reappared yet. I keep a eye open
[10:08] <Ng> tjaalton: somewhat subjectively, it feels like this sluggish X I'm seeing is very much tied to system load - I'm applying the morning round of updates atm and even opening menus is crawling. Obviously load will hurt X, but it feels subjectively much worse than Intrepid
[10:14] <tjaalton> Ng: you had a G45?
[10:15] <Ng> tjaalton: that machine isn't jaunty yet, this is on my X3100 laptop
[10:15] <Ng> GM965, whatever
[10:25] <Ng> sunbird isn't being wildly gentle with the CPU, but it's not munching all of one core, let alone both, and there don't seem to be an unusual number of interrupts or other vmstat type things
[10:25] <tjaalton> mine works fine, so I'm not sure what's wrong
[10:28] <Ng> hmm
[10:28] <mnemo> intel 2.6.2 is out, that's nice.. i wonder what kind of stuff they fixed in it
[10:32] <Ng> tjaalton: EXA, right?
[10:32] <tjaalton> Ng: yes
[10:33]  * Ng hmms, the log looks normal, it's showing direct rendering
[10:33] <Ng> out of interest, can I make it use UXA?
[10:34] <tjaalton> yes
[10:34] <Ng> I'm not expecting it to be better, but I'm curious to try it
[10:36] <Ng> http://pastebin.com/f1ecd0032 is my log - the only bits I see that I'm not used to seeing are the dumps of PIPEBSTAT before and after something
[10:36] <Ng> and dmesg has errors from i915.ko about vblank counts for disabled pipe 0, presumably because I'm using pipe B
[10:46] <Ng> tjaalton: (would that be with Option "AccelMethod" "UXA" ?)
[10:46] <tjaalton> Ng: (yes) :)
[10:48] <Ng> ta
[10:48] <Ng> I hope I can figure out what's straining X, this is going to be very hard to live with for 6 months ;)
[10:54] <tjaalton> heh :)
[10:54] <tjaalton> try a livecd too
[11:11] <Ng> tjaalton: yeah good plan
[11:17] <popey> Ng: normal or extra effects in compiz?
[11:18] <Ng> normal
[11:18] <Ng> I reset my settings with ccsm yesterday, but I've not tested it with a fresh user or a livecd yet
[11:18] <popey> if i have extra (on my 945) and banshee is starting up, resizing a terminal window results in one frame of wobbled sides followed by the full screen terminal
[11:19] <popey> in fact with one core flat out (thanks banshee) i get much the same
[11:19] <popey> if i close banshee (getting a core back) then it seems to recover somewhat
[11:20] <popey> still not stellar
[11:33] <tjaalton> hmm, suspend/resume works again, even with UXA
[22:31] <tjaalton> bryce: filed bug 334578, could you confirm and subscribe u-a? :)
[22:32] <bryce> tjaalton: sweetness, yes I'll confirm
[22:32] <bryce> tjaalton: btw will you be merging the xserver 1.6 once alpha-5 freeze is over?
[22:33] <tjaalton> bryce: sure, it needs some other bits too though
[22:33] <tjaalton> at least the new libdrm
[22:34]  * bryce nods
[22:56] <RAOF> Is bug #328944 or either of your radars?
[22:56] <tjaalton> RAOF: yes, will get in together with libdrm 2.4.5
[22:57] <RAOF> Cool.
[22:57] <RAOF> I wasn't sure what wanted to happen there, otherwise I would have done it :)
[22:57] <tjaalton> I never got to uploading it
[22:59] <RAOF> The version of libdrm with libnouveau-drm patched on?  I noticed :)
[23:00] <tjaalton> I bet
[23:02] <RAOF> Eh.  It's not exactly critical.
[23:02] <RAOF> I wonder if nouveau gallium builds again...
[23:03] <tjaalton> I know, that's why you haven't been pushing it since ;)