[00:22] <TimU> Scratch that.  Starting fresh from "apt-get source ..." and running debuild seems to be working.  I must have horked it somehow while trying to enable debugging.
[07:12] <hyperair> does anybody here have experience with ati cards and backlight setting?
[08:18] <RAOF> bryce: I think the libdrm changes are ready for upload.  I've added the LP foo to the changelog, pushed to git again.
[08:36] <tjaalton> RAOF: ok, I can take care of it
[08:36] <RAOF> Thanks.
[08:36] <tjaalton> superm1: 529d1d720 made it in 7.4
[08:37] <superm1> tjaalton, spectacular.  i sent an email to that ML and when i got turned away to that vmware guy's email. so sounds like he made it happen :)
[08:37] <tjaalton> superm1: your email got through to the list today :)
[08:50] <tjaalton> building mesa 7.4 now, and if it's ok I'll upload it to my ppa
[08:50] <tjaalton> and file a FFe for it
[10:01] <tjaalton> ok, mesa is building now https://edge.launchpad.net/~tjaalton/+archive/ppa
[17:08] <bryce> tjaalton: let me know the bug id for the FFe so I can help push it through
[17:14] <sbeattie> bryce: low-priority, but do you know what's up with bug 351076? They all provide xserver-xorg-video-2.1 which xserver-xorg-core lists as a conflict; are they incompatible and should be removed, or is there some other fix/update that needs to happen?
[17:14] <ubottu> Launchpad bug 351076 in xserver-xorg-input-void "Package xserver-xorg-input-aiptek does not install (Januty beta)" [Undecided,New] https://launchpad.net/bugs/351076
[17:19] <tjaalton> sbeattie: there are a few input drivers that won't build against xserver 1.6. on my list of things to fix..
[17:19] <tjaalton> bryce: it doesn't exist yet :)
[17:19] <bryce> tjaalton: ok.
[17:20] <bryce> actually I think I ran across a "please upgrade to mesa 7.4" bug we could probably reuse
[17:20] <tjaalton> yeah
[17:20] <bryce> sbeattie: what timo said :-)  I can confirm I was able to reproduce the problem.  
[17:22] <tjaalton> i was hoping to just sync them from debian, but it might take a week or two until 1.6 is in unstable
[17:22] <tjaalton> only a handful of drivers are updated against it in experimental
[17:23] <tjaalton> there are maybe six drivers to fix
[17:26] <sbeattie> okay, I just wanted to make sure you guys were on top of it; I dislike having uninstallable packages in the archive.
[17:27] <tjaalton> the archive rebuild was a great reminder ;)
[17:30] <Ng> this is so weird, the guy who sites next to me just upgraded to jaunty, he also has a 965 and his is smooth as in Intrepid. I downgraded from xorg-edgers to jaunty packages and mine is back to being highly affected by CPU. our Xorg.0.logs are very similar
[17:30] <Ng> could something like mtrr be at fault I wonder?
[17:31]  * Ng never been entirely sure how to check the sanity of mtrrs
[17:48] <bryce> tjaalton: hmm, a week or two might cut things a bit close; what needs done to update them specifically?  Just packaging stuff or does it need code changes as well?
[17:51] <jcristau> tjaalton: if you can prep them in git i could upload to experimental probably.
[17:51] <jcristau> no time to do it myself this week though
[17:51] <tjaalton> bryce: new upstream versions, likely
[17:54] <tjaalton> jcristau: ok, I'll have a look tomorrow
[18:08] <bryce> tjaalton: do you happen to know how .pot files get added to xkeyboard-config?  (bug 349341)
[18:09] <ubottu> Launchpad bug 349341 in xkeyboard-config "launchpad xkeyboard-config .pot is obsolete: 1.4 whereas binary is 1.5" [Undecided,Confirmed] https://launchpad.net/bugs/349341
[18:09] <bryce> tjaalton: it seems once in the past there was a .pot file included, but the packaging doesn't seem to produce it 
[18:13] <Ng> UXA doesn't seem to help either, which is interesting
[18:13] <Ng> the only thing that seems to help is flipping up to the edgers packages, but that brings occasional X hangs, and more frequent segfaults on resume
[18:16] <tjaalton> bryce: don't remember off-hand, but + can check once I get home
[18:17] <bryce> ok thanks
[18:18] <tjaalton> Ng: now is a good chance to try mesa  7.4 from my ppa ;)
[18:19] <Ng> tjaalton: sure, I'll downgrade to jaunty -intel stuff again
[18:21] <Ng> tjaalton: in your personal ppa?
[18:21] <tjaalton> Ng: yep
[18:22] <Ng> ok, installed, rebootin'
[18:26] <Ng> tjaalton: well it doesn't seem to have improved things, but it hasn't made them any worse :)
[18:29] <tjaalton> Ng: eexcellent :)
[18:30] <Ng> is the new mesa so we can sneak in a newer -intel? ;)
[18:32] <tjaalton> a cunning plan, you say? :)
[18:32] <Ng> I have a sticky tail and a "Fox" name badge for such a plan ;)
[18:35] <bryce> Ng: sort of late in the process to be upping -intel... the mesa update is mainly for bug fixes
[18:36] <tjaalton> yeah :)
[19:21] <superm1> tjaalton, at some point were you looking into pulling the patch referred to in  http://www.phoronix.com/scan.php?page=news_item&px=NjYwMw into ubuntu?  it certainly would make the things jockey has to do less necessary and great from a user perspective
[19:24] <tjaalton> superm1: yes I was
[19:25] <superm1> tjaalton, what happened with it?
[19:25] <tjaalton> doesn't work if you have an xorg.conf
[19:25] <tjaalton> so we'd need to not ship one first
[19:26] <superm1> but it's still an improvement over the current situation then no?
[19:26] <jcristau> loading with nv > failing to load because nvidia doesn't work
[19:26] <jcristau> so no
[19:27] <superm1> oh it totally fails if no xorg.conf is around
[19:27] <tjaalton> different codepath
[19:27] <tjaalton> um
[19:28] <jcristau> FatalError("no screens found"); -> unhappy user :)
[19:28] <tjaalton> I mean, it doesn't work if there _is_ an xorg.conf
[19:28] <superm1> so is the failure coming from extra checks going on in the NV driver or in X libraries?  
[19:31] <tjaalton> I'm not sure what jcristau meant, but yes there are issues :)
[19:31] <jcristau> it's something like that: if (no xorg.conf) { make a list of possible drivers; for each one, if its preinit works, break, happiness, if not try the next one } else { find the first candidate; if its preinit fails, fatalerror }
[19:32] <superm1> so why the logic to fail if the first candidate doesn't work when there is an xorg.conf?  is there a good reason for that?
[19:32] <jcristau> (there's no good reason for it to work like that, but nobody's fixed it, so..)
[19:32] <tjaalton> ah, right :)
[19:33] <superm1> ah
[19:33] <tjaalton> anyway, a couple of episodes of "my name is earl" now.. ->
[19:33] <superm1> tjaalton, b4 you run, do you have the commit handy? i'd like to play around with it a bit when i get some time
[19:34] <jcristau> superm1: 66fb2530
[19:34] <superm1> thanks jcristau 
[19:34] <jcristau> np
[20:37] <Ng> bryce: I figured as much ;(
[20:41] <bryce> Ng: if you know of specific patches/fixes that should  be backported, I'd be game
[20:45] <Ng> bryce: I don't suppose there are any magic tools that'll help me build the driver between whatever is in jaunty and current git?
[20:48] <bryce> sort of a wip but there is scripts in xorg-pkg-tools that does it
[20:49] <bryce> Ng: but basically apt-get build-dep xserver-xorg-video-intel, grab the git snapshot, autogen.sh, copy over the debian directory, and debuild to get a deb as usual
[20:50] <bryce> assuming you want to keep everything maintained by the packager; if you don't care, build/install from git should work too, although there may be some ./configure differences
[20:53] <Ng> bryce: I'll see what I can turn up. I guess a good initial thing would be to just diff the sources, but my guess is there will be a fair number of changes
[20:59] <bryce> Ng: indeed.  git log and git show can be helpful
[21:12] <tjaalton> bryce: 'cd po; make xkeyboard-config.pot' would generate the potfile
[21:12] <bryce> tjaalton: mm, ok, any reason not to make the package do that automatically?
[21:13] <bryce> Subject: Announcing the Next Ubuntu Hug Day! - April 02 2009
[21:13] <bryce> This week's HugDay target is *drum roll please* xorg-server and xserver-xorg-video-intel!
[21:13] <bryce>  * https://wiki.ubuntu.com/UbuntuBugDay/20090402
[21:13] <tjaalton> bryce: I guess the build has changed
[21:14] <jcristau> it looks like the pot file is not usually needed for the build, and isn't shipped in the upstream tarball.
[21:18] <tjaalton> yeah, so I'm not sure how launchpad wants it..
[21:21] <tjaalton> bryce: that's going to be a busy session ;)
[21:21] <bryce> hope so