| jbarnes | compiz may have rotation plugins | 00:01 |
|---|---|---|
| jbarnes | but you wouldn't get rotated cursor movement in that case | 00:01 |
| jbarnes | but maybe that's ok | 00:01 |
| jbarnes | no other branches available afaik | 00:01 |
| bryce | ok thanks | 00:01 |
| Sarvatt | tjaalton: thanks! sorry about that, I had to run out again right after I got in. the problem with the fglrx/nvidia patches were that if someone has an xorg.conf without a driver explicitly being named it would default to the first one in the list then fail | 00:29 |
| Sarvatt | like if they had a default xorg.conf | 00:30 |
| Sarvatt | or used the default as a template to add an option | 00:31 |
| Sarvatt | when i start up with the default xorg.conf only it tries to load fglrx and aborts after on my powerpc machine | 00:36 |
| hyperair | Sarvatt: i just recompiled mesa-utils with -O0, and the segfaults stopped happening. | 05:51 |
| hyperair | i mean mesa | 05:51 |
| hyperair | i said to soon. it just crashed. | 05:58 |
| hyperair | ...and i still can't get the line number. what the hell? | 06:01 |
| tjaalton | Sarvatt: right.. that one | 06:34 |
| * hyperair has discovered that gdb allows one to continue running a segfaulted X. | 09:41 | |
| Ng | hmm wasn't there a jaunty PPA which has newer intel drivers, but not quite as crack-fuelled as xorg-edgers? | 10:23 |
| tjaalton | xorg-updates? | 10:24 |
| Ng | https://edge.launchpad.net/~ubuntu-x-swat/+archive/x-updates/ :) | 10:25 |
| Ng | ta | 10:25 |
| tjaalton | libdrm 2.4.12 merged & uploaded | 12:20 |
| bryce | tjaalton, who is this raphael character and why is this the first I've noticed him? | 12:27 |
| tjaalton | raphael geissert, openchrome maintainer in debian | 12:27 |
| tjaalton | recently became a DD too | 12:27 |
| tjaalton | IIRC | 12:28 |
| bryce | I'm failing to understand why he's being pissy about an svn update in ubuntu | 12:28 |
| tjaalton | a bit over-sensitive imho | 12:29 |
| bryce | yeah, ok, tempest in teapot, I'm going back to bed | 12:30 |
| tjaalton | heh, night | 12:30 |
| tjaalton | I'll disable the fglrx/nvidia autoconfig patches in order to get the new xserver uploaded | 13:06 |
| tseliot1 | tjaalton: what are the fglrx/nvidia autoconfig patches? | 13:21 |
| tjaalton | tseliot1: they modify the tables to try those before nv/ati | 13:22 |
| tseliot1 | tjaalton: aah, nice | 13:23 |
| tjaalton | just that they fail when there's a minimal xorg.conf | 13:23 |
| tseliot1 | because of the depth? | 13:24 |
| tjaalton | so they can't be used until the logic is fixed | 13:24 |
| tjaalton | and I'm not sure that's happening anytime soon | 13:24 |
| tseliot1 | I meant defaultdepth | 13:24 |
| jcristau | no, not because of that | 13:24 |
| tjaalton | without a conffile the server builds one of it's own, and fills it with multiple device entries | 13:25 |
| tjaalton | but that's not happening when you have the conf | 13:25 |
| tjaalton | it should be fixed to add those after the one(s) from the conf, I think | 13:26 |
| tjaalton | or by some other means | 13:26 |
| jcristau | i think it needs something like duplicating the screen and device sections, and setting the Driver for each device section according to the matches | 13:27 |
| tseliot1 | yes, I noticed that behaviour (when no conf file is available) | 13:27 |
| tjaalton | uploaded | 13:36 |
| tjaalton | jbarnes: bug 394273, driver specific or in the xserver? pitti thinks it's the server, and I'm not sure about it anymore :) | 15:30 |
| ubottu | Launchpad bug 394273 in xorg-server "[i945] defaults to non-native screen resolutions (mode change from KMS)" [Unknown,Confirmed] https://launchpad.net/bugs/394273 | 15:30 |
| jcristau | tjaalton: aiui the kernel driver will use the lid information for lvds | 15:33 |
| jcristau | whenever that patch lands | 15:33 |
| jcristau | but otherwise, the decision on what mode(s) to use when more than one output is enabled is done by the server, and currently it tries to use a mode that works on all outputs | 15:44 |
| tjaalton | but apparently it did work before kms | 15:45 |
| jcristau | where's the log for that? | 15:46 |
| tjaalton | gzipped in that bug | 15:46 |
| tjaalton | oh | 15:46 |
| tjaalton | not the old one afaik | 15:46 |
| jcristau | gzipped attachments ftl :) | 15:47 |
| tjaalton | yes, and apparently he's been using the packages from edgers | 15:47 |
| tjaalton | so it's xserver master | 15:47 |
| tjaalton | hmm no, it's 1.6.1foo | 15:48 |
| jcristau | i don't think the server's heuristic has changed since 1.5.x | 15:48 |
| tjaalton | right | 15:49 |
| maxb | The standard Ubuntu orange stripy progress bar seems to be causing Xorg to use a pathologically huge amount of my CPU in karmic (nvidia). I don't remember this always being the case | 16:34 |
| jbarnes | tjaalton: oh not sure about that one... lemme look | 17:06 |
| jbarnes | ah yeah, jcristau is right, there's a lid patch for that | 17:08 |
| jbarnes | should land soon | 17:08 |
| jbarnes | in the -next tree | 17:08 |
| tjaalton | jbarnes: thanks for confirming, so I'll move it to the kernel then | 17:09 |
| jbarnes | yeah it's the kernel that tries to match the modes | 17:10 |
| Sarvatt | xserver needs the libselinux build dep dropped for ubuntu | 18:24 |
| Sarvatt | we've still got the --disable-xselinux universally, and its stuck in a dep wait because its not in main | 18:26 |
| Sarvatt | oh its waiting on libaudit-dev | 18:28 |
| bryce | are we at 2.6.31-4 for the kernel now? | 18:33 |
| bryce | yep looks like, cool | 18:34 |
| bryce | jbarnes, Sarvatt: http://people.canonical.com/~bryce/drivers.svg | 18:44 |
| bryce | http://people.canonical.com/~bryce/upstream.svg | 18:45 |
| maxb | *blink* wow | 18:49 |
| tjaalton | Sarvatt: hrm, wonder why meld didn't show that. that dep was dropped the last time | 19:06 |
| tjaalton | well, maybe it's time to really get those in main | 19:06 |
| Sarvatt | hmm, thinking it might be a good idea to make a script or package in edgers that reverts everything so people can revert it easier | 20:16 |
| Sarvatt | any suggestions on the best way to do that? i was just thinking of grepping sources for edgers and commenting it out then apt-get updating and making it do a sudo apt-get install <package>/distro for every package in there | 20:20 |
| bryce | Sarvatt, hmm, seems like there ought to already be a "manage ppa subscriptions in sources.list" type app out there | 20:23 |
| bryce | Sarvatt, in any case, it would probably be smart to slap a GUI on it for maximum user easiness | 20:24 |
| bryce | tjaalton, thanks for syncing all those packages | 20:25 |
| bryce | that was on my todo list for today :-) wee | 20:25 |
| jbarnes | bryce: haha nice | 20:25 |
| Sarvatt | i wish apt recorded more details on packages like the source of it instead of just saying http://ppa.launchpad.net for all PPA packages | 20:25 |
| bryce | Sarvatt, btw, mvo has typically been extraordinarily helpful in apt matters; he might also be able to steer you towards existing code for managing this stuff | 20:26 |
| jbarnes | bryce: nice note too, thanks | 20:27 |
| jbarnes | I'll have to reply with something similar when I get back from vacation | 20:27 |
| bryce | jbarnes, :-) Yeah I feel like we've hit a solid milestone with -intel | 20:27 |
| Sarvatt | whoa, those graphs are nuts bryce, you've been busy! :D | 20:28 |
| bryce | I think I'm going to try putting this level of focus on -ati next | 20:30 |
| bryce | Sarvatt, hopefully we can pull in the kms stuff for -ati into karmic, and then use the same approach to get -ati bug reporters to test | 20:31 |
| virtuald | Is there any way I can use an ICM file extracted from a windows driver archive to correct the colors on my second display? | 20:36 |
| virtuald | tried to use xcalib but it didn't work out | 20:37 |
| mvo | Sarvatt: I'm happy to talk about issues with apt - maybe tomorrow? its getting late here in my TZ :) | 20:50 |
| tjaalton | bryce: np, requestsync makes it easy to do | 21:04 |
| Sarvatt | ahh not going to be as easy as i thought, forgot that it would go and install the things that arent currently installed if i do that for every package so i should at least have it parse the installed package list and only do things currently installed somehow | 21:17 |
| bryce | Sarvatt, if it's too tricky, a starting point might just be to list the steps on a wiki page. We can point people to that, and use it as a starting point for figuring out how to script it up | 21:47 |
| virtuald | how do i see if i run kms or some other fb? | 22:24 |
| virtuald | now i see it is, drm printk's when i switch to console | 22:35 |
| virtuald | though it switches to resolutions i don't want | 22:36 |
| virtuald | and i have to configure x to get the right resolutions | 22:36 |
| virtuald | is that a bug? | 22:36 |
| virtuald | i even have to add a modeline to get 1152x864@75Hz on my second monitor | 22:38 |
| TwoToneSpirit | Hello everyone. I'm wondering if two users, sitting at different terminals, can use the same instance of X and share the work environment (ie, two cursors and mouse pointers). | 23:57 |
| TwoToneSpirit | My dream really is to just have one gnome workspace that is shared between two computers - I'd drag stuff there to leave for my girlfriend and she'd do the same for me. Or we'd work together on that workspace. | 23:58 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
