[10:09] hi, guys. i upgraded to jaunty and enabled the proprietary ati driver and now my x is borked. can you help me fix it, please? [10:09] Guest21892: trying booting the safe mode [10:10] there is a menu to fix the X configuration [10:10] crevette: there is no safe mode in jaunty. it gives an option to fix x but that didint' do it. [10:11] crevette: i meant there's no safe mode for x. [10:12] ah [10:12] i'm in irssi now. how do i switch between rooms? [10:13] humm, I didn't used irssi for years ... [10:13] Guest21892: ^n ^p, /window, alt-N, ... [10:14] jcristau: that worked. thank you. [10:17] does anyone have any ideas about disabling the ati driver? [10:17] dpkg --purge xorg-driver-fglrx [10:18] tjaalton: that will disable the proprietary driver enabled? [10:18] Guest21892: it will purge it from your system [10:18] tjaalton: thanks!!! i'll try it. [10:24] tjaalton: ok. i purged the ati driver. now when i startx i get a message saying 'no screens found'. [10:24] put the logfile somewhere [10:26] but first make sure /etc/X11/xorg.conf is sane.. move it aside and try without one [10:26] tjaalton: ok: pastebin.com/f267af216 [10:27] how did you end up with the radeonhd driver? [10:27] tjaalton: i don't know what you mean by move it aside. [10:27] rename it, but it doesn't matter here [10:27] it's probably empty anyway [10:27] someone in #ubuntu told me to try it. [10:28] right [10:28] so dpkg --purge xserver-xorg-video-radeonhd then [10:28] I guess that'll make it work again, if you managed to install it in the first place.. [10:29] jaunty on your system, that is [10:30] i upgraded to jaunty, everything was ok, then i enabled the ati driver, that's when things went bad. [10:30] after i purge do i have to change my xorg.conf? [10:30] I don't think it supports that device [10:30] anymore [10:31] no [10:31] ok. i'll try. thanks. [10:31] neither does radeonhd for that matter [10:31] but -radeon does [10:33] should i install radeon? [10:33] it should be installed [10:33] by default [10:33] ok. i'll try startx. wish me luck. [10:45] tjaalton: ok. i did what you said and now x comes up but my keyboard, mouse, and touchpad don't work. [10:45] binMonkey: same deal, logfile thanks [10:45] i am jazzed that at least i get a gui up. [10:46] ok. thanks. [10:49] tjaalton: ok, here it is: pastebin.com/f457e6482 [10:49] tjaalton: if it matters, i tried dpkg-reconfigure xserver-xorg. [10:50] it does nothing of interest [10:50] try rebooting [10:50] or restart HAL and try again [10:51] and remove the fglrx module (rmmod fglrx) [10:52] how do i restart hal? [10:53] just reboot ;) [10:53] ok. wish me luck. [10:56] WOOOOOTTT!!!!!!!! [10:57] tjaalton: it's working. [10:57] THANKS A MILLION FOR YOUR HELP! [10:57] np [10:57] don't break it again [10:58] i won't. i hate to be a pest, but is there a driver for the 200m that i can use in jaunty? [10:58] you already have it [10:58] the radeon? [10:58] yes [10:58] what more do you want? [10:59] it's the only one supporting your device [10:59] ok. i guess i should be thankful. [10:59] and i am. [11:00] I'm pretty sure that the restricted manager didn't suggest to install fglrx and you forced the install, am I right? [11:00] i'm not sure. i used 'hardware drivers' in the main menu. [11:00] what does it say now? [11:01] hold on. [11:01] i see what you're saying. i did ignore the warning. [11:02] but it did work really well in 8.10. [11:02] but you're not using 8.10 anymore [11:02] I don't think it should even offer it [11:02] yeah. i just didn't know that jaunty didn't support it. had i known... [11:03] what does the warning say? [11:03] this is my old laptop that i'm using. can i change a video card without having to worry about other hardware? [11:04] umm, laptop? [11:04] no [11:04] it's built-in [11:05] i closed it but it was something like 'these drivers have no public source code. they represent a risk, ubuntu cannot fix them, etc.' [11:05] that's a generic warning [11:06] dang. i guess the only game i'll play for a while is nethack. [11:06] file a bug, by running 'ubuntu-bug restricted-manager' [11:06] it shouldn't offer the driver at all [11:06] the free driver should support 3D just fine [11:07] i think i agree. especially if it's not supported. [11:07] i tried googleearth. it's a little jerky but i can live with it. [11:09] youtube looks ok. [11:10] i got two good things out of this: i learned about pastebinit and i learned a little bit about screen. [11:12] thanks again for the help. have a good night. [11:21] * bryce_ waves [11:21] o/ [11:21] tjaalton: btw pitti said that 2.7.1 caused some freeze issues and a kernel panic on his laptop; he'll file a bug this evening and we'll chat with jbarnes_BCN when we see him [11:22] bryce_: ah, ok.. [11:22] has worked fine here, but IIRC he has a 945 [11:23] tjaalton: looking at the changelogs, seems most likely to be something in the upstream code; the debian and ubuntu packaging sounds pretty sane, so guessing it's one of keith's patches, but really hard to say [11:24] bryce_: btw, in case you missed it.. debian dropped the pci-ids patch, and it pointed out a flaw in the xserver patch 143_default_to_vesa. I've disabled that one for now [11:24] yep, saw that [11:24] ok, good [11:24] tjaalton: there was one person who said his hardware got misdetected as -vesa with that change [11:24] encouraged him to file a bug [11:24] also, there are a bunch of upstreamable patches in the xserver, have you planned to push those? [11:25] probably should go upstream [11:25] you mean the little segfault fixes? [11:25] generally all of those that I did at least have a bug reported upstream [11:26] yeah, those ones. at least they weren't in 1.6.1.901, might be in master [11:27] well, I'd found that upstream (i.e. whot) didn't like the approach I was using to solve them [11:27] usptream would rather leave the crash in until the root cause was solved, and not just paper over the issue [11:27] oh, right.. [11:27] in the distro though, I was happy to paper over the issue since crashing is bad ;-) [11:28] so... I made sure the bugs were registered upstream but didn't bother putting in the patch if I thought upstream would rather figure out the root cause than paper over. [11:28] yeah [11:28] in one or two cases they did that, and the better patches are already upstream [11:29] I chatted with pitti about this approach, and he agreed it was better for us to stop the crashing than to wait for the final solution [11:29] yep, makes sense [11:30] I hope to have more time this release to do more of these. We've got a ton of crash reports that could probably be solved this way [11:36] I haven't done much bug triaging for a while, but I've seen the bug count getting higher again.. [12:53] tjaalton: indeed, I'm worried it's going to get out of control again [13:38] Is it possible to run 2 seperate X servers (not screens) on 2 heads, on 2 graphics cards? [13:38] in theory, yes [13:39] any more information i can gleam on that? [13:39] google for multiseat [13:39] i currently use seperate X screens and the bugs make me want to hurt people, i figure completely seperate X screens might be better [13:39] ah, i don't exactly want multiseat though, but i admit its close [13:40] i only want one mouse/keyboard [13:40] i guess that's a good starting point though, thanks [13:41] err. with only one mouse/kbd, how do you think this can work? [13:41] hopefully by moving my mouse across (similar to multiple X screens) [13:41] but obviously not moving windows across [13:41] well. no. [13:42] why not? [13:42] there must be some way of switching what seat your on [13:43] the seats have their own mouse/keyboard.. [13:43] what happens if you try when you only have 1 keyboard/mouse? [13:43] then you have no mouse/keyboard on the other head [13:43] and theres no way to switch which keyboard/mouse controls which head? [13:45] even if there is no way to switch, i can still do it [13:45] could use synergy (It's like a KVM switch, but software, and without the video switch) [13:46] and just tell it that the networked computer is localhost \o/ [13:47] any other ideas/suggestions about it? [13:47] not that I know of [13:48] hehe [13:48] the reason is because i actually want to run the DE (in my case gnome) twice, completely seperately [13:48] because the way gnome handles multi X screens is appauling and nobodys going to fix it, and i'm fed up with waiting :) [13:49] most DE's suffer from the same issues, though [13:49] you could fix it yourself... [13:49] i had a look at the code, but not being a C programmer i couldn't even figure out what file i was supposed to look in [13:50] and the bug report is 3 years old, and the developer is "too busy" so, this seems like the next best bet really [14:32] bryce_: you seem to have something left to push to xorg.git [14:32] the changes in 5ubuntu20 [14:44] tjaalton: nope [14:45] oh [14:45] something is wrong on my end [14:45] ah, there it is [14:55] tjaalton: ideas on if removing x's hal dependency is feasible in karmic's timeframe? [14:56] bryce_: are you on the "improving boot speed" session?-) [14:56] just lost the stream [14:56] yes [14:57] I haven't looked at how to use dbus directly, the xserver supports that [14:57] don't know if there are limitations [14:57] scott was suggesting keith was looking at using udev in place of hal [14:58] oh, even lower [14:58] he had also mentioned that kms would help move a chunk of startup time out [14:58] yes [14:58] that'd run earlier on during kernel startup, in parallel with module startup [14:59] guess we ought to resurrect the timing patch (and fix the hang bugs that were in it) [15:00] end of session [15:00] yeah, got the stream back just in time for that [15:01] using udev directly means losing solaris and bsds, so i'm not sure how that can fly upstream.. [15:01] true [15:03] hal is going to be deprecated anyway .... [15:04] crevette: well if there's no sensible replacement.. [15:04] http://lists.freedesktop.org/pipermail/xorg/2009-May/045590.html [15:05] jcristau: the deal is to use libudev apparently and others have to find their solutions [15:05] crevette: and i'm saying that 'use libudev' is not a sensible replacement [15:05] adn DeviceKit is a daed-end too :) [15:16] jcristau, #ifdef __LINUX ? [15:18] bryce_: that's where an abstraction layer comes in === ripps_ is now known as ripps [23:49] hi, guys. after some drama last night that tjaalton helped me fix, i've become obsessed with my ati 200m card. if i post my xorg.conf can you guys tell me what to do to speed it up a little?