[02:36] hey guys I am not sure about this... which aspect of ubuntu would be responsible for right click drive icons with no unmount available? [02:43] coz_: That would be something on the desktop; probably nautilus. [02:45] RAOF, ah ok thanks I was getting confused about who would be responsible :) [03:03] Hey, asac. Bouncy network? [03:23] i think he switched to your old isp :) [03:32] Man, anything touching i965g in any way is a one-way ticket to GPU hang. [03:32] anyone know if anyone has started debianizing mesa/demos? kind of hesitant to update mesa in edgers now with it gone [03:33] i'm not even putting that in libgl1-mesa-dri-gallium on edgers its so broken :D [03:33] Not at this point, no. [03:34] There's no release with mesa/demos yet, is there. I'm not sure if Debian want a snapshot, even in experimental. [03:34] However... asking in #debian-x would make that clearer :) [03:36] i guess it wont remove the old mesa-utils on upgrade and it doesnt really change much so its not that big a deal [03:42] Yeah. It shouldn't use any non-public ABI, which means that it'll survive just fine. [03:42] Once there's a release we can split out the demos, and there will be much rejoicing. [03:43] Man, why doesn't debian have dbgsym infrastructure yet? :( [03:52] does debian still require locally built and uploaded binary packages for amd64? [03:53] or am I crazy and did they never do that? i haven't looked into it as much as I should have [03:54] I don't think they ever* required that. [03:54] * For values of “ever” extending multiple years back. [03:55] You *are* required to locally build and upload a binary package, from what I can gather. [04:34] Sarvatt: Hey, did you ever end up packaging libkms? [10:16] hi Bernardo [10:19] RAOF: Hi! [10:24] Has there been any progress on the Mesa package with OpenGL ES support? [12:59] bryceh: you declined Bug #563100 a while ago, believing it to be a bug in the proprietary nvidia driver. upstream, it was declared an X bug, and there is a 1-line patch to fix it. can you give it another look? [12:59] Launchpad bug 563100 in xorg-server (Ubuntu) (and 1 other project) "Mouse movement corrupted with Xinerama enabled (affects: 33) (dups: 5) (heat: 232)" [Medium,In progress] https://launchpad.net/bugs/563100 [13:00] bryceh: also, others on lp are adamant that there's not really a proper workaround since nouveau can't handle having 2 graphics cards in use at once [13:03] do you expect ati multi-gpu configurations to work? [13:08] no idea. have they ever? [13:09] oh wait thats not in context of the bug i was just talking about for nvidia users [13:09] (since you just /join'd) [13:09] ignore me :) im not an X person [13:12] * akoya-6805 ignores maco [13:13] maco, interesing...not x person either but I have noticed an issue with cursor on both single an dual monitor systems here... not responding well or jumping about the screen particularly on single monitor [13:28] Howdy [13:28] you guys running the ubuntu-x-swat ppa? [14:02] akoya-6805: it should work with fglrx [14:06] not if one card is "legacy" ;) [14:09] #radeon guys said i'need the vga arbiter... [14:11] akoya-6805: you can't use fglrx and radeon at the same time, therefore, when I said "fglrx", I excluded this use case [14:12] if you use Lucid, the arbiter is available. I'm not sure if radeon supports multiple GPUs [14:13] fair enough [14:13] sure it does. [14:13] with kms it's generally expected to work [14:14] (accourding to #radeon adamk) [14:14] I've never tried myself. I guess it's just RandR that doesn't make use of multiple gpus [14:15] http://pastebin.ubuntu.com/439310/ [14:15] see for yourself, although that is with ums [14:16] ok [14:17] akoya-6805: so what's the problem? [14:17] or were you just asking? [14:19] i guess my question had been answered ;) [14:19] but:now i have to troubleshoot kms [14:19] reading on that now... === |thade| is now known as Alexia_Death_ [17:08] hi guys. Is there somebody awake here atm ? [17:10] what an amusing article on phoronix about the openchrome driver, or whatever it's being called now [17:11] https://bugs.launchpad.net/ubuntu/+source/firefox/+bug/562802 => any idea whats causing that ? seems to only happen with nvidia driver, not with nouveau [17:11] Launchpad bug 562802 in firefox (Ubuntu) "xorg use cpu 100% sometime with firefox (affects: 4) (heat: 24)" [Undecided,Incomplete] [17:24] Hi, anyone know how I'd hide the mouse pointer permanently? [17:25] It's for a touch screen device, and there's no mouse going to be connected. So the mouse pointer just gets in the way [17:57] chrisccoulson: have you seen this? http://cvs.fedoraproject.org/viewvc/devel/xorg-x11-server/xserver-1.7.0-randr-gamma-restore.patch?view=markup [17:57] regarding the gnome-screensaver failing to restore gamma the second time problem [17:58] Sarvatt, that looks like it might fix it [18:01] Sarvatt - could you point RAOF to that when he's online? [18:01] yepyep [18:03] thanks [18:13] mesa dropped 3MB off the tarball size by dropping demos, i was deleting progs/objviewer before making the tarball before too so it's probably closer to 10MB+ :) [18:32] Sarvatt, did that adjustment to powermizer fix your gpu? [18:32] i need closure on that anecdote [18:44] somebody has any idea on my issue ? [19:00] nobody ? :D === rickspencer3_ is now known as rickspencer3 [19:32] bjsnider: i haven't had time to test it, will test it now [19:36] mvo, Sarvatt, I've put in a suggestion that the xorg-edgers ppa-purge might be worthwhile to add to python-software-properties: https://code.edge.launchpad.net/~bryceharrington/software-properties/rm-apt-repository/+merge/25988 [19:37] mvo, Sarvatt, and rename it to 'rm-apt-repository' so it is matchy to 'add-apt-repository' [19:52] bryceh: nice, thanks for that! [19:52] bjsnider: no it didn't fix it unfortunately :( [19:57] that's odd [19:57] what are you saying, that mobile chip users have to use an old blob? [19:58] i'm trying 173 out now to see if it fixes it, 195 and 256 for sure are screwed up [19:59] 256 even moreso, metacity is unusable with it [19:59] with 195 just gl/glx are screwed up [20:02] http://sarvatt.com/downloads/nvidialol.png [20:02] thats what 256 looks like [20:02] didn't had issues with 256 btw, installed it yesterday [20:02] are you on a laptop? [20:02] yep [20:03] what GPU? [20:03] 01:00.0 VGA compatible controller: nVidia Corporation G71 [GeForce Go 7900 GS] (rev a1) [20:03] it's only the panel that's a bit screwed up [20:03] dupondje, the opengl version you're getting from glxinfo is 2.1 right? [20:03] bjsnider: yeah, unfortunately theres no way to fix it so it always looks like that, 195 gets that same corruption maybe 1/10 boots [20:04] bjsnider: can't check anymore, removed it again, as it didn't fix the issue I had with firefox [20:05] dupondje: tried unpatching cairo? [20:05] rebooting into 173 now, lessee if it works [20:05] Sarvatt: nope [20:06] dupondje: does it happen with an upstream firefox build? [20:06] or just the ubuntu ones [20:06] i'd try rebuilding firefox without the cairo patches [20:07] like have you tried firefox 3.6.whatever straight from mozilla? [20:10] Sarvatt: tried daily from mozilla ppa, and same issue, gonne try a version directly from mozilla [20:10] yeah the daily ones have the same patches [20:12] compiz looks lovely - http://sarvatt.com/downloads/nvidialolcompiz.png [20:12] 173 isn't any better :( [20:15] sarvatt: http://www.openbsd.org/songs/song39.ogg :p [20:19] Sarvatt, wait a minute now...let's think about this. 173 isn't any better? that code hasn't been touched in a very long time. why couldn't this be a hardware problem? [20:20] i'm not putting it past being a local problem i'm missing (old libs hanging around or something), setting up a clean lucid liveusb with persistant storage to try the blob out there now [20:21] and lol virtuald :) [20:21] i did use to install the drivers from nvidia.com manually for years on this thing [20:21] that's right because you said it doesn't happen on a livecd [20:22] it didn't start having this problem until the move to alternatives in lucid [20:23] well more specifically when we moved to 195, it worked fine when it was first moved to alternatives [20:23] this is why i clean install the root partition every 6 months [20:23] sometime in early feb/late january [20:23] rolling upgrades work in theory but not in practice [20:23] i dont see *anything* it could be though [20:24] bjsnider: tell that to the people who've been upgrading for 10+ years. [20:24] well, they always seem to have problems they're willing to live with [20:24] *shrug* [20:25] i've been using ubuntu as a rolling release switching to the new dev release as soon as it opens since intrepid at least with no problems [20:25] apparently not [20:25] not sure thats actually the problem here :) [20:26] hmm what's /usr/lib/nvidia/pre-install [20:26] Sarvatt: same issue with firefox straight from mozilla :( [20:28] dupondje: ah ok so its not the cairo patches.. [20:28] scrolling on launchpad is a disaster [20:28] Xorg going to 100% cpu, and lagging like hell [20:29] happen in other browsers too? [20:29] cant even use firefox at the moment to test it since all my machines are on maverick and firefox is broken because of the client side decorations :D [20:33] its better in Chromium it seems [20:34] max 30% cpu load or so [20:34] do you have a quad core? [20:35] dual [20:36] is it any better in a newer firefox or anything? 3.7? [20:36] same issue it seems [20:36] but gonne try to switch to nouveau now, as I didn't had that issue on it before I switched to nvidia [20:36] or is there any firefox release where you can't reproduce it so you can rule out things besides firefox being borked? [20:38] brb [20:41] and indeed [20:41] nouveau has no single issue [20:41] oh you just had the problem with the blob? why am I not surprised? :D [20:42] was it just with compiz enabled too? [20:42] no compiz enabled [20:43] not on nvidia nor nouvea [20:43] u [20:43] :) [20:45] so fast with nouveau :) its a dream :D [20:47] sadly nouveau doesn't have power control, it kills my battery :D [20:55] Sarvatt: we can conclude that the nvidia driver is just 'broken' ? :( [21:23] 32 bit livecd - compiz isn't garbled but causes lots of pauses with the WAIT in xorg.0.log, glxinfo still locks up the system for 10 seconds or so.. argh :( [21:23] time to try karmic out again [21:36] throwing another wrench in it - when i deactivated nvidia-current and activated nvidia-173 it left nvidia-current installed and set up, no wonder if wasn't any different :) [21:41] hey, is there a semi-official ppa with updates for X drivers? [21:41] like newer version of nvidia 195 driver for lucid [22:03] wow, that sucks.. it actually is a hardware problem i'm having. karmic is screwed up too and so is windows [22:03] kklimonda: x-updates [22:04] https://launchpad.net/~ubuntu-x-swat/+archive/x-updates [22:07] i think tseliot put 195.36.24 in -proposed too if you dont want to jump to 256 [22:09] Sarvatt: either I'm blind or there is no 195.36.24 there - only a beta driver. I know that 195.36.24 may get uploaded to lucid-proposed at some point but now I don't see it anywhere. [22:10] i still have the sources around for the 195.36.24 that was in x-updates before if you want it [22:10] let me find a PPA to throw it in [22:11] urg, nouveau 3D is working fine, what the heck [22:11] Sarvatt: I don't need them - I'm just writing a small howto on updating sound and graphics drivers for stable Ubuntu releases. Right now on my LoCo forum there are a lot of people suggesting compiling alsa from sources and installing nvidia drivers from nvidia.com [22:22] well that place is x-updates, I may have jumped the gun updating to 256 early but the same could be said about upgrading to 195.36.24 because I did it while those were beta too :) the 256 ones will probably be branded official instead of beta in a week given the track record for the past year worth of updates [22:24] All we need is nouveau ;) [22:27] yup nouveau 3D actually works \o/ [22:27] cant get 3D working right with the blob in karmic or lucid all the sudden for some reason [22:28] Sarvatt: with git version or what ? :) [22:28] yea xorg-edgers [22:29] sweet :) [22:29] Power Management would be sweet also but alot of improvement has been done already [22:30] need to use the 2.6.34 kernel in the ppa on lucid to use it though, not installed automatically [22:31] or Maverick ? [22:31] yea the one in the PPA is the maverick kernel copied over to lucid [22:32] nouveau hasn't been working in maverick stock for a few weeks now, needs to be rebuilt.. [22:34] its a dilemma again :) go Maverick with issues, but bleeding edge :) or stay Lucid :p [22:36] if you use gnome i would just use lucid+ppas for awhile, maverick is all kinds of messed up thanks to the gtk client side decorations stuff being enabled :) [22:40] new kernel version is being pushed to the ppa also it seems :) [22:44] yeah just saw it was uploaded to maverick and copied it over. pcie aspm finally enabled in a ubuntu kernel \o/ [23:00] what the heck is going on there - http://launchpadlibrarian.net/49099476/buildlog_ubuntu-maverick-i386.xserver-xorg-video-ati_1:6.13.99%2Bgit20100525.428125c0-0ubuntu0sarvatt_FAILEDTOBUILD.txt.gz [23:00] just happening on lucid.. [23:00] err maverick [23:02] Sarvatt: pkg-config bug [23:02] bugs.debian.org/583005 [23:02] happening in debian too? [23:02] ahh [23:03] gcc doesn't like -fvisibility\=hidden too much. [23:17] hah actually reset my 945 successfully while hung, echo "wedged : 0" | sudo tee /sys/kernel/debug/dri/0/i915_wedged [23:21] hmm gnome-control-center's appearance preferences gives a warning that the blob isnt installed and needs to be to activate desktop effects even though it works fine on xorg-edgers, wonder if thats in the desktop effects integration patch or jockey [23:22] looks like its the jockey-gtk --check-composite call [23:57] RAOF: you were asking if I packaged up libkms? http://sarvatt.com/downloads/patches/libkms.patch [23:57] it still doesn't build though [23:58] Sarvatt: Thanks. I thought that while I was futzing about with egl/gles/openvg I might as well get the kms framebuffer drivers working, too. [23:59] i've got libegl here too - http://sarvatt.com/downloads/patches/egl.patch [23:59] the egl/gles stuff needs a seperate branch of mesa 7.8 though i think?