[00:19] looks like the latest librsvg2-2 introduced a regression, I see bad rendering in my Lucid VM [00:21] http://img.photobucket.com/albums/v390/Looce/img-launchpad/ubuntu1004-badrendering.png === nhandler_ is now known as nhandler === jtechidna is now known as JontheEchidna === gw_ is now known as wise_crypt [08:54] LucidFox: seems like a likely enough thing to happen ^^ [13:41] I would like to know if anyone is up to date with ati drivers on ubuntu. Proprietary ones [13:41] They keep freezing my laptop, and a lot of people have troubles with them on ubuntu. [13:41] Where should I adress this, here or do I need to take it to Ati website? [13:45] ikorm: !ATI https://help.ubuntu.com/community/BinaryDriverHowto This guide and its subpages describe how to install the proprietary binary/restricted drivers provided by video card manufacturers. [13:45] ikorm: I think take to ATI support, not to here. [13:46] ikorm: unless it is a request to update a driver in the repo [13:49] anyone in here worked with python-apt? [13:50] i needex to know how to add a ppa with it === LucidFox_ is now known as LucidFox [14:34] Hi everyone. where is the file that holds the setting for session type GNOME | BLACKBOX. I want to be able to edit this setting via bash shell script, not via the login screen session drop list. [15:08] doko: could you check something with gas? [15:09] doko: https://bugs.launchpad.net/ubuntu/+source/binutils/+bug/587186 [15:09] Launchpad bug 587186 in gcc-4.4 (Ubuntu) "libc6 upgrade fails: illegal instruction" [High,Triaged] [15:19] rlameiro: you could look how apt-add-repository does it. [15:24] thanks geser, i will [15:53] my apologies if this is not the appropriate channel for these kinds of questions [15:54] i'm looking for the most graceful manner to disable app-indicators in order to resolve https://bugzilla.gnome.org/show_bug.cgi?id=621382 [15:54] Gnome bug 621382 in general "Ubuntu Lucid - indicators need to be killed before starting gnome-shell" [Normal,Unconfirmed] [15:54] killing seems like a pretty rude solution - i'm not using ubuntu myself, so any help would be very much appreciated [16:42] fmuellner: killing indicator-application-service [16:42] should be enough [16:42] kklimonda: thanks! [16:43] kklimonda: would that process be restarted with gnome-panel? [16:44] fmuellner: hmm.. let me see. [16:45] fmuellner: yes, killing gnome-panel, then killing indicator-application-service and starting gnome-panel again does restart indicator-application-service itself. [16:45] kklimonda: cool, thanks a lot! [17:32] pitti: are you working to the cdbs merge? === yofel_ is now known as yofel [19:53] Hmph. I wish there existed a timezone aware ssh hack :) [19:53] e.g. Based on source IP geolocation set the timezone in the session === dendro-afk is now known as dendrobates === dendrobates is now known as dendro-afk [21:13] it's time for plymouth to go. [21:14] it can't even do its job without specific drivers anymore [21:14] o rly? [21:14] https://bugs.launchpad.net/bugs/593408 [21:14] Launchpad bug 593408 in plymouth (Ubuntu) "nothing is visible on screen during boot sequence with fbcon disabled -- really annoying when the system decided to run a periodic fsck" [Undecided,New] [21:15] fbcon has always been broke on this system so don't get any bright ideas [21:16] so get it fixed. [21:16] fbcon isn't a "specific driver" [21:16] it's a generic one. [21:17] fyi, i'm not using nouveau on my desktop, and it displays the terrible looking lowres splash, but still works. [21:17] I got tired of the crashes [21:18] fbcon-vesa is misaligned [21:18] fbcon-vesa is probably displaying at 640x480 [21:18] how about filing a real bug report about that? i'm sure the nouveau people will like more information. [21:18] which you can perform a (separate) monitor auto-adjustment for [21:18] monitor auto-adjust isn't going to work in this case [21:19] what you're saying is that "X technology doesn't work without Y technology, which should work for everyone, but due to a few bugs can't work for me. so let's remove X technology instead of fixing Y technology" [21:19] i'm sure you understand what's wrong with what i just highlighted. [21:19] nouveau is in active development yes, and they need hardware dumps [21:19] hyperair: I think "plymouth uses fbcon which is more than the old bootsys that just needed a VGA text console, which causes regressions" is fairly valid of a complaint. [21:20] of course, "we should remove plymouth" isn't a solution [21:20] well since the hardware is actually bad [21:20] nouveau isn't going to be able to fix it [21:20] Does Nouveau work while in X/GNOME/whatever? [21:20] it's unstable [21:21] If the hardware is bad... what are the Nouveau guys to do about it, in software? [21:21] *boggles* [21:21] jdong_: oh how i wish we can take such an outlook on everything new on ubuntu that causese regressions. but yes i agree. [21:21] the motherboard designers put two devices on the same address and wrote special windows drivers to make it work [21:22] jdong_: in particular, i'm annoyed that plymouth doesn't give me my usual verbose initialization messages. [21:22] the nouveau project can implement that split too, even if not right now [21:22] I can't afford to wait [21:23] seriously I have so few services that I don't need the parallel boot sequence [21:23] how hard would it be to check [ -x /sbin/plymouthd] || run each boot service attached to the console in turn [21:24] * CynthiaG thinks she misunderstands Plymouth versus Upstart, or that you do [21:24] Plymouth is just the splash screen that messages get written onto, and Upstart handles the parallel booting... at least Upstart did that in Karmic [21:24] CynthiaG: In karmic that was right [21:24] in lucid it doesn't seem to be [21:25] can't you apt-get purge plymouth, though? [21:25] I tried. [21:25] it wants to remove e2fsprogs [21:26] huh [21:26] e2fsprogs depends on plymouth? :| [21:26] sounds like a very ugly kludge [21:27] e2fsprogs doesn't depend on plymouth [21:27] however, mountall does [21:27] Ah [21:27] e2fsprogs depends on sysvinit-tools which depends on upstart which depends on plymouth [21:27] so the bug should be on mountall. [21:28] er no [21:28] upstart doesn't depend on plymouth [21:28] Confusing dependency chain is confusing? :( [21:28] upstart depends on mountall [21:29] mountall depends on plymouth [21:29] mountall is the buggy one [21:29] it should function without plymouth [21:29] so file a bug [21:29] hyperair @ un214 or me? [21:29] whoever [21:29] * hyperair isn't really interested in the issue. [21:30] whoever's interested, file a bug. [21:30] I wouldn't be such a great source of information on the issue because I'm not having it :\ [21:30] heh [21:30] CynthiaG: you can force the issue by adding alias fbcon off to /etc/modprobe.conf [21:31] or blacklist fbcon [21:31] un214: even in a VM, you think? [21:31] a VM of the LiveCD [21:31] hyperair: I have another bug for blacklist fbcon doesn't work [21:31] huh [21:31] believe me I tried that one [21:31] weird. [21:33] actually, never mind, I've just created a virtual hard drive and will try what you say === lifeless_ is now known as lifeless [21:43] I was talking to some others earlier and they say the dependency is reall -- w/o plymouth no boot program can talk to the console [21:50] I really wish they had bolted the parallel init on top of sysvinit -- it'd make it a lot easier for me to find out what scripts run at boot time in what order [21:52] what order... in a parallel init system? :) [21:53] that statement hurts my head [21:54] I need to know the critical ones bucause if you can't fix this I get to replace all the bootscripts [21:54] like saying I want different process contexts, but they all share the same memory [21:55] besides, there exists a partial order (http://en.wikipedia.org/wiki/Topsort) [21:55] the only ordering is between jobs that depend on each other... which you can see in the init script [21:56] un214: creating /etc/modprobe.conf with the line "alias fbcon off" in it does absolutely nothing [21:56] oh that's right they moved it [21:56] create a file in /etc/modprobe.d instead? [21:56] I actually ended up adding it to /etc/modprobe.d/blacklist.conf [21:56] psusi: don't say depend on each other or keybuk is gonna come in here screaming at you ;-) [21:56] jdong, hey... care to sponsor the defrag package? ;) [21:57] psusi: whoa, where's this defrag package?? [21:57] jdond_: there is such a thing as co-depends [21:57] lol... it may not be 100% accurate, but still a useful generalization concept ;) [21:57] un214: I'm not arguing with that, it's not proper terminology for Upstart :) [21:57] but yes, you can do a topological sort by Upstart events :) [21:58] jdong, didn't see my post the other day to the motu mailing list? launchpad.net/e2defrag is the project.... has link to bzr repo and source tarball for the release I made the other day... binary build in my ppa [21:58] I used to have a script that did that [21:58] psusi: neeeeat; no unfortunately I've been swamped with work lately [21:58] and I've even tested it on up to 2tb disks [21:58] sweet [21:58] made some changes to allow that without needing 6gb of ram [21:59] un214: /etc/modprobe.d/blacklist.conf +alias fbcon off still does absolutely nothing for me, in the VM [21:59] CynthiaG: did you run update-initramfs afterwords [21:59] and in the process ran into a kernel bug... I was trying to go over 2tb and found that lvm snapshots silently wrap around at 2tb [21:59] un214: oh, no I didn't [21:59] I'll do that [22:00] you can use snapshots of a zero target to create a virtual disk of huge size with a smaller backing store... was going to use that to try making a 4tb disk to defrag, but it's all hosed up because of the wraparound [22:02] aha un214, now it did something! the boot screen was green, and I can see absolutely nothing instead of the login screen [22:02] ok you hit the situation but not the bug [22:02] that reminds me... I need to poke Keybuck about merging my ureadahead changes [22:02] hit Alt-F1 to get login [22:02] I heard the drum-like thing that serves as the login ready indication [22:02] Ubuntu 10.04 LTS abra tty1 || abra login: [22:02] somehow I got that green screen broken [22:02] and I still haven't heard from Ted about my e2fsck patch.... [22:03] that's what virtual terminal 1 says [22:03] CynthiaG: my bug is somehow I'm no longer getting the green screen [22:03] hmm [22:04] What are you getting instead? [22:04] blackness until login prompt [22:04] well, in any case, it's not the Ubuntu logo and the five circles [22:04] which is to be expected since blacklisting fbcon [22:05] I'll bet that if I got fsck to ask me any questions during boot it hangs forever [22:05] However, I only get a login prompt for GNOME (vt7) after I do like you said and switch to vt1 [22:05] I run KDE and don't have the problem [22:06] (*dm login not working) [22:06] it would probably hang forever if fsck needed to ask you anything, yes [22:06] but I think the on-boot fsck is non-interactive [22:06] until it needs to boot you to a maintenance shell, that is. And then I don't know what it does, but it should switch to a sane terminal [22:07] CynthiaG: last I checked maintenance boot doesn't use plymouth [22:08] I don't mean a maintenance boot (single-user), I mean fsck having unrecoverable errors and throwing you into a maintenance shell from a normal boot :) [22:08] who knows what that's gonna do [22:08] Errors such as "mount time in the future" from Karmic [22:08] My money's on wrong console acrtive [22:09] Those threw me to a maintenance shell during the alphas [22:09] that shouldn't be an error -- it means the system clock is off until it can sync to network [22:09] I know, but they were still considered unrecoverable in Karmic [22:10] ewww [22:10] Aye [22:13] Off-topic fun fact: Apparently a Lucid install needs to read 157 MB of data between boot and the GNOME desktop. [22:15] I'm gonna play a hunch and see if I get usable feedback with chmod -x plymouth [22:16] test will take awhile [22:19] what does marking plymouth non-executable do? [22:20] prevents it from being run, without necessarily removing the package [22:20] because removing plymouth the package would also delete e2fsprogs because of a dependency [22:22] and how will that help him see messages that otherwise plymouth is unable to print?? [22:23] s/he is testing a bug that is making plymouth unable to print things anyway [22:23] https://bugs.launchpad.net/bugs/593408 [22:23] Launchpad bug 593408 in plymouth (Ubuntu) "nothing is visible on screen during boot sequence with fbcon disabled -- really annoying when the system decided to run a periodic fsck" [Undecided,New] [22:24] and with fbcon enabled, it's apparently misaligned enough that still nothing is visible, or something [22:24] well not allowing plymouth to run isn't gonna cause the boot system to print anything either [22:25] *boggles* :) [22:25] I wonder what that person wants to achieve now [22:26] well chmod -x /sbin/plymouthd resulted in the boot feedback showing up again [22:27] In a console? [22:27] Er. Text console. [22:27] yup, right on tty1 [22:27] I was unaware of such a fallback except for init scripts [22:27] So that's what you wanted to achieve. [22:27] not quite [22:27] stuff like mountall IIRC directly called plymouth APIs to output [22:28] I've got to go find the options being used to call fsck and change them [22:28] CynthiaG, sounds about right... that's why I've been working on using defrag to pack all that data in order at the start of the disk so it gets pulled in faster ;) [22:28] psusi: So e2defrag is also a boot-sequence optimiser? [22:28] CynthiaG: psusi is combining it with ureadahead to make it such :) [22:28] if you don't want a gui boot console, just pass "text" on the kernel command line and it will remain in 25 line mode [22:29] Oho! That sounds awesome [22:29] jdong: looks like I get to go break mountall then [22:29] I'll bet he's the one calling fsck anyway [22:29] or probably just find the 9.04 version anyway :) [22:29] CynthiaG, not exactly... but you can give it a list of files that should be given priority when packing... such as a list pulled from the ureadahead pack file ;) [22:29] psusi: tried that doesn't work [22:30] psusi: Sounds even more awesome :D [22:42] ugh goodbye mountall [22:43] replacing it with fsck -a && mount -a [22:43] you mean fsck -p === jtechidna is now known as JontheEchidna [22:43] no actually I mean fsck -A [22:44] boot time fsck is run in prone mode... not to mention that the rest of the system won't boot without mountall [22:44] s/prone/prune [22:44] preen [22:44] err, yea, that one ;) [22:44] shit... I'm getting old. [22:44] you want to fix mountall to not depend on plymouth? [22:45] no [22:45] remember, plymouth is broken for me [22:46] you should probably fix it then [22:47] can't determine why its broken [22:47] * psusi watches defrag do its thing [22:50] \o/ [22:57] anyway reading the plymouth source made the problem obvious [22:58] /bin/plymouth was damaged [23:00] I'm still probably gonna make a mountall-noplymouth as this thing's bloody stupid [23:05] cjwatson, have you seen my e2fsck patch in bug #556621? Wondering if you had any thoughts on it. [23:05] Launchpad bug 556621 in e2fsprogs (Ubuntu) "lazy_itable_init not on by default" [Wishlist,In progress] https://launchpad.net/bugs/556621 [23:10] cjwatson, also could you possibly help move bug #568050 along by uploading the fix to lucid-proposed? This was a serious regression that prevents installation for effected users so I'd like to see it make it in for the 10.04.1 respin [23:10] Launchpad bug 568050 in parted (Ubuntu) "Ubuntu 10.04 can't create partition on fakeraid" [High,In progress] https://launchpad.net/bugs/568050