[00:01] RAOF: I don't understand the question. [00:02] Darxus: Those tests are presumably enabled in the Ubuntu packages, yes? If so, when the package is built for Ubuntu they'll be run, and presumably *pass*. You could check out the buildlog of the Ubuntu GTK package to see if that's happening. [00:03] RAOF: Ah, yes, interesting, thanks. [00:03] But it sounds like my problem is just that I'm using dpkg-buildpackage -rfakeroot instead of a build target. [00:04] Yeah, just booting into quantal and running it there, without the chroot, got past that problem :/ [00:37] Uploaded and rebult: https://launchpad.net/~darxus/+archive/wayland-gtk-quantal [00:37] Thanks for your help. [00:59] Would it work to just run the build through dbus-launch? [00:59] (To fix my problem with building in a chroot.) [01:48] /12 [01:50] ugh! === charles_ is now known as charles [04:21] Good morning [05:23] good morning [05:24] bonjour didrocks, ça va? [05:25] gutent morgen pitti! Ça va bien, et toi? [05:25] Je suis bien, merci! [05:25] I hope to get udisks2 to succeed today, on https://jenkins.qa.ubuntu.com/view/Quantal/view/AutoPkg%20Test/ [05:26] this page becomes a real joy to look at :) [05:26] (tabix is fixed upstream already) [05:26] hehe, you are in the green business, right? :) [05:26] J'aime la couleur vert! [05:27] (or is it "verte" in this case, as couleur is female?) [05:30] verte dans ce cas, oui :) [05:56] Damnit! gdb, give a reasonable backtrace! [05:57] Is anyone else seeing semi-regular crashes in libdbus, with a backtrace full of ??s that can't be retraced? [05:57] I'm starting to suspect some form of memory corruption. [05:57] I think I looked at backtraces like that [05:58] They look something like http://paste2.org/p/2200544 ? [06:01] gwibber-service is fond of crashing there, as is tomboy & banshee. [06:20] RAOF: I think you got the right bug # :) [06:21] RAOF: I see a compiz task, anything needed on their end? [06:21] I see the xserver-xorg-video-intel to be fix released [06:22] didrocks: I don't believe so, no. I'll invalidate the compiz tasks. [06:23] RAOF: hum, is it really fix released into the distro? [06:24] I find all the status change and revert quite puzzling :) [06:24] didrocks: I'm unsure; tjaalton knows more. I think there was at least *one* fix released to the distro; I don't know if that's all of it. [06:24] let me check that [06:25] I think jasoncwarner is more or less up to date [06:25] so if he's still experiencing it… [06:40] tjaalton: any hint? [06:41] didrocks: reopened, no idea as of yet [06:42] tjaalton: ok, thanks :) [06:42] I confirmed I have it regularly [06:42] how regularly? [06:42] like, after 5/6 suspend [06:42] ok [06:42] I didn't get it in the last week because I didn't suspend as much before rebootin [06:42] rebooting* [06:43] I tried to reproduce it by turning the screen off/on with xrandr, but I'm hitting _other_ bugs that way [06:43] and different ones depending on the intel generation.. [06:43] talk about frustration [06:43] "nice" [06:43] tjaalton: turning the screen off/on, I didn't get it, really suspsend/resume [06:44] ok, wonder what it's doing differently then.. [06:44] yeah, I'm not sure, maybe turning the screen off/on can triggers it as well, just never experienced it that way [06:45] it's not just the dpms cycle then, different bug that way [06:45] but I'll script suspend/resume and try to hit it while xtracing compiz.. [06:45] you, and resume, basically, you end up either with: [06:45] - black screen, and if you kill gnome-screensaver, sometimes the background [06:45] - or just the background [06:46] you have to kill compiz and starts metacity [06:46] oh if you just have the background, then it's the same I get with xrandr [06:46] restarting compiz doesn't work [06:46] or maybe not [06:46] glxgears works [06:46] a vt-change resumes it [06:46] no, it doesn't have any impact here [06:46] this with snb, 965gm hangs properly [06:46] but some other games, more gl demanding don't [06:47] I really have to reboot to put back the driver in a good "gl state" [06:47] hum, I have an x220, let me see what chipset [06:47] yeah that metacity thing is what I've mostly seen [06:47] snb [06:47] same as on my t420s [06:47] 00:02.0 VGA compatible controller: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller (rev 09) [06:47] if this is snb :) [06:47] still running precise there, and seeing this [06:47] yep [06:47] yeah, I got it once or twice in precise [06:47] but really rare [06:47] a lot more in quantal [06:48] another sandybridge on quantal, 965m on quantal [06:48] thumper said he's seeing it on every login or so [06:48] and I couldn't reproduce it with the new unity stack [06:48] ok, should be put the bug on the release tracker list? [06:48] just the normal way, ie. not that frequently [06:48] yeah [06:48] seems quite important [06:48] doing it then, thanks! [06:48] yes, oem-important too [06:49] tjaalton: when I'm in that state [06:49] can I give you any relevant info? [06:49] don't think so, we have the trace from compiz being hung, xserver waiting [06:49] ok [06:49] need to know what compiz was doing [06:49] and ttbomk xtrace should tell that [06:50] ttbomk? [06:50] to the best of my knowledge :) [06:50] oh, didn't know that one :) [06:50] ok, I'll try that [06:58] good morning everyone [06:58] after using my desktop for a while, all the indicators become very slow, I mean when I click on an indicator it takes several seconds to display the menu. Is it a known issue ? [06:59] jibel: not that I noticed/saw. Maybe check with larsu when he's online? [06:59] hey chrisccoulson [06:59] if not, we can add it to the release list [06:59] bug* [06:59] didrocks, ok, will do. thanks [06:59] hey didrocks, how are you? [07:00] chrisccoulson: I'm good, thanks! yourself? how was your flight? [07:00] didrocks, uneventful ;) [07:00] heh, no strike, nothing then? :) [07:00] chrisccoulson: as you are with the thunderbird guys, can we remove the chat buttons please please please? ;) [07:00] heh [07:01] and also I got bug 1047117 with this morning's updates [07:01] Launchpad bug 1047117 in libgnome "package libgnome2-bin not installed failed to install/upgrade: trying to overwrite /usr/share/man/man1/gnome-OPEN.1.gz , which is also in package libgnome2-0 2.32.1-2ubuntu1" [High,Confirmed] https://launchpad.net/bugs/1047117 [07:05] ah, let me look :) [07:06] I see seb128 runs away on vacation after breaking post-beta1! :-) [07:18] jibel: fixed [07:19] didrocks, fantastic ! [07:48] pitti: do you think the recent updates in https://code.launchpad.net/~mvo/aptdaemon/support-change-credentials-on-add-repo/+merge/112098 are sufficient? I would love to merge/upload this soon(ish) [07:49] mvo: guten Morgen, wie gehts? [07:49] * pitti looking [07:49] pitti: guten morgen! gut, danke [07:49] pitti: und selbst ? [07:50] didrocks: btw, do you use the screen locker when suspending? [07:50] tjaalton: right, but even trying to kill it from tty1, doesn't change anything (perhaps just from the black background to the wallpaper, but I'm unsure) [07:51] meaning, compiz is locked [07:51] mvo: gut, danke! [07:51] didrocks: yeah I know, you can unlock it and then you see the mouse cursor changing. I was just thinking of how to best reproduce it [07:51] ran 11 cycles without the lock screen, now trying with it [07:52] let's hope that will help triggering it :) [07:53] indeed [07:53] mvo: MP updated [07:53] mvo: odd, I didn't get mail with your recent reponse [07:54] meh, testing with fwts doesn't lock the screen when it suspends [07:54] pitti: cool, thanks, I will fix the remaining bits next [07:55] mvo: as for the integration test caes with apt, that's harder to set up of course (writing an apt conf file and running apt-get, etc.); but as long as you tested it manually, that's fine [07:55] I don't think it needs to block this, it was just a question [07:56] pitti: thanks, its a good point, I'm not sure I will manage to find time for it currently but I do think it would be really good to have === tkamppeter_ is now known as tkamppeter [08:11] didrocks, oh, are you getting compiz hangs when resuming from suspend too? [08:12] hey chrisccoulson, how are you? made it to Warsaw? [08:12] chrisccoulson: yeah, seems we can start an union :) [08:12] hey pitti, i'm good thanks. how are you? [08:12] I'm fine, thanks! [08:12] i got to warsaw ok :) [08:12] didrocks, yeah, i keep getting that here :( [08:13] chrisccoulson: do you have any particular way to trigger it more so that tjaalton can debug? [08:13] didrocks, not really. it's quite random [08:13] chrisccoulson: what hw do you have? [08:13] tjaalton, intel [08:13] which generation :) [08:14] sandybridge, ivy-, ironlake, 965, earlier? [08:19] hi pitti, good morning and a quick question: I've got a 12.04 pc where jockey fails to install the nvidia driver. What's the best way to manually install nvidia and uninstall nouveau? [08:19] dpm: how does it fail? [08:19] dpm: "sudo apt-get install nvidia-current" should suffice [08:23] dpm: Careful; there are reasons other than Jockey bugs why Jockey might not offer to install the driver :) [08:23] that's why I ask -- whether it doesn't detect it, or whether it tries to install but the pacakge fails [08:25] pitti, at the end of the process, jockey tells me that it could not install the driver and that I should look at /var/log/jockey.log (will look at it in a minute). Installing the driver manually works (the installation part), but then I get into a black screen. Which is weird, as this is a fresh install on a computer where 12.04 + nvidia has been working before [08:26] perhaps you can pastebin jockey.log [08:28] pitti, can't get into the pc right now, it's got wireless network only and I cannot get into it through the command line, might take me a few minutes [08:29] dpm: wifi should be fine; we configure system connections by default now, so if you are in a VT you shold still have wifi === 16WAA23L4 is now known as jhernandez [08:30] ah, cool, let me try [08:34] didrocks: 30 cycles with the screenlocker, and before that maybe 20 without [08:35] waow, seems that you are in a luckier hw [08:35] chrisccoulson: yo uhave an x220 as well? [08:35] tjaalton: isn't sandybridge the mixed nvidia/intel one? [08:35] man, I can't believe I can't get into the GRUB menu, I've been hammering the shift key for 2 minutes [08:35] tjaalton: I only have the intel card here [08:35] didrocks: no, sandybridge is a codename for the intel part. you can have it with optimus or without [08:36] ah ok ;) [08:37] and I tested this with a dell v1450 on quantal, will try with the thinkpad on precise next [08:37] tjaalton: maybe you can try on a live quantal on the thinkpad? [08:38] RAOF, so on that computer I've got several issues: when the nouveau driver is used, the system becomes extremely slow, to a point that's unusable (mouse moves in increments that are minutes long). Any ideas what it could be, or how to fix that until I can install nvidia? [08:39] didrocks: that's possible yes [08:40] tjaalton: I think you will have higher chance to trigger it that way [08:41] and beta1 live image is soooo shiny :) [08:42] :) === lifeless_ is now known as lifeless [08:43] didrocks, chrisccoulson: do you use amd64 or i386? how much memory? [08:44] tjaalton: amd64, 8Go of memory [08:44] nice [08:44] GB* [08:44] quite new, yeah ;) [08:44] hmm no i386 images at all [08:45] i'm all amd64 as well [08:46] pitti, http://pastebin.ubuntu.com/1190411/ (the pc has got an integrated ati graphics card using fglrx, not sure if that interferes with the nvidia card I'm trying to get working now). Right now even after having installed nvidia-current manually, it seems to be loading nouveau or something else [08:47] dpm: hm, this log doesn't even have an attempt of installing the driver [08:48] dpm: guess you need to disable the integrated one from bios [08:48] dpm: so it might very well conflict with the ati card; can you disable that in the BIOS? what does lspci show? [08:50] tjaalton, pitti, unfortunately, I can't. But this system has been working in the past, with the nvidia card and the 'non-disableable' ati integrated chip [08:50] dpm: If you're using fglrx, no other video driver will work. [08:50] aaah [08:50] dpm: Same as nvidia; if you're using nvidia, no other video driver will work. [08:50] * dpm uninstalls fglrx [08:50] That should fix it :) [08:52] oh, missed that part.. :) [08:54] RAOF, just out of interest, though, any clue about nouveau's slowness (to the point that it's unusable)? That happened also during the LiveCD install (thus no fglrx installed), I had to remove the nvidia card and do the install with the integrated ati one [08:55] (it all works after uninstalling fglrx, thanks for the help, guys!) [09:10] dpm: Without logs it's not clear what's happening; if I had to guess, I'd suspect that something went dolally and you're using the llvmpipe software renderer rather than nouveau. [09:27] hiya [09:27] can somebody take a look at https://code.launchpad.net/~logan/ubuntu/quantal/notify-osd/debian-merge/+merge/118465? [09:28] also https://code.launchpad.net/~bcurtiswx/ubuntu/precise/telepathy-glib/0.18.2-0ubuntu1/+merge/117176 maybe [09:52] chrisccoulson, ping [09:56] pitti: may I suggest simplifying the NEWS entries for pygobject? I think it lists all changes, even the ones that are, well, useless to know about :-) [09:57] pitti: (ChangeLog is there for people interested in all the details) [09:57] (hi!) [09:57] vuntz: hm, I already take out stuff like "bump configure version"; I do like to describe the actual changes in succint form, with bug refs [09:57] vuntz: oui, et bonjour! ça va? [09:59] bien, bien :-) [09:59] pitti: "Update .gitignore"? :-) [09:59] vuntz: right, I forgot to take that out [10:00] pitti: "tests: Replace deprecated assertEquals() with assertEqual()" [10:00] pitti: I also see at least three lines about pep8 in the 3.3.1 entry [10:01] pitti: of course, what you do is already much better than a "git shortlog" (which many tarballs have as NEWS...) [10:02] vuntz: still, thanks for pointing out; I'll apply more care next tiem [10:03] it's never easy to find the right level of details, especially since pygobject is developer-oriented (compared to apps) [10:03] anyway. I'm glad to see so much love given to pygobject [10:03] pitti: you rock :-) [10:04] oh, btw, an alternative is to provide a short paragraph highlighting the important bits for distributors [10:05] (and keep the longer list, which can be useful for hackers) [10:53] didrocks: heh, locking the screen doesn't work on the beta1 live :) [10:54] jibel: known? ^ [10:54] ah, on the live [10:54] yeah, I think you have no ubuntu password [10:54] so desactivated [10:54] right [10:54] you should be able to force it in g-c-c [10:54] it's activated there [10:54] by default [10:54] interesting [10:55] idea [10:55] tjaalton: create another user in the live system [10:55] log with it [10:55] you should have a normal session then [10:56] yep, setting the passwd wasn't enough [10:56] starting with a fresh user will hopefully do it [10:57] same thing [10:57] but I'll just trace it from precise, should be the same bug :) [10:57] it=compiz [10:58] hoping you will trigger it easily on it, I just had it once there [10:58] (on precise) [10:59] we'll see [10:59] probably easier if I have two sessions open [11:01] ha, first cycle :D [11:01] \o/ [11:01] so I'd say something has changed for the better.. but still [11:03] well, I never experienced that on precise apart from one time, so it's really random and day-of-the-week related :) === MacSlow is now known as MacSlow|lunch [11:05] * MCR1 hates day-of-the-week related bugs, because he always hits those while noone else can reproduce them... [11:05] screen lock had nothing to do with it, had it off for this session [11:07] ok, at least, it's safe then :) [11:07] hehe, yes [11:12] interesting that switching to another running session after resume immediately suspends the machine again [11:12] both precise and quantal === MacSlow|lunch is now known as MacSlow [12:04] didrocks, not to my knowledge [12:05] urgh, today's updates broke my netboot again :( [12:05] *netbook [12:08] bug 1047306 [12:08] Launchpad bug 1047306 in unity "[Quantal] [Intel Atom] Unity 2012/09/07 updates broke it all" [Critical,Confirmed] https://launchpad.net/bugs/1047306 [12:20] JohnLea, bug 668382 [12:20] Launchpad bug 668382 in software-center "Unclear warning before removing critical packages" [Medium,Triaged] https://launchpad.net/bugs/668382 [12:21] mpt; bug 1047315 [12:21] Launchpad bug 1047315 in ayatana-design "Dash - All applications which are a dependency of Ubuntu Desktop should not have an uninstall button in their preview " [High,Fix committed] https://launchpad.net/bugs/1047315 [12:22] jibel: hum, unity didn't change since beta1 [12:22] jibel: do you have the list of things you updated? [12:22] meaning, you didn't test beta1 on that netbook? [12:23] didrocks, yes pasting a list of this morning's updates [12:23] didrocks, to question 2 - no, I need this machine [12:24] so, it can be a previous unity version which is guilty, right? [12:24] didrocks, no, it was broken, then fixed, then broke again today [12:24] "nice" [12:25] jibel: well, not today, but on Monday I guess [12:25] (you didn't update since monday, right?) [12:25] as it's the last unity upload [12:25] didrocks, bug 1042211 was my original report [12:25] Launchpad bug 1042211 in mesa "[quantal] [regression] [i915] Corrupted display, desktop and menus don't repaint correctly using Mesa 9.0 (8.0.4 works)" [Critical,Incomplete] https://launchpad.net/bugs/1042211 [12:26] didrocks, something broke between the 4th and 7th [12:26] jibel: please paste the list of your updates [12:26] I see no unity/compiz/nux uploads meanwhile [12:26] so maybe a mesa/xorg thing [12:30] didrocks, apt history https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1047306/+attachment/3298087/+files/history.log [12:30] Launchpad bug 1047306 in unity "[Quantal] [Intel Atom] Unity 2012/09/07 updates broke it all" [Critical,Confirmed] [12:31] jibel: yeah, so no unity/nux/compiz [12:31] but mesa [12:31] I think we know who to turn to :) [12:32] jibel: do you have time to rollback some mesa package to confirm? [12:32] jibel: anyway, I'm adding that and will ping RAOF/bryceh about those [12:33] didrocks, ack, I'll downgrade the mesa stack [12:33] jibel: thanks, keep me in touch [12:33] * didrocks targets for quantal === ralsina is now known as pyar === pyar is now known as ralsina [12:38] tjaalton: ^ [12:38] you might be interested as the upload :) [12:38] uploader* [12:38] should I assign this to you? [12:40] meh [12:40] maybe, but it's weekend now, nothing upstream unless 13b8eb64523a70af30b576271886a737f51dbe40 would fix anything [12:40] now/soon [12:40] tjaalton: If jibel confirms the downgrade is working, should we rather revert? [12:41] the chipset is quite common [12:41] and breaking that for the week-end… [12:41] that would reopen bug 1046933 [12:41] Launchpad bug 1046933 in mesa "glsl/linker: array buffer overrun [CVE-2012-2864]" [Undecided,New] https://launchpad.net/bugs/1046933 [12:41] though that could be added back [12:42] tjaalton: the CVE commit would be easy to backport you think? [12:42] a oneliner [12:42] not really big apparently, yeah :) [12:43] tjaalton: can you handle it or do you want me to do it? [12:44] either way [12:44] did beta already have the same unity/compiz? [12:45] ok saw the backscroll [12:45] tjaalton: yeah, I confirmed this [12:45] look at the list of upgrade [12:45] attached to the bug [12:45] seems that mesa is the only obvious possible offender, right? [12:46] ok [12:49] tjaalton: don't put too much pressure on you, if you can't catch what's wrong, we can go the easy way on the revert + the patch [12:49] hoping that jibel will confirm that with the older mesa stacks, it works [12:49] we'll see [12:51] tjaalton: should he revert to 9.0~git20120821.c1114c61-0ubuntu1 or 9.0~git20120821.c1114c61-0ubuntu2? [12:51] as 9.0~git20120821.c1114c61-0ubuntu2 was never published in beta1 [12:52] tjaalton: ignore my question :) [12:52] it wasn't? [12:52] ah [12:52] was puzzled by the "superseeded" date on launchpad :) [12:52] well, three commits to src/mesa/drivers/dri/i915 since the previous snapshot.. [12:52] could have been worse… :) [12:53] "mesa: Use a new, more specific hook for shader uniform changes." [12:53] sounds plausible [12:53] shader making an app segfault, sounds too familiar :) [12:56] I asked earlier, maybe now somebody can take a look at https://code.launchpad.net/~logan/ubuntu/quantal/notify-osd/debian-merge/+merge/118465? [12:56] and https://code.launchpad.net/~bcurtiswx/ubuntu/precise/telepathy-glib/0.18.2-0ubuntu1/+merge/117176 maybe :) [12:56] I can build a test package and if that doesn't work, a revert is possible unless Sarvatt can figure out the cause [12:57] dholbach: I'll find someone to review them [12:57] tjaalton: sure [12:57] * dholbach hugs didrocks [12:57] * didrocks hugs dholbach [12:57] robru: hey! [12:59] (repost from #ubuntu-unity): hey everyone, I have just noticed that classic desktop in 12.10 has wrong order of window control buttons -> in unity that order is [x] [-] [o] while in classic mode they are [x] [o] [-] -- is this by design? [13:00] didrocks, tjaalton I downgraded mesa to 9.0~git20120821.c1114c61-0ubuntu2 and the machine is back to a working state [13:00] thanks jibel :) [13:00] zyga: that's probably my fault, is there a bug # for that? [13:01] jibel: ok do you have a beefy machine to build mesa on? [13:01] I use gnome-shell too much :| [13:01] i guess atom is 32bit? [13:01] jbicha, not that I know, I wanted to ask before reporting it [13:01] jbicha, on gnome-panel? or on metacity? [13:02] I can fix it without a bug [13:02] k, thanks! [13:02] :) [13:02] tjaalton, there are 64bit versions of atom [13:02] jibel: are you on amd64? [13:03] I actually have one too, but not with that generation of gfx [13:04] jibel: ping? I can build mesa for you in 5min if you do [13:04] tjaalton, i386 [13:04] duh [13:04] why would i run 64bit on a netbook with 2G of ram [13:05] hmm wonder how long it takes to build a i386 schroot [13:05] jibel, to get more registers, atom has no out of order execution or register renaming IIRC and really suffers from performance as a side effect [13:06] yeah, performance on a netbook [13:07] ok mk-sbuild is running.. [13:08] don't tell bad thing about netbooks… It happened that unity built faster on one of mine even if I dputed at the same time to a ppa :) [13:08] (and it was inside a pbuilder) [13:09] pitti, can you upload cups-filters to Debian and to Quantal? Thanks. [13:10] larsu, hi, is there a way yet to connect to the messaging menu entry of an application? (not a source) [13:18] jibel: building [13:28] jibel: test packages being copied to http://koti.kapsi.fi/~tjaalton/lp1047306 [13:29] should be enough to just get liblg1-mesa-dri and -glx [13:34] jibel: ping? please test [13:37] sbte, what do you mean? Any application can appear in the messaging menu [13:37] ok I'm gone -> [13:38] didrocks: I'll check here in 30min [13:38] tjaalton: ok, thanks :) [13:38] larsu, yes, but if the entry is clicked, I don't know what signal is emitted [13:38] if any [13:39] sbte, ah! no, you don't get a signal, the messaging menu will execute the "Exec" line from the .desktop file [13:39] (same for the shortcut actions) [13:40] tjaalton, thanks. testing [13:40] larsu, why not let the application handle it? [13:43] sbte, you can handle it, as long as your application is single-instance (which it really should be if you want to integrate with the messaging menu) [13:43] larsu, it is [13:43] sbte, why exactly do you need it? [13:44] larsu, to unhide it [13:44] sbte, do you use GApplication? (or GtkApplication) [13:44] if it is single instance, you can wake up the main instance and unhide it [13:44] larsu, no, we're stuck with gtk2 because of a bugged gtk3 version in precise [13:45] sbte, precise doesn't have the new libmessaging-menu [13:45] hey kenvandine! [13:45] sbte, you're using libunique then? [13:45] hey didrocks [13:46] larsu, no dbus [13:46] kenvandine: how are you? [13:46] good [13:46] and you? [13:46] have a good vacation? [13:46] yeah, were great! thanks :) [13:46] small question on notify-osd [13:46] you missed all the fun :) [13:46] sbte, oh you're doing it manually. Well, just unhide your main instance when the application is started again. Where's the problem? [13:46] heh, I heard about that [13:47] I prepared a lot before leaving for the gsettings transition [13:47] that was the fun on itself TBH :) [13:47] hehe [13:47] larsu, we also support multi-instance emesene [13:47] does notify-osd has some merge-upstream like branches, or not? [13:47] in which case the messaging menu would fail I guess [13:47] it will start a new instance [13:47] didrocks, no [13:48] sbte, how should that work with the messaging menu? We don't support putting the same application into the menu multiple times. [13:48]