[00:13] bdmurray: curious - maybe it's fixed by the freetype upload which changes various bits surrounding font metrics [00:27] jdstrand: still around? [00:28] roaksoax: yes [00:29] micahg: Did you ever get a chance to look into blueman? [00:31] jdstrand: quick question about creating a user for MAAS. [00:31] jdstrand: so the daemons already run as 'maas' user, WSGI for MAAS code also runs as 'maas' user. Do we need to create a home dir for the user? or can we not have a home dir? [00:33] ScottK: not yet [00:34] roaksoax: if it doesn't need a home dir, then no. You can specify /nonexistent as the home dir, but be sure to use --no-create-home [00:34] OK [00:35] jdstrand: awesome, thanks [00:35] np [01:04] ScottK: blueman needs a merge, but the dh_python2 conversion was wrong, I'll file a Debian bug [01:04] they also transitioned to python-gi [01:19] Thanks. [02:06] I'm trying to build an app for my ppa that won't build on launchpad, but does with pbuilder-dist lucid . It has a build-depends as follows: [02:06] Build-Depends: [02:06] debhelper (>= 6), [02:06] python, [02:06] python (>= 2.6.6-3~) | python-support [02:06] it doesn't build because it can't get the python without trying to get the python-support instead. Is what I want reasonable? [02:08] jsjgruber-x-p: #ubuntu-packaging is better for random PPA build failures [02:09] ok, thanks [02:33] @pilot out === infinity changed the topic of #ubuntu-devel to: Precise: UI and feature freeze Archive: frozen | Dev' of Ubuntu (not support or app devel) | build failures -> http://bit.ly/HaWdtw | #ubuntu for support and general discussion for hardy -> oneiric | #ubuntu-app-devel for app development on Ubuntu http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: === jbicha is now known as Guest73631 === Guest73631 is now known as jbicha_ [03:19] Morning folks. [04:53] Good morning [04:54] morning pitti [04:54] Morning pitti === Whoopie_ is now known as Whoopie [05:54] Hi gang, I'm trying to sort out our (Studio) bzr branches in better shape, and Micah kindly pointed out some errors with fields in debian/control. What's the proper URI for Vcs-Bzr field? Should it be 'lp:ubuntustudio-default-settings', 'lp:~ubuntustudio-dev/ubuntustudio-default-settings/UbuntuStudio' or the long form [05:54] 'https://code.launchpad.net/~ubuntustudio-dev/ubuntustudio-default-settings/UbuntuStudio'? [05:59] astraljava: Most seem to use the http:// URI. [06:03] infinity: Ok, I'll fix ours to comply with that, then. Thank you! === smb` is now known as smb [07:17] anyone seen this ? [07:17] walking through installer from yesterday's iso [07:17] it just hang as soon as i enter the user info and hit continue [07:18] (i did have encrypted home) [07:18] er.. did select. [07:18] walking through installer from yesterday's iso [07:19] good morning [07:21] RAOF, happy birthday! :) [07:27] Happy birthday RAOF. [07:28] what's supposed to replace pcretest in the new multiarch pcre3 packages (this was dropped 2 releases ago)? [07:28] dholbach, msvb: Thanks. [07:31] micahg: why do you need a replacement? [07:31] slangasek: I've got package testing for UTF-8 support in PCRE [07:31] ah [07:32] can you just bundle the source as part of your test? [07:32] or put it in the pcre package's own build-time test suite? [07:32] of the pcretest script? probably [07:32] (cf. the original bug report requesting pcretest, bugs.debian.org/162998) [07:36] * micahg wonders why this works in Debian [07:59] micahg: Sure. [07:59] So. [08:00] Why did GLIB drop that feature, or is the test for it just broken? [08:00] Cause either that's a glib regression, or goffice is doing something stupid. [08:01] (Or is it a bug and not a feature that it's looking for? But then it wouldn't make sense to include pcre if glib isn't buggy?) [08:01] Confusing. [08:01] hmm, according to the docs it should be there [08:02] ah, it's the include most likely has changed [08:02] micahg: Could it... Yeah. [08:02] micahg: I was just about to suggest single include madness. === doko_ is now known as doko [08:02] pitti, ping [08:03] hello doko [08:03] micahg: I bet throws errors. [08:03] pitti, the jockey-gtk crash did hit me yesterday [08:06] what is "the" jockey crash? [08:06] doko: is there a bug # for this with details? [08:08] pitti, bug 804662 [08:08] Launchpad bug 804662 in jockey (Ubuntu) "jockey-gtk crashed with TypeError in _execute_child(): execv() arg 2 must contain only strings" [High,Fix released] https://launchpad.net/bugs/804662 [08:09] @pilot in [08:09] doko: can you please file it again? it seems to be a slightly different crash === krnekhelesh is now known as nik90 [08:12] infinity: /me bangs head on desk (I forgot to try a rebuild :), was the single include nonsense) [08:19] hmm - guess the bot knew that I'm actually piloting tomorrow [08:19] * jamespage goes for more coffee [08:23] < Riddell> shell scriptage help needed: http://paste.kde.org/456044/ [08:23] that breaks on first line if more than one .pot file exists [08:23] how do I get it to test for .pot file existing without breaking on more than one? [08:24] Riddell, wouldn't for file in po/*pot; sed "s/charset... work? [08:25] only if you set shopt nullglob [08:25] if no files match po/*pot, you try to sed a file literally named "po/*pot" [08:26] Riddell: for file in po/*pot; do if [ "$file" != 'po/*pot' ]; then sed [...]; fi; done [08:26] so patch pilot bot is missing because the srever went down yesterday for some unkown reason. I need tsimpson to resstart that one (and tell me how to do it), so whoever the patch pilot today is, please adjust the topic manually. [08:26] bryceh: not if the directory is empty [08:27] jamespage: could you direct my earlier comment to whomever is the patchpilot today? [08:27] ls -l po/*pot 2> /dev/null && for file in po/*pot; do echo $file ; done [08:29] if [ "$(ls -A po/)" ]; then this seems to do it nicely [08:30] every time I use bash I end up wanting to rewrite in python :) [08:30] ah, but that's not bash [08:30] if you were using bash, you *could* do shopt nullglob, and then you wouldn't have to check ;) [08:38] hey, i'm trying to run the installer (several times now) on a new laptop [08:38] if [ "$1" = deconfigure ]; then [08:38] :; # blah, do something useful with ldso [08:38] fi [08:38] when i come to the point of adding a new user. [08:38] This is why I shouldn't read other people's maintainer scripts... [08:38] as soon as i click continue, ubiquity goes to 100% cpu [08:38] i'm choosing a separate /home partition and to format it. [08:38] is this known? [08:42] smoser: Are you perhaps running into bug #966294? [08:42] Launchpad bug 966294 in ubiquity (Ubuntu Precise) "Ubiquity loops forever from ubiquity_webcam_play" [High,In progress] https://launchpad.net/bugs/966294 [08:43] RAOF, that could be [08:50] RAOF, thank you! [08:50] i removed the ucvideo.ko module [08:50] (and then rm'd it for good measure) [08:50] and i'm moving on [09:18] Riddell: [09:18] +if [ "$(ls -A po/)" ]; then [09:19] Riddell: I think it would be more robust to use [ -n "$(ls ..)" ] [09:19] Riddell: also, there seems to be a missing "fi" in http://launchpadlibrarian.net/101498692/pkg-kde-tools_0.14.2ubuntu4_0.14.2ubuntu5.diff.gz ? [09:23] pitti: fix uploaded [09:36] where can I find old isos? (specifically I'm looking for precise beta 1 isos) [09:53] (tried old-releases.u.c but it doesn't have precise isos) [09:56] Adri2000: we archive them to tape I think, but they aren't available anywhere public (or even easily Canonical-accessible) [09:56] pitti, how much memory is required to setup the virtual device in udisks test ? The test fails with 'cannot allocate memory' when modprobing scsi_debug. The highest dev_size_mb I could set with 2GB of virtual memory was 112 [09:57] jibel: that's a good question, I'm not really sure about it [09:57] jibel: I would expect that 2.5 GB should be enough then, as it requests 300 MB [10:03] cjwatson: argh :/ ok... (wanted to debug an X regression post beta 1, would have been useful in order to track down the culprit packages) [10:04] pitti, ok, I'll start with 2.5 and increase until it passes. [10:05] jibel: merci [10:05] jibel: I'm preparing a new udisks uplaod with the missing dependencies [10:08] pitti, ok, thanks. postgresql test failed but it looks like a bug in autopkgtest which expand '+' to '%2B' in the package name [10:11] yeah, I wondered about that, it didn't even get to the tests there [10:14] yay, precise-adt-ubiquity passed [10:15] cool! [10:18] hello devs [10:19] jibel: https://jenkins.qa.ubuntu.com/view/Precise/job/precise-adt-upower/lastSuccessfulBuild/ looks weird -- did that succeed actually, or was this a test environ failure? [10:19] when does beta 12.10 start? [10:20] sam-c: https://wiki.ubuntu.com/QReleaseSchedule [10:21] sam-c: Its release schedule isn't finalised yet [10:22] the details haven't; you can reasonably expect it not to be too far off the above wiki page though, maybe with a week or two's shift here and there [10:23] pitti, it succeeded but publication to the public jenkins is slow [10:23] jibel: ah, I see; great to see it works now! [10:24] so it syncs results from the parent job first then children jobs many minutes later [10:24] thanks lets hope 12.04 is a success! === MacSlow is now known as MacSlow|lunch [11:08] jibel: nice, it's there now === tkamppeter_ is now known as tkamppeter [11:27] infinity, will you address bug 929219 before the release? [11:27] Launchpad bug 929219 in eglibc (Ubuntu Precise) "chromium-browser, gvfsd-http and others using eglibc crash with SIGSEGV in __nscd_get_mapping() or gethostbyname2_r()" [High,Triaged] https://launchpad.net/bugs/929219 [11:28] doko: Yeah, testing an upload now. [11:28] doko: Just getting a few more fixes in at the same time. [11:29] thanks === mbiebl_ is now known as mbiebl === ogra is now known as Guest6312 === Guest6312 is now known as ogra_ [11:46] Am I going blind or is there no mapping (via python-apt) from a package in the cache to the path in /var/apt/cache/archives? === GirlyGirl_ is now known as GirlyGirl [11:50] ev: there is no obvious one, iirc you need to create a acquire object, load the package into it and then you can get it via the destfilename :/ [11:50] ow [11:50] give me a sec (or some more and I can try to create a short example) [11:50] mvo: cheers [11:51] (background: I'm trying to see if any Conflicts or Replaces are already in /var/cache/apt/archives) [11:52] I'm not sure I understand, why is this relevant? I guess its something with ubiquity :) ? [11:52] it looks like apt.apt_pkg.parse_depends(candidate.record['Conflicts']) will get me half way there, then presumably apt.apt_pkg.check_dep() with the destfilename version parsed out? [11:52] mvo: apport-retrace speedup... [11:53] mvo: https://code.launchpad.net/~ev/apport/reunpack_to_sandbox/+merge/101643 - we unpack the debs we need for retracing to a sandbox using dpkg -x. With my speedup change we cache this sandbox between runs. [11:53] But now we need to be careful to handle replaces/depends and virtual packages therein === MacSlow|lunch is now known as MacSlow [11:57] oh and handling that is easy enough for the simple case. But when the Replaces and Conflicts packages have versions, then I need to parse that out (using parse_depends I guess, as above) and compare that against something in /var/cache/apt/archives - and this is where I'm stuck :) [11:57] ev: I see, interessting! here is a small example http://paste.ubuntu.com/926289/ [11:58] ev: fetcher.items should always only have 1 entry with this code (auto_fix and auto_inst false) so the loop is not really needed [12:01] lo === glenn is now known as Guest64148 [12:01] mvo: is there anything in python-apt that would then rip the version out of that path? So if I was trying to get the versions of all the packages in /var/cache/apt/archives [12:02] I dug around and didn't see anything [12:02] how can i make sure that my package build from, would also use my dependency package from source? === Guest64148 is now known as glenn__ [12:03] how can i make sure that my package build from source, would also use my dependency package from source? [12:03] its a deb file on my hd [12:03] is that easy? :) [12:05] only thing im finding is the control file === _salem is now known as salem_ [12:26] apw: do we still need the bcm43xx stuff, given that we automatically install wl these day? [12:26] s [12:27] does the data in the command-not-found-data package normally get updated just before release? [12:27] pitti, not sure they 100% overlap to be honest [12:27] jamespage: yes, usually before every milestone [12:28] pitti, thought that was the case. [12:32] apw: ah, that was it: https://launchpad.net/ubuntu/+source/linux-firmware/1.77 [12:33] apw: /lib/firmware/phanfw.bin is the biggest file in it (1.8 MB), so that caused quite some growth [12:34] anyway, I suppose we can't just drop that, but at some point we need a way to split that into smaller packages [12:34] in 12.10 we'll get bigger images, but we need to squeeze something for 12.04 [12:35] pitti, well the problem is they are mostly firmware for things we care about, video, network, camera, that sort of thing [12:35] yes, I know [12:35] pitti, it is possible that we have some old firmware in here, stuff that can never be loaded [12:35] right, that's what I thought, hence my question [12:35] pitti, will discuss some potentials for ripping out for now with tim [12:35] in a 27 MB compressed hog there must be _something_ we can get rid of [12:37] apw: thanks! [12:37] I'll have a look at wallpapers [12:38] sladen: so, wallpapers grew a bit, and by yet unknown means we need to free 1.2 MB [12:39] sladen: do you want me to shrink down some wallpapers, or do you want to have a look? [12:40] pitti, it's Cimi who handled that update no? [12:41] not sure, sladen talked to me about it last [12:41] ok [12:42] sladen: e. g. the original Shoes.jpg (1.0 MB) looks exactly the same to my eye as my reduced Shoes2.jpg with 570 kB [12:42] micahg, do you care for the lubuntu seeds ? seems audacious is blocking it on armhf and given that nobody seems to be able to fix the gles issue that blocks it i was wondering if you could do an arch specific seed hack to omit the audacious requirement on arm arches for now [12:43] Cimi, hey [12:43] sladen: e. g. the original Shoes.jpg (1.0 MB) looks exactly the same to my eye as my reduced Shoes2.jpg with 570 kB [12:43] pitti, ^ meet Cimi, artwork master ;-) [12:43] Cimi: hello [12:43] we both know each other seb128 :) [12:43] Cimi: so is it you or sladen that I should ask about ubuntu-wallpapers? [12:43] :-p [12:43] pitti, I did the package/compression, sladen uploaded it [12:44] pinky, which is the big file to you? [12:45] Cimi: there is no particular file, I'm just looking at ls -lSr [12:45] and starting with the biggest one [12:45] Cimi: trying to reclaim the growth of the latest version [12:45] http://paste.ubuntu.com/926344/ [12:45] (assuming that I am pinky) [12:45] pitti, they all seem pretty well compressed [12:46] Sand.jpg compresses less well [12:46] Cimi: but compressing Shoes.jpg (see above) and perhaps some others would help a lot [12:46] pitti, which walls are yoi looking at??? [12:46] pitti, do you look at a different set? [12:46] apt-get source ubuntu-wallpapers.. [12:46] pitti, there is no shoes or sand in the wallpapers [12:47] http://paste.ubuntu.com/926344/ [12:47] ah, we don't ship Shoes in u-w-precise, I see [12:47] pitti, ubuntu-wallpapers has all the old wallpapers as well but those are not installed by default [12:47] it's just in the source [12:47] pitti, they are in i.e -oneiric [12:48] u-w is 2.8 [12:48] less than oneiric [12:49] I think I did a great job here [12:50] Cimi: right, I missed that the source has -oneiric etc. [12:50] so, any other ideas what we can throw out again? [12:50] pitti, we have 2.8 now [12:50] pitti, really, how much do you want to shrink here? [12:50] no, I mean in general, not (just) wallpapers [12:51] i386 grew by 2 MB since beta-2, and we need to shrink it by 1.2 MB [12:52] * pitti looks at telepathy-salut, that grew by 0.2 MB [12:52] libreoffice-common grew by 0.6 MB [12:52] Sweetshark: ^ any idea for a diet? [12:54] pitti, ubuntu-wallpapers deb was 2.4MB in oneiric and is 2.4MB in precise [12:54] pitti, I did my best to keep the same size [12:54] Cimi: again, I'm not blaming you for anything [12:55] Cimi: you did a great job there [12:55] pitti, ok [12:55] I just asked the channel for ideas what we can shrink again :) [12:55] pitti, only solution I have is to remove wallpapers [12:55] pitti, which works but it's not ideal with the community [12:55] it's not ideal for Ubuntu users either [12:56] users like the selection ;-) [12:56] at this point few things that we can do are "ideal" [12:56] seb128, which wallpaper do you like most? [12:58] pitti, we're not shipping the gnome wallpapers, right? [12:58] * apw wonders if ubuntu users like to have firmware for their hardware more or less than pretty backgrounds [12:58] pitti, ok we may be able to strip some bnx2 firmware pretty easily but we have near zero way of knowing if thats going to cause problems [12:59] apw, I would never sacrifice HW compatibility for stupid walls [12:59] pitti: hmmm, 3.5.1-1ubuntu5 already seemed to be bigger [12:59] libreoffice-common (Δ 0.6 MB - 3.5.1-1ubuntu1: 20.9 MB 3.5.2-2ubuntu1: 21.5 MB) [13:00] maybe libreoffice ships stuff we can compress [13:00] pitti, all JPG/PNG can be compressed [13:00] even loseless [13:00] pitti: ha! [13:01] pitti: it was readding the ugly presentationm templates again! [13:01] Sweetshark: win! [13:01] (between 3.5.1-1ubuntu4 and 3.5.1-1ubuntu5) [13:01] Sweetshark: ah, so that's not the .6 MB from above? [13:01] oh wait, it could very well be [13:01] cimi: pitti: wallpapers is ~2.7 MB. we were 2.5 MB last time, but we snarfed the extra after Kaleo uploaded a u-f-f without shipping Medium [13:01] yes, it is [13:02] sladen, looks like 2.4 from the website [13:02] pitti: 3.5.2 didnt change from 3.5.1-1ubuntu5 in size. [13:02] sladen, unless default wallpaper stais in a separate package from community [13:02] Sweetshark: that's /usr/lib/libreoffice/share/template/common/layout ? [13:03] Cimi: yes, default wallpaper is now in u-w, and community wallpapers in u-w-p [13:03] sladen, ah ok, found mistake [13:04] pitti: any new packages, or just bulging at the waist in the existing packaging? [13:05] sladen: just the latter [13:05] we added gtk2-engines-pixbuf since beta-2 (0.1 MB) [13:05] pitti, I can probably remove the dep from light-themes [13:05] no you can't [13:06] it was added for a reason, it spammed xsession-errors with warning without it [13:06] we still have gtk2 apps [13:06] pitti: no, /usr/lob/libreoffice/share/template/en-US/presnt/*.otp [13:06] seb128, is used by light-themes gtk2 [13:06] seb128, for old gnome-panel, old gedit, old ubuntuone gtk [13:07] seb128, if I remove those, it should stop spamming [13:07] Cimi, nothing for firefox, tb? [13:07] seb128, no [13:07] seb128, let me try [13:08] hmm, did dpkg --set-selections stop working in precise ? [13:08] * pitti runs through http://people.canonical.com/~ubuntu-archive/germinate-output/ubuntu.precise/desktop.depends again with a fine comb [13:08] * ogra_ cat get it to install anything [13:08] *cant [13:11] oh, silly me ... nevermind [13:12] seb128: I suppose we need rarian-compat still? [13:12] seb128, pitti lp:~cimi/light-themes/remove-pixbuf-dep-shrink-size [13:13] I thought we need taht? [13:13] seb128, you could try removing pixbuf and running apps [13:14] seb128: it pulls in docbook-xml [13:14] pitti, is that a runtime thing? I though it was for builds [13:14] seb128: yes, that's what I was wondering [13:14] seb128: rarian-compat is explicitly seeded [13:14] pitti, that seems wrong [13:17] seb128: the only place where this would possibly be relevant is yelp, right? [13:17] pitti, yes [13:17] pitti, will old style documentation [13:17] pitti, maybe check with jbicha? [13:17] or maybe mterry knows about it [13:18] rarian-compat plus docbook-xml are 430 kB [13:18] I know he has been packaging the new yelp-xsl, etc [13:18] seb128, pitti are all documentation packages assets compressed with optipng/optijpeg? [13:18] Cimi: yes, pkgbinarymangler does that [13:18] ok [13:18] Cimi: for all packages, not just documentation [13:18] Cimi: we also optimize SVGs [13:18] * mterry reads [13:18] all that bought us some 8 MB in past cycles [13:18] pitti, watch out from PNG [13:19] mterry, do you know if rarian-compat is a build-time only thing or use by yelp at runtime? [13:19] pitti, I tried optipng with something different than -o0 [13:19] mterry, i.e if we need it on the CD [13:19] pitti, and was clearly degrading the image [13:19] seb128, AFAIK just build time. [13:19] ok, what I though as well [13:19] pitti, let's try dropping it :p [13:19] I didn't have scrollkeeper or rarian-compat installed for some months and no issue [13:20] but I don't use documentation a lot [13:20] it's a hard dep of ubunt-desktop [13:20] seb128, did you have a look at my branch? [13:20] * pitti tries to purge it [13:21] ... and yelp works just fine [13:21] seb128, mterry ^ [13:21] pitti, try old style documentations, those which still have an omf [13:21] cool [13:21] ok, I'll unseed it [13:21] pitti, the new world order doesn't use rarian-compat for sure [13:21] i.e yelp-xsl yelp-tools are used nowadays [13:22] e. g. gnome-search-tool [13:22] seb128, lp:~cimi/light-themes/remove-pixbuf-dep-shrink-size [13:22] Cimi, that looks fine to me [13:23] kenvandine, hey, what do you think about lp:~cimi/light-themes/remove-pixbuf-dep-shrink-size ? [13:23] seb128: works just fine [13:23] kenvandine, we are looking for CD space [13:23] * kenvandine looks [13:23] seb128, kenvandine only think will break is ubuntuone control panel gtk [13:23] pitti, ok, I though it was a build time thing so that seems to confirm it, great, kill it! [13:23] seb128, buttons will be unthemed [13:23] seb128, but "who cares in 12.04" [13:23] seb128: *happy* [13:23] we have the qt one [13:23] Cimi, we deprecated the gtk panel [13:23] seb128, indeed [13:23] pitti, ;-) [13:24] how much space does it save? [13:24] kenvandine: 0.1 MB, not much [13:25] MP https://code.launchpad.net/~cimi/light-themes/remove-pixbuf-dep-shrink-size/+merge/101735 [13:25] we added gtk2-engines-pixbuf since beta-2 (0.1 MB) [13:25] but it's half of what's left to do, assuming that LibO gets fixed [13:25] pitti, my branch also removes some small assets so it should be slightly more [13:25] ah, nice [13:25] looks fine to me [13:26] we need to get rid of that stuff sooner or later anyway [13:26] let's do it now [13:26] Cimi, i approved it [13:26] you merge to trunk and i'll upload [13:26] kenvandine, I'll merge [13:27] pitti, hi, now that Maverick is EOL, I think these packages from this tracking bugs will not go anymore to -updates, bugs 967065 and 967066, correct? I'll mark this bugs as invalid if so [13:27] Launchpad bug 967065 in Kernel SRU Workflow "linux-ti-omap4: 2.6.35-903.33 -proposed tracker" [Undecided,In progress] https://launchpad.net/bugs/967065 [13:27] Launchpad bug 967066 in Kernel SRU Workflow verification-testing "linux-mvl-dove: 2.6.32-425.44 -proposed tracker" [Undecided,In progress] https://launchpad.net/bugs/967066 [13:30] kenvandine, when you package, remember to remove pixbuf dep [13:30] hey guys - I'm curious, does whoopsie[1784]: segfault at 0 ip 00000000004050a2 sp 00007fff367e5488 error 4 in whoopsie[400000+9000] [13:30] mean anything to anyone? [13:30] Cimi, done [13:30] Cimi, is there a bug report related to this? [13:30] kenvandine, no [13:30] seb128, ^ [13:30] you love bug reports [13:31] Cimi, kenvandine: that's fine, just upload ;-) [13:31] just checking [13:32] herton: ah, thanks [13:35] ev: ^ (whoopsie segfault) [13:37] hah, I found another 330 kB [13:37] $ cat /usr/share/hwdata/pci.ids /usr/share/hwdata/usb.ids | gzip -9 | wc -c [13:37] 334112 [13:38] we can symlink them to the versions from usb-utils/pci-utils [13:43] seb128, pitti, Cimi: light-themes uploaded [13:43] kenvandine, thanks [13:43] pitti, the .deb is 8K smaller [13:43] kenvandine: plus teh dropped dep? [13:44] kenvandine: thanks [13:44] yeah [13:44] kenvandine: so why was the dep added in teh first place? [13:44] someone added it because it caused messages to get spammed to the console [13:44] kenvandine, pitti: some of the old gtk2 theming file were using the pixbuf [13:45] that was leading the .xsession-errors spamming [13:45] it was bug 762167 [13:45] Launchpad bug 762167 in light-themes (Ubuntu) "missing dependency on gtk2-engines-pixbuf" [Low,Fix released] https://launchpad.net/bugs/762167 [13:47] hallyn: you should have a /var/crash/*whoopsie*.crash file [13:48] hallyn: please run ubuntu-bug /var/crash/*whoopsie*.crash and file a new bug [13:48] ev: no, it wasn't me [13:48] ev: it was in a qemu-kvm bug which had no qemu-kvm msgs in syslog, only whoopsie ones [13:48] i've marked it as affecting whoopsie, hold on lemme find the bug # in lp [13:49] ev: bug 959964 [13:49] Launchpad bug 959964 in qemu-kvm (Ubuntu) "qemu-system-x86_64 crashed with SIGABRT" [Medium,Incomplete] https://launchpad.net/bugs/959964 [13:50] I'm just wondering ifthat points to anything wrong on the host that would explain both qemu-kvm and whoopsie userspace errors [13:51] ah, phew [13:52] mterry: oh, we need to disable the whole gnome-keyring test suite, not just the test(s) which cause the hang? [13:52] looks like they hit the memory corruption bug in whoopsie from march [13:52] it's not a new one [13:52] mterry: can you reproduce the hang, actually? I had assumed it was due to the buildds running hardy kernels [13:52] but it wont explain the qemu issue [13:56] drat :) [13:56] I was hoping the qemu issue was some kernel resource problem causing pthread clone failure [13:56] thanks [13:56] Sweetshark: can you make the change to drop them again? [14:00] pitti, no I can't reproduce [14:00] pitti, I disabled the whole suite. I believe I saw different tests hanging (?) [14:00] pitti, but it's possible I could have been more targetted [14:02] mterry: I'm a bit worried about testing SRUs and point releases, on different arches [14:02] mterry: do you know if this ever affected armel/armhf? [14:02] if it is only affecting i386/amd64, there is even more reason to believe it's due to the hardy kernel [14:03] pitti, I only know of it happening on i386, but I don't have the full history of problems with the package [14:03]