=== Chipzz_ is now known as Chipzz === elmo_ is now known as elmo === dendro-afk is now known as dendrobates === zyga is now known as zyga-afk === dendrobates is now known as dendro-afk [01:50] the distribute log for oneiric is a mess [01:50] https://launchpad.net/ubuntu/+source/distribute/+changelog [01:50] can i propose that oneiric simples get the same version as in precise? [01:56] Cas-: what do you mean the log is a mesS? [01:57] i mean that the version 0.16.6 is just patched and patched instead of just pulling release versions [01:58] we don't update release version in stable releases unless they are totally broke beyond fixing [01:58] backports are possible for some applications [01:58] we i think this is one of them [01:58] Cas-: is there something wrong with the version in precise? [01:58] err...oneiric [01:58] yes [01:59] bug #? [01:59] https://bugs.launchpad.net/ubuntu/+source/distribute/+bug/938786 [02:00] Launchpad bug 938786 in distribute (Ubuntu) "Import issue with namespace packages in pkg_resources" [High,New] [02:01] basically a fix from 0.16.7 was applied to 0.16.6 that was actually broken and then fixed in 0.16.8 [02:01] Cas-: as soon as the current package in -proposed migrates to -updates, we can upload this fix [02:03] the list of changes for distribute is very small http://pypi.python.org/pypi/distribute#changes so thats why i suggested oneiric should match precise version of 0.6.24 [02:04] yeah, we generally don't do that, are there other specific fixes needed for oneiric besides this one? [02:04] not that i have specifically encoutered [02:06] ok, so let's just do one more SRU then, I"ll give you a bug task, but I'm not comfortable uploading the change, I'd prefer someone with a python background review it, do you want to prepare a debdiff? (it can be sponsored on your behalf then and go straight in the sponsorship queue) [02:07] the only change that matters is the tab indent in pkg_resources.py [02:08] i dont have a setup for debdiff [02:09] bzr branch works also [02:14] well it's late here i'll need to sort it 2mw === bradm_ is now known as bradm === dendro-afk is now known as dendrobates [05:10] Good morning === bladernr_ is now known as bladernr_afk === tjaalton_ is now known as tjaalton [06:36] is LP being upgraded or so? it's horribly unstable ATM [06:38] pitti: we're looking at it [06:38] pitti: no, its not mid-deploy (and that doesn't cause instability anyhow); we're seeing broken connections to the db master and currently divide-and-conquering the path [06:38] lifeless: ah, thanks [06:39] we've just tried something; tell us if its better worse ? [06:46] I did three changes now, they all succeeded [06:46] before I needed to try each bug state change some 5 times [06:46] cool [06:46] thanks [06:49] bzr commit worked as well [06:57] ev: FYI, I fixed the ui.py test failures in apport trunk now [07:09] cjwatson, slangasek, StevenK: it seems /srv/launchpad.net/ubuntu-archive/ubuntu-germinate/ does not get updated any more; any hint how I could debug this? is that being called by soyuz or by some cronjob which I fail to see? [07:10] pitti: hmm, I suppose this would be related to cjwatson's changes this cycle to speed up publishing, but I'm afraid I don't know the details [07:11] it just stopped two days ago [07:11] That's what I was thinking. [07:11] three now [07:11] ah [07:11] so it's still supposed to be done as part of the publisher run [07:11] and I figure that's rather bad, as I assume live CD builds etc. depend on this? [07:11] but there's been some restructuring [07:11] (not to mention breakign http://people.canonical.com/~ubuntu-archive/component-mismatches.txt etc.) [07:12] well, that output impacts the task fields in the archive [07:12] right, so our recent seed changes will probably not take effect [07:13] * slangasek nods [07:17] infinity: hmm, wubi livefs build failure, "mount: unknown filesystem type 'ext4'" - is that related to the changes I saw you making? [07:33] * ogra_ wonders where infinity's live-build upload went [07:56] I'm not sure if it's useful to say this, or if it's done regularly anyway: libhighgui2.1 and libcv2.1 binary packages are NBS and can now be removed === smb` is now known as smb [08:08] Adri2000: I did so an hour ago or so [08:08] after sitplus finally built on powerpc [08:08] Adri2000: I'm watching this regularly, yes [08:13] good morning [08:14] good morning dholbach [08:14] hey diwic [08:17] pitti: cool. btw, I can't find anymore the list of removed packages (no removals.txt at http://people.canonical.com/~ubuntu-archive/), where is it? [08:17] Adri2000: I don't know whether there is a replacement [08:19] it was useful in some cases :) [08:23] Adri2000: do you need a list of removed packages or the reason (and date) when a package got removed? the later can be found in LP [08:24] geser: even binary package removals can be found in LP? how? === d1b_ is now known as d1b [08:30] Adri2000: sure, e.g. https://launchpad.net/ubuntu/oneiric/amd64/libcv2.1 should mention it once it's removed (don't have a better example right now) [08:31] ok, https://launchpad.net/ubuntu/precise/amd64/libcv2.1 mentions it so it's good :) [08:32] ah right, forgot to replace the release in the url [08:40] also, e. g. https://launchpad.net/ubuntu/+source/libaws/+changelog [08:41] or https://launchpad.net/ubuntu/+source/libaws/+publishinghistory (with more data) === tkamppeter_ is now known as tkamppeter [09:22] Anyone taking care of bug 911436? It leads to crashes of the CUPS daemon, especially when the CUPS package gets updated and therefore the daemon restarted. [09:22] Launchpad bug 911436 in p11-kit (Ubuntu) "https crashed with SIGSEGV in lookup_or_create_bucket()" [Critical,Triaged] https://launchpad.net/bugs/911436 [09:26] slangasek, you have set bug 911436 to "Triaged" a while ago, are you working on solving it? [09:26] Launchpad bug 911436 in p11-kit (Ubuntu) "https crashed with SIGSEGV in lookup_or_create_bucket()" [Critical,Triaged] https://launchpad.net/bugs/911436 [09:29] pitti: cheers! [09:29] lamont, infinity: temporarily stealing some buildds for i386 for the langpack upload, as it's all quiet otherwise [09:29] (just FYI) [09:30] ev: so AFAICS from yesterday's discussion we still need a way to disable apport's LP bug filing, but not whoopsie submission post-release, right? [09:31] how can i, in build conflicts, conflict against a specific range of versions? [09:31] (in particular, libglew released packages using 1.5.2 code as "1.5.7.is.1.5.2", which makes it kind of tricky to depend on what you'd normally call libglew1.5-dev (>= 1.5.4) ) [09:44] I need a quick FFe for wayland, which now has a real upstream release instead of the git snapshots we've had before. it's also a build-dependency of the new mesa bugfix release I'm preparing [09:45] nothing else depends on it, so risk of regression is negligible [09:46] tjaalton: eek, that would mean wayland needs to go into main? [09:46] it is already [09:46] ok sorry, I'll rephrase [09:46] eek :) [09:46] the new mesa build-depends on this release :) [09:46] tjaalton: so, otherwise this sounds fine [09:46] it has that build-depend already [09:46] *has had [09:48] so the bugfix mesa release needs this new wayland to build libwayland-egl [09:48] mesa will fix a bunch of crashers, at least bug 926379 [09:48] Launchpad bug 926379 in mesa (Ubuntu) "compiz crashed with SIGSEGV in intel_miptree_release() (Needs mesa-8.0.1)" [High,In progress] https://launchpad.net/bugs/926379 === zyga-afk is now known as zyga [10:04] pitti: so I take it that it's ok to update wayland too? [10:04] tjaalton: yes, fine for me [10:04] thanks! [10:04] tjaalton: you might want to coordinate with bryce, he did the wayland updates before [10:05] these are maintained in pkg-xorg git, so yeah he should know what's happening :) [10:06] Sarvatt did the previous upload [10:09] pitti: there's also the default wayland compositor, weston, that could be synced from experimental (NEW)? [10:09] tjaalton: no idea about that; that seems less urgent and more invasive, and should probably get an FFE bug [10:10] yes less urgent [10:10] I'll leave it for now [10:15] slangasek: Potentially. I'll look at it in the morning. [10:19] pitti, yes the merge is fine by me; I have no plans to look at or work on wayland this cycle [10:20] pitti, also since no one really uses the compositor I don't really see any harm in letting that be sync'd in. It's really only existing for testing purposes, for people who want to observe compositors crashing a lot. [10:20] *chuckle* [10:26] what is whoopsie and why does it use 89% of 8G of RAM? :) [10:26] ev: ^ [10:26] dholbach: ev's new crash reporting daemon [10:26] gah [10:27] the joys of C programs? [10:27] :) [10:27] I did run it through memcheck [10:27] dholbach: do you have any .upload files in /var/crash? [10:28] 28294 whoopsie 20 0 37.0g 7.0g 1116 S 0 90.4 0:33.84 whoopsie [10:28] ev, no, just a bunch of .crash files [10:29] interesting [10:29] did you elect to report any crashes using the new apport UI? [10:29] yes [10:30] dholbach: could you pastebin a ls -la of /var/crash [10:31] curious to know if there's anything particularly large in there [10:31] ev, http://paste.ubuntu.com/853789/ [10:31] ev, maybe it's struggling with the one owned by daniel.daniel? [10:32] dholbach: well, it's not supposed to touch anything at all if there isn't a matching .upload file for it [10:32] aha [10:33] dholbach: could you possibly do sudo stop whoopsie; CRASH_DB_URL=http://localhost /usr/bin/whoopsie -f > whoopsie.log 2&>1 for a bit [10:33] sure [10:33] and see if it starts growing out of control [10:33] much appreciated [10:33] * ev goes memleak hunting [10:34] ev, [10:34] aniel@daydream:~$ sudo stop whoopsie; CRASH_DB_URL=http://localhost /usr/bin/whoopsie -f > whoopsie.log 2&>1 [10:34] whoopsie stop/waiting [10:34] Trace/Breakpoint ausgelöst (Speicherabzug geschrieben) [10:34] daniel@daydream:~$ [10:35] hum, let me try to find a translation for this :) [10:35] in any case it crashed [10:35] whoopsie crashed with signal 5 in __libc_start_main() [10:36] "Trace / breakpoint is triggered (dump written)" [10:37] dholbach: if you would be so kind, CRASH_DB_URL=http://localhost gdb --args /usr/bin/whoopsie -f [10:37] r then bt [10:37] I guess I will need some debug symbols - right now it's useless [10:38] :) [10:40] dholbach: https://bugs.launchpad.net/ubuntu/+source/whoopsie-daisy/+bug/939392 for the memleak bug [10:40] Launchpad bug 939392 in whoopsie-daisy (Ubuntu) "Large memleak in whoopsie" [Undecided,New] [10:40] ev, http://paste.ubuntu.com/853800/ [10:41] ah, whoops [10:41] I'll fix that, but... [10:41] sudo CRASH_DB_URL=http://localhost /usr/bin/whoopsie -f >whoopsie.log 2&>1 [10:42] http://daisy.ubuntu.com/ doesn't seem to exist/ [10:42] ? [10:42] pitti: not yet - IS is working on that [10:43] ev, I'll let you know if anything interesting turns up [10:43] dholbach: cheers, and sorry about this [10:43] no, no - it's fine === soren_ is now known as soren === cking_ is now known as cking [11:04] hello [11:06] something has happened such that my local apt is broken and I can no longer upgrade [11:06] jml: did you try sudo apt-get -f install? [11:06] (withouth the '?') [11:06] pitti: hah. of course. :) [11:06] jml: what does it say? [11:07] pitti: http://paste.ubuntu.com/853832/ [11:07] E: Internal Error, No file name for libc6 [11:07] erk [11:07] and mvo is on holiday === Claudinu` is now known as Claudinux [11:08] jml: perhaps try this: [11:08] sudo rm /var/lib//apt/lists/* /var/cache/apt/*cache* [11:08] sudo apt-get update === Claudinux is now known as Guest45614 [11:08] sudo apt-get -f install [11:09] pitti: ok. will try that [11:09] jml: perhaps save the caches before, for later investigation [11:09] oops :) [11:10] pitti: /var/lib/apt/lists/partial still exists. Should I remove that? [11:10] jml: keep the dir [11:10] you can remove the files in it [11:10] o [11:10] k [11:10] jml: by gut feeling is that the caches were broken, the apt lists are usually okay [11:11] modulo memory/fs corruption of course [11:12] pitti: still get the same error on apt-get -f install [11:13] jml: can you please show me "apt-cache policy libc6"? [11:14] pitti: http://paste.ubuntu.com/853840/ [11:15] that looks fine [11:15] jml: does "sudo apt-get install --reinstall libc6" work? [11:15] i. e. does it know about the package name ther? [11:16] * jml tries [11:16] jml: oh, wait [11:16] jml: perhaps it thought it already downloaded the libc6 .deb, but didn't [11:16] jml: rm /var/cache/apt/archives/* might helP? [11:17] yay [11:17] boo [11:17] googling suggests that it might be related to /var/lib/dpkg/status [11:18] I don't know how to regenerate status, though [11:18] http://paste.ubuntu.com/853844/ [11:19] there should be a backup file [11:20] see bug #859188 for a similar problem [11:20] Launchpad bug 859188 in apt (Ubuntu) "can't apt-get install --reinstall when foreign-arch version is present (E: Internal Error, No file name for libgl1-mesa-glx)" [High,Fix released] https://launchpad.net/bugs/859188 [11:20] geser: 'status-old' is identical to 'status', fwiw. [11:21] try "apt-get install --reinstall libc6:i386 libc6:amd64" [11:22] geser: $ sudo apt-get install -y --reinstall libc6:i386 libc6:amd64 [11:22] Reading package lists... Done [11:22] Building dependency tree [11:22] Reading state information... Done [11:22] You might want to run 'apt-get -f install' to correct these: [11:22] The following packages have unmet dependencies. [11:22] libc-dev-bin : Depends: libc6 (< 2.14) but 2.15-0ubuntu2 is to be installed [11:22] libc6 : Depends: libc-bin (= 2.15-0ubuntu2) [11:22] libc6:i386 : Depends: libc-bin:i386 (= 2.15-0ubuntu2) [11:22] libc6-dev : Depends: libc6 (= 2.13-24ubuntu4) but 2.15-0ubuntu2 is to be installed [11:22] libnih1 : Depends: libc6 (< 2.15) but 2.15-0ubuntu2 is to be installed [11:22] E: Unmet dependencies. Try 'apt-get -f install' with no packages (or specify a solution). [11:22] sorry [11:22] I honestly hit C-c on the URL in the browser [11:22] hmm [11:22] there are some more backups in /var/backups/dpkg.status.* [11:23] (I could swear that something about recent Ubuntus introduces a lag on keyboard input when switching windows/desktops) [11:25] well, here's the diff between current status and the first one that has a difference: http://paste.ubuntu.com/853852/ [11:25] is this an upgrade precise -> precise? [11:25] jml: that libnih1 is old [11:25] I'm wondering why it doesn't want to upgrade the other packages [11:26] especially since he already has libc6 2.15 [11:26] geser: yes, it is. [11:26] jml: apt-cache policy libnih1? [11:26] http://paste.ubuntu.com/853853/ [11:28] does it upgrade libc6 if you leave out the --reinstall? [11:28] jml: sudo apt-get install --reinstall libc6 libnih1 libc-dev-bin libc6:i386 [11:29] and try adding libc-dev-bin libc6-dev libnih1 to the list of packages to install [11:29] * pitti ^5s geser [11:31] hmm [11:32] pitti: the --reinstall command just errors out with unmet dependencies... [11:33] jml: you can try adding all the unmet ones until it works [11:33] or you get a meaningfull error [11:34] jml, I had that problem when upgrading precise a while ago and found in some bug (don't remember number) that "sudo dpkg -i /var/cache/apt/archives/libc...." was the right way to resolve the problme [11:34] http://paste.ubuntu.com/853861/ [11:34] after that you can continue with "-f install" and the "dist-upgrade" [11:34] after I add packages, I get the original libc6 error again :\ [11:35] bah :( [11:35] diwic: oddly, I don't have a libc6 deb in my apt/archives folder [11:35] bad luck [11:36] :-/ [11:36] apt-get download libc6 [11:36] ... and sudo dpkg -i libc6_*.deb [11:36] (it'll land in the current folder, not the apt cache) [11:37] as this seems to be a multi-arch issue (as mentioned in that bug) perhaps slangasek knows how to get out of it (he commented on that bug) [11:39] ok [11:40] so the libc6 install failed because libc-bin wasn't at the right version, but I installed (dpkg -i) libc-bin from apt/archives and then tried installing the downloaded libc6. That seemed to work, and apt-get -f install seemed to finish successfully [11:40] (although weirdly, the config system used readline rather than the pretty aubergine TUI) [11:40] who maintains ubuntu-seeds? [11:41] jml: great! [11:41] hrw: the "owners" of the respective derivatives, e. g. ~ubuntu-core-dev, ~xubuntu-dev, and so on [11:41] and apt-get dist-upgrade is downloading a lot of stuff now :) [11:41] pitti: thx [11:41] pitti, geser, diwic: thanks so much! [11:42] jml, np - but it has happened on two machines here so I assumed it was generic enough that someone would actually fix it, but we'll see... [11:44] diwic, TheMuso: WDYT about bug 930703 ? [11:44] Launchpad bug 930703 in pulseaudio (Ubuntu) "wishlist: demote pulseaudio-esound-compat to Suggests" [Undecided,New] https://launchpad.net/bugs/930703 [11:44] I think it's about time [11:46] pitti, I don't know of any application using the esound interface; that said, I tend to err on the side of caution at this part of the cycle [11:47] pitti, but I guess such applications would then bring this in through some kind of dependency, huh? [11:48] diwic: I think we should promote libesd0's Suggests: pulseaudio-esound-compat to Recommends: [11:49] then everythign which still actually needs esound will still get it [11:49] but we avoid installing it by default [11:49] not that it takes much space or so, but it offers a wildly obsolete API [11:50] pitti, that sounds okay to me [11:54] jibel: pitti suggested that I talk to you about adding a check for $(dpkg-query -W -f='${Conffiles}\n' | grep obsolete | awk {'print $1'}) to https://jenkins.qa.ubuntu.com/view/Precise/ [11:55] jibel: hm, that's perhaps something which could be added to ./AutoUpgradeTester/post_upgrade_tests/test_lts_upgrade_system.py ? [11:56] jibel: OOI, for changes to these scripts, is it enough to commit, or does it need an upload? [11:56] jibel: IOW, are you running from bzr or from the pacakge? [11:56] jibel: for reference, you can check how I use that stanza in src: upgrade-system [12:01] ev: before you continue to hack on apport, please pull trunk into your branch (your branch is merged) [12:01] will do [12:01] ev: I made quite intrusive structural changes to the tests [12:01] cool, will have a look [12:01] ev: the tests are now all in test/ and all look alike [12:01] yay! [12:01] ev: this also makes it possible to run them against the installed system libs, too [12:02] hm, thehre are now 4889 LOC in apport/ and 8462 LOC in test/ [12:03] :D [12:03] pitti, hi [12:03] hey tkamppeter [12:03] pitti, heard of apport exploding when trying to submit bugs? crashing python in fact: [12:03] *** glibc detected *** /usr/bin/python: double free or corruption (fasttop): 0x000000000378bf60 *** [12:04] apw: not on my radar; sounds like a bad transfer annotation somewhere in GTK [12:04] apw: are you able to reproduce this? [12:04] pitti, i tried to submit an xorg bug, and its failed twice, not sure its identicle symptoms tho [12:05] apw: ah, not a crash, just "apport-bug xorg"? [12:05] what did you click? [12:05] pitti, ok so ubuntu-bug xorg, clicked the first two entries in the first dialog, the last one in the second, then hit 'Send' without changing anything in the third [12:06] last two are identicle heap corription thingies [12:06] pitti, I got the Ubuntu-branded test page today. I will put it into the cups-filters package. I will put it into debian/local/ (where the apport hook is) and replace the upstream page by this one in the case that the package is built for Ubuntu. Is this OK? [12:06] apw: second question is about lightdm log files; yes or no? [12:07] pitti, sorry i also said yes to that one [12:07] tkamppeter: sure; could you do this in debian/rules with if dpkg-vendor --is ubuntu; then ... ? [12:07] tkamppeter: then we can keep it in sync with Debian [12:07] so i was saying its 'reproducible' and 'happened more than once'; they could have my logs; and i would 'do anything' to get it fixed [12:08] apw: meh, works fine here [12:09] * apw ubuntu-bug ubuntu-bug's to see what happens [12:09] ok that seems to work ... so i can file a bug about not being able to file a bug === _salem` is now known as _salem [12:10] apw: the .crash report might be more useful, though [12:10] full GTK stack trace and such [12:10] so will there be a crash report for the apport crash? [12:11] i'd assume it suppressed there to prevent circles [12:11] no, it shouldn't [12:12] apport has plenty of crash reprots [12:12] most of which are not very useful of course, but still, they are there [12:13] hmmm, so i ran it three times, and got 3 aborts on my terminal, but only one maybe in /var/crash, does it do local dup detection ? [12:13] apw: the file name is by program name and user id, so yes [12:13] didrocks: are you planning on uploading activity-log-manager 0.9.2 today? I only mention it because it should Breaks and Replaces on whoopsie < 0.1.10, as its taking ownership of the GNOME Control Center stuff. [12:14] pitti, ok so i have a crash here, so how do i tickly apport to make a bug out of it [12:14] apw: usually it pops up by itself, but you can apport-bug /var/crash/*apport*.crash [12:15] ev, would i expect all my reports in /var/crash to be readable by woopsie? [12:15] pitti, seems i have like 10 in there, and i can't be bothered to handle them all right now [12:15] apw: just the _usr_share_apport_apport-gtk.crash one [12:16] ev: yeah, it's on my schedule :) [12:16] pitti, i ran the command you suggested, and have a dialog saying, "The application Report a problem ... has closed unexpectly." and saying 'leave closed' and 'relaunch' [12:16] pitti, how do I add the new test page to the package? If I put it into debian/local/ I get [12:16] ev: thanks for the update! Will add the Breaks:/Replaces: [12:16] dpkg-source: info: using source format `3.0 (quilt)' [12:16] dpkg-source: error: unwanted binary file: debian/local/default-testpage.pdf [12:16] dpkg-source: error: detected 1 unwanted binary file (add it in debian/source/include-binaries to allow its inclusion). [12:16] tkamppeter: well, it pretty much says what it wants :) [12:17] ev: no need for a FFe? the whoopsie part was already in precise? [12:17] pitti, now I see. [12:17] didrocks: correct; the whoopsie part was already in precise and cjwatson said this would be a UIFe [12:18] apw: yes, should be [12:18] ev: ok, and we will even upload before UIF, so all good! thanks :) [12:18] cheers! [12:18] I'll sort an upload of whoopsie without the control center stuff now [12:18] ev, ok something isn't hnouring that, as most are, but i suspect one which was triggered by me running ubuntu-bug is not [12:19] apw: is it 0600? or not group owned by whoopsie? [12:19] ev, just got an error trying to submit a bug 'cannont connect to crash database' [12:20] ev, its apw:apw 660 [12:20] -rw-r----- 1 apw apw 32993298 Feb 23 11:58 _usr_share_apport_apport-gtk.1000.crash [12:20] huh? /var/crash should be g+s [12:21] I take it that it's not? [12:21] pitti, is debian/source/include-binaries a directory for the binaries or a list of the binaries? [12:21] drwxrwsrwt 2 root whoopsie 4096 Feb 23 12:17 /var/crash [12:21] tkamppeter: I think a file [12:21] ev, yes it is [12:25] pitti, it is working now, it is indeed a file listing the exceptions. [12:25] apw: I'm baffled as to how that happened then [12:26] ev, its only the default ownership, so i guess somethign must have changed it, in apport [12:27] oh, I didn't realize you could ignore it [12:27] * ev digs [12:27] you can chgrp anything in there you own [12:27] so i could go and do it myself, as they are my files [12:30] pitti, ok finally got it filed: bug #939466 [12:31] Error: Launchpad bug 939466 could not be found [12:35] apw: thanks; need to wait on retracer [12:41] pitti, I have pushed a new version of cups-filters. Can you upload it to Debian and Ubuntu, it contains the new test page and therefore needs to get uploaded before UI Freeze today. [12:42] pitti, and when you are at it, can you also upl;oad CUPS to Debian and Ubuntu? I have fixed auto-config for PS printers in it. [12:49] ev: whoopsie seems to be missing the package description. [12:52] Q-FUNK: fixing [12:54] ev: or well, it has the short description, but the long one doesn't appear. === MacSlow is now known as MacSlow|lunch [13:00] slangasek, hi [13:08] ev, nothing interesting in ~/whoopsie.log, but in ~/1 there's http://paste.ubuntu.com/853954/ -- whoopsie doesn't seem to be leaking right now [13:09] cjwatson, mdz: message in ubuntu-devel-announce for review/approval [13:09] dholbach: I think I may have found part of the problem. If it fails to parse a report it will keep trying every time it processes the report queue (every hour or so) [13:09] I've fixed that in trunk and will push an upload shortly [13:10] cool [13:10] thanks for your work on this ev [13:10] thanks for your help in debugging it [13:10] I just tried not to stand in the way too much ;-) [13:10] dholbach: I don't suppose I could convince you to tar up /var/crash and send that my way? I'd be interested to know why it's failing to parse one of those reports, if that's the case. [13:11] haha [13:14] ev, I'll first need to find out how to do it in a way where the other coworking space people don't kill me :) [13:15] dholbach: sure :) whenever you can. Doesn't need to be right now [13:19] ev, ok, I'm uploading it using trickle right now - I'll be walking the dog now and let you know where you can find it later on [13:21] dholbach: cheers! [13:29] tkamppeter: yes, can do [13:30] hmm, any idea why is cdrom-detect/try-usb=true added to the kernel cmdline when usb disk is being generated using usb-creator-gtk? [13:30] ev, new apport looks nice. I'd include mpt too, but he's not here [13:30] or rather what is checking for this option? [13:30] mterry: cheers! [13:30] pitti: Hi. As result of fontconfig package now have information about fonts-nanum, I think bug #835304 is not invalid now. What do you think about it? [13:30] Launchpad bug 835304 in ttf-nanum (Debian) "contained fontconfig setting files force to make it default font" [Unknown,Fix released] https://launchpad.net/bugs/835304 [13:31] ev: mterry | pitti, did you do the revamp of the apport dialog? Looks nice [13:31] ev: FYI :) [13:32] :) [13:32] join #ubuntu [13:33] ev: btw, I now put Package:, ExecutablePath: and Stacktrace: on the top of the details list, so that it's easier to see the affected package [13:33] ev: having "ApportVersion" at the top is not exactly the most interesting thing :) [13:34] pitti: brilliant! [13:34] heh [13:34] it's not enough yet to alleviate bug 938707 , but an improvement at least [13:34] Launchpad bug 938707 in apport (Ubuntu) "Hard to see which program crashed for "Internal error" reports" [Medium,Confirmed] https://launchpad.net/bugs/938707 [13:38] ev: ping "Notify Evan Dandrea to remove the disclaimer from Ubiquity's first page " [13:39] Riddell: on it [13:39] elmo: ping "Notify James Troup to remind mirrors to check free disk space " [13:41] tjaalton: https://launchpad.net/ubuntu/+source/mesa/7.11-0ubuntu3.1 [13:41] tjaalton: any idea where the armhf build went to? [13:42] pitti, Q-FUNK auto-upgrade testing is running from bzr so a commit is enough or file a bug against u-m, attach the unittest script you want to run and assign it to me. [13:42] tjaalton: argh, unping; picked the wrong one [13:43] jibel: I have no idea of how to implement the scripts or where they would go. [13:44] jibel: I was simply asking pitti if LP has any tags for reporting packages that leave obsolete configs in place after an upgrade (as detected using the snippet above). [13:44] jibel: he recommended talking to you because it could be a good idea to automate such checks. [13:52] Q-FUNK, and he is right, it's a good idea to automate such checks :) [13:53] Q-FUNK, I looked at upgrade-system [13:53] Q-FUNK, the idea is to run $( dpkg-query -W -f='${Conffiles}\n' | grep obsolete | awk {'print $1'} ) post-upgrade [13:53] Q-FUNK, and check that the list is empty ? [13:54] jibel: essentially, yes. if it's not empty, to auto-file bugs against the packages concerned, indicating which files are concerned. [13:55] you can see what it does if you try: sudo FLAUSCH=jah upgrade-system [13:55] infinity, lamont: FTR, buildds back to normal [13:56] jibel: these days, due to Debian making automated runs of piuparts, there's fewer of these orphan configs than bebore -- but upgrading LTS to LTS+1 leaves a lot of cruft. === MacSlow|lunch is now known as MacSlow [13:59] jibel: currently, on my ubuntu+1 desktop, I only found 3 packages with such obsolete configs. on my LTS+1 host, it's close to 50 packages. [14:01] Q-FUNK, bug auto-filing, we are not here yet, but I can add the test to the daily auto-upgrade tests. That will give a good picture of the status of main and universe. [14:01] jibel: assuming that you would generate an automated bug report, packages that have such an obsolete config would need to call dpkg-maintscript-helper via debian/*maintscript to move or delete. [14:04] tkamppeter: cups and cups-filters uploaded === bladernr_afk is now known as bladernr_ [14:08]