#title #ubuntu-meeting Meeting Meeting started by slangasek at 16:01:58 UTC. The full logs are available at http://ubottu.com/meetingology/logs/ubuntu-meeting/2013/ubuntu-meeting.2013-01-16-16.01.log.html . == Meeting summary == *lightning round *Bugs ''LINK:'' https://jenkins.qa.ubuntu.com/view/Precise/view/Upgrade%20Testing%20Dashboard/job/precise-upgrade-lucid-universe/ARCH=amd64,LTS=lts,PROFILE=universe,label=upgrade-test/ (bdmurray, 16:35:58) ''LINK:'' https://errors.ubuntu.com/bucket/?id=%2Fusr%2Fbin%2Flsb_release%3AIOError%3A%3Cmodule%3E%3Amain%3Acheck_modules_installed%3Agetoutput%3Agetstatusoutput (bdmurray, 16:46:09) *AOB Meeting ended at 17:03:39 UTC. == Votes == == Action items == * (none) == People present (lines said) == * slangasek (39) * bdmurray (38) * ev (36) * cjwatson (33) * ubottu (25) * jodh (22) * stgraber (22) * ogra_ (21) * xnox (13) * doko (8) * meetingology (3) * barry (3) * stokachu_ (1) == Full Log == 16:01:58 #startmeeting 16:01:58 Meeting started Wed Jan 16 16:01:58 2013 UTC. The chair is slangasek. Information about MeetBot at http://wiki.ubuntu.com/meetingology. 16:01:58 16:01:58 Available commands: #accept #accepted #action #agree #agreed #chair #commands #endmeeting #endvote #halp #help #idea #info #link #lurk #meetingname #meetingtopic #nick #progress #rejected #replay #restrictlogs #save #startmeeting #subtopic #topic #unchair #undo #unlurk #vote #voters #votesrequired 16:02:34 o/ 16:03:00 * stgraber waves 16:03:26 [TOPIC] lightning round 16:04:11 $ echo $(shuf -e barry doko stgraber jodh ev bdmurray slangasek ogra cjwatson xnox stokachu) 16:04:14 bdmurray stokachu stgraber xnox jodh doko barry slangasek cjwatson ev ogra 16:04:45 bdmurray: you're on :) 16:04:46 rolled out errors branch allowing for source package searching 16:04:46 pushed errors branch fixing issue with width of the frequency bar 16:04:47 pushed daisy branch creating a counter for release:src_package to facilitate regression detection 16:04:49 discussion with evan regarding error tracker and phased updates 16:04:59 updated https://wiki.ubuntu.com/ErrorTracker/PhasedUpdates 16:05:00 tested team view change to errors in canonistack 16:05:00 submitted an RT to have webstats for errors.ubuntu.com 16:05:00 discovered, researched and reported bug 1098250 regarding suspendresume crashes 16:05:01 bug 1098250 in Apport "unable to send suspend resume crash from raring" [High,Incomplete] https://launchpad.net/bugs/1098250 16:05:03 investigation into inability to report distribution upgrade crashes from quantal to raring 16:05:12 created a test case for bug 1077253 regarding apport 16:05:13 ubuntu-release-upgrader branch fixing bugs 1072828, 1068874, 1093697, 1098001, 1067542 16:05:13 bug 1077253 in apport (Ubuntu Quantal) "possible for ExecutableTimestamp to not be for the binary in ExecutablePath" [High,Triaged] https://launchpad.net/bugs/1077253 16:05:16 verification of Q SRU for aptdaemon bug 1060505 16:05:17 bug 1072828 in ubuntu-release-upgrader (Ubuntu Quantal) "[REGRESSION] release upgrader tool became untranslated" [High,Triaged] https://launchpad.net/bugs/1072828 16:05:18 uploaded P,Q SRU for aptdaemon bug 1078544 16:05:19 bug 1067542 in ubuntu-release-upgrader (Ubuntu) "possible for some do-release-upgrade crashes to be unreportable" [High,Fix released] https://launchpad.net/bugs/1067542 16:05:20 uploaded udev to R and a P SRU for bug 1099278 16:05:20 resolved an issue with arsenal not completely removing bug tasks 16:05:21 bug 1093697 in Ubuntu Translations "check-new-release-gtk appears untranslated" [Medium,Fix committed] https://launchpad.net/bugs/1093697 16:05:22 bug 1068874 in ubuntu-release-upgrader (Ubuntu) "does not need to enable apport" [Medium,Fix released] https://launchpad.net/bugs/1068874 16:05:23 bug 1098001 in ubuntu-release-upgrader (Ubuntu) "DistUpgradeController calls apport with a source package as an argument" [High,Fix released] https://launchpad.net/bugs/1098001 16:05:24 bug 1060505 in aptdaemon (Ubuntu Quantal) "Google Chrome is a "package of bad quality"" [Medium,Fix committed] https://launchpad.net/bugs/1060505 16:05:25 bug 1078544 in aptdaemon (Debian) "python-aptdaemon: upgrading marks auto-installed packages as manual" [Unknown,New] https://launchpad.net/bugs/1078544 16:05:26 bug 1099278 in udev (Ubuntu Precise) "Network devices with a MAC with prefix 2c:c2:60 should not be have a persistent rule" [Medium,Triaged] https://launchpad.net/bugs/1099278 16:05:49 ⌁ done ⌁ 16:06:19 - Working on a lot of internal SEG projects this month, nothing public facing to report this week. done. 16:06:32 - Upstart (BLUEPRINT: foundations-r-upstart-user-session-enhancements) 16:06:32 - Working on the session socket code, going through reviews/fixes cycles now. 16:06:35 - Poking at a libnih bug preventing the upstart-event-bridge from working as a user. 16:06:38 - Talks on shutdown/reboot/logout sequence. 16:06:41 - Container (BLUEPRINT: servercloud-r-lxc) 16:06:43 - Code reviews. 16:06:46 - LXC upstream now supports user namespaces. 16:06:48 - Ported libcap to bionic, got LXC to build agsinst it. 16:06:51 - Re-implemented utmpx.h within LXC when building with bionic. 16:06:53 - Other 16:06:56 - Tracked down a cups-browsed bug causing a lot of upstart job respawns when avahi isn't enabled. 16:06:59 - Some follow-up on a weird biosdevname issue. 16:07:01 - Upgraded to a 3.8 kernel, helped track down a regression on ThinkPad x230/t430, fix sent upstream by Seth. 16:07:04 - Helped Serge track down my weird mount/loop device bug (loop devices aren't freeed under weird conditions). 16:07:07 - Got some new monitors with fancy DisplayPort chaining, so I've been fixing a few gnome-control-center bugs and helping debug kernel bugs. 16:07:10 - Networking (BLUEPRINT: foundations-r-networking) 16:07:13 - Some e-mails/bug activity on isc-dhcp-server-ldap not actually having LDAP support in 4.2. 16:07:16 - TODO 16:07:18 - Go through the isc-dhcp and resolvconf bugs. 16:07:21 - Try to apply redhat's fiberchannel patchset to our isc-dhcp. 16:07:23 (DONE) 16:07:56 * usb-creator-fastboot: 16:07:56 - made fastboot work to flash nexus7 images without sudo 16:07:56 - made usbcreator to pop-up when nexus7 is plugged in (in a branch) 16:07:56 - todo fastboot backend (reusing recipes from nexus7-installer) 16:07:56 * lp:~xnox/upstart/user-log-dir 16:07:57 - user upstart logging, work in progress (some review comments to address) 16:07:59 * patch pilot tuesday (report sent) 16:08:01 * unbreak devscripts, usb-creator 16:08:03 * upload ubiquity and minor merges (dmraid, pyxdg) 16:08:05 * discussed with server team that we want to keep clustered lvm 16:08:07 support (dropped in debian) 16:08:09 .. 16:08:21 - Blueprint: 16:08:21 - Upstart setenv+getenv MP raised. Working on '--global' option. Further 16:08:22 feedback appreciated: 16:08:22 https://code.launchpad.net/~jamesodhunt/upstart/setenv+getenv/+merge/142939 16:08:24 - Reviewed lp:~xnox/upstart/user-log-dir. 16:08:27 - Reviewed lp:~stgraber/upstart/upstart-session-socket. 16:08:31 - Added Shutdown details to Enhanced Sessions Spec - awaiting feedback: 16:08:34 https://wiki.ubuntu.com/FoundationsTeam/Specs/RaringUpstartUserSessions#Shutdown_Details 16:08:37 - Started work on preliminary shutdown code. 16:08:41 - SRU: 16:08:45 - Investigating ARM build failure for SRU bug 980917. 16:08:47 bug 980917 in upstart (Ubuntu Precise) "Failed to create pty - disabling logging for job [SRU]" [Medium,Fix committed] https://launchpad.net/bugs/980917 16:08:48 - Other: 16:08:51 - Investigating a customer issue. 16:08:54 - Investigated ureadahead bug 1085766 16:08:54 - Fixed bug 1096531 - awaiting feedback: 16:08:55 bug 1085766 in ureadahead (Ubuntu) "/var/log/upstart/ureadahead.log contains garbage" [High,Fix released] https://launchpad.net/bugs/1085766 16:08:57 https://code.launchpad.net/~jamesodhunt/upstart/bug-1096531/+merge/143314 16:08:57 bug 1096531 in upstart (Ubuntu) "After touch /forcefsck and reboot: Assertion failed in log_clear_unflushed" [High,In progress] https://launchpad.net/bugs/1096531 16:09:01 🀟 16:09:04 16:09:46 - OpenJDK 7 security update 16:09:46 - get creduce built and packaged, using it to scale down preprocessed source files 16:09:46 - got multiarch patches into the 4.6 and 4.7 gcc branches 16:09:46 - GCC 4.8 update 16:09:46 - buildd baby sitting 16:09:47 - merged remaining GCC bits from the cross toolchains back into GCC 16:09:49 - updated the armhf, arm64 and powerpc cross compilers 16:09:51 (done) 16:10:01 oneconf: re-enable py2 support and bug 1088536; also enabled autopkgtests. ported lazr.config to py3; released lazr.config, lazr.delegates, lazr.smtptest. looked into bug 1097783. python security issues. distribute bug #349 (crash on setup.py with BOM). done. 16:10:04 bug 1088536 in oneconf (Ubuntu) "Updating packaging for Python 3" [High,In progress] https://launchpad.net/bugs/1088536 16:10:06 bug 1097783 in genshi (Ubuntu) "re module apis return longs now" [High,Triaged] https://launchpad.net/bugs/1097783 16:10:07 Error: Launchpad bug 349 could not be found 16:13:46 jodh: ARM> yeah, I've been retrying it from time to time, to no avail :-/ 16:13:54 Geez! Hurry up, slangasek. I want to make a tea. 16:14:06 ev: :P 16:14:24 just make a strong tea then :) 16:15:44 * travel last week through Wednesday, then was out sick over the weekend + Monday with the flu, so little to report 16:15:47 * have been working to help the Kylin team get oriented so they can meet alpha2 for 13.04 16:15:50 * recovery mode has regressed due to the recent mountall SRU, so working on fixing that (bug #1099349) 16:15:52 bug 1099349 in mountall (Ubuntu Quantal) "mountall 2.36.3 hangs in recovery-menu (lvm; mountall 2.36 has no issue)" [High,Triaged] https://launchpad.net/bugs/1099349 16:15:53 (done) 16:16:34 Build fixes: pyparted (with Tim), sivp, gnumeric 16:16:35 12.04.2: Switched to X enablement stack for relevant images; cleaned up various associated issues due to having to use metapackages. 16:16:38 Phased updates: Reviewed spec and discussed with ev/bdmurray/wgrant; fixed up some glitches in update-manager; started on Launchpad implementation. 16:16:41 Cross-building: Further work on backporting architecture qualifier support to Debian buildds; met up with Wookey for status check. 16:16:44 Continued work on removing nested functions from GRUB. 16:16:46 Hard thinking about the situation where the language and location selected during installation don't identify a valid locale (bug 1094872). 16:16:49 bug 1094872 in ubiquity (Ubuntu Raring) "Installation information cause letter format instead of DIN A4" [High,Triaged] https://launchpad.net/bugs/1094872 16:16:49 Looking at entropy pool issues at first boot (bug 1098299). 16:16:51 bug 1098299 in installation-report (Ubuntu) "entropy pool should be seeded earlier in boot process" [High,In progress] https://launchpad.net/bugs/1098299 16:16:52 .. 16:17:29 - Continued work on the prodstack error tracker deployment. 16:17:29 - I've moved my testing over to EC2 from canonistack. Between the resource 16:17:29 limits (I couldn't spawn 15+ instances) and frequently getting instances in 16:17:29 the error state, canonistack was really slowing me down. 16:17:29 We need to fix this though. EC2 is expensive ($15 for two days usage), and 16:17:29 the error tracker infrastructure now cannot fit on canonistack without 16:17:29 modifying people's resource limits, which hinders our ability to test and 16:17:30 on-ramp developers. 16:17:30 - Fixing bugs in charms along the way. Pushed one fix to jjo for Cassandra, 16:17:31 and filed bug 1099917 for the other. 16:17:31 - Fixed bug 1100245 in juju with https://code.launchpad.net/~ev/juju/fix-1100245/+merge/143483 16:17:32 - A few bugs to iron out, but we're close. Webops is supposed to be picking 16:17:32 bug 1099917 in cassandra (Juju Charms Collection) "Cassandra charm changes cluster name incorrectly" [Undecided,New] https://launchpad.net/bugs/1099917 16:17:33 this up and throwing it at stagingstack, but they're firefighting. 16:17:33 - Code review for Brian. Discussions on the implementation of phased updates. 16:17:36 bug 1100245 in juju "exceptions.UnboundLocalError: local variable 'settings' referenced before assignment" [High,Triaged] https://launchpad.net/bugs/1100245 16:18:20 done ? 16:18:43 he paused for breath 16:18:58 doesnt sound like breath :) 16:19:02 more like chatting :P 16:20:18 cjwatson: thanks :) We have fixed the issue in newer releases so I'll cherry-pick the relevant bits and make a .2 release for precise. 16:20:18 - Code review for Brian. Discussions on the implementation of phased updates. 16:20:19 - Started implementing "silent error types" after discussion with Martin: 16:20:19 https://wiki.ubuntu.com/ErrorTracker#error 16:20:19 (DONE) 16:20:36 done: 16:20:37 * started fiddling with porting the nx7 images to xz compression 16:20:37 * looked into porting update-notifier to pkexec (then only xdiagnose will keep gksu on the cd) 16:20:37 * started on fiddling with tegra3 gstreamer codecs, seems there is half a community port to gst 1.0 already 16:20:37 todo: 16:20:38 * discuss livefs builder situation with infinity (carried over from last week) 16:20:40 * fix plymouth vs console-setup racing somehow (likely needs cjwatson advice) to allow initrd-less booting 16:20:43 .. 16:21:38 cool 16:21:42 [TOPIC] Bugs 16:21:46 bdmurray? 16:22:15 bug 1093819 looks interesting 16:22:17 bug 1093819 in Wubi "Wubi installed 12.10 amd64 without configuring i386" [Undecided,New] https://launchpad.net/bugs/1093819 16:23:05 I have not tested it myself though 16:23:34 cjwatson: any idea about that one? where are we running dpkg --add-architecture now? 16:24:01 apt-setup 16:24:04 generators/01setup 16:25:02 controlled by apt-setup/multiarch preseeding, which defaults to "i386" on amd64 and empty on all other arches 16:25:23 is it possible preseeding isn't working right under wubi? 16:25:24 don't see wubi preseeding that 16:25:27 and there is dpkg maint script to add i386. 16:25:31 ah 16:25:36 slangasek: it would have to be actively preseeding it *off* 16:26:10 should be easy to look in the preinstalled image and check, I guess 16:26:31 oh yes, of course, preinstalled image so it might not run apt-setup 16:26:40 * cjwatson forgets the exact details of the new setup - ev? 16:26:57 even so, dpkg.postinst should take care of it 16:27:37 oh hi 16:27:56 it's been a while for me too :), but: 16:28:14 compressed disk image unpacked, then small initramfs tomfoolery 16:28:21 for initial setup 16:29:44 how much of ubiquity does it go through, if anyy? 16:29:45 *any 16:29:51 none 16:29:58 ok, so it's purely about the preinstalled image 16:30:13 lupin -> scripts/local-bottom/lupin_setup 16:30:26 * cjwatson peers at cdimage 16:30:41 oh great, it's a tar of a root.disk so I can't easily inspect it locally ... 16:30:50 * ev whistles 16:31:23 hey man, I just make the tea around this place. Any name you find in bzr blame that looks similar to mine is purely coincidental. 16:31:49 so, er, somebody with vaguely fast networking might like to download/unpack/loop-mount http://releases.ubuntu.com/12.10/ubuntu-12.10-wubi-amd64.tar.xz and see what its dpkg metadata looks like 16:32:55 so then an automatic upgrade test is failing in bug 1096022 16:32:58 bug 1096022 in update-manager (Ubuntu) ""E:Error, pkgProblemResolver::Resolve generated breaks" during lucid->precise universe upgrade of amd64" [Medium,New] https://launchpad.net/bugs/1096022 16:33:10 bdmurray: can you follow up on that download/unpack/loop-mount after the meeting, then? 16:33:52 slangasek: yes, downloading now 16:34:01 cool 16:34:16 bdmurray: hum, so this is a *new* upgrade failure in the jenkins autotester? 16:34:38 slangasek: that's the impression I got 16:35:05 do we know what date it started happening? Maybe we can trace it to a particular SRU's publication 16:35:50 hrm, actually it looks intermittent 16:35:58 https://jenkins.qa.ubuntu.com/view/Precise/view/Upgrade%20Testing%20Dashboard/job/precise-upgrade-lucid-universe/ARCH=amd64,LTS=lts,PROFILE=universe,label=upgrade-test/ 16:36:39 oh, well then :P 16:36:46 Okay, I'll look at that in more depth. 16:37:05 Who could I talk to about understanding the pkgProblemResolver logs though? 16:37:30 * slangasek raises his hand 16:37:41 great 16:37:57 somebody in my timezone even! 16:38:38 I ran across bug 1077551 again and it seems worrisome to me 16:38:41 bug 1077551 in aptdaemon (Ubuntu) "Installing Machinarium removes the desktop" [Medium,New] https://launchpad.net/bugs/1077551 16:39:08 something's strange about that jenkins build, though; why is Dec 27, 2012 the oldest run of the job? 16:39:34 did we lose history? 16:39:37 there was a comment from glatzor regarding 'We already don't allow to remove "essential" packages by using aptdaemon, e.g ubuntu-minimal. Perhaps we should add some more to the list (e.g. unity or even ubuntu-desktop)?' 16:41:02 I'm not sure that's something that should be a policy in aptdaemon 16:41:17 update-manager has that kind of policy, and maybe software-center should too 16:43:35 so I'll follow up with the software center team then? 16:45:06 sounds good to me 16:46:03 then looking at errors.ubuntu.com this has a high instance count 16:46:09 https://errors.ubuntu.com/bucket/?id=%2Fusr%2Fbin%2Flsb_release%3AIOError%3A%3Cmodule%3E%3Amain%3Acheck_modules_installed%3Agetoutput%3Agetstatusoutput 16:47:02 err, what ? 16:48:00 * ogra_ gets a weird bluescreen here 16:48:04 hmm the instance I looked at has a ProcCmdline of /usr/bin/python -Es /usr/bin/lsb_release -a 16:48:22 clicking it gets me to http://www.homestarrunner.com/sbemail118.html 16:50:02 ev, is that on purpose or some bad joke i dont understand ? 16:50:10 bdmurray: very strange; why would we start seeing failures from lsb_release in 12.04, 9 months after release? 16:50:15 it means there's an error in the code 16:50:24 and it's not a bad joke! :) 16:50:28 ogra_: it could be a great joke you don't understand. 16:50:37 yeah! 16:50:45 well, i got from the SSO page directly to that bluescreenish page 16:50:54 pretty scary 16:51:02 telling me i have a virus 16:51:03 in the past I've seen failures from lsb_release which were due to people building derivatives and misconfiguring /etc/lsb-release 16:51:04 yup, digging at your error now 16:51:24 not sure that fits this case though 16:51:28 i just get an sso page loop 16:51:49 i get that when i try it in a second tab 16:52:00 barry: I did too until I clicked all the checkboxes :) 16:52:12 yeah, you need to send your team 16:52:16 I have a branch that makes this a bit more clear 16:52:26 ogra_: could you try again? 16:52:33 just trying to figure out which error is yours 16:52:53 * barry goes click-n-clack 16:52:54 ev, works now 16:53:06 damn, though it looks like it's unicode fun 16:53:11 and ECHILD is just a bizarre thing to happen there 16:53:55 that kind of looks like "somebody else reaped my child process before I could" 16:54:14 bdmurray: btw - have you seen bug 1100198? 16:54:16 bug 1100198 in apport (Ubuntu) "source_linux.py logic fails to detect 3.8.0-0-generic as an Ubuntu kernel" [Undecided,New] https://launchpad.net/bugs/1100198 16:54:35 jodh: no, I'll have a look 16:57:25 bdmurray: do you want to open a bug report for the lsb_release crash, and we can add it to the stack? It doesn't look urgent 16:58:11 slangasek: sounds good 16:58:28 anything else? 16:58:30 there might have an apport sru that changed this 16:58:36 hmm 16:59:21 anyway, that is all 16:59:27 ok 16:59:29 [TOPIC] AOB 16:59:35 anything else else? 17:00:04 I was wondering about an idea for the build-dep-substitution problem in cross-builds 17:00:39 (e.g. that when you see Build-Depends: gcc-4.6, you probably actually want Build-Depends: gcc-4.6:native, gcc-4.6-$HOST_TRIPLET) 17:01:26 What would people (esp. doko, infinity) think about a scheme whereby packages that need this kind of substitution done have some field in their package metadata (Cross-Available: yes or something) to say so? 17:01:57 It would beat writing a big list into sbuild and dpkg-checkbuilddeps, having to look at the apt cache in dpkg-checkbuilddeps, or any of the other heuristics I've been able to think of 17:02:57 cjwatson: maybe we can take that to #ubuntu-devel for follow-up? 17:03:03 sorry, I'm being pulled into meetings for the client sprint :) 17:03:31 Fair enough, or I might just take it to debian-devel since I believe there's an appropriate thread there at the moment 17:03:36 ok 17:03:39 #endmeeting Generated by MeetBot 0.1.5 (http://wiki.ubuntu.com/meetingology)