#title #ubuntu-meeting Meeting Meeting started by cjwatson at 16:02:07 UTC. The full logs are available at http://ubottu.com/meetingology/logs/ubuntu-meeting/2012/ubuntu-meeting.2012-11-21-16.02.log.html . == Meeting summary == *lightning round ''LINK:'' http://paste.ubuntu.com/1374772/ (xnox, 16:09:58) *bugs Meeting ended at 16:32:14 UTC. == Votes == == Action items == * (none) == People present (lines said) == * cjwatson (37) * ogra_ (30) * stgraber (28) * ev (28) * xnox (25) * ubottu (18) * doko (13) * jodh (12) * stokachu (12) * meetingology (3) * barry (2) == Full Log == 16:02:07 #startmeeting 16:02:07 Meeting started Wed Nov 21 16:02:07 2012 UTC. The chair is cjwatson. Information about MeetBot at http://wiki.ubuntu.com/meetingology. 16:02:07 16:02:07 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 16:02:14 #topic lightning round 16:02:17 $ echo $(shuf -e barry cjwatson ev doko ogra_ jodh stgraber xnox stokachu) 16:02:20 stgraber ev barry stokachu jodh cjwatson xnox ogra_ doko 16:03:08 - Upstart 16:03:08 - Spent some time talking with James and Dmitrijs talking about the plan for the Session Init work 16:03:11 - Implemented a basic upstart-dconf-bridge at: lp:~stgraber/upstart/upstart-dconf-bridge 16:03:14 - Container 16:03:16 - Reviewed a bunch of merge proposals and patches on the upstream mailing-list. 16:03:20 - Had a look at the early implementation of the syslog namespace by Serge. 16:03:22 - Wrote a script to run Steam under LXC: http://www.stgraber.org/2012/11/16/running-steam-in-a-lxc-container/ 16:03:25 - QA Tracker 16:03:28 - Did a bunch of optimization work on how the tracker loads its various plugins. 16:03:31 - Redesigned landing page (milestones list) to be filtered by javascript and show milestones grouped by series. 16:03:34 - Added a new testsuites list per balloons' request 16:03:36 - Fixed a bunch more php errors, now replaced by nice 404s :) 16:03:39 - Other 16:03:42 - Helped the security team by extracting all my SecureBoot keys out of my system (ended up being much more tricky than expected). 16:03:44 - sssd was broken by a new ldb upload, breaking authentication on most of my systems. 16:03:47 - The new linux-libc-dev broke libnl breaking the sssd rebuild I had to do. 16:03:50 - Had to patch libnl to cope with some defines rename in the 3.7 kernel, then managed to rebuild sssd and login on my systems again. 16:03:54 - Uploaded the new sudo upstream release including sssd support. 16:03:56 - Installer 16:03:59 - Fixed casper which was trying to boot an ltsp chroot instead of a the livefs on Edubuntu. 16:04:02 - Investigated bug 1079266 and bug 1077598. Was suspecting a Wubi 10.04 to Wubi 12.04 upgrade issue but that's not the case. 16:04:05 Tried to reproduce with identical package list, still without any luck. Will have to ask for more information... 16:04:06 Launchpad bug 1077598 in casper (Ubuntu) "duplicate for #1079266 package plymouth 0.8.2-2ubuntu30 failed to install/upgrade: dependency problems - leaving unconfigured" [Undecided,Confirmed] https://launchpad.net/bugs/1077598 16:04:07 Launchpad bug 1077598 in casper (Ubuntu) "package plymouth 0.8.2-2ubuntu30 failed to install/upgrade: dependency problems - leaving unconfigured" [Undecided,Confirmed] https://launchpad.net/bugs/1077598 16:04:07 - Release 16:04:10 - Filed RT ticket to get scripting access to https://help.ubuntu.com/community/UbuntuHashes from nusakan. 16:04:13 - TODO 16:04:16 - Look at jodh's upstart specs. 16:04:18 (DONE) 16:04:27 - Short week. I took holiday on Monday to move into my new house. \o/ 16:04:28 - Removed the dependency on archive.admin.canonical.com so hazmat can 16:04:28 debug the juju issue in the daisy and rabbitmq charms. 16:04:28 - Attended an Acunu webinar on Hailo's use of Cassandra, specifically 16:04:28 Analytics. Lots of really good points on how to best decide when to NIH a 16:04:28 tool or pick up a product. 16:04:34 - Uploaded fixed activity-log-manager to quantal and precise (LP: #993056). 16:04:35 - Fixed up my branch to add openid authentication to 16:04:35 http://errors.ubuntu.com/oops-local and submitted it to webops. 16:04:35 - Finally finished verifying the USB disk for the shop. 16:04:35 - Picked back up my compiz branch for spawning apport when applications hang. 16:04:39 Launchpad bug 993056 in activity-log-manager (Ubuntu Quantal) ""Privacy" > "Diagnostics" > "Send error reports" can't be turned on or off" [High,Fix committed] https://launchpad.net/bugs/993056 16:04:41 - Started looking at package installation failures as an error type again. 16:04:41 We're collecting these right now (5883 total reports on 09/11/12), and we are 16:04:41 in fact bucketing them: http://tinyurl.com/blsqa3a 16:04:41 - Mail to Steve and Matthew on conffile prompts as errors. 16:04:49 - Fix to apport for fsync'ing upstart reports, per our email discussion: 16:04:50 https://code.launchpad.net/~ev/apport/fsync-upstart-crashes/+merge/135134 16:04:50 - UDS survey. 16:04:50 - Teaching daisy retracer code to use a squid cache, rather than locally 16:04:50 caching the debs. The previous behavior was a significant part of the 16:04:55 retracer disk space growth. With this change, the webops team should be 16:04:55 getting fewer weekend pages to clear the caches. 16:04:55 - Attempting to fix having separate application thread crash dialogs: 16:04:55 https://bugs.launchpad.net/ubuntu/+source/apport/+bug/1033902 16:04:55 The simple approach is to see if the pid is still around by the time 16:04:55 apport-gtk gets the error report, but I want to make sure that's not racey. 16:04:58 (done) 16:05:00 Launchpad bug 1033902 in apport (Ubuntu) "Application thread crash shows application crash error alert" [Undecided,Confirmed] 16:05:09 tox packaging for debian (in the new queue now, will be synced to ubuntu when that lands). worked on virtualenv 1.8.2, but that isn't compatible with python 3.3, so worked with upstream to get a 1.8.3 which is Python 3.3 compatible (might even land today). distribute 0.6.30, for Python 3.3 compatibility. python-oauthlib 0.3.4. python bug 16514 (sys.path import regression in Python 3.3). more investigation on bug #1061149. merge 16:05:09 proposal for bug #1077083 (piston-mini-client -> oauthlib). preliminary review of pykde4 plugin system for porting to py3. dmb meeting. upgraded main dev box to raring. off this thu and fri for usa thanksgiving holiday. done. 16:05:13 Launchpad bug 16514 in hfsutils (Ubuntu) "hfsutils: new changes from Debian require merging" [Medium,Fix released] https://launchpad.net/bugs/16514 16:05:14 Launchpad bug 1061149 in linux (Ubuntu) "boot occasionally hangs while "Checking battery state..."" [Medium,Confirmed] https://launchpad.net/bugs/1061149 16:05:15 Launchpad bug 1077083 in piston-mini-client "Switch from python-oauth to python-oauthlib" [High,In progress] https://launchpad.net/bugs/1077083 16:05:32 bug 932860 needs sru review 16:05:33 bug 1077095 needs sru review 16:05:33 bug 967091 needs sru review 16:05:33 (done) 16:05:35 Launchpad bug 932860 in appmenu-gtk (Ubuntu Quantal) "Broken (or missing) multiarch support" [High,In progress] https://launchpad.net/bugs/932860 16:05:37 Launchpad bug 1077095 in appmenu-gtk (Ubuntu Quantal) "Please blacklist "IBM Notes"" [Low,In progress] https://launchpad.net/bugs/1077095 16:05:38 Launchpad bug 967091 in libvdpau (Ubuntu Precise) "Wrong tint in flash when it uses video acceleration" [Undecided,Confirmed] https://launchpad.net/bugs/967091 16:06:43 * upstart: 16:06:43 - fixed bug 1079710. 16:06:43 - fixed bug 1079715 and raised MP. 16:06:43 - completed work on SRU bug 980917 and uploaded to precise-proposed. 16:06:46 - spec wrangling with stgraber + xnox (and a loooong meeting ;-): 16:06:47 Launchpad bug 1079710 in upstart "Chroot jobs which produce output get logged to /var/log/upstart *outside* the chroot" [Medium,Fix committed] https://launchpad.net/bugs/1079710 16:06:48 Launchpad bug 1079715 in upstart "'telinit u' run from within a chroot causes a crash" [High,New] https://launchpad.net/bugs/1079715 16:06:50 Launchpad bug 980917 in upstart (Ubuntu Precise) "Failed to create pty - disabling logging for job [SRU]" [Medium,Fix committed] https://launchpad.net/bugs/980917 16:06:54 stgraber: SB keys> Have you been able to use this to get anywhere with debugging shim? 16:06:59 https://wiki.ubuntu.com/FoundationsTeam/Specs/RaringUpstartUserSessions 16:06:59 * TODO: 16:06:59 - rework branch and discuss dbus_connection_open_from_fd() with D-Bus upstream. 16:06:59 - resolve Upstart namespacing design (as currently being discussed on 16:07:06 the mailing lists). 16:07:06 - start work on implementing spec above with stgraber + xnox. 16:07:08 ⍾ 16:07:17 lol mario bros? 16:08:17 ev: congrats on the new house 16:08:33 Still working on Secure Boot backports to 12.04.2. debian-installer mostly done (in ppa:cjwatson/ppa), but GRUB is reporting spurious out-of-memory errors, so I've had to sidetrack into investigating that. 16:08:38 Worked on cross-autobuilder. 16:08:40 * Mainly involved teaching myself lots about juju. 16:08:42 * This now works and, with four sbuild units attached, can build - or fail to build - quantal+raring main for armhf in around a day. 16:08:43 cjwatson: not yet, but now that I have a working efivars filesystem, I should be able to add entries to the db while in setup mode, which should let me do local builds of the shim and use that for testing. Will add that to my todo, can hopefully spend some time on it this week 16:08:45 * About 27% of quantal main cross-builds; raring has regressed slightly, at 26.7%. 16:08:49 * Next step is to arrange to publish the results. 16:08:51 Patch pilot half-day on Friday. 16:08:54 Detailed discussion with Martin and Jean-Baptiste of plans for proposed-migration/autopkgtest integration. 16:08:57 Fixed various build failures: debian-installer, mpdris2, prelude-manager. 16:08:59 Fixed most of python-apt's autopkgtest failures. Martin fixed the last bits. 16:09:03 Fixed a mismerge of a cross-build-dependency handling patch in apt. 16:09:05 Marked gnome-desktop3-data Multi-Arch: foreign, spotted in cross-autobuilder results. 16:09:08 Rearranged cdimage to make it safe to build for multiple series in parallel (rather belatedly). Spent some of today tracking down a consequential failure. 16:09:11 .. 16:09:57 * investigating bug 988055 16:09:57 reproduced on lucid and got valgrind output with debug symbols 16:09:58 http://paste.ubuntu.com/1374772/ 16:09:58 Reading that I am a bit lost between rpcsecgss, libgluegss & krb5. 16:09:58 Each layer seems to allocate memory on initial connection and I am 16:09:58 failing to trace where it is not freed. 16:09:59 Launchpad bug 988055 in nfs-utils (Ubuntu Lucid) "rpc.gssd memory leak" [High,Triaged] https://launchpad.net/bugs/988055 16:10:00 * Devices FS: flashbench is now in raring & profiled my sd-cards/memory sticks. 16:10:03 got back f2fs i386/amd64 from ogasawara, the arm builds failed. 16:10:04 TODO benchmark aligned f2fs on i386/amd64 & investigate arm FTBFS 16:10:08 * ongoing upstart user sessions discussions 16:10:10 * finished clucene transition (3 packages) britney output should be 16:10:12 smaller now 16:10:14 * bug 837064 / rt 55554 is in progress of being deployed. A deployment 16:10:15 Error: Launchpad bug 837064 could not be found 16:10:17 was attempted on the lucid-based host, which failed because my patch 16:10:18 was targeting quantal's psycopg2. Now fixed & pending redeployment. 16:10:20 xnox: join the clue on that rpc mess 16:10:21 * mdadm merge is in progress. 16:10:22 * plymouth needs a rebuild against (now in raring) 16:10:24 libdrm_nouveau2. But it FTBFS bug 1081613. 16:10:25 Launchpad bug 1081613 in plymouth (Ubuntu Raring) "FTBFS in raring-proposed (libdrm_nouveau 2)" [Medium,Confirmed] https://launchpad.net/bugs/1081613 16:10:26 .. 16:10:45 club* 16:10:54 stokachu: i have a rabit hole diagram in my notepad. I am suspecting it will be rpcsecgss in the end. 16:11:37 done: 16:11:37 * nexus7 images are available under http://cdimage.ubuntu.com/daily-preinstalled/current/ (still a bit buggy but soon to be fully usable) 16:11:37 * various ac100-tarball-installer fixes (need to drop the arch name soon and make it more generic) 16:11:37 * started working on an ubuntu-default-settings-nexus7 package 16:11:37 * more plymouth debugging 16:11:39 todo: 16:11:40 xnox: if i get some free time ill keep looking into it as well 16:11:41 * set up a regular meeting for discussing the remaining WIs of https://blueprints.launchpad.net/ubuntu/+spec/desktop-r-reduced-power-ram 16:11:46 * flash-kernel fixes 16:11:48 .. 16:13:04 cjwatson: thanks! 16:13:41 - mostly a GCC week 16:13:41 - getting the target specific bits into upstream, only outstanding are alpha, ia 16:13:41 64, m68k 16:13:41 - packaged gcc-4.8, updated all local patches, new runtime libraries libatomic a 16:13:41 nd libasan 16:13:41 cjwatson: i really need to get the appmenu-gtk m-a sru'd as soon as possible if anyone could help me out 16:13:43 - found out the 4.8 build is broken when making -Wformat-security the default 16:13:45 - found out the 4.8 build is broken with -fstack-protector on armhf and powerpc 16:13:47 (PR53475) 16:13:49 - prepared backports of the multiarch patches for upstream inclusion 16:13:51 - llvm-3.2 in raring, fixed clang ftbfs 16:13:53 - binutils update 16:13:53 the other 2 are not so important 16:13:55 (done) 16:14:03 stokachu: fair enough, I'll take another look 16:14:12 thanks :D 16:14:42 s/bits/multiarch bits/ 16:14:49 will have to refresh my memory of the previous go-around 16:15:05 #topic bugs 16:15:18 Brian's not here, but anything else people want to particularly raise? Otherwise we can finish early :-) 16:15:20 cjwatson: i basically used your suggestion but did it in a way that conf files were not conflicting during parallel installs 16:17:35 ok, I'll take a look 16:17:41 * ogra_ has bug 1080747 16:17:42 Launchpad bug 1080747 in livecd-rootfs (Ubuntu) "Set a build stamp for pre-installed images" [Medium,New] https://launchpad.net/bugs/1080747 16:18:09 i'm not really sure what to do with it, the livefs builder doesnt know anything about the build name when rolling the rootfs 16:18:50 You'll need a .disk directory somewhere or similar 16:18:52 i could just randomly dump a unix timestamp into some file i guess but that seems ugly 16:18:56 It doesn't and can't belong in the livefs 16:19:07 well, i have nothing processing that 16:19:25 its either rootfs or initrd i can put stuff in 16:19:38 and by design they come finished out of live-build 16:19:44 well in that case I guess the livefs builder may have to invent its own build ids ... 16:19:50 thats my prob with this bug :) 16:20:06 so unix timestamp into some fiel then, ok 16:20:09 *file 16:20:11 thx 16:20:22 unix timestamp is a bit unfriendly. I'd use YYYYMMDD HH:MM:SS or some such 16:20:27 k 16:20:37 that way it at least stands a chance of corresponding approximately to something on cdimage 16:20:47 heh, yeah 16:21:07 try to make it look as much like other media-info files as you can, if possible 16:21:24 tehoretically cdimage should know what it will become in the end, no ? 16:21:43 we coudl look it up in advance and hand it to the live builder as BuildLiveCD arg 16:21:57 (not for now, but as a general fix) 16:22:46 ... as long as ubuntu-bug ubiquity works and hands over a sensible id that we can identify all is fine. 16:23:09 i'll use the std apport path /var/log/installer/media-info 16:23:11 not with the way buildlive is currently (mis)designed 16:23:17 so apport will even pick it up 16:23:25 it runs before cdimage has picked a build id 16:23:44 cjwatson, well, good for the next redesign then :) i didnt mean we need it now, a hack will do for me 16:27:03 anything else? 16:27:15 not here 16:27:36 bug 1081613 that I mentioned, but it's not urgent. 16:27:37 Launchpad bug 1081613 in plymouth (Ubuntu Raring) "FTBFS in raring-proposed (libdrm_nouveau 2)" [Medium,Confirmed] https://launchpad.net/bugs/1081613 16:27:50 yeah, I was hoping to constructively ignore that until Steve is back ;-) 16:27:54 simply britney prints 8000 package uninstallable because of libdrm 16:28:12 we need to update plymouth to 0.8.5 anyway 16:28:28 so if anyone wants short britney output.... feel free to update plymouth =) 16:28:40 slangasek said it wouldnt be an easy merge though, seems we package our own upstream 16:29:07 there was a proposal with 0.8.6.1 but it's not a clean debdiff. 16:32:13 OK, sounds like that's everything for now, thanks all 16:32:14 #endmeeting Generated by MeetBot 0.1.5 (http://wiki.ubuntu.com/meetingology)