[15:00]  * slangasek waves
[15:00] <stokachu> o/
[15:00]  * ogra_ shores
[15:00] <jodh> \o
[15:00] <doko> hi
[15:00] <slangasek> #startmeeting
[15:00] <meetingology> Meeting started Wed Apr 10 15:00:50 2013 UTC.  The chair is slangasek. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
[15:00] <meetingology> 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
[15:01] <slangasek> [TOPIC] lightning round
[15:01] <slangasek> $ echo $(shuf -e barry doko stgraber jodh ev bdmurray slangasek ogra cjwatson xnox stokachu)
[15:01] <slangasek> cjwatson stokachu xnox stgraber jodh doko ev ogra barry bdmurray slangasek
[15:01] <slangasek> cjwatson: and... go!
[15:03] <cjwatson> Minor cdimage publishing fixes and other general helping out with beta-2.
[15:03] <cjwatson> Continuing to push the Haskell transition up the hill in spare moments.  Almost there now.
[15:03] <cjwatson> Fixed bug 1158750 (ubiquity generated non-UTF-8 locales in some cases).
[15:03] <cjwatson> Spent most of the rest of the week working on an app installer prototype at Steve's request, including discussion with Stuart.
[15:03] <cjwatson> ..
[15:03] <cjwatson> (Sorry, was off in another window poking at a bug in the last item)
[15:03] <stokachu> Top priority: bug 1158465
[15:03] <stokachu> Need Sponsor/SRU: bug 1013798, bug 1027086, bug 857983, bug 859600, bug 1069570
[15:03] <stokachu> Need ETA for: bug 1057358
[15:03] <stokachu> Backport request: 1068399
[15:03] <stokachu> ...
[15:06] <ogra_> xnox sent a mail
[15:06] <slangasek> he did - and stgraber is off today
[15:06] <stokachu> i'd really love to have bug 1158465 done today
[15:07] <slangasek> jodh:
[15:07] <jodh> * blueprints:
[15:07] <jodh>   - foundations-r-upstart-user-session-enhancements:
[15:07] <jodh>     - Submitted MP for basic Upstart python3 module + unitests:
[15:07] <jodh>       lp:~jamesodhunt/upstart/python-upstart-module
[15:07] <jodh>     - Currently working through review comments - thanks barry!
[15:07] <jodh> * other:
[15:07] <jodh>   - almost finished work for bug 1157713 (need to finish grinding through
[15:07] <jodh>     autoconf+libtool to plug the bits together).
[15:07] <jodh>   - investigating bug 1124384: Behaviour is still mystifying. Hit a few
[15:07] <jodh>     problems with LXC in the process and wrote a quick set of DEP-8
[15:07] <jodh>     tests to avoid future re-occurence:
[15:07] <jodh>     https://code.launchpad.net/~jamesodhunt/ubuntu/raring/lxc/dep-8-tests/+merge/157938
[15:07] <jodh> ⥺
[15:08] <doko> - python 2.7.4 and 3.3.1 releases
[15:08] <doko> - GCC 4.6.4 and 4.7.3 release candidates
[15:08] <doko> - ftbfs in main down to four
[15:08] <doko> - ftbfs in universe below 500
[15:08] <doko> - eclipse, ecj, swt-gtk updates and fixes
[15:08] <doko> - Tcl/Tk multiarch fixes
[15:08] <doko> - Twisted source package split into one for Py2 and one for Py3
[15:08] <doko> - python-defaults fixes for -proposed
[15:08] <doko> - buildd watch
[15:08] <doko> (done)
[15:08] <doko> do I need FFe's for the final GCC releases?
[15:08] <xnox> stokachu: bug 1158465 seems it's already in -proposed, but missing the comment/status change on the bug report. Can someone from SRU team publish the comment?
[15:08] <slangasek> stokachu: 1158465 is needing a sponsor?  perhaps someone on the cloud team could/should sponsor this?
[15:09] <xnox> wrong number. The bug 1057358 needs a comment that it is in -proposed already.
[15:09] <stokachu> slangasek: the only cloud guys i know are in CTS and do not have rights to sponsor
[15:10] <slangasek> jodh: so stgraber and xnox are both off today; are there any MPs you're currently blocked on getting a review of?
[15:10] <jodh> slangasek: well, it'd be good to get further input on https://code.launchpad.net/~jamesodhunt/upstart/allow-multiple-cmdline-confdirs/+merge/156512
[15:10] <ev> - Wrote a script to poke at the 'retracing' row of the Indexes CF to see how
[15:10] <ev>   many problems are actually seen as being retraced in the DB vs how many
[15:10] <ev>   actually are. Followed up with an RT:
[15:10] <ev>   http://rt.admin.canonical.com/Ticket/Display.html?id=60566
[15:10] <ev>   I also made some small improvements to the retracing code to avert this sort
[15:10] <ev>   of thing in the future and log when it happens. This should mean quite a few
[15:10] <ev>   more problems showing up as retraced, but it shouldn't skew our daily counts.
[15:10] <ev> - Resurrected the BucketVersions rebuild code.
[15:10] <ev> - Looked into getting apport working on the nexus 7. Rebuilt the initrd
[15:10] <ev>   disabiling debuggerd, setting the proper ulimits, and correctly setting the
[15:10] <ev>   core pipe handler, but no dice.
[15:10] <ev>   - Talked through the kernel team with it on ogra's advice, but I didn't get
[15:10] <ev>     much of a response. The android kernel seems to completely ignore
[15:11] <ev>     /proc/sys/kernel/core_pattern. gcore works just fine. Suggestions welcome.
[15:11] <ev> - Fought weird data in the crashdb encountered in our script to fix bug
[15:11] <slangasek> jodh: ok, queuing it here
[15:11] <ev>   1077122 (RT 60205). Looks like FS corruption:
[15:11] <ev>   http://paste.ubuntu.com/5680293/
[15:11] <ev>   Fixed, and after about two days of running, we finally have data to quickly
[15:11] <ev>   weight errors by how long its been since the first one for that system for
[15:11] <ev>   that release, which should repair the spikes in the average errors per
[15:11] <ev>   calendar day graph.
[15:11] <ev> - Looking into why the retracer amd64 failed queue (indicates apport-retrace
[15:11] <ev>   crashing) significantly climbed over the past few days. Improved logging
[15:11] <ev>   around apport-retrace failing while running the failed queue against apport
[15:11] <ev>   trunk tip.
[15:11] <ev>   - Upon investigating further, it looks like the firewall is still preventing
[15:11] <ev>     a large number of retraces from looking up the Contents.gz file. I built a
[15:11] <ev>     new version of apport to dig at this further, but then did some digging on
[15:11] <ev>     one of the retracer machines and confirmed that it's the HTTP proxy we're
[15:11] <ev>     using.
[15:11] <ev> - More investigative work into how we're bucketing crashes, what we're missing,
[15:11] <ev>   and how the algorithm can be improved. I've been doing this via the script
[15:11] <ev>   that's meant to repopulate BucketVersions with per-release
[15:11] <ev>   sorted-by-dpkg-version data. This dovetailed nicely with the KernelOOPs work.
[15:11] <ev> - Colleague reviews.
[15:11] <ev> - Implemented a new URL for finding a bucket by its bug number
[15:11] <ev>   (https://errors.ubuntu.com/bug/1094218), based on the earlier work Brian did
[15:11] <ev>   (RT 60640, deployed).
[15:11] <ev> - Worked with Emily on the layout of errors.ubuntu.com.
[15:11] <ev> - Meetings with Tom and James to discuss bringing up a secondary Cassandra ring:
[15:11] <ev>   - We want to be able to upgrade Cassandra to 1.2, but James is worried about
[15:11] <ev>     the risks without a hot-standby ring in place.
[15:11] <ev>   - We want to deploy Hadoop on Cassandra and open up access to data scientists
[15:11] <jodh> slangasek: ta
[15:11] <slangasek> doko: is there something we could have done differently to catch the python-minimal module issue before it hit the archive?
[15:11] <ev>     within Canonical, but James and Tom don't want to do this until we have a
[15:11] <ev>     way to keep that load off the primary cluster, like with a secondary one.
[15:11] <ev>   - We need to add more nodes anyway, as we're running out of disk space again.
[15:11] <ev>   - We can (presumably) save a lot of space by using compression, but James
[15:11] <ev>     doesn't want to look into turning this on retroactively until after the
[15:11] <ev>     13.04 release.
[15:11] <ev>   - We came up with a number of options, each with its own set of problems and
[15:11] <ev>     cost, that we intended to put into a write up for approval by Robbie/Rick.
[15:11] <ev>   - However, James came up with a plan to build the secondary ring in
[15:11] <ev>     Prodstack, so we're going to build it out there, upgrade to 1.2, and if all
[15:11] <ev>     goes well, upgrade the primary ring to 1.2:
[15:11] <ev>     https://rt.admin.canonical.com/Ticket/Display.html?id=60652
[15:11] <ev> - Looking into why we're not getting a lot of KernelOOPS reports for Andy.
[15:11] <ev>   - Turned out to be a bug where we're not bucketing them properly. In the
[15:11] <ev>     process of writing test cases for this and fixing by refactoring the
[15:11] <ev>     submission code.
[15:11] <ev> - Set up a meeting with Legal to create an NDA for https://errors.ubuntu.com \o/
[15:12] <ev> - Got a new version of django-openid-auth rolled out to production to fix:
[15:12] <ev>   - https://errors.ubuntu.com/oops-local/2013-04-09/04522.errors.ubuntu.com1
[15:12] <ev>   - https://errors.ubuntu.com/oops-local/2013-04-09/77302.errors.ubuntu.com1
[15:12] <ev>   - https://bugs.launchpad.net/errors/+bug/1073680
[15:12] <ev> (done!)
[15:12] <ogra_> phew
[15:12] <stokachu> strongbad!
[15:12] <ogra_> done:
[15:12] <ogra_>  * revisited the livecd-rootfs implementation of building android images, sadly the build has to source some shell from the android build system which is full of bashisms (about 3000 lines of bad bash script)
[15:12] <ogra_>    pondering to fix the build system or to just use #!/bin/bash in the scripts
[15:12] <ogra_>  * armhf image tests and fixes (ac100, omap4)
[15:12] <ogra_>  * reviewed several merge proposals
[15:12] <ogra_>  * some bug work
[15:12] <ev> computer over!
[15:12] <ogra_>  * added experimental raring images to the phablet sync process on cdimage
[15:12] <ogra_> todo:
[15:12] <ogra_>  * finish the livecd-rootfs android build implementation
[15:12] <stokachu> haha
[15:12] <ogra_>  * go forward on the cdimage side of the android cross build stuff
[15:12] <ogra_>  * finally start with the emulator work for ubuntu touch
[15:12] <doko> slangasek, better review? I think I messed with the backport, for (py3, not py2)
[15:12] <ogra_>  * look at universe armhf only ftbfs
[15:12] <ogra_> ..
[15:12] <slangasek> stokachu: seems that utlemming is the maintainer of the walinuxagent package, I imagine he would be willing and able to sponsor for you
[15:13] <stokachu> slangasek: sponsor his own MP?
[15:13] <doko> slangasek, it was python3-minimal, not python-minimal
[15:14] <barry> bug 1163609 (really bug 1124508).  bug 1063350.  still working on bug 1078697.  reviewed jhunt's upstart branch.  more research on click packages, on hold now for results of dpkg prototype. dmb meeting. done.
[15:14] <slangasek> stokachu: sure; no rule against it (and in fact, since MPs are mostly useless for SRUs, he should Just Upload™)
[15:15] <stokachu> slangasek: i asked utlemming and he doesn't have rights to upload
[15:15] <slangasek> stokachu: wat!
[15:15] <bdmurray> updated daisy branch based on evan's feedback
[15:15] <bdmurray> updated oopsrepository branch per evan's changes
[15:15] <bdmurray> added test for oops-repository srcversbuckets cf
[15:15] <bdmurray> worked on daisy and errors branches to create a BugToCrashSignature_CF
[15:15] <bdmurray> merge my branch to daisy and errors for bug to crash signatures work
[15:15] <slangasek> stokachu: ok, well I'm due to take a patch pilot day this week, so I'll have a look today
[15:15] <ev> woo!
[15:15] <bdmurray> updated daisy branch based on evan's feedback
[15:15] <bdmurray> updated oopsrepository branch per evan's changes
[15:15] <bdmurray> added test for oops-repository srcversbuckets cf
[15:15] <bdmurray> worked on daisy and errors branches to create a BugToCrashSignature_CF
[15:15] <bdmurray> merge my branch to daisy and errors for bug to crash signatures work
[15:15] <stokachu> slangasek: thanks it is _really_ urgent to get done today if at all possible
[15:15] <ogra_> geez, you did double the work !
[15:16] <slangasek> ogra_: my advise is to switch to bash rather than create a delta with the android upstream buildsystem to try to remove bashisms
[15:16] <bdmurray> setup error tracker in canonistack again
[15:16] <bdmurray> pushed branch fixing errors bug 1064398 and daisy bug 1056264
[15:16] <bdmurray> code review for evan regarding hashes
[15:16] <bdmurray> wrote pycassa query to use SourceVersionBuckets to find new crashes in packages
[15:16] <bdmurray> iso testing bug triage
[15:17] <bdmurray> ogra_: there new stuff now
[15:17] <bdmurray> bug triage of ubuntu-release-upgrader, update-manager, and ubiquity bug reports
[15:17] <bdmurray> review of update-manager bug day work
[15:17] <bdmurray> uploaded ubuntu-release-upgrader fix for bug 1127451
[15:17] <bdmurray> fixed bug 801288 regarding hugday
[15:17] <bdmurray> investigation into http://launchpad.net/bugs/1162469
[15:17] <bdmurray> investigation into policykit password issue
[15:17] <bdmurray> research into bug 1069072
[15:17] <bdmurray> update-notifier: modified avahi disabled notification to run as an upstart user session
[15:17] <ogra_> slangasek, well a simple s/^function // got checkbashisms from ~1500 complaints down to 300 :)
[15:17] <bdmurray> done
[15:18] <ogra_> but there are a lot of arrays in there i dont want to rework
[15:18] <slangasek> ogra_: I'm sure the last set won't be nearly so easy ;p
[15:18] <ogra_> yeah
[15:18] <ogra_> the bad thing is they are all suffixed with .sh ... but have no shebang at all
[15:18] <slangasek> bdmurray: did you mention that you've successfully identified an SRU regression in -proposed now using errors? \o/
[15:19] <ogra_> so it blindly assumes you have bash when building
[15:19] <ogra_> (everything is just sourced)
[15:20] <bdmurray> slangasek: no, not explictly
[15:20] <slangasek> bdmurray: ah well, kudos for it, I think this is a big deal :)
[15:20] <slangasek>  * finished up reviews... everyone's getting a promotion and/or fired and/or getting a raise, tune in for the exciting conclusion of this story next month
[15:20] <slangasek>  * helped get Final Beta out last week, noticed some gaps in our checklist now that alphas are optional, working to fix the process so things don't fall through the cracks
[15:20] <slangasek>  * attending to partner archive (new fluendo gstreamer plugins needed)
[15:20] <slangasek> (done)
[15:20] <bdmurray> slangasek: thanks!
[15:20] <slangasek>  * sponsor upload of likewise-open for bug #1162836
[15:21] <slangasek>  * landed my first patch to compiz (bug #763148), should land in the archive any day now if the unity regression tests manage to pass
[15:21] <slangasek>  * reviews of the frozen queue
[15:21] <slangasek> any further questions over status?
[15:21] <slangasek> doko: when you say "better review", do you mean in the queue?
[15:22] <slangasek> doko: queue reviews are expensive; I'd rather we think about whether this could have been caught automatically with the right kinds of pre-acceptance testing
[15:23] <doko> slangasek, well, both by myself, and by the queue handler, but the diff moving stuff from binary-indep to binary-arch wasn't that helpful
[15:23] <doko> which you can't always catch, if you you unpack one package before the other
[15:24] <slangasek> ev: btw, on the android kernel question, ogra_'s actual suggestion was to talk to ricardo (== rsalveti), who I see hasn't been active on #ubuntu-kernel
[15:24] <slangasek> ev: so perhaps there's a useful lead there for you still
[15:25] <ev> ohhhh
[15:25] <ev> slangasek, ogra_: thanks!
[15:25] <ev> will do
[15:25] <ogra_> yeah, in #ubuntu-touch
[15:26] <rsalveti> sorry, what is the question, trying to find at the backlog
[15:26] <slangasek> doko: right... but we do have quite a bit of setup around upgrade testing, so given that python is critical to update-manager, perhaps it makes sense to always do an upgrade test when we upload python3?  (automatically in QA, not you personally)
[15:26] <slangasek> rsalveti: apport not working on Android kernel on nexus7 because it seems to be ignoring the core handler setting
[15:26] <slangasek> [TOPIC] Bugs
[15:27] <rsalveti> slangasek: right, thanks
[15:27] <bdmurray> I was able to recreate bug 1160441 the other day
[15:27] <stokachu> my top priority bug was addressed but could i confirmation on my other bugs posted
[15:27] <bdmurray> stokachu: I'll look 1013798 today
[15:28] <stokachu> thanks
[15:28]  * slangasek reposts the list of stokachu's bugs:
[15:28] <slangasek> 08:03 < stokachu> Need Sponsor/SRU: bug 1013798, bug 1027086, bug 857983, bug 859600, bug 1069570
[15:28] <slangasek> 08:03 < stokachu> Need ETA for: bug 1057358
[15:28] <slangasek> 08:03 < stokachu> Backport request: 1068399
[15:28] <slangasek> man I hate libgcrypt
[15:28] <stokachu> slangasek: fortunately this fix is in python-gnutls
[15:29] <stokachu> i think seb128 may be best for 857983
[15:30] <slangasek> bdmurray: hmm, so that calendar code is in ubiquity?
[15:30] <bdmurray> I think the regional format was set incorrectly
[15:30] <stokachu> 1057358 needs the sru message about it being in -proposed
[15:30] <slangasek> bdmurray: I guess I'd check to see what's in /etc/default/locale + ~/.pam_environment in these various scenarios
[15:31] <slangasek> stokachu: bug #859600 is currently blocked by another SRU in -proposed, feel free to harrangue the submitter of that bug to do his verification (I sent a ping on the bug yesterday)
[15:32] <cjwatson> bdmurray: 1160441> could be another consequence of bug 1158750 which I fixed
[15:32] <bdmurray> cjwatson: okay, I'll test it again then
[15:33] <cjwatson> a bunch of stuff seems to have bitrotted in terms of ability to cope with non-UTF-8 locales
[15:33] <doko> https://launchpad.net/ubuntu/+source/python-gnutls/1.2.4-1ubuntu1 that one is built
[15:33] <stokachu> slangasek: whats the other bug #?
[15:33] <cjwatson> bdmurray: but this would be quick to verify by looking at /etc/default/locale
[15:33] <slangasek> stokachu: bug #1094496
[15:34] <cjwatson> bdmurray: or it's true we could still be missing some locale category or something; it's all distressingly hardcoded ...
[15:34] <slangasek> stokachu: (https://launchpad.net/ubuntu/+source/gnome-keyring/, precise-proposed, expand details)
[15:34] <bdmurray> cjwatson: I already deleted the virtual machine
[15:34] <stokachu> slangasek: ok
[15:34] <bdmurray> Is bug 1160783 something for cyphermox or us?
[15:35]  * cjwatson updates the bug
[15:35] <bdmurray> ev: thanks for errors.ubuntu.com/bug/1160783 !
[15:35] <cjwatson> bdmurray: us I suspect
[15:36] <slangasek> stokachu: fixed the missing info on bug #1057358
[15:36] <bdmurray> actually there aren't that many instances of it
[15:36] <stokachu> slangasek: awesome thanks
[15:36] <slangasek> per-bug errors pages - yes, woot
[15:38] <ev> bdmurray: sure thing :)
[15:38] <ev> it was an easy one to pick off
[15:38] <bdmurray> moving on to bug 1164754 thumper upgraded from Q to R and did not have gnome-control-center-unity installed but I'm still waiting on log files
[15:38] <bdmurray> any ideas though?
[15:39] <stokachu> my low hanging fruit one is bug 1068399
[15:40] <slangasek> bdmurray: well, unity only recommends gnome-control-center-unity, g-c-c-u isn't a dep or recommend of ubuntu-desktop...  is an ubuntu-meta upload needed?
[15:40] <slangasek> (not sure it should ever show up in Task: ubuntu-desktop without the metapackage first being uploaded?)
[15:41] <cjwatson> it would show up in Task, though that won't affect upgrades
[15:41] <cjwatson> gnome-control-center-unity isn't mentioned in the desktop seed or anything
[15:41] <slangasek> right, so it seems like it should get seeded?
[15:41] <ogra_> oh, that reminds me that i also have an outstanding meta upload
[15:42]  * ogra_ nearly forgot about that
[15:43] <slangasek> ogra_: perhaps coordinate with bdmurray wrt g-c-c-u?
[15:43] <ogra_> happy to do the seed change for him
[15:43] <xnox> yes we need to have gnome-control-center-unity as it replaces the /stock/ appearance panel with many unity settings.
[15:44] <bdmurray> ogra_: great thanks!
[15:45] <bdmurray> I did some digging into bug 1102593 but didn't see how it could be occurring
[15:46] <slangasek> does the submitter have teamview installed?
[15:46] <slangasek> that's always a good way to get bugs
[15:47] <bdmurray> so if somebody could have a 2nd look at it with me that'd be great
[15:47] <barry> bdmurray: i'm looking now
[15:48] <bdmurray> barry: thanks
[15:48] <bdmurray> then bug 1163773
[15:48] <bdmurray> I saw a few similar reports regarding upgrading kernel packages
[15:49] <bdmurray> like bug 1163275
[15:51] <slangasek> cjwatson: any ideas on 1163773?
[15:53] <cjwatson> slangasek: could be a dup of bug 1163142 et al; commented
[15:54] <slangasek> ok, cool
[15:54] <cjwatson> certainly looks like a similar ps structure
[15:54] <slangasek> bdmurray: anything else?
[15:54] <cjwatson> I was expecting that bug to have a bazillion unmarked dups
[15:55] <bdmurray> stokachu: it looks like somebody got to bug 1013798, I'll review them in the sru queue then
[15:55] <stokachu> bdmurray: awesome, thanks man
[15:55] <bdmurray> cjwatson: I'll poke around for dupes
[15:55] <bdmurray> slangasek: bug 1058884
[15:58] <slangasek> strange that the file has the wrong checksum
[15:59] <slangasek> was that corruption, or just a wrong version?  (Why did you suspect the file was wrong?)
[16:00] <bdmurray> slangasek: it seemed like either the file was wrong or the hardware was having issues
[16:01] <slangasek> hmm
[16:01] <jodh> some of those bugs mention py3clean - .py* files being removed at the same time as being read?
[16:02] <slangasek> who could follow up on this?  barry?
[16:02]  * slangasek looks at the clock
[16:03] <barry> slangasek: bug 1058884?
[16:03] <slangasek> barry: yes
[16:03] <barry> yikes.  that's odd.  okay, i'll look at it
[16:04]  * barry was already looking at bug 1102593 ;)
[16:04] <slangasek> barry: thanks :)
[16:04] <slangasek> [TOPIC] AOB
[16:04] <slangasek> anything else?
[16:04] <jodh> Coverity !!
[16:05] <slangasek> :)
[16:06] <ogra_> crazy talk
[16:06] <slangasek> #endmeeting
[16:06] <meetingology> Meeting ended Wed Apr 10 16:06:19 2013 UTC.
[16:06] <meetingology> Minutes (wiki):        http://ubottu.com/meetingology/logs/ubuntu-meeting/2013/ubuntu-meeting.2013-04-10-15.00.moin.txt
[16:06] <meetingology> Minutes (html):        http://ubottu.com/meetingology/logs/ubuntu-meeting/2013/ubuntu-meeting.2013-04-10-15.00.html
[16:06] <slangasek> thanks, all!
[16:06] <ogra_> thanks !
[16:07] <barry> thanks!
[16:07] <jodh> ta