[15:00]  * slangasek waves
[15:00]  * sil2100 waves back
[15:01]  * barry just wavers
[15:01]  * bdmurray waivers
[15:01] <slangasek> #startmeeting
[15:01] <meetingology> Meeting started Thu Aug 13 15:01:31 2015 UTC.  The chair is slangasek. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
[15:01] <meetingology> Available commands: action commands idea info link nick
[15:02] <slangasek> [TOPIC] Lightning round
[15:02] <slangasek> $ echo $(shuf -e barry doko bdmurray slangasek caribou infinity sil2100 robru cyphermox pitti tdaitx)
[15:02] <slangasek> caribou tdaitx sil2100 robru pitti doko bdmurray cyphermox infinity slangasek barry
[15:02] <slangasek> no caribou here today
[15:02] <slangasek> tdaitx: you first then
[15:02] <tdaitx> Past week
[15:02] <tdaitx> - Transition manager:
[15:02] <tdaitx> * learning about ben and britney
[15:02] <tdaitx> * checking how to fix the bad dependencies
[15:02] <tdaitx> * got python-apt to work
[15:02] <tdaitx> * Laney solved the same issue in a different approach
[15:02] <tdaitx> * discovered the glories of grep-dctrl
[15:02] <tdaitx> - FTBFS:
[15:02] <tdaitx> * investigated notmuch; was failing due dtach running with TERM="unknown"
[15:02] <tdaitx> * found ceph fix on github, tested, and debdiff'ed it (LP: #1483403)
[15:02] <tdaitx> * opened Debian #795178 for ceph as well
[15:02] <tdaitx> - OpenJDK
[15:02] <tdaitx> * tracking down LP: #1482924
[15:02] <tdaitx> * OpenJDK 6/7 supports TLS 1.1/1.2, but TLS 1.0 is the default
[15:02] <tdaitx> * patched and tested OpenJDK 7 to default to TLS 1.2
[15:02] <tdaitx> Next week
[15:02] <tdaitx> - figure out what needs to be backported from OpenJDK 8 into 7 and 6
[15:02] <tdaitx> - backport TLS 1.2 to OpenJDK 7 and 6
[15:02] <tdaitx> Waiting/On hold
[15:03] <tdaitx> - TCK validation for OpenJDK 7 (doko)
[15:03] <tdaitx> Updates
[15:03] <tdaitx> - was pointed to IS wiki pages and got ssh/sshebang to work
[15:03] <tdaitx> - applied for a new passport (will collect it tomorrow, then start the US visa process)
[15:03] <tdaitx> (done)
[15:03] <sil2100> o/
[15:03] <sil2100> - Brain overheating, need to move away to some icy place srsly
[15:03] <sil2100> - Landing team work, silo coordination, preparing landing e-mails
[15:03] <sil2100> - Find root cause of failing armhf image builds - fixed by copying timezonemap binaries to overlay
[15:03] <sil2100> - Sending out a quick guideline e-mail for landings
[15:03] <sil2100> - Creating an uploading manual langpack packages for china images
[15:03] <sil2100> - Investigate system-image importer crashing on the g++5 channel delta generation
[15:03] <sil2100>   * Custom tarballs corrupted, chased it down with ogra, tyhicks and jdstrand to an issue in apparmor
[15:03] <sil2100> - Updated CI Train bileto and sync documentation
[15:03] <sil2100> - Fixing commitlog case when cdimage build does not have a new rootfs build associated
[15:03] <sil2100> - i386 rc-proposed not-building:
[15:03] <sil2100>   * Fixed issue with thumbnailer hard-depping on a -gles version that didn't exist
[15:03] <sil2100>   * Looking into and getting a head-ache from packages ignoring the 'provides' fields
[15:03] <sil2100>   * Probably related to the new apt in vivid being more strict
[15:03] <sil2100> - Launchpad overlay translation enablement
[15:03] <sil2100>   * Copying over few missing touch packages to overlay to enable 15.04 exports
[15:03] <sil2100>   * Requesting batch copies, coordinating imports, requesting auto-exports/auto-uploads
[15:04] <sil2100> - Starting to wrap my head around britney1 and britney2 code
[15:04] <sil2100> (done)
[15:04] <robru> * Bileto:
[15:04] <robru>  - fixed various small UX papercuts (dashboard jumpiness, SSO teams checked by default, dialog text when abandoning a request)
[15:04] <robru>  - achieved 100% test coverage for python backend
[15:04] <robru>  - allow API searches against multiple fields
[15:04] <robru>  - pagination in both backend and frontend
[15:04] <robru> * Cupstream2distro:
[15:04] <robru>  - Added check to reject building/configuring superceded merges.
[15:04] <robru>  - generated new API token in wake of minor leak, and squashed the leaky code (with regression tests)
[15:04] <robru>  - get some stuff that was cowboyed in production and commit it to trunk
[15:04] <robru> ✓ done like dinner
[15:04] <pitti> - g++ 5 transitions and related FTBFS and autopkgtest fixes on 4 days
[15:04] <pitti> - eat ice cream every day to compensate for the blood and sweat from above
[15:04] <pitti> - Port sessioninstaller to Python 3 (#1440368)
[15:04] <pitti> - systemd: Fix "systemctl link" (#1480310), debug screen locking during dist-upgrade (#1473800, ongoing)
[15:04] <pitti> - britney: Fix handling of unbuilt reverse deps (#1483481), took a few iterations as it's quite hairy
[15:04] <pitti> EOT
[15:05] <doko> pitti, do you expense that?
[15:05] <pitti> is anybody not? :-)
[15:05] <pitti> (also, it's vital food now -- EOVERHEAT)
[15:05] <doko>  - transitions ...
[15:05] <doko> - was at GNU Cauldron last week(end)
[15:06] <doko> (done)
[15:06] <bdmurray> implemented the display of package pockets on the package versions table in errors
[15:06] <bdmurray> tested staging update of errors showing package pockets
[15:06] <bdmurray> submitted RT to have production/staging versions of errors/daisy updated to show package pockets
[15:06] <bdmurray> fixed an issue with the display of the pocket for packages for Ubuntu RTM 14.09
[15:06] <bdmurray> submitted RT to have production/staging versions of errors/daisy updated to fix RTM pockets
[15:06] <bdmurray> reported and fixed errors LP linking bug LP: #1483826
[15:06] <bdmurray> review of retracing failures for ways to improve
[15:06] <bdmurray> updated meta-release* for Trusty point release
[15:06] <bdmurray> wrote retracer missing executable log finding code
[15:06] <bdmurray> discussed with stub adding another dse node
[15:06] <bdmurray> uploaded apport SRU for T, V fixing LP: #1470572
[15:06] <bdmurray> ubiquity, update-manager bug triage
[15:07] <bdmurray> merged whoopsie merge proposal
[15:07] <bdmurray> review of a bug control application
[15:07] <bdmurray> sponsored some patches
[15:07] <bdmurray> created a proper fix for bash-completion bug LP: #1358154
[15:07] <bdmurray> ✔ done
[15:08] <bdmurray> cyphermox is next right?
[15:09] <cyphermox> ok
[15:09] <cyphermox>  - debugging samba ppc64el hangs on trusty (LP: #1472584)
[15:09] <cyphermox>  - NM fixes for wifi scanning and ofono - ipv6
[15:09] <cyphermox>  - looked into UTF-8 issues with console-setup in ubiquity install
[15:09] <cyphermox>  - uploaded multipath-tools SRU for trusty
[15:09] <cyphermox>  - partman-base fix for rpmb mmcs (LP: #1456443)
[15:09] <cyphermox>  - ipmitool SRU (LP: #1481780)
[15:09] <cyphermox>  - capser lightdm autologin fix
[15:09] <cyphermox> (done)
[15:10] <slangasek> sil2100: what 'provides' fields did you have being ignored?
[15:11] <slangasek> no infinity, so
[15:11] <slangasek>  * gcc-5 transition: things are going well, gcc-defaults got unblocked this week and many individual libraries have made it through
[15:11] <slangasek>   * we still have the main mass of entangled library transitions to push in (the one that requires proposed-migration to hint several hundred packages at the same time and fails because it can't figure out how to make 10,000 binary packages installable), but this is being worked on
[15:11] <slangasek>   * miscellaneous bug fixes uploaded, various leaf packages removed from wily
[15:11] <slangasek>  * travel:
[15:11] <slangasek>   * DebConf15 starts this weekend; I'll be out from tomorrow
[15:11] <slangasek> (done)
[15:11] <slangasek>   * LinuxCon, Plumbers, etc are next week so I'll be there from Wednesday afternoon to Friday
[15:12] <slangasek> barry:
[15:12] <barry> looked at buildbot build failures (debian bug #794300).  need upstream help; bug filed.
[15:12] <barry> worked on django for py35.  will require an update to 1.8.3 from experimental, which is undergoing its own stalled transition.  created ppa pythoneers/py35django18 to track this, which is stacked on py35asdefault.  django itself builds but several of the other django related/dependent packages in the archive ftbfs
[15:12] <barry> long and difficult work on pytables (LP: #1478146)  3.2.1-0ubuntu1 builds now but apparently segfaults its dep-8 tests.  still investigating that.
[15:12] <barry> --done--
[15:12] <sil2100> slangasek: well, that's what it seems it's doing... nothing changed in deps and suddenly the -gles versions of UITK are not installed when apps are depending on UITK, even though the -gles ones have 'Provides' of the UITK ones in packaging
[15:13] <slangasek> sil2100: can you give me an exact package name / provides?
[15:13] <sil2100> slangasek: so suddenly everything went to hell and things that apt could install correctly suddenly causes a 'conflicts' and ubuntu-touch became uninstallable
[15:14] <sil2100> qtdeclarative5-ubuntu-ui-toolkit-plugin-gles prodives and conflicts with qtdeclarative5-ubuntu-ui-toolkit-plugin
[15:15] <sil2100> A whole batch of packages dep on qtdeclarative5-ubuntu-ui-toolkit-plugin, and on the i386 versions -gles should be installed instead - in the past it just worked if the build system pulled in qtdeclarative5-ubuntu-ui-toolkit-plugin-gles
[15:15] <slangasek> so one thing you may be seeing is the effect of the random ordering of things through seeds; if something $somewhere changed that caused dependencies to be traversed in a different order, and you were previously forcing qtdeclarative5-ubuntu-ui-toolkit-plugin-gles but now qtdeclarative5-ubuntu-ui-toolkit-plugin is getting seen first
[15:15] <slangasek> sil2100: where does the failure show up?
[15:16] <sil2100> slangasek: on the i386 cdimage builds but also locally when I try it in a chroot... there was no seeds change, but there was a new apt uploaded, so maybe now it traverses the deps differently
[15:16] <sil2100> Anyway, still looking into that, want to see what actually changed in apt
[15:17] <slangasek> ok
[15:17] <slangasek> any other questions re: status?
[15:17] <sil2100> Might be something else... but I checked all changes in our touch packages, experimented a lot with deps in the chroot and nothing, so I started looking into the apt direction
[15:17] <sil2100> Timeline-wise it seems possible
[15:18] <sil2100> slangasek: when infinity will be back? :)
[15:18] <slangasek> good question
[15:18]  * sil2100 wants to molest him a bit about britney pop-star stuff
[15:19]  * bdmurray shakes head
[15:21] <slangasek> sil2100: oh, and you know there's a follow-up SRU to apt that's currently in vivid-proposed which hasn't been released yet?  there was a regression that was critical-path for 14.04.3, but considered less relevant to vivid; perhaps there's *some* interaction there
[15:21] <slangasek> sil2100: "oops I did it again" is not an endorsed engineering practice in Ubuntu
[15:22] <barry> despite it being the best pop song of the last 50 years
[15:22] <bdmurray> if it were it would be "toxic"
[15:22]  * pitti takes notes to not break britney a third time :)
[15:23] <slangasek> ok
[15:23] <slangasek> anything else? :)
[15:23] <bdmurray> sil2100: What's up with https://bugs.launchpad.net/ubuntu/+source/livecd-rootfs/+bug/1471903?
[15:25] <sil2100> bdmurray: was waiting for infinity to take a look at it, slangasek agrees with this proposition but Adam seemed to have some concerns
[15:26] <sil2100> slangasek: ;)
[15:26] <sil2100> slangasek: I'll look into that apt upload, I saw it but didn't know yet what it was about
[15:26] <bdmurray> sil2100: okay, thanks
[15:26] <slangasek> sil2100: have you asked infinity to look at it?
[15:27] <sil2100> slangasek: sent an e-mail yesterday + some -proposed stuff, but it seems he wasn't around yesterday as well
[15:28] <slangasek> sil2100: he was around, but that's not the same as replying to an email
[15:28] <slangasek> sil2100: by -proposed do you mean you uploaded an SRU?
[15:29] <sil2100> No, -proposed migration questions (regarding britney and not only)
[15:29] <slangasek> ok
[15:30] <slangasek> anything else (x2)?
[15:30] <slangasek> everyone still has enough to work on at http://people.canonical.com/~ubuntu-archive/transitions/ ?
[15:31] <cyphermox> in anything else,
[15:31] <barry> oh yes :)
[15:31] <cyphermox> apparently today's image is broken in not-EFI
[15:31] <cyphermox> doesn't boot.
[15:31] <slangasek> I would like to see the main transition un-stick by the end of the week; if you aren't sure which pieces to help with for that, you can check with Laney or doko on #ubuntu-release
[15:32] <slangasek> cyphermox: yes, I've just seen a bug report from davmor2 come through about this.  have you reproduced the failure?
[15:32] <slangasek> (and who uploaded what?)
[15:32] <cyphermox> working on it
[15:32] <cyphermox> I've reproduced the non-failure on EFI
[15:33] <cyphermox> and now I've reproduced the failure
[15:33] <slangasek> there is a new syslinux in wily as of 20 hours ago
[15:33] <slangasek> possibly the first build with gcc5
[15:33] <cyphermox> that's the most likely culprit, in fact
[15:34] <slangasek> definitely the first build with gcc5
[15:34] <davmor2> cyphermox: just added a comment to my bug yesterday image boots fine so it is only todays build that is broken
[15:34] <barry> slangasek: can you rename this one: http://people.canonical.com/~ubuntu-archive/transitions/html/python3.3-5.html
[15:34] <cyphermox> I'll check that out, I'll roll my own CD to see
[15:34] <slangasek> cyphermox: want to build-depend on gcc-4.9 as a short-term workaround while debugging further?  since this is C not C++ there's no compatibility problem with rolling back temporarily
[15:35] <slangasek> cyphermox: and handing you bug #1484571
[15:35] <cyphermox> slangasek: yeah, that could be a good temporary solution so people don't get unbootable systems for the time being
[15:35] <slangasek> barry: done
[15:36] <slangasek> cyphermox: it's not "people get unbootable systems", syslinux is only used on the CDs - but we need the CDs to be bootable to make sure the g++ transition isn't leaving the OS on the floor.  So please revert syslinux to gcc-4.9 asap and we can continue debugging after
[15:36] <barry> slangasek: thanks.  does it take a while to show up on http://people.canonical.com/~ubuntu-archive/transitions/ ?
[15:37] <slangasek> barry: ~1h
[15:37] <barry> slangasek: ack
[15:37] <slangasek> (possibly less now that I've been trying to prune the done transitions from the list)
[15:37] <cyphermox> slangasek: I think this would also likely break people trying to use usb-creator to write iso images to USB, even if the iso was good.
[15:38] <slangasek> cyphermox: sure; it won't break bootability of an existing system, at least, but is still critical :)
[15:38] <cyphermox> of course, fixing that now :)
[15:39] <slangasek> thanks
[15:39] <slangasek> anything else? (x3)
[15:40] <slangasek> #endmeeting
[15:40] <meetingology> Meeting ended Thu Aug 13 15:40:54 2015 UTC.
[15:40] <meetingology> Minutes:        http://ubottu.com/meetingology/logs/ubuntu-meeting/2015/ubuntu-meeting.2015-08-13-15.01.moin.txt
[15:40] <slangasek> thanks!
[15:41] <barry> thanks!
[15:41] <pitti> thanks everyone!
[15:41] <pitti> see some of you next week at debconf!
[15:43] <barry> pitti: have fun!
[17:28] <Quatravius> Hi