[15:00]  * slangasek waves
[15:00]  * infinity grunts.
[15:01] <slangasek> #startmeeting
[15:01] <meetingology> Meeting started Thu Sep 25 15:01:12 2014 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
[15:01] <slangasek> [TOPIC] lightning round
[15:01] <slangasek> $ echo $(shuf -e barry doko stgraber jodh bdmurray slangasek cjwatson caribou infinity mvo bhuey sil2100 robru)
[15:01] <slangasek> robru infinity mvo cjwatson stgraber doko bdmurray bhuey barry sil2100 slangasek jodh caribou
[15:01] <slangasek> robru: you're first!
[15:02] <slangasek> robru: too slow!
[15:02] <slangasek> infinity:
[15:02] <infinity> Typing...
[15:02] <robru> ah sorry, not ready anyway
[15:02] <slangasek> :)
[15:03] <infinity> - Helped with some evil partner package reviews
[15:03] <infinity> - Meetings (several!) with people about $customer doing silly things with firmware
[15:03] <infinity> - Working on updating PPC packages
[15:03] <infinity> - Working on Utopic Beta 2 release
[15:03] <infinity> [...]
[15:03] <infinity> POssibly also stuff, and/or things.
[15:04] <slangasek> infinity: '(done)' ?
[15:04]  * mvo was wondering the same
[15:05] <infinity> slangasek: Oh, sure.
[15:05] <mvo> apt:
[15:05] <mvo> - Debug regression (Bug#762160)
[15:05] <mvo> - Release  cve-2014-6273 (http.cc buffer overflow)
[15:05] <mvo> - Work on droping privs for the apt download methods
[15:05] <mvo> - Work on feature/acq-trans for more robustness in acquire
[15:05] <mvo> click:
[15:05] <infinity> slangasek: That was the [...]
[15:05] <mvo> - Debug pycurl/gnutls issue with search.apps.ubuntu.com and send RT
[15:05] <mvo>   ticket, blocks using pycurl for searches
[15:05] <mvo> - lp:~mvo/click/click-check-libs (find non sdk libs and warn about them)
[15:05] <mvo> - lp:~mvo/click/lp1245826-multiple-hookfiles (support multiple files
[15:05] <mvo>   for a given hook)
[15:05] <mvo> - lp:~mvo/click/acquire (more work on acquire support)
[15:05] <mvo> - lp:~mvo/click/repository (ported to licurl, stuck with gnutls issue)
[15:05] <mvo> - lp:~mvo/click/sso (ported to libcurl, stuck with gnutls issue)
[15:05] <slangasek> infinity: heh :)
[15:05] <mvo> - upload click-bin-path hook (to add stuff to PATH) to utopic
[15:05] <mvo> python-apt:
[15:05] <mvo> - Review/merge Colins ubuntu-rtm fix (thanks!)
[15:05] <mvo> software-properties:
[15:05] <mvo> - Review/merge Colins patch, upload new version
[15:05] <mvo> system-image:
[15:05] <mvo> - Discussion about leaner download backend
[15:05] <mvo> - lp:~mvo/ubuntu-system-image/pycurl
[15:05] <mvo> (done)
[15:05] <cjwatson> GRUB update in unstable, adding x32 and ppc64el support (ppc64el support replaced our previous cross-compiler hack).
[15:05] <cjwatson> man-db release to add high-precision timestamp support, speed up mandb / man -K, and stop using debconf in the trigger.  Trawled through Ubuntu bugs about the last part, clearing out >140, and prepared SRUs (bug 1372673).
[15:05] <cjwatson> Fixed livecd-rootfs to tolerate dpkg-dev no longer being on the image.
[15:05] <cjwatson> netcfg SRU for bug 1350302.
[15:05] <cjwatson> Fixed add-apt-repository on ubuntu-rtm.
[15:06] <sil2100> Eeek!
[15:06] <cjwatson> Fixed UbuntuKylin image builds to contain the proper Ubiquity slideshow (bug 1372731).
[15:06] <cjwatson> Various ad-hoc copies to reduce the number of uninstallables reported by proposed-migration for ubuntu-rtm/14.09.
[15:06] <cjwatson> Tracked down a click autopkgtest regression (fortunately never hit -proposed).
[15:06] <cjwatson> Devirtualised/otherwise-configured 10 new silos in each of ubuntu and ubuntu-rtm for the CI Train folks.
[15:06] <cjwatson> Proposed branch for bug 1364327.  Took ages because I first had to get my local autopkgtest/qemu environment sorted out, which involved upgrading to utopic ...
[15:06] <cjwatson> Working on an OpenSSL backport stack for ppc64el.  In patch hell; send ice cream.
[15:06] <cjwatson> ..
[15:06] <stgraber>  - A lot more internal work and meetings..
[15:06] <stgraber>  - Figured out all the details for upcoming trips (vacation, LinuxCon,
[15:06] <stgraber>    Plumbers, Client sprint, Server sprint).
[15:06] <stgraber>  - Released LXC 1.0.6 with 85 bugfixes by 24 individual contributors.
[15:06] <stgraber>  - Went through the LXC patch backlog, reviewed and committed everything.
[15:06] <stgraber>  - Worked on fixing the LXC init script situation in preparation for 1.1~alpha2
[15:06] <stgraber>    (to be released tomorrow ideally).
[15:06] <stgraber>  - Implemented a fix for uid/gid consistency on system images (touch only at
[15:06] <stgraber>    the moment).
[15:06] <stgraber> (DONE)
[15:06]  * sil2100 is here if anything
[15:06] <sil2100> Did I miss my turn?
[15:07] <cjwatson> sil2100: nope, you're after barry
[15:07] <infinity> cjwatson: Oh, nice, so we're going to drop the cross-hack in Ubuntu too, I assume?
[15:07] <sil2100> Thanks
[15:07] <bdmurray> updated daisy (r529) handle AssertionError raised by apport on invalid keys in a report fixes OOPS-ce96293a10a1987c516e4ed968fa715e
[15:07] <bdmurray> submitted RT to get retracer log files from staging retracers
[15:07] <bdmurray> submitted RT 75272 regarding nodetool access
[15:07] <bdmurray> had daisy retracers really updated to r528
[15:07] <bdmurray> pushed merge proposal with daisy-retracer charm change to create version_info.py
[15:07] <bdmurray> pushed daisy r531 which logs the version of daisy to the retracer log files
[15:07] <cjwatson> infinity: already have.  but it doesn't quite work apparently, needs -mno-vsx -mno-altivec or asm hacks, pfsmorigo is on it
[15:07] <bdmurray> updated and tested daisy charm change to generate version_info.py
[15:07] <bdmurray> worked with thedac to remove coredumps from swift as there are more items in swift than the retracing queue
[15:07] <bdmurray> research into if duplicate OOPSes are an issue in the Error Tracker and how to fix it
[15:07] <bdmurray> created a graph of .new missing versus foreign origin packages
[15:07] <bdmurray> submitted apport bug 1372612 regarding Ubuntu RTM and origin issue
[15:07] <bdmurray> reported and testing of apport bug 1373154 regarding ExecutableTimestamp and MarkForUpload being True
[15:07] <bdmurray> SRU verification of bug LP: #1309447
[15:07] <bdmurray> ✔ done
[15:07] <cjwatson> infinity: (so, cough, it's possible the final beta may not be bootable on ppc64el ...)
[15:08] <infinity> cjwatson: Oh.  Fun.
[15:08] <infinity> cjwatson: Good thing I hadn't tested yet.
[15:10] <slangasek> bhuey: hi, your turn
[15:10] <bhuey> ok
[15:10] <bhuey> -read both ubuntu-policy.pdf and packaging tutorial documents
[15:10] <bhuey> -got X11 working in all containers for TCK testing
[15:10] <bhuey> -successfully ran TCK inside containers for precise/trusty/utopic
[15:10] <bhuey> -general clean up of container configurations TCK instances
[15:10] <bhuey> -begin general creation of container creation scripts so that the entire test environment csn be replicated
[15:10] <bhuey> -began looking at bugs that needs to be fixed by Oct 20th
[15:10] <bhuey> (done)
[15:10] <barry> debuntu: LP: #1371291, debian bug #762142, syncpackage python-concurrent.futures (to fix ftbfs), LP: #1367907, LP: #1290847 (both for python34 sru, awaiting doko's feedback).  reviewed py3 swiftclient packaging and uploaded websocket-client 0.18.0 for internal users.
[15:10] <barry> phone: system-image 2.4-0ubuntu1 slogging through ci-train for utopic and rtm.  internal discussions.  testing/review/excitement over mvo's pycurl branch.  LP: #1369717.  LP: #1370586.
[15:10] <barry> other: git'n it
[15:11] <barry> -- done --
[15:11] <sil2100> o/
[15:11] <slangasek> so now that bhuey has read policy, the rest of our meeting today can be to quiz him on it ;)
[15:11] <sil2100> - Filling in the standing FFe for utopic Ubuntu Touch packages (LP: #1371635)
[15:11] <sil2100> - Refactoring some of the commit-log syncing scripts
[15:11] <sil2100> - Landing team work, preparing landing e-mails
[15:11] <bhuey> slangasek: shit
[15:11] <sil2100> - CI Train maintenance and feature:
[15:11] <sil2100>   * Testing binary copy sync functionality
[15:11] <sil2100>   * Improve error verbosity in some cases for sync rebuilds
[15:11] <sil2100>   * Cleaning up left-over jenkins jobs
[15:11] <sil2100>   * Refactoring parts of the build job to enable better testing
[15:11] <sil2100>   * Add basic unit tests for the sync functionality
[15:11] <sil2100>   * Make the ~rtm addition to the version number optional
[15:11] <sil2100>   * Fixing a new small bug that caused the spreadsheet to malfunction
[15:11] <sil2100> - Reviewing and testing bugfixing merge for appmenu-qt5
[15:11] <sil2100> - Fixing/unbricking testing device
[15:11] <sil2100> - Writing announcement e-mails
[15:12] <sil2100> - Making sure the documentation of RTM landings is up-to-date
[15:12] <sil2100> - Debugging issues with commit-log generation scripts
[15:12] <sil2100> - Discussions regarding getting rid of the spreadsheet (finally!)
[15:12] <sil2100> (done)
[15:12] <robru> ok, i'm ready now, can i go?
[15:12] <slangasek> robru: sure
[15:12] <robru> Train:
[15:12] <robru> - Adapted 'citrain' (silo installation script) to work on RTM
[15:12] <robru> - Updated citrain spreadsheet, jenkins backend, and dashboard to recognize the 10 new silos
[15:12] <robru> - Began test-drive reimplementation of cupstream2distro/silomanager.py, with the intention to fix the utter disgusting lack of encapsulation, such that other bits of citrain no longer need to have knowledge of the custom json format used to maintain each silo's state (this will make it easier/possible to back citrain onto the airline ticket system when the
[15:12] <robru> time comes)
[15:12] <robru> - Fixed the deploy job to correctly assign pre-prod silos, making it possible to test things outside of production
[15:12] <robru> - Stopped the prepare job from being able to delete packages from the PPA, which it was often doing erroneously and deleting packages that had been through hours of scarce QA approval
[15:12] <robru> - Implemented a generic signal / unhandled exception handler, which is capable of a) logging and b) saving the signal/exception message to the silo status, such that the dashboard can now report when a job exits uncleanly, so we no longer have the situation where the dashboard reports 'Building...' even if something has actually gone horribly wrong
[15:12] <robru> - Cleaned up the use of subprocess module, so that the testsuite no longer spews random crap on STDOUT, making it a lot easier to see a clean, successful test run
[15:12] <robru> NFSS:
[15:12] <robru> - Added veebers' new js module to lp:ubuntu-assets and got it approved by oatman
[15:12] <robru> (done)
[15:12] <slangasek> sil2100: FFe> did everything work out wrt touch packages in the auto-accept list?
[15:13] <infinity> bhuey: A binary package leaves my dput travelling three versions per epoch, when is it appropriate to use Conflicts versus Breaks, and does the answer change when drunk?
[15:13] <sil2100> slangasek: I think it did, since we didn't see any further problems
[15:13] <barry> infinity: binNMU
[15:14] <cjwatson> infinity: consider a spherical manual page
[15:14] <infinity> cjwatson: I actually ell oh elled.
[15:14] <infinity> cjwatson: And, for some reason, I read that in Bill Bailey's voice.
[15:16] <slangasek>  * business meeting up in Seattle yesterday, the jet lag is horrible
[15:16] <slangasek>  * review of a number of impending partner packages
[15:16] <slangasek>  * working to move a number click apps out of the rootfs into a default "custom" tarball (bug #1367332)
[15:16] <slangasek>  * progress on a console-setup merge with Colin's help; the package may be uploadable this year
[15:16] <slangasek>  * created new silos at LT's request
[15:16] <slangasek>  * LT/CI meetings to discuss next steps on CI Airline (e.g., ditching the spreadsheet)
[15:16] <slangasek> (done)
[15:16] <bhuey> infinity: yes, only when also taking opiates with alcohol
[15:16] <jodh> * upstart:
[15:16] <jodh>   - investigating bug 1371651.
[15:16] <jodh>   - A respectful prod to all upstart devs to take a look at the
[15:16] <jodh>   following if they get a chance :-):
[15:16] <jodh>   https://code.launchpad.net/~jamesodhunt/upstart/bug-1360208/+merge/234869
[15:16] <jodh> * system-image
[15:16] <jodh>   - Good progress - root is now r/o with a writeable overlay.
[15:16] <slangasek> if a ~ falls in the middle of your version, does it make a sound
[15:16] <jodh>   - More tweaks to lp:~ubuntu-foundations-team/+junk/create-ubuntu-core-image.
[15:17] <jodh>   - Split lp:~ubuntu-foundations-team/+junk/system-image-rootfs into 2:
[15:17] <jodh>     - lp:~ubuntu-foundations-team/+junk/ubuntu-core-config [1]
[15:17] <jodh>     - lp:~ubuntu-foundations-team/+junk/initramfs-tools-ubuntu-core [2]
[15:17] <jodh>   - Currently testing recent changes to [2].
[15:17] <jodh> * misc:
[15:17] <jodh>   - procenv is now running for lucid as well as all newer supported releases:
[15:17] <jodh>     https://code.launchpad.net/~jamesodhunt/+recipe/procenv-daily-ubuntu-lucid
[15:17] <jodh> ⭈
[15:18] <slangasek> jodh: bug #1371651> do we know what /changed/ to cause this problem?
[15:18] <slangasek> (it was the kernel, wasn't it!)
[15:19] <jodh> slangasek: no - I've asked jibel to check packages differences, but I don't think it's upstart.
[15:19] <slangasek> I agree
[15:19] <infinity> slangasek: No one seems entirely sure who to blame just yet, though the kernel seems a likely candidate.
[15:19] <slangasek> plymouth has been uploaded recently (twice in the past 8 days), but those changes should also not have had an effect
[15:20] <infinity> (Or, the whole process is just amazingly fragile, and the fault lies in all the players together, plus it was Wednesday)
[15:20] <jodh> well as I noted on the bug, we do see plymouth crashing in the logs but that might be due to fb issues which are also causing the failure of upstart to emit the plymouth-ready event.
[15:20] <slangasek> right, in which case --> kernel
[15:20] <jodh> slangasek: right, jibel is trying to track down apw I believe.
[15:20] <slangasek> has anyone asked apw about it?
[15:21] <slangasek> ok :)
[15:21] <infinity> The part where "it works with systemd" (while also obviously not being directly an upstart bug) kinda points to the fragility argument.
[15:21] <jibel> slangasek, apw is working on it
[15:22] <jodh> infinity: but has anyone tried uselessd yet? :)
[15:22] <slangasek> infinity: on systemd we don't have any flicker-free boot integration
[15:23] <apw> the state the machine ends up in when this is reproduced seems to indicate everything is in order, and running service lightdm start sorts it out (i believe)
[15:23] <apw> w
[15:23] <apw> which leans towards a userspace race in my eyes, but i am waiting on some installs
[15:24] <slangasek> apw: if the plymouth-ready signal isn't being emitted because of a kernel-induced plymouth crash, that's not really a matter of a race
[15:24] <slangasek> anyway, seems like this is in hand so we can discuss it on better channels post-meeting
[15:24] <slangasek> [TOPIC] AOB
[15:24] <slangasek> anything else today?
[15:24] <slangasek> I've been failing to tag people to do the weekly talk-about-what-you're-working-on thing
[15:24] <slangasek> does somebody want to volunteer to do one next week?
[15:25] <jodh> apw: fwics, if you remove "quiet splash", the problem is not recreatable but would value other trying it (I've asked a few times on the bug :-)
[15:25] <bdmurray> cjwatson: regarding bug 1372673, it looks like you found most of the duplicates. Did you check other packages though?
[15:26] <cjwatson> bdmurray: nope, I don't have a mechanical search for it and some of the problems are subtly different
[15:26] <cjwatson> so I had only finite time
[15:26] <cjwatson> bdmurray: I'm certain I found all the duplicates on man-db - the others that look similar are actually cropping up in postinst configure instead which is different
[15:27] <slangasek> no volunteers? :)
[15:27] <cjwatson> you have to hunt for debconf failures immediately following "Processing triggers for man-db" or translations thereof
[15:27] <slangasek> you don't have to prepare anything... (in fact it's better if you don't!...) you just have to talk about what you're working on
[15:28] <robru> slangasek: i volunteer for the week of the washington sprint so this meeting won't be so painfully early for me ;-)
[15:28] <bdmurray> cjwatson: okay, I may have a look around and then check with you
[15:28] <cjwatson> ok sure
[15:28] <slangasek> robru: invest in coffee
[15:28] <robru> slangasek: just say NO to drugs ;-)
[15:28] <cjwatson> our coffee machine wasn't working this morning /o\
[15:29] <slangasek> robru: you can balance it out with bhuey's opiate-laced alcohol
[15:29]  * ogra_ listens up 
[15:29] <bdmurray> I've moved away from machines to an aeropress now.
[15:29] <cjwatson> slangasek: guess I can if everyone else is running away; it's been a while for me
[15:29] <cjwatson> (next week)
[15:29] <slangasek> cjwatson: thanks :)
[15:30] <slangasek> #action cjwatson to do a what-I'm-working-on presentation next week
[15:30] <meetingology> ACTION: cjwatson to do a what-I'm-working-on presentation next week
[15:30] <slangasek> bdmurray: this is only impressive if you also have a button to deliver the prepared cup of coffee to your desk via pneumatic tube
[15:31] <cjwatson> slangasek: have you seen Wallace and Gromit or is that too much of a British thing?
[15:31] <slangasek> cjwatson: yes, quite familiar with it :)
[15:31] <slangasek> (the last movie even had a theatrical release in the US!)
[15:31] <slangasek> er, I think
[15:32] <slangasek> oh right, Chicken Run was technically not W&G, just Aardman
[15:32] <bdmurray> I agree with your thought
[15:32] <slangasek> anyway, speaking of which
[15:32]  * slangasek slinks away from the desk to get more coffee
[15:32] <slangasek> #endmeeting
[15:32] <meetingology> Meeting ended Thu Sep 25 15:32:56 2014 UTC.
[15:32] <meetingology> Minutes:        http://ubottu.com/meetingology/logs/ubuntu-meeting/2014/ubuntu-meeting.2014-09-25-15.01.moin.txt
[15:32] <slangasek> thanks, all :)
[15:32] <barry> \o/
[15:33] <mvo> thanks
[15:34] <stgraber> thanks!