[01:07] <tedg> robru, By chance are you still around?
[01:31] <robru> tedg, hey, just got back. what's up?
[01:33] <kenvandine> robru, he wants a silo for a ubuntu-app-launch fix
[01:34] <robru> tedg, ok you got silo 2
[01:34] <kenvandine> i just hassled tedg on IM :)
[01:35] <tedg> Heh
[01:35] <robru> kenvandine, hey hey. how was malta? such a shame we couldn't connect
[01:35] <kenvandine> good, yeah missed you guys
[01:35] <tedg> robru, Thanks!
[01:36] <robru> tedg, you are welcome!
[01:36] <kenvandine> i did get to hang out with seb, larsu and dsert though
[01:36] <kenvandine> and of course tedg :)
[01:36] <robru> kenvandine, i wish I could go to these things for both weeks
[01:36] <kenvandine> tedg, that was some damn good carbonara!
[01:36] <tedg> robru, You say that the first couple times ;-)
[01:36] <tedg> kenvandine, It was.
[01:37] <kenvandine> ha... i do not miss the 2 week trips, not at all :)
[01:37] <kenvandine> tedg, thanks for the fix!
[01:37] <tedg> kenvandine, The L'Artist restaurant was the best there though.
[01:37] <kenvandine> i didn't get to try that place
[01:37]  * kenvandine goes back to writing API docs for bacon2d :)
[01:38] <kenvandine> which is not nearly as much fun as game hacking :)
[01:38] <tedg> Oh, ogra uploaded a fix...
[01:39] <kenvandine> ah... distro patch? :)
[01:40] <tedg> Yeah...
[01:41] <kenvandine> when?
[01:41] <kenvandine> i don't see a change mail for it
[01:42] <tedg> kenvandine, https://launchpad.net/ubuntu/+source/ubuntu-app-launch
[01:42] <tedg> kenvandine, Not for this, for a packaging issue.
[01:42] <kenvandine> ah
[01:44] <robru> tedg, oh right, you have to sync distro back to trunk there.
[01:44] <tedg> On it.
[01:45] <robru> cool
[01:46] <tedg> Who can review this? https://code.launchpad.net/~ted/qa-regression-testing/ubuntu-app-launch/+merge/222115
[01:47] <robru> tedg,  these people? ;-) http://bazaar.launchpad.net/~ubuntu-bugcontrol/qa-regression-testing/master/changes
[01:47] <tedg> Ewww, security people :-)
[01:59] <tedg> Hmm, clock isn't starting for me now. "(process:3727): GLib-GIO-ERROR **: No GSettings schemas are installed on the system"
[02:10] <imgbot> [02:12] <tedg> Uhg, missed 66 I guess.
[02:14] <robru> yeap
[02:28] <robru> camako, hey
[02:29] <camako> robru, hey so is there anything I need to do at this point?
[02:32] <robru> camako, haha we're talking in two channels, bah ;-)
[02:33] <robru> tedg, http://ci.ubuntu.com/smokeng/utopic/touch/mako/65:20140604.1:20140530/8401/music_app/ is this related to UAL?
[02:36] <tedg> robru, I don't think so. Noticing that clock and terminal are also broken.
[02:43] <tedg> robru, Okay, nope, I think at least terminal is mine.
[02:43] <robru> tedg, ah
[02:46] <tedg> robru, Rebuilding that silo with the fix.
[02:46] <robru> tedg, cool
[03:06] <plars> ugh
[03:06] <plars> robru: something's wrong still
[03:06] <plars> robru: I'm testing 65 at home with *just* music_app and it's working
[03:06] <plars> unfortunately... I really hate that stupid song it plays
[03:06] <robru> plars, what do you mean by "just music_app"?
[03:07] <robru> plars, I think you can mute the phone while the test runs ;-)
[03:07] <plars> robru: not running all the tests, just that one since it seems to fail everywhere
[03:07] <robru> plars, oh, so you're suggesting that this is an infrastructure problem?
[03:07] <plars> robru: that would involve getting out of the chair :)
[03:07] <plars> robru: no
[03:07] <plars> robru: but it could be like the other problems we've seen over the past couple of days where a package is messing things up
[03:09] <robru> plars, right I see.
[03:09] <robru> plars, anything I can do to help with that?
[03:13] <plars> robru: going to try a few things, but I'm exhausted right now
[03:13] <plars> I'll stay up as long as I can and try to sort it out
[03:14] <robru> plars, alright, let me know if there's a specific test you want me to run on my mako or whatever, i've got a couple spare cycles.
[03:14] <plars> robru: the changes list for 65 was enormous
[03:15] <robru> plars, yeah, sorry abou that. are you bisecting it?
[03:15] <plars> robru: no way
[03:15] <plars> robru: I'd be up for the next few days
[03:15] <robru> plars, I blame UAL for everything. revert that one first ;-)
[03:15] <plars> :)
[03:22] <tedg> I think a couple of the apps issues were based on not having all the env vars set.
[03:23] <tedg> They were falling back to running as legacy apps, which doesn't set all the confinement variables.
[03:33] <tedg> robru, I didn't change the tracepoints yet, was letting the application startup guys get off the ground before throwing them a twist.
[03:33] <robru> tedg, ah sorry, didn't realize those were *all* tracepoints, I was just concerned by the length of that list
[03:35] <robru> tedg, thanks for clarifying
[03:35] <tedg> robru, almost all, most are inert, but going to fix them anyway. Stop the next person from getting confused.
[03:35] <plars> tedg: does https://jenkins.qa.ubuntu.com/job/utopic-touch-flo-smoke-daily/144/artifact/clientlogs/calendar_app/_usr_lib_python3_dist-packages_autopilot_run.py.32011.crash/*view*/ mean anything to you?
[03:35] <plars> tedg: specifically: DuplicateSignature: upstart-app-launch-invalid-appid
[03:35] <plars> the apps that are crashing seem to all have this
[03:37] <tedg> plars, It's basically a report saying that it can't find the app.
[03:39] <plars> tedg: and http://paste.ubuntu.com/7592099/
[03:40] <plars> tedg: is that the issue you were describing where not all the environment variables are set?
[03:41] <tedg> plars, Hmm, not sure on that one.
[03:41] <tedg> plars, That may just be from the app not starting though.
[03:41] <plars> robru: hmm, I think I'm able to reproduce this now at home
[03:41] <tedg> plars, It seems like that error is in autopilot, so my guess would be that it can't find widgets because there isn't a window.
[03:42] <robru> plars, oh really?
[03:42] <plars> wait, maybe not, too early
[03:42] <veebers> tedg, plars: I overheard (incompletely) that there are some issues with upstart, autopilot and applications. I may be of service if it has something to do with autopilot
[03:42] <plars> I thought it was farther along
[03:43] <tedg> veebers, Not sure it's an autopilot issue specifically, but here's the trace: http://paste.ubuntu.com/7592099/
[03:43] <plars> nope, it's still working here
[03:43] <tedg> To really break things you're going to have to wipe.
[03:43] <tedg> The problem is that the cache dir in ~/.cache/upstart-app-launch will stick around.
[03:44] <veebers> tedg: hmm, all I can tell from that is that the app failed to launch :-) Is this issue (issues?) happening outside of the autopilot tests?
[03:44] <veebers> i.e. is autopilot launching (or attempting to) the application correctly
[03:44] <tedg> veebers, Yes, it is. So I don't think it's an AP issue.
[03:45] <veebers> tedg: ah ok, I'll butt-out then :-) let me know if I can help out if there are AP issues
[03:56] <plars> robotfuel: interesting, I've seen that off and on with a few devices lately
[03:56] <plars> robotfuel: not on the one that was current when you said that (64 I think?) but with 65, and with some earlier ones too
[04:01] <robotfuel> plars: are there any workarounds that you know of?
[04:01] <plars> robotfuel: holding the volume down button while turning it on
[04:01] <plars> robotfuel: I was getting suspicious that we just had some devices starting to go bad
[04:04] <robotfuel> plars: interesting I had to do it 2 times, the phone went to recovery from the ubuntu-device-flash script after holding the volume button down, but it still wanted it in bootloader
[04:06] <plars> robotfuel: oh, well if you specify --bootloader to ubuntu-device-flash it want's it in fastboot mode
[04:06] <plars> robotfuel: but it won't try to reboot it to fastboot, you have to do that part
[04:10] <tedg> K, so silo 2 seems to fix the issue of the click packages falling back to legacy.
[04:10] <tedg> (and a related bug stemming from that)
[04:10] <tedg> Clock is still broken for me though.
[04:12] <tedg> Seems to be getting a GSettings error.
[04:15] <tedg> robru, plars, so I think that silo2 fixes a bunch of the errors, not sure about all though.
[04:15] <plars> well, I guess we'll see how 66 looks and see if silo2 provides any relief
[04:15] <plars> I'm going to run the full set of tests from home with some hope of reproducing whatever's going on
[04:16] <plars> in isolation, I can't seem to reproduce, but it could just be something that doesn't happen every time
[04:18] <tedg> Makes sense. When would we have results for 67?
[04:18] <tedg> 1000 UTC?
[04:21] <tedg> I'm going to head to bed, if someone could publish 2 that'd be awesome. It definitely makes things better.
[04:21] <plars> tedg: after it builds
[04:21] <plars> tedg: 66 is the latest image right now
[04:21] <plars> and it's looking bad for unity8 at the moment
[04:22] <plars> I'm watching the carnage right now but I'm going to need sleep shortly... hopefully I won't have nightmares about this all night
[04:23] <tedg> Heh, I hope not!
[04:23] <tedg> 'night folks
[04:23] <camako> robru, now status says "Migration, one pkg at least is not available at the dest.....". Do I need to intervene?
[04:26] <plars> robru: well, I'm going to go try to get some sleep. expect unity8 failures in this image 66 at the very least. Not sure about the failures from 65 yet since I didn't manage to reproduce but I'll have a full run going tonight
[04:26] <plars> robru: I'll be mostly off tomorrow, but will try to get back to my computer as soon as I can (probably mid-late afternoon)
[04:47] <camako> Can anyone from CI team clarify what this means : "Can't merge: One package at least is not available at the destination......"
[05:57] <Mirv> camako: did you get that sorted out? it means some package hasn't migrated yet to release pocket.
[07:47]  * sil2100 sighs
[07:53] <sil2100> So, the tests seem to be somewhat broken again - not sure if that's showing off a real bug or just some new problems
[07:56] <sil2100> I wonder if the big landing of the UAL has something to do with it, as the apps fail on application start
[07:56] <ogra_> sil2100, the two apps are UAL
[07:56] <ogra_> missing the package rename in autopilot
[07:56] <ogra_> no idea what uniy8 is though
[07:56] <sil2100> ogra_: you think so? As I see other apps working fine without the rename
[07:56] <ogra_> Attempting to launch application 'com.ubuntu.music_music_1.3.476' with URIs '' via upstart-app-launch
[07:57] <sil2100> I see message "with URIs '' via upstart-app-launch" in others that pass
[07:57] <ogra_> yes, i think so :)
[07:57] <ogra_> oh ?
[07:57] <ogra_> hmm
[07:57] <sil2100> ogra_: if you open the console log and grep for 'via upstart' you'll see that it's used multiple times
[07:58] <ogra_> ** (process:3263): WARNING **: Unable to find keyfile for application 'com.ubuntu.music_music_1.3.476'
[07:58] <sil2100> So I thought that it's just a leftover name
[07:58] <ogra_> could be indeed
[07:58] <ogra_> _usr_lib_python3_dist-packages_autopilot_run.py.32011.crash
[07:58] <ogra_> both crashed apps have that one
[07:59] <sil2100> hmmm
[07:59] <sil2100> Maybe screen unlock didn't work?
[07:59] <sil2100> Oh, crash
[07:59] <ogra_> yeah
[08:00] <sil2100> First I thought about screen unlock failures, as it makes sense in the failure rate... i.e. if it fails for an application test suite, then all tests fail - for unity8 only certain tests fail as the unlock happens on each test
[08:00] <ogra_> sil2100, hmm, could it be that your fix for the test setup yesterday used an old branch ?
[08:01] <sil2100> hmm, what do you mean?
[08:01] <ogra_> i see "ubuntu_filemanager_app" there ... that was changed to filemanager a while ago
[08:01] <ogra_> looks like it rolled back a few versions or so
[08:02] <sil2100> I was branching cleanly and pushing an MR, so hm
[08:02] <ogra_> well, something is weird here
[08:02] <sil2100> psivaa: hey, you around? :)
[08:02] <psivaa> sil2100: yep, just coming in and looking at the results :/
[08:04] <sil2100> psivaa: so we're wondering, do you have any idea why ubuntu_filemanager_app is there as per ogra_'s comment ? ^
[08:04] <sil2100> psivaa: and do you think it could be the screen unlock failing?
[08:05] <sil2100> (I remember Saviq or someone mentioning it as a possible problem)
[08:05] <psivaa> sil2100: looking, but the filemanager is really weird
[08:05] <ogra_> yeah
[08:06] <ogra_> could the device pull from the wrong branch ?
[08:07] <ogra_> there is no mention at all of ubuntu_filemanager_app anywhere in the current code
[08:08] <Saviq> sil2100, I doubt there's any issue with unlocking, we'd see a *really* red result if that was the case
[08:08] <Saviq> oh yay, crashes
[08:08]  * Saviq retraces
[08:09] <Saviq> (process:10790): GLib-GIO-ERROR **: No GSettings schemas are installed on the system
[08:09] <Saviq> huh?
[08:09] <psivaa> ogra_: sil2100: filemanager was run as ubuntu_filemanager_app accidentally in one of the reruns as far as i see. i'm running it as filemanager now.
[08:10] <ogra_> psivaa, ah, ok,, that helps with the confusion at least
[08:10] <sil2100> Saviq: where's that from?
[08:10] <sil2100> psivaa: ok, thanks
[08:10] <Saviq> sil2100, unity8 logs
[08:10] <Saviq> sil2100, https://jenkins.qa.ubuntu.com/job/utopic-touch-mako-smoke-daily/245/artifact/clientlogs/unity8/unity8.log/*view*/
[08:10] <Saviq> sil2100, that's what unity8 crashes with, at least
[08:12] <sil2100> hmm
[08:13] <sil2100> Maybe one of the devices has some broken state with packages missing and it causes unity8 and python-autopilot crashes?
[08:14] <ogra_> The following NEW packages will be installed:
[08:14] <ogra_>   gir1.2-upstart-app-launch-2 libpython-stdlib libpython2.7-minimal
[08:14] <ogra_>   libpython2.7-stdlib libupstart-app-launch2 python python-autopilot
[08:14] <ogra_>   python-contextlib2 python-dbus python-dbus-dev python-decorator python-evdev
[08:14] <sil2100> ogra_: in which test case is that? python2...
[08:14] <ogra_> https://jenkins.qa.ubuntu.com/job/utopic-touch-mako-smoke-daily/245/consoleFull
[08:15] <ogra_> looks like mediaplayer-app
[08:16] <sil2100> On 66 only one test failure on mediaplayer-app, but still strange it's pulling in that
[08:16] <Saviq> ogra_, sil2100 "mediaplayer_app unity8 dialer_app calendar_app sudoku_app ubuntu_filemanager_app ubuntu_weather_app online_accounts_ui" is the list of suites that were run in that job
[08:16]  * ogra_ checks if thats the same device the two broken apps were run on 
[08:16] <Saviq> doesn't include music_app
[08:17] <ogra_> mako-11
[08:17] <Saviq> ogra_, music was on mako-02
[08:17] <Saviq> there goes that theory
[08:17] <ogra_> calendar-app is mako-11
[08:18] <Saviq> why oh why don't we get the .crash files preprocessed!!!
[08:18] <ogra_> Saviq, and unity8 was mako-02 too
[08:18] <Saviq> ogra_, huh https://jenkins.qa.ubuntu.com/job/utopic-touch-mako-smoke-daily/245/consoleFull ?
[08:18] <ogra_> oops, sorry
[08:18] <ogra_> 11 too
[08:19] <sil2100> Ok, the theory also has a problem since mako-11 also ran tests which passed fine
[08:19] <ogra_> before mediaplayer ?
[08:19] <ogra_> or after it installed all that cruft (which was never removed)
[08:19] <sil2100> Well, unity8 was ran first from what I saw
[08:20] <sil2100> And it failed, right?
[08:20] <ogra_> right, not saying that we are not seeing two bugs
[08:20] <sil2100> Ah, no
[08:20] <sil2100> Actually...
[08:20] <ogra_> psivaa, is there a way to find the order in which which tests were run on which device
[08:20] <ogra_> (so many witches in that sentence :P )
[08:21] <sil2100> But yeah, so the first thing ran was mediaplayer_app, then unity8 and then the others... so still, weather should have failed right?
[08:21]  * ogra_ gets coffee
[08:21] <psivaa> ogra_: yep, let me take a look
[08:21] <ogra_> sil2100, unless it gets along with the old AP tests still
[08:21] <sil2100> ogra_: that's actually true
[08:21] <sil2100> But, why didn't we see it sooner then?
[08:22] <ogra_> because UAL only landed in 65 ?
[08:22] <sil2100> Ah, right! The gir1.2-upstart-app-launch-2 that's now installed
[08:22] <ogra_> the same apps fail in 65 and 66
[08:23] <sil2100> Ok, this starts to make sense
[08:23] <ogra_> (question is ... were they run in the same order there)
[08:25]  * sil2100 is checking the deps
[08:27] <psivaa> so with 66, mako-02 ran gallery_app, notes, messaging, music, calc, shorts, uitoolkit in the order and music terribly failed
[08:29] <psivaa> and with 66 mako-11 ran mediaplayer, unity8, dialer,calendar, sudoku, ubuntu_file_manager, weather, online_accounts_ui in the order
[08:30] <sil2100> psivaa: could you give me a link to the logs from the mako-2 run?
[08:30] <psivaa> sil2100: https://jenkins.qa.ubuntu.com/job/utopic-touch-mako-smoke-daily/244/consoleText
[08:30] <psivaa> brt for the meeting
[08:31] <popey> can someone reproduce bug 1326694
[08:31] <popey> also, landing call
[08:31] <Mirv> did you discuss https://code.launchpad.net/~ted/ubuntu-app-launch/confine-click-apps/+merge/222114 yet? ready for publishing from ted
[08:31] <sil2100> Yeah, trying to log...
[08:59] <ogra_> popey, no memory leakage on my flo over night ...
[09:00] <ogra_> (on #64 that was)
[09:04] <jibel> psivaa, bug 1326707 add any info you may have from smoke tests.
[09:04] <psivaa> jibel: ack, thanks
[09:05] <sil2100> HAH!
[09:05] <sil2100> ogra_, psivaa: I found something
[09:05] <ogra_> tell us
[09:07] <Chipaca> How's the train doing? Can I ask for silos &c?
[09:07] <ogra_> you can ask for silos but it is not clear if they can land
[09:08] <sil2100> ogra_, psivaa: actually, it's less relevant than I thought, sadly, but it seems that different AP versions are used for different tests
[09:08] <ogra_> there are still lots of issues with the upstart-app-launch rename that killl the tests
[09:08] <sil2100> ogra_: did you kick a new image?
[09:08] <Chipaca> that's ok, landing is overrated :-p
[09:09] <sil2100> Chipaca: give us some additional minutes and we'll get back to you :)
[09:09] <ogra_> sil2100, yeah, the bot should pick it up any minute
[09:10] <imgbot> [09:15] <Chipaca> sil2100: I'll go preparing a row :)
[09:16] <Chipaca> or maybe I won't? has the spreadsheet changed?
[09:17] <Laney> I keep getting to the old one
[09:17] <Laney> Chipaca: http://wiki.ubuntu.com/citrain should take you to the current spreadsheet
[09:18] <Chipaca> Laney: thanks :)
[09:20] <Chipaca> ok, row #34 ready for a silo whenever a silo is ready for row #34 :)
[09:25] <Chipaca> wow! that's a lot of green :D
[09:34] <sil2100> Chipaca: ok, tried to assign, but the MP is not a valid MP ;)
[09:35] <sil2100> Chipaca: could you modify that?
[09:35] <Chipaca> d'oh!
[09:36] <Chipaca> sil2100: there. sorry.
[09:37] <dbarth> sil2100: o/ hiya, i have line 27 ready if you have silos available
[09:37] <sil2100> dbarth: hey!
[09:38] <Saviq> sil2100, ogra_, are you any closer to knowing what happened? it looks like something that landed causes havoc in unity8 https://jenkins.qa.ubuntu.com/job/generic-deb-autopilot-runner-mako/1112/?
[09:38] <Saviq> we get 100% crash rate in ap tests on device
[09:39] <Saviq> oh noes, my ssh!
[09:39] <sil2100> uh
[09:39] <sil2100> We didn't know it's so serious
[09:40]  * Saviq tries to retrace
[09:41] <ogra_> http://paste.ubuntu.com/7593682/
[09:41] <t1mp> Saviq: I don't know if it is related, but I also got 100% failure in autolanding with a simple change: https://jenkins.qa.ubuntu.com/job/generic-deb-autopilot-runner-mako/1106/?#showFailuresLink
[09:41] <ogra_> -legacy UAL renaming
[09:41] <t1mp> ^the MR is this https://code.launchpad.net/~elopio/ubuntu-ui-toolkit/fix1326072-create_fake_xauthority/+merge/221953
[09:41] <Saviq> t1mp, sounds related indeed
[09:42] <Saviq> t1mp, so we got broken Qt it seems
[09:43] <Saviq> unless it's the gsettings thing again
[09:43] <ogra_> http://people.canonical.com/~ogra/touch-image-stats/65.changes
[09:43] <ogra_> thats the changeset
[09:43] <jibel> sil2100, where on the wiki do you document procedures of the CI team ? I want to document the escalation process for unstable tests we discussed during the sprint.
[09:43] <ogra_> doesnt look like there was much Qt in it
[09:43] <Saviq> (process:3579): GLib-GIO-ERROR **: No GSettings schemas are installed on the system
[09:43] <ogra_> right
[09:43] <Saviq> ogra_, yeah, I take that back
[09:43] <Saviq> it's this fucker
[09:43] <ogra_> ++
[09:43] <sil2100> jibel: there is no documentation for that existing, the only procedures have been written in an annoucement e-mail on the phone ML
[09:44] <Saviq> so gsettings-desktop-schemas?
[09:44] <jibel> sil2100, ok
[09:44] <ogra_> Saviq, yeah, i wonder where that came from
[09:44] <sil2100> hmmm
[09:44] <ogra_> ah, greeter
[09:44] <brendand> sil2100, how did you plan to escalate AP failures to QA?
[09:45] <ogra_> err
[09:45] <ogra_> no, that was touch-schemas
[09:45] <Saviq> ogra_, no we have that for long
[09:45] <Saviq> https://launchpad.net/ubuntu/+source/gsettings-desktop-schemas/3.12.0-1ubuntu1
[09:45] <ogra_> well, it got updated
[09:45] <Saviq> looks like a debian sync?
[09:45] <ogra_> hrmpf
[09:45] <sil2100> brendand: as per the e-mail, we proceed as discussed during the sprint - we assess the test if it's reproducible manually and then either forward it to QA or to the developers in the bug
[09:46]  * Saviq flashes 64 and upgrades one by one
[09:46] <ogra_> seb128, do you know anything about https://launchpad.net/ubuntu/+source/gsettings-desktop-schemas/3.12.0-1ubuntu1 ? smells like we are missing somthing ...
[09:46] <ogra_> (process:3579): GLib-GIO-ERROR **: No GSettings schemas are installed on the system
[09:46] <ogra_> or Laney ^^
[09:47] <sil2100> It doesn't seem to be anything landed throught the citrain:
[09:47] <sil2100> http://people.canonical.com/~lzemczak/landing-team/65.commitlog
[09:47] <Laney> erm
[09:47] <Laney> how do you do that?
[09:47] <seb128> ogra_, seems like the schemas compiler didn't run?
[09:47] <brendand> sil2100, yes but how?
[09:47] <brendand> sil2100, by email, bug tag?
[09:48] <sil2100> brendand: it will be marked in the bug and in the landing e-mail, and generally inform QA people available so that it can be passed down further
[09:49] <sil2100> brendand: if there are any propositions, I'm all ears
[09:50] <ogra_> seb128, when should that run ? postinst ?
[09:50] <Laney> it's a trigger
[09:50] <seb128> ogra_, trigger
[09:50]  * ogra_ checks the image build logs 
[09:50] <seb128> ogra_, ls /usr/share/glib-2.0/schemas
[09:50] <ogra_> seb128, hard to do, thats in the lab :)
[09:50] <ogra_> Saviq, do you have the error locally ?
[09:50] <seb128> do you have a gschemas.compiled
[09:51] <Saviq> ogra_, just flashing
[09:51] <cjwatson> Saviq: Merge rather than sync (our terminology is that syncs are verbatim copies)
[09:51] <Laney> http://people.canonical.com/~ubuntu-archive/livefs-build-logs/utopic/ubuntu-touch/latest/livecd-armhf.out
[09:51] <Saviq> cjwatson, ok, that
[09:52] <ogra_> hmm, i dont see any specific errors
[09:52] <sil2100> ogra_: I looked throught the diff you made and I think I see a problem
[09:52] <ogra_> sil2100, tell me
[09:53] <sil2100> ogra_: +    def launch_ubuntu_application(self <- this shouldn't be renamed
[09:53] <sil2100> ogra_: as per the merge for the python3 rename:
[09:53] <sil2100> ogra_: https://code.launchpad.net/~ted/autopilot/ubuntu-app-launch/+merge/220909
[09:54] <sil2100> ogra_: you can see that it was left as launch_upstart_application, just that the underlying bones got modified
[09:54] <ogra_> i wonder if that was an oversight though ... i thought tedg wanted to remove all mentioning of upstart ... but yeah
[09:55] <sil2100> ogra_: as you can see, he basically did a s/Upstart/Ubuntu , case-sensitive
[09:55] <ogra_> sigh, but yeah, seems like it is a lot more than just a search/replace
[09:55] <sil2100> Since any other methods that were generally lowercase didn't get changed
[09:55] <ogra_> right
[09:56] <ogra_> let me redo that then
[09:56]  * sil2100 just sounded like a complete non-technical person
[09:56] <sil2100> ;p
[10:00] <jibel> sil2100, brendand FTR https://wiki.ubuntu.com/CI/UnstableTests
[10:01] <Saviq> what a week, btw...
[10:03] <brendand> jibel, i'm dubious about the idea that a failure that isn't reproducible should be reassigned to infrastructure
[10:03] <brendand> jibel, but ok for it to be owned by qa
[10:04] <Saviq> ogra_, so no, of course the tests pass fine locally :|
[10:04] <jibel> brendand, it's now on the wiki, open for discussion and improvements :)
[10:04] <Saviq> ogra_, oh wait actually
[10:05] <Saviq> ogra_, yeah, I get it on 66
[10:05] <Saviq> nitctl: No such variable: XDG_DATA_DIRS
[10:05] <Saviq> soudns related
[10:06] <ogra_> not really ... unless you read from Pictures/Music/Video etc etc
[10:06] <ogra_> XDG_RUNTIME_DIR would be worrying :)
[10:06] <Saviq> hmm ok
[10:06] <brendand> jibel, how do you comment on the wiki? i never knew you could do that
[10:07] <ogra_> sil2100, http://paste.ubuntu.com/7593792/ a little shorter now :)
[10:07] <ogra_> Saviq, any dbus errors ?
[10:08] <Saviq> ogra_, not that I saw, no
[10:08] <ogra_> k
[10:08] <ogra_> thats good at least
[10:08]  * Saviq flashes 94 and will upgrade one by one
[10:08] <ogra_> well, try to resinstall the schemas package ... see if it spills an error from the trigger
[10:12] <ogra_> Saviq, hmm, i dont see a unity8-greeter landing, i thought you and mterry had changed the wrapper script again in your silo
[10:12] <Saviq> ogra_, http://people.canonical.com/~ogra/touch-image-stats/65.changes
[10:13] <ogra_> yay for descriptive changelogs ... sigh
[10:13] <ogra_>  * Bump version so ubuntu-touch-session can reference this one
[10:13] <Saviq> ogra_, yeah, sorry about that...
[10:13] <ogra_> thats the changelog entry ... i see the changes looking at the diff now
[10:14] <Saviq> gawedasduibidfbdsgsgasdkaw ^W
[10:14] <ogra_> +1
[10:16]  * sil2100 checks the diff
[10:17] <ogra_> psivaa, can you check if gsettings-desktop-schemas is installed on mako-11
[10:17] <ogra_> (hoping it is still in the state it was when the tests were run)
[10:17] <psivaa> ogra_: ack, 1 sec
[10:18] <ogra_> and also as seb128 said  ls /usr/share/glib-2.0/schemas
[10:18] <ogra_> see if they are there
[10:19] <ogra_> the console log at least doesnt show a removal ...
[10:20] <psivaa> ogra_: gsettings-desktop-schemas: Installed: 3.12.0-1ubuntu1
[10:20] <ogra_> thanks
[10:20] <ogra_> hmm
[10:21] <psivaa> and there are a lot of xmls listed under /usr/share/glib-2.0/schemas/
[10:21] <seb128> psivaa, what about a gschemas.compiled
[10:22] <psivaa> seb128: that's there. let me paste the whole list
[10:22] <Laney> where do you see this error?
[10:22] <ogra_> Laney, smoke testing
[10:22] <Laney> direct link?
[10:22] <psivaa> ogra_: seb128: http://paste.ubuntu.com/7593881/
[10:23] <ogra_> Laney, https://jenkins.qa.ubuntu.com/job/utopic-touch-mako-smoke-daily/245/artifact/clientlogs/unity8/unity8.log/*view*/
[10:24] <Laney> psivaa: does gsettings list-recursively work?
[10:24] <ogra_> (as phablet i suppose)
[10:24] <sil2100> ogra_: ok, looks better I see... tedg doesn't rename FakeUpstartBase classess, but I guess those are used only internally with autopilot autopilot tests
[10:24] <ogra_> k
[10:25] <sil2100> So it seems ok
[10:25] <ogra_> so upload ?
[10:25] <sil2100> Give me another minute ;)
[10:25] <ogra_> heh, k
[10:27]  * ogra_ meanwhile fixes the changelog accordingly
[10:27] <sil2100> uh
[10:27] <psivaa> Laney: http://paste.ubuntu.com/7593883/ is the output but as root
[10:27] <ogra_> psivaa, try as phablet too ... the tests run as the user
[10:28] <sil2100> ogra_: another thing... I think we cannot rename UpstartApplicationLauncher to UbuntuApplicationLauncher ;p
[10:28] <sil2100> ogra_: since I think it's imported from the UAL
[10:28] <ogra_> oh, why ?
[10:28] <sil2100> As per the merge tedg made ;p
[10:28] <sil2100> It's probably defined as such and hm
[10:28] <psivaa> ogra_: it works too when as phablet
[10:28] <ogra_> right, just making sure
[10:28] <sil2100> Since I didn't see it defined locally in autopilot, so I guess it's imported somehow?
[10:28] <sil2100> Still digging
[10:29] <Laney> this is very suspicious
[10:29] <ogra_> sil2100, in testcases.py iirc
[10:29] <seb128> Laney, what? the gsettings no schemas thing?
[10:29] <Laney> yes
[10:29] <seb128> yeah, doesn't make sense to me
[10:29] <Laney> what if you re-run those tests now on the same machine?
[10:29] <sil2100> Ah!
[10:29] <sil2100> Sorry, missed that, but yeah.. we cannot rename that
[10:30] <sil2100> ogra_: it seems to be exported, and python3-autopilot didn't get that renamed
[10:30] <sil2100> (this would require changing all application tests)
[10:30] <sil2100> ogra_: so indeed, it's much more complicated than just regexing :|
[10:30] <ogra_> sil2100, but -legacy uses py2
[10:30] <ogra_> doesnt it ?
[10:31]  * Mirv gets into the -gles fork packaging fun
[10:31] <sil2100> ogra_: yeah, but python2 and python3 versions of AP need to use the same function/class names
[10:31] <sil2100> ogra_: so if it stays as UpstartApplicationLauncher in python3, it has to stay the same in python2
[10:32] <sil2100> ...I think
[10:32] <ogra_> i dont get that ... this package ships UbuntuApplicationLauncher
[10:32] <sil2100> https://code.launchpad.net/~ted/autopilot/ubuntu-app-launch/+merge/220909
[10:32] <sil2100> Look at this change and grep for UpstartApplicationLauncher
[10:33] <sil2100> It's still there, it's not renamed, so the python3 one stays with the UpstartApplicationLauncher name, doesn't rename it to Ubuntu
[10:33] <ogra_> hmm
[10:33] <cjwatson> sil2100: or  try: import foo; except ImportError: import bar  presumably
[10:34] <cjwatson> (more awkward but you certainly can)
[10:34] <sil2100> cjwatson: what do you mean?
[10:34] <cjwatson> I mean it's entirely possible to write bilingual Python 2/3 code where the same functionality is imported under different names
[10:35] <imgbot> [10:35] <imgbot> [10:35] <popey> \o/
[10:35] <cjwatson> just replying to your train of thought above
[10:35] <sil2100> cjwatson: right, but it's best to keep both python 2 and 3 versions in sync, exporting the same names for both versions
[10:35] <cjwatson> yeah, all other things being equal
[10:35] <cjwatson> certainly if both 2/3 versions were there before, no reason not to do both
[10:36] <cjwatson> I've run across cases in porting where the python3 version had no users previously and so it was better to improve the API there even though it had to be kept the same in python2 for compatibility
[10:38]  * ogra_ thinks we really should lave it to ted ... i feel a bit uneasy not knowing exactly why some bits were renamed and why others werent at all in the code 
[10:38] <Saviq> ok I can easily repro the gsettings issue when upgrading... let's try again to pinpoint the package
[10:39] <ogra_> hmm, the change of 67 doesnt have MIr
[10:39] <ogra_> sigh
[10:39]  * ogra_ guesses we should have checked -proposed :P
[10:40]  * ogra_ does so now 
[10:40] <sil2100> tedg: heeeey!
[10:41] <sil2100> tedg: wake up!
[10:41] <Saviq> unlikely ;)
[10:41] <ogra_> unity-system-compositor, unity-mir, mir and platform-api are all sitting in -proposed ... damned
[10:41] <Saviq> unless he's jetlagging :D
[10:43] <cjwatson> mir/arm64 build failed, which should have shown up in citrain
[10:43] <ogra_> yeah, i see it on excuses
[10:44] <cjwatson> and it has a couple of rdeps now
[10:44] <sil2100> Let me check the PPA
[10:44] <sil2100> robru was the publisher so I don't know much about this landing
[10:44] <ogra_> well, not that important then ... the Mir team can care for that ... we just wasted an image build :P
[10:44] <cjwatson> I think the android-properties fix just needs a bit more fixing
[10:45] <ogra_> it will stay in -proposed and we have other issues
[10:45] <sil2100> HOW DID HE PUBLISH THIS?
[10:45]  * cjwatson tries to remember if that was my fault
[10:45] <sil2100> I mean, sorry for the caps, but how could it gotten published with failures in amd64?
[10:45] <robru> hey guys, I'm up due to jetlag
[10:45] <cjwatson> the failure that is, not the publishing
[10:45] <cjwatson> sil2100: arm64, not amd64 :)
[10:45] <robru> sil2100, it was arm64, not amd64
[10:45] <cjwatson> I *think* it must have been forced
[10:47] <sil2100> Right, typo, those are close
[10:47] <sil2100> Anyway, it was a valid arch for mir
[10:47] <sil2100> robru: did you get some archive admin ACKing on releasing without this arch?
[10:48] <robru> sil2100, cjwatson: so, this one was quite strange. I saw the build had failed in the silo, but the citrain build job was acting really funny. it was claiming that the build was ongoing, it couldn't see that the build had failed. so I tried rebuilding a couple times, no luck. camako had indicated that the silo was highly tested, and I tested it myself, it looked good on device so I was hoping that publishing it would "unstick" the
[10:48] <robru> failing arm64 build somehow
[10:48] <cjwatson> Afraid not
[10:48] <cjwatson> I'm working on fixing it now
[10:48] <robru> sil2100, no, no ACK for releasing without that arch, I was hoping that arch would rebuild in -proposed, as I've seen that happen when publishing in the past.
[10:48] <sil2100> robru: ah... right, whenever a build job is stuck it might mean that there's a dep-wait somewhere
[10:48] <sil2100> And the build is failing
[10:49] <cjwatson> Right, the build failure's definitely real; looks like a two-liner to fix but I'll test-build it
[10:49] <cjwatson> well, marginally more than two, http://paste.ubuntu.com/7593994/
[10:49] <robru> sil2100, look at the PPA though, there's lots of depwaits on the Often Ignored arches, citrain didn't block on those. it just blocked only on the single arm64 "ongoing build". quite a strange build log if you review it
[10:50] <cjwatson> in case of doubt the best way to check is to run "rmadison -s utopic -S <source package name>" and check what architectures it's built for there; those are the architectures that proposed-migration is going to require
[10:50] <sil2100> robru: yeah, but that's happening every time and it's not normal - whenever a build is stuck on building something it means one of the arches that are required didn't build
[10:51] <sil2100> cjwatson: right, useful command
[10:51] <robru> cjwatson, thanks
[10:51] <cjwatson> reductions are sometimes possible but require manual removal of the no-longer-supportable binaries and I'd usually only do that if the reverse-dependency chain is tiny/absent
[10:51] <sil2100> cjwatson: I usually did the long way, like... LP
[10:51] <cjwatson> i.e. minimum fallout
[10:51] <robru> sil2100, yeah, I know it didn't build, but I didn't understand why. I was hoping it would just rebuild itself in -proposed
[10:51] <sil2100> robru: it's usually much better to poke upstreams in this case
[10:51] <infinity> robru: If it won't build in the PPA, it won't magically work in -proposed.
[10:51] <cjwatson> ... usually
[10:52] <infinity> robru: So, copying out in that case is definitely not the answer.
[10:52] <sil2100> robru: right, as our PPAs have -proposed enabled and the same builders
[10:52] <cjwatson> There are weird edge cases sometimes :)
[10:52] <infinity> cjwatson: Well, unless timing means some rdep is fixed between A and B. :P
[10:52] <sil2100> So it's like -proposed, just safer ;p
[10:52] <robru> infinity, cjwatson: yes, I saw weird edge cases some months ago that made me think this
[10:52] <cjwatson> Anyway, should be easy to sort out as soon as am1 finishes branching this.
[10:53] <sil2100> cjwatson: thanks!
[10:53] <popey> sil2100: ogra_ seems 67 doesn't have new unity/mir?
[10:53] <sil2100> popey: no ;)
[10:53] <cjwatson> popey: See the conversation immediately above.
[10:53] <popey> \o/
[10:54] <popey> oh, ☻
[10:54] <popey> Summarise it for me in one word.
[10:54] <sil2100> popey: as discussed ^ heh, we had a premature ejacu...image build
[10:54] <Saviq> ogra_, confirmed, it's url-dispatcher that causes the gsettings dying
[10:54] <sil2100> No one of us actually checked if it moved out of -proposed ;p (bummer)
[10:54] <robru> popey, "no"
[10:54] <robru> ;-)
[10:54] <popey> expected ☻
[10:54] <sil2100> Saviq: oh?
[10:55] <ogra_> Saviq, lovely
[10:55] <sil2100> Saviq: ok, I probably missed out some of the discussion, but how is that so?
[10:55] <sil2100> Was that the url-dispatcher landing from tedg yesterday?
[10:55] <sil2100> I'm saying we revert that
[10:56] <Saviq> my apt history.log: http://pastebin.ubuntu.com/7594050/
[10:56] <Saviq> no idea where the --reinstalls come from ;)
[10:56] <Saviq> looks like from rootfs
[10:57] <Saviq> but I upgraded gsettings-desktop-schemas and url-dispatcher + ubuntu-app-launch
[10:57] <Saviq> and that's it, gsettings dead
[10:57] <Saviq> well, the weird thing it's actually only happening under autopilot... the phone booted fine...
[10:58] <sil2100> Saviq: so, those two in combination cause it, or simply the upgrade of url-dispatcher is enough?
[10:58] <ogra_> Saviq, did you try *only* updating gsettings-desktop-schemas
[10:58] <Saviq> sil2100, can't upgrade url-dispatcher alone, it deps on the ohter
[10:58] <ogra_> ?
[10:58] <Saviq> ogra_, yes
[10:58] <ogra_> and thats behaving fine ?
[10:58] <Saviq> yup
[10:58] <ogra_> ok, yeah, then i agree
[10:58] <Saviq> see the start-date difference? that was me rebooting and testing
[10:59] <ogra_> ah, right
[10:59] <Saviq> let me flash again and upgrade the minimal amount of pkgs (like only ubuntu-app-launch)
[11:03]  * sil2100 needs to jump out to the shop
[11:05] <cjwatson> sil2100,robru: https://code.launchpad.net/~cjwatson/mir/fix-android-properties-check/+merge/222163 (still building here and I have to go out shortly, but it's certainly getting a heck of a lot further).  Will presumably need review from somebody like kdub
[11:06] <robru> cjwatson, want me to silo that?
[11:06] <ogra_> cjwatson, we're not in a hurry with that one
[11:07] <cjwatson> robru: Happy to wait for a mir reviewer - e.g. with their processes I'm not sure if it needs to be merged into lp:mir/devel before going into a silo
[11:07] <ogra_> fixing sll the UAL rename fallout has priority anyway
[11:07] <cjwatson> ogra_: Yup, just whack-a-moling tech debt
[11:07] <ogra_> it is actually good that mir is stuck there so we dont taint fixing work
[11:07] <robru> cjwatson, fair enough, thanks for the branch
[11:11] <sil2100> ogra_: I'll try to do the autopilot rename in the meantime
[11:11] <ogra_> brave
[11:11] <sil2100> That's my third name
[11:11] <ogra_> haha
[11:11] <ogra_> whats your second ?
[11:12] <sil2100> It's a secret -_^
[11:13] <Chipaca> ogra_: do you know where bugs (a kernel crash) in ubuntu-emulator should be reported?
[11:16] <Saviq> sil2100, ogra_, confirmed http://pastebin.ubuntu.com/7594147/
[11:16] <Saviq> sil2100, ogra_, that's enough to get gsettings in a bad state
[11:17] <ogra_> why does it remove upstart-app-launch ?
[11:17] <ogra_> (why does it have to ... why is it installed)
[11:18] <Saviq> ogra_, probably unity8-autopilot before that
[11:18] <ogra_> i dont see it in the list there
[11:18] <Saviq> hmm indeed
[11:18] <Saviq> ogra_, I started from image 64, btw
[11:19] <ogra_> oh, ok
[11:19] <ogra_> that would explsin it
[11:19] <ogra_> *explain
[11:19] <Saviq> ogra_, thought so
[11:19] <Saviq> that was my whole idea, 64 (working) and work from there
[11:19] <sil2100> I think I've got something here
[11:19] <sil2100> With the autopilot thingy
[11:19] <Saviq> *no* idea how that affects it, since none of those packages actually touch gsettings or schemas...
[11:20] <Saviq> so it must be some interaction... but dunno how to test without installing unity8-autopilot... that's what /me is interested in :P
[11:26] <ogra_> Saviq, well, there is a dependency chain in unity8-autopilot ...
[11:26] <sil2100> ogra_: http://paste.ubuntu.com/7594185/
[11:26] <Saviq> ogra_, yeah, but installing *just* that is fine
[11:26] <sil2100> ogra_: this should do it
[11:26] <ogra_> unity8-autopilot -> ubuntu-ui-toolkit-autopilot -> python-autopilot -> gir1.2-upstart-app-launch-2
[11:26] <Saviq> ogra_, I mean after installing that I can run unity8's ap tests fine
[11:27] <Saviq> ogra_, only after upgrading UAL does it break
[11:27] <ogra_> Saviq, lets see what happens after we landed sil2100's fix above
[11:27] <Saviq> k
[11:27] <ogra_> that gir1.2-upstart-app-launch-2 causes havoc
[11:27] <sil2100> ogra_, Saviq: I don't guarantee it will work, but I worked with tedg's merge + common sense and grep magic, should be safe
[11:28] <ogra_> sil2100, well, more broken than broken cant happen anyway :P
[11:28] <ogra_> lets upload this and let tedg deal with the fallout
[11:29] <ogra_> (merging back fixing subsequent issues) ....
[11:29] <ogra_> so we dont lose more time
[11:30] <ogra_> the debdiff looks ok to me but i must admit i dont know much about that code ... comparing it to the non -legacy merge it looks good though
[11:30] <sil2100> ;D
[11:30] <sil2100> ogra_: is this enough or you want the src package as well?
[11:31] <sil2100> If we're cool then I jump out to that store for a moment, we're missing lunch essentials
[11:31] <ogra_> thats enough i have the source pkg here i can apply the patch
[11:31] <sil2100> ogra_: thanks :)
[11:33] <ogra_> uploaded
[11:36]  * ogra_ finally finds the time to upgrade his flo to 67
[11:40] <ogra_> ricmm, doesnt looks good for papi v2 ...
[11:40] <ogra_> *look
[11:40] <kgunn> sil2100: choo-choo told me mir0.2.0 was migrating, but looks like its waiting on u-s-c in migration pocket...can you check if its stuck ?? PES guys are wanting it to land
[11:40] <ogra_> the UAL rename causes more and more afllout :(
[11:41] <ogra_> kgunn, it failed to build on arm64 ... there was a lot of discussion in the scrollback
[11:42] <ogra_> (about 1h ago)
[11:43] <ogra_> WOAH !
[11:43] <ogra_> we have outlook in the store !!
[11:43] <ogra_> lol
[11:58] <popey> ogra_: ☻
[11:59] <robru> kgunn, cjwatson kindly put together a fix for mir on arm64, if you can review it and incorporate it into the silo, then that can progress: https://code.launchpad.net/~cjwatson/mir/fix-android-properties-check/+merge/222163
[12:08] <ogra_> popey, the click installled fine here btw
[12:08] <ogra_> on 67
[12:08] <ogra_> (this was an OTA though, not a fresh install)
[12:09] <mardy> fginther: hi! Any progress on adding the online-accounts stuff to http://162.213.34.64:8080/gaps/project/ ?
[12:19] <popey> ogra_: can you uninstall?
[12:22] <ogra_> popey, aha, no, i cant
[12:22] <ogra_> damned ... i'm doomed to have outlook installed forever now !
[12:22] <ogra_> :)
[12:23] <popey> bug 1326694
[12:23]  * ogra_ confirms
[12:24] <ogra_> popey, i cant confirm your ram issue on flo though
[12:24] <ogra_> it didnt pile up any additional ram over night
[12:25]  * sil2100 is preparing lunch
[12:26] <ogra_> hmm
[12:26] <sil2100> I see autopilot in -proposed :)
[12:26] <ogra_> looks like copy pasting from paste.u.c messed up UTF-8 completely
[12:26]  * tedg scans backlog, decides that dream was better
[12:26] <ogra_> sorry for mangling your name
[12:26] <sil2100> Saviq: thostr_1: ^ bug 1326694
[12:27] <ogra_> tedg, looks like you forgot to port autopilot-legacy
[12:27] <tedg> sil2100, Can you give me an update of what's current?
[12:27] <sil2100> tedg: yeah, so it seems you updated the py3 autopilot, but py2 was left unchanged ;)
[12:27] <thostr_1> sil2100: will take care of it
[12:27] <tedg> Oh, I didn't realize there were two.
[12:27] <ogra_> there seems to be another issue with gstettings schemas with the new url-dispatcher
[12:27] <sil2100> tedg: so, I tried to do the same changes and we released it to the archive
[12:27] <tedg> sil2100, Great, thanks!
[12:27] <ogra_> which we dont know yet if it is related to -legacy or not
[12:27] <sil2100> tedg: yeah... right now the py2 bindings are part of the autopilot-legacy source ;)
[12:28] <sil2100> tedg: could you take a look at the debdiff and check if I did it good ;p?
[12:28] <tedg> Sure
[12:28] <sil2100> I tried using your merge for py3 in this process
[12:28] <popey> ogra_: ok
[12:29] <sil2100> tedg: http://paste.ubuntu.com/7594185/
[12:29] <tedg> Heh, stole it from LP :-)
[12:29] <sil2100> ;)
[12:30] <tedg> Just to be curious, I thought all the tests were py3 already? no?
[12:30] <ogra_> tedg, i think that legacy part iis sorted now ... would be good if you could help Saviq to look into why gsettings break with the new url-dispatcher
[12:30] <ogra_> (which in turn breaks half of the unity8 tests)
[12:31] <tedg> Hmm, okay. Saviq what are you looking into?
[12:31] <ogra_> tedg, it *might* be related to -legacy pulling in upstart-app-launch stuff it shouldnt ... but we are not sure
[12:32] <tedg> The change there was tiny, I'd be surprised it was the root cause.
[12:32] <tedg> In URL dispatcher
[12:32] <ogra_> Saviq, could pretty well nail it down to installing that package and the world fell apart
[12:33] <ogra_> but that indeed doesnt necessarily mean it is url-dispatcher itself
[12:33] <Saviq> tedg, yeah, I took image 64, installed unity8-autopilot, stuff's fine
[12:33] <ogra_> the gsettings schmas got merged newly from debian for example
[12:33] <Saviq> tedg, then, installed ubuntu-app-launch, and boom, gsettings in autopilot die :|
[12:34] <tedg> Is there an error printed anywhere?
[12:37] <tedg> So are we talking UAL name change? or URL Dispatcher?
[12:39] <ogra_> https://jenkins.qa.ubuntu.com/job/utopic-touch-mako-smoke-daily/245/artifact/clientlogs/unity8/unity8.log/*view*/
[12:40] <ogra_> (process:10716): GLib-GIO-ERROR **: No GSettings schemas are installed on the system
[12:40] <ogra_> thats the error
[12:40] <tedg> I'm getting that with the clock app on my flash from last night.
[12:40] <tedg> Should be able to backtrace that
[12:40] <ogra_> and i guess we are talking u-a-l since unity8 already failed before the last url-dispatcher upload
[12:42] <fginther> mardy, I'll follow up with alesage, there was an MP to get most of this done, but it got stuck in review. We'll unwedge it.
[12:43] <tedg> ogra_, Those errors are all coming from different processes, is that because we're restarting unity8?
[12:44] <ogra_> yeah, i think we do
[12:46] <tedg> Wonder if there's an env var we're getting wrong.
[12:46] <tedg> I guess also, have there been any SDK changes?
[12:46] <ogra_> well, did you change any env vars with the UAL change ?
[12:47] <tedg> Not on purpose.
[12:47] <ogra_> http://people.canonical.com/~ogra/touch-image-stats/65.changes is where it started
[12:47] <brendand> ogra_, are you guys looking at the music app not starting?
[12:47] <ogra_> (teh sdk-libs in there is just an unrelated metapackage rebuid)
[12:47] <ogra_> brendand, nope
[12:47] <tedg> brendand, I think that's fixed in silo2
[12:48] <ogra_> brendand, does it not start ?
[12:48] <tedg> brendand, If you'd like to double check that silo, that'd be awesome.
[12:49] <brendand> tedg, doing
[12:51] <mardy> fginther: thanks! :-)
[12:51] <ricmm> robru: hi
[12:51] <ricmm> robru: I see platform-api 2.0.0 in -proposed
[12:51] <robru> ricmm, hey
[12:51] <ricmm> robru: what is it doing there? silo 007 hasnt been published
[12:52] <ogra_> sil2100, seems that Mir landing was quite a bit prematures
[12:52] <ricmm> do you know what silo that came from?
[12:52] <ogra_> -s
[12:52] <sil2100> robru: SLEEP!
[12:52] <robru> sil2100, can't
[12:52] <robru> ricmm, it came from silo 8
[12:53] <sil2100> ricmm: there was some Mir landing last night, robru knows more
[12:53] <ogra_> sil2100, platform-api 2.0 should not be used before ricmm's silo landed ... seems the Mir landing pulled from trunk or so ... given the package version number
[12:53] <robru> ricmm, don't worry, it's blocked in proposed, it's not about to land, it's blocked by mir build failure on arm64, cjwatson has a fix, kgunn is reviewing
[12:54] <robru> ogra_, oh, awesome
[12:54] <sil2100> geh, yesterday I didn't even see the Mir landing
[12:54] <ricmm> I dont get it
[12:54] <ricmm> I dont see 2.0.0 in the chanelog of silo 8's MRs
[12:54] <ricmm> ogra_: got a link to the source upload?
[12:54] <ricmm> I want to see the diff
[12:54] <ogra_> https://launchpad.net/ubuntu/utopic/+source/platform-api/2.0.0+14.10.20140603.3-0ubuntu1
[12:54] <robru> ricmm, good point
[12:55] <ogra_> https://launchpad.net/ubuntu/+source/platform-api/2.0.0+14.10.20140603.3-0ubuntu1
[12:55] <ogra_> there are the diffs ...
[12:55] <sil2100> I actually thought we'll just land ricmm's platform-api v2 today in the morning
[12:55] <ogra_> looks like only packaging changes
[12:56] <ogra_> but why is it called 2.0.0 then ?
[12:56] <robru> that is really curious
[12:56] <ogra_> the upstream tarball is 2.0 too
[12:56] <brendand> tedg, looks good to me
[12:57] <tedg> brendand, Great, thanks for testing it.
[12:57] <ricmm> ogra_: looks like jenkins decided to update version to 2.0.0
[12:57] <ricmm> but thats the only change, the rest I dont care
[12:57] <ricmm> but seriously, I want to ask something
[12:57] <camako> robru, I incorporated cjwatson's fix, do we need to start a rebuild?
[12:57] <ogra_> phew
[12:57] <ricmm> why are we even landing stuff like Mir when everything is broken?
[12:57] <ogra_> ricmm, it wasnt
[12:58] <robru> ricmm, because I thought everything got fixed? we're not in TRAINCON-0
[12:58] <ricmm> but its clearly not fixed
[12:58] <robru> camako, yes, did you merge cjwatson's diff into an existing MP? or just add the MP to the silo?
[12:58] <camako> robru, the former
[12:58] <robru> camako, cool, then a rebuild will handle ti
[12:58] <ogra_> ricmm, http://ci.ubuntu.com/smokeng/utopic/touch/mako/64:20140604:20140530/8385/ ... it landed after we got these test results
[12:59] <sil2100> ricmm: TRAINCON-0 was dropped, we had good test results and good image state when that was announced, but then some other migrations happened and caused todays problems
[12:59] <tedg> robru, sil2100, can we publish silo 2, I'm not sure it fixes the Unity8 issue, but it should fix a couple of the apps tests.
[12:59] <tedg> Be nice to get in the next image
[12:59] <ogra_> ricmm, the prob is that it landed together with UAL changes too ... which shouldnt have happened
[12:59] <robru> tedg, it's up to sil2100 now, i'm not supposed to be awake currently
[12:59] <sil2100> tedg: ok, so... it's also a potential fix for the things we're encountering?
[13:00] <ogra_> ricmm, and the UAL changes broke the world again ... (fix pending in proposed)
[13:00] <ogra_> sil2100, seems like the music app should be better at least
[13:00] <ogra_> if i understand the discussion right
[13:00] <tedg> sil2100, It doesn't fix any gsettings stuff (not sure on that one yet) but it does get click apps running as click apps, which gives them a consistent environment.
[13:00] <sil2100> Ok, then let's risk it
[13:00] <ogra_> ++
[13:01] <robru> camako, ugh, I see your build failure, hang on
[13:04] <ricmm> I'm still not sure why 2.0.0 got picked by jenkis as the right version to play
[13:05] <ricmm> is it possible there is a bug in the CI train scripts?
[13:05] <robru> ricmm, yeah that's a mystery to me
[13:05] <ricmm> platform-api 2.0.0 was in silo 008 before
[13:05] <ricmm> then it was flushed and move to 7
[13:05] <ricmm> perhaps jenkins is seeing the PPA history
[13:05] <robru> ricmm, oh yeah, that's exactly it.
[13:05] <ricmm> and actually doing last+1 even if it was deleted?
[13:05] <sil2100> ricmm: I checked lp:platform-api/devel and I don't see 2.0.0 there, so it's hm
[13:05] <robru> ricmm, didn't know about that history
[13:05] <ricmm> thats pretty awful
[13:05] <sil2100> Let me check the code
[13:05] <robru> ricmm, technically this is a bug in launchpad, not citrain
[13:06] <robru> cjwatson, we hit this once before I think, just remembering now ^^
[13:06] <ricmm> I rather put my money on ci train bug, I'm sure you can filter results on !deleted
[13:06] <sil2100> But we remove all packages from the PPA, not leaving any trace
[13:06] <robru> sil2100, not true, there's a trace ;-)
[13:07] <ricmm> sil2100: package history always remain, the real issue is when pulling the last known version
[13:07] <ricmm> to construct the changelog
[13:07] <ricmm> theres probably a missing check for ==deleted :)
[13:07] <ricmm> which just caused me to change every single one of my (14?) branches to update the build deps
[13:08] <robru> ricmm, sorry to hear that
[13:08] <robru> sil2100, https://launchpad.net/~ci-train-ppa-service/+archive/landing-008/+builds?build_text=&build_state=all here's the build history, clearly shows deleted packages
[13:08] <sil2100> Then it's a serious issue, it shouldn't be like that, deleted packages shouldn't be considered in such a thing by default
[13:09] <sil2100> Looking at citrain now
[13:10] <robru> sil2100, I have vague memories of raising this with didier some months ago, hit some snag where I'd accidentally bumped a version I didn't want to, then tried to delete the package and lower the version, couldn't do it, jenkins kept finding the higher version number, so I had to just roll with the higher version.
[13:11] <robru> camako, https://ci-train.ubuntu.com/job/landing-008-1-build/98/console ok this looks like it's on it's way now, will require re-testing of course.
[13:11] <sil2100> Yeah...
[13:12] <sil2100> ricmm, robru: so, I see Didier was simply fetching the list of sources, while LP API has a property declaring the state of the package
[13:12] <camako> robru, thanks... sure will retest
[13:12] <robru> camako, you're welcome
[13:12] <sil2100> ricmm, robru: so I'll try fixing that for the future, but we'll have to do something with the Mir landing in the PPA
[13:12] <ricmm> well its already in proposed
[13:12] <ricmm> either you delete that and make it see its 1.2.0 again, or I dont know
[13:12] <ricmm> one way is to flush it to a silo other than 008/007
[13:12] <sil2100> ricmm: we can drop it if it's very bad
[13:13] <ricmm> I dont know how bad it might be, it just needs a round of checking my silo 007 branches for build-dep discrepancies
[13:13] <ricmm> as I dont want things linking to the incorrect platform-api
[13:13] <sil2100> Mir is blocked in -proposed anyway, so I guess we could drop papi and push a new one
[13:13] <ricmm> because now ther will be a papi 2.0 that actually ships 1.2 sonames
[13:13] <sil2100> Right...
[13:13] <ricmm> so, that in itself is broken, even if it doesnt cause world breakage
[13:14] <ricmm> if dropping platform-api from proposed and manually uploading 1.2.0 from their branch is doable, I'd say do that
[13:15] <sil2100> Let's see what we can do
[13:15] <robru> ricmm, not sure if that's possible... I think the genie might be out of the bottle on this one, version '2.0.0' is just ruined.
[13:16] <robru> would need an archive admin to give a definitive answer on the feasability of that
[13:16] <ricmm> name one
[13:16] <ricmm> I dont know them
[13:16] <ricmm> maybe we can ask right away :)
[13:16] <robru> ricmm, well cjwatson or infinity
[13:16] <ricmm> cjwatson: what do you think colin?
[13:17] <sil2100> robru: no worries
[13:17] <sil2100> I'm sure we can find a way
[13:18] <sil2100> I hope
[13:18] <sil2100> cjwatson: if you're around, if we drop a platform-api 2.0.0 from -proposed, will we still be able to upload a version lower than that?
[13:21] <cjwatson> robru: sorry, on a call, can I have a one-line summary?  (is your question the same as sil2100's?)
[13:21] <robru> cjwatson, yes
[13:21] <cjwatson> robru,sil2100: if you do that, you can never reuse the identical version, but you can go backwards
[13:21] <sil2100> \o/ ok, so there's still hope
[13:21] <sil2100> cjwatson: thanks!
[13:22] <sil2100> ricmm: ^
[13:22] <cjwatson> it's *possible* that it only works via copies rather than direct uploads, I don't quite remember
[13:22] <cjwatson> try it, worst case use a different silo, and it can be handled in -proposed
[13:23] <cjwatson> robru: regarding the citrain part, I'm pretty sure I fixed it to skip deleted packages immediately after that incident
[13:23] <ricmm> cjwatson: after this incident? or a previous one
[13:23] <cjwatson> previous
[13:23] <ricmm> it totally went through this time
[13:23] <jhodapp> sil2100, can you publish line 31 for me, silo 18?
[13:23] <cjwatson> "that" incident
[13:23] <ricmm> darn
[13:23] <cjwatson> means not this one
[13:24] <ricmm> well it could mean "that" one up there ;)
[13:24] <robru> ricmm, yeah, "that one up there", that I referenced happening some months ago :-P
[13:24] <ricmm> well even worse then, if it has already happened
[13:24] <ricmm> sil2100: do you have a plan then?
[13:24] <robru> yeah, not sure how this happened again
[13:25] <sil2100> ricmm: so, let's play it safe and do this:
[13:25]  * ogra_ goes for a walk to calm down again 
[13:25] <sil2100> ricmm: let's ask cjwatson to drop papi from -proposed, then I'll assign the preprod silo 000 for the platform-api merge from that Mir landing
[13:26] <sil2100> ricmm: we'll build it there, it should be correct version
[13:26] <ricmm> sil2100: alright, I trust you to do the right thing
[13:26] <sil2100> ricmm: then we publish that silo (it will make a copy, so we *should* be safe)
[13:26] <ricmm> lemme know how it goes, if all fails I'll fix it in my branches
[13:26] <ricmm> 1-2 days *after* UAL is fixed/Mir lands/ etc
[13:26] <ricmm> win 18
[13:27] <sil2100> cjwatson: once you're after meetings, can you drop the platform-api 2.0.0+14.10.20140603.3-0ubuntu1 from -proposed? :)
[13:27] <sil2100> Damn, I'll have to copy all the mir packages to the preprod silo as well...
[13:27] <sil2100> Oh, wait, no
[13:27] <sil2100> I don't have to, phew
[13:28] <jhodapp> or robru, it's a simple fix for QMediaPlaylist support that helps fix the dialer that Bill's team really needs
[13:28] <sil2100> Since all prereqs are in -propsoed now
[13:28] <robru> jhodapp, oh sorry
[13:28] <robru> sil2100, but dont' forget that mir is rebuilding, the one in -proposed is broken
[13:28] <jhodapp> robru, np
[13:29] <sil2100> Crap...
[13:29] <robru> jhodapp, silo 18 isn't marked as testing: pass
[13:29] <robru> sil2100, so you'll need to republish silo 8 again. might be easier to just turf that whole silo and then reassign it in a different one
[13:29] <sil2100> robru: right, if a rebuild is needed then it's even easier now
[13:29] <robru> sil2100, well, it's already rebuilding.
[13:30] <cjwatson> sil2100: could you give me a reason string that I can copy and paste? :)
[13:31] <jhodapp> robru, it needs to be marked as testing or something else?
[13:31] <sil2100> cjwatson: let me think ;) "Removing platform-api 2.0.0 as the version number is incorrect: the package in reality is still 1.2.0, but has been bumped by CI Train to 2.0.0 due to an error. Reupload with correct version number needed."
[13:31] <sil2100> cjwatson: would that be ok?
[13:31] <cjwatson> OK, the previous fix I got applied to citrain was actually for something slightly different
[13:32] <robru> jhodapp, you know the spreadsheet right? if you tested the silo, please mark 'yes' in the "testing pass?" field on the silo page
[13:32] <sil2100> kgunn: hi!
[13:32] <jhodapp> robru, oh right, I always forget about that sub-tab
[13:32] <robru> jhodapp, yeah, heh
[13:32] <cjwatson> The problem then was that even after deleting a package you couldn't go back; so I made it take the most recently created upload, not just the one with the highest version
[13:32] <sil2100> kgunn: are you handling the rebuild of the mir landing right now?
[13:32] <robru> sil2100, also inform camako about this since he's working on this  mir landing quite closely.
[13:32] <cjwatson> But it didn't occur to me that people might delete and then not upload a new thing
[13:32] <sil2100> cjwatson: ah, ok ;) THanks
[13:32] <cjwatson> robru,ricmm: ^-
[13:32] <ricmm> sil2100: I think its a bit early for kgunn
[13:32] <sil2100> camako: hi!
[13:33] <robru> cjwatson, ahhhh that makes more sense
[13:33] <sil2100> camako: so... I would need to stop the Mir rebuild that's happening and re-do it from scratch
[13:33] <ricmm> cjwatson: right, so in this case silo 007 was busted (the vm, or whatever was misbehaving)
[13:33] <ricmm> cjwatson: so the landing was migrated to 008
[13:33] <jhodapp> robru, ok ready
[13:33] <ricmm> so that stale deleted 2.0.0 was left in the PPA itself
[13:33] <sil2100> camako: we actually need to assign a different silo for it... is that ok? We can do the build there
[13:34] <ricmm> cjwatson: so do you think we can actually drop safely from -proposed and re-push platform-api with 1.2.0 ?
[13:34] <cjwatson> ricmm: mm, so probably an artifact of the general way that silos are an awful hack then
[13:34] <cjwatson> ricmm: yes
[13:34] <cjwatson> ricmm: it's *possible* that you won't be able to use the same silo to prepare it
[13:34] <ricmm> cjwatson: yup, sparks off of the larger silohack
[13:34] <sil2100> ricmm: yes, cjwatson said it's possible - but we must make sure we won't reuse the invalid name again in the future
[13:34] <ricmm> ok
[13:34] <cjwatson> sil2100: yeah, you won't anyway due to the embedded date
[13:34] <ricmm> I think thats why lukasz is offering to use 000 for the platform-api Mir landing
[13:34] <ricmm> instead of the busted silo
[13:34] <sil2100> Right, true
[13:34] <cjwatson> right, you could copy source+binaries over from the existing one if that's useful
[13:34] <sil2100> camako: are you around?
[13:35] <robru> sil2100, he was earlier, not sure
[13:35] <cjwatson> but I know that this works as far as the primary archive is concerned because <small>we might have had to do it before because I screwed up</small>
[13:35] <robru> cjwatson, what was that? I can't read it
[13:35] <cjwatson> good :)
[13:35] <robru> ;-)
[13:36] <sil2100> camako, kgunn: anyway, I stop the build of silo 008
[13:36] <sil2100> ;)
[13:36] <tedg> Saviq, I'm confused with this GSettings thing, does gsettings only get used with autopilot?
[13:36] <cjwatson> so I've queued that removal, it'll happen with the next publisher run
[13:37] <robru> sil2100, ok so hey, amongst all this crazy papi stuff, do you think it's ok if I publish this tiny bugfix from jhodapp? I'm afraid to publish anything right now, seems the last 2 or 3 times I hit publish something unexpected goes wrong...
[13:37] <cjwatson> https://launchpad.net/ubuntu/+source/platform-api/+publishinghistory
[13:37] <robru> cjwatson, thanks
[13:38] <sil2100> cjwatson: thank you :)
[13:38] <cjwatson> yeah, thinking about it, the wind-versions-backwards hack of course only works with copies
[13:38] <sil2100> I'll assign 19 for that since I see it's clean
[13:39] <sil2100> So we'll use a silo for this
[13:40] <Saviq> tedg, no, that's the thing...
[13:40] <Saviq> tedg, but *something* causes the environment to get screwed up probably
[13:40] <cjwatson> oh, except getSourceAncestry only looks for PUBLISHED/PENDING - oh I'm so confused, anyway this approach will work
[13:40] <Saviq> tedg, like I can see XDG_DATA_DIRS: no such variable in the console
[13:43] <tedg> Saviq, So you think that someone messes up the global environment, and then when unity8 gets restarted it doesn't have the variable.
[13:43] <tedg> ?
[13:43] <cjwatson> I think indeed citrain should be filtering to only Published/Pending publications in most places it calls getPublishedSources, although it has to be careful since unfortunately it isn't possible to pass a list of acceptable statuses, you have to filter the result
[13:43] <sil2100> robru: anyway, let's wait a moment with landing stuff
[13:43] <sil2100> cjwatson: right, I'll do the modification, since now getPublishedSources is used directly, without any filtering
[13:44] <robru> sil2100, figured, that's why I asked ;-)
[13:44] <robru> sil2100, so seeing as kgunn and camako aren't around, we should probably just kick the new build for them, so it's ready to test when they get back. should I do it or are you on it?
[13:44] <kgunn> robru: i'm around
[13:45] <robru> oh cool
[13:45] <sil2100> robru: I'm on it
[13:45] <kgunn> what do you need ?
[13:45] <robru> also cool
[13:45] <sil2100> I was waiting for the spreadsheet to update
[13:45] <sil2100> kgunn: we're cool for now, just know that we flush silo 008 in a moment ;)
[13:45] <kgunn> i know there's some mess up with papi2.0 but hadn't quite been abckle to figure from scroll ba
[13:45] <robru> kgunn, well we moved your landing into silo 19, so you can kick a build, but don't worry, sil is doing it
[13:45]  * ogra_ notes AP-legacy is done with autopkg testing 
[13:45] <kgunn> sil2100: ack
[13:45] <sil2100> Your silo is in 18 now
[13:45] <sil2100> 19
[13:45] <cjwatson> sil2100: there are a few places which pass status="Published", which are probably fine - presumably they're intending to exclude pending publications
[13:45] <sil2100> ogra_: \o/
[13:45] <ogra_> 20?
[13:45] <kgunn> ok
[13:45] <kgunn> 21
[13:45] <robru> sil2100, if you use my backend page, you can kick the build before the spreadsheet updates (no pointless waiting)
[13:45] <ogra_> :)
[13:47] <kgunn> sil2100: do i need to do a full build? or watch build ?
[13:47] <sil2100> kgunn: I did a full build
[13:47] <kgunn> ack...
[13:47] <sil2100> kgunn: ok, quick update:
[13:47] <kgunn> sil2100: bottom line, do we need to retest ?
[13:47] <sil2100> kgunn: so... the mir landing published yesterday was broken, as it was FTBFS for arm64
[13:47] <robru> kgunn, yeah, it will require a retest (needs to retext cjwatson's fix anyway)
[13:47] <sil2100> kgunn: so it got stuck in -proposed
[13:48] <cjwatson> I can confirm my fix actually builds on arm64
[13:48] <kgunn> camako: ^ supposing you already knew
[13:48] <robru> cjwatson, excellent
[13:48] <cjwatson> shouldn't expect it would take much testing, just smoke I guess
[13:48] <kgunn> ack
[13:48] <sil2100> kgunn: it was already rebuilding with a fix, but then ricmm noticed that due to a bug in citrain the platform-api built in your PPA and released fetched the 2.0.0 version number
[13:48] <kgunn> we beat the hell out of it
[13:48] <sil2100> kgunn: while it's not correct
[13:48] <kgunn> we'll focus on n4
[13:48] <cjwatson> since build system only
[13:49] <sil2100> Right
[13:49] <kgunn> yeah...we're still confident
[13:49] <robru> sil2100, build in 19 failed due to versions in -proposed already, had to FORCE_BUILD it
[13:50] <sil2100> robru: no need for the two of us doing the same thing, just wanted to press the button when I saw you doing it ;)
[13:50] <sil2100> We're stepping on our toes!
[13:50] <robru> sil2100, ah sorry, noticed it a while ago and thought you hadn't noticed ;-)
[13:51] <robru> sil2100, can I flush 8 and merge 2? ;-)
[13:51] <sil2100> 2 is merging, you can flush 8 ;)
[13:52] <robru> ok
[13:52]  * sil2100 is still waiting for autopilot-legacy
[13:55] <sil2100> ogra_: it seems to be migrated!
[13:55] <sil2100> ogra_: python-autopilot | 1.4.1+14.10.20140430-0ubuntu4   | utopic/universe  | all
[13:55] <sil2100> ogra_: should we kick a new image, or are we waiting for some other fix?
[13:56] <sil2100> robru: why you can't sleep? ;)
[13:57] <robru> sil2100, massive jetlag. I passed out at around 9PM last night... woke up at 3AM... dunno
[14:01] <tedg> ogra_, I'm not seeing XDG_DATA_DIRS in my upstart global environment, is that not set as part of session init?
[14:02]  * tedg reboots to be sure
[14:04] <alesage> fginther, hiya
[14:06] <fginther> alesage, can we work on getting https://code.launchpad.net/~allanlesage/cupstream2distro-config/enable-webcred-coverage/+merge/219606 merged today? There were just a few comments
[14:07] <alesage> fginther, yep I'm there, will refresh
[14:15] <tedg> sil2100, I think we might have a fix for the unity8 stuff. If you haven't done a build, let's wait.
[14:15] <sil2100> tedg: ogra_ is not around anyway, so yeah - just give us a ping :)
[14:16] <tedg> sil2100, Saviq is preparing a patch.
[14:16] <tedg> It seems to be related to XDG_DATA_DIRS not being set for the session.
[14:17] <ogra_> sil2100, that means another 3h or so ?
[14:18] <sil2100> ogra_: oh, you're back!
[14:18] <ogra_> yep
[14:18] <sil2100> In this case we can build an image now, otherwise we won't get any test results indeed... since unity8 builds and migrates long
[14:18] <ogra_> /etc/X11/Xsession.d/60x11-common_xdg_path
[14:18] <ogra_> thats what sets XDG_DATA_DIRS
[14:19] <ogra_> sil2100, right
[14:19]  * ogra_ triggers one 
[14:19] <sil2100> tedg: so let's target the next image for your fix, thanks for working on it!
[14:19] <ogra_> triggered  ...
[14:20] <ogra_> sil2100, fyi there are ned langpacks building (soon or already) they will show up in the diff
[14:21] <tedg> ogra_, it's not showing up in initctl list-env --global
[14:21] <tedg> Not sure why that is
[14:23] <ogra_> tedg, hmm, confirmed ...
[14:24] <ogra_> tedg, does /usr/share/upstart/sessions/xdg-dirs.conf exist ?
[14:24] <ogra_> seems thats not run then
[14:24] <Saviq> ogra_, truth be told that's what mterry (Xsessions.d parsing) disabled, right?
[14:25] <ogra_> Saviq, no
[14:25] <Saviq> hmm
[14:25] <ogra_> Saviq, only lanuching of dbus from there
[14:25] <ogra_> (at least he should)
[14:25] <mterry> ogra_, no, we don't run that dir anymore I don't believe
[14:25] <tedg> ogra_, Yes it does exist.
[14:25] <Saviq> ogra_, nooo
[14:25] <Saviq> ogra_, https://code.launchpad.net/~mterry/unity8/dbus-x11/+merge/221894
[14:26] <Saviq> but really, my testing would need to have been so broken
[14:26] <mterry> Saviq, that wasn't the one
[14:26] <mterry> ogra_, https://code.launchpad.net/~mterry/ubuntu-touch-session/no-lightdm-session/+merge/221891
[14:26] <ogra_> Saviq, right, that shoulld only disable launching dbus from Xsession ... we still rely on lightdm to process the other scripts from there
[14:26] <Saviq> ah right sorry
[14:27] <ogra_> and we likely still need some of them
[14:27] <Saviq> aand we're back on that train
[14:27] <mterry> We do?
[14:27] <Saviq> mterry, sounds like xdg dirs would be one that we do
[14:28] <ogra_> mterry, not sure ... but i would guess we havent gotten rid of everything from there yet
[14:28] <Saviq> mterry, and yeah, there's more there that sound useful
[14:29] <ogra_> i dont see anything beyond the xdg stuff in my install though
[14:29] <mterry> I thought we set XDG_*_DIRS from an unusually verbose /etc/environment but I see it's missing there, yeah
[14:29] <ogra_> ssh gaent is surely not used atm
[14:30] <mterry> Yeah, I just looked too, and xdg dirs and ssh agent are only two that aren't X-specific or session-setup specific but upstart does that for us
[14:30] <ogra_> right
[14:31] <Saviq> mterry, at least XDG_DATA_DIRS isn't set up by upstart
[14:31] <mterry> Right, I was saying xdg dirs and ssh agent are the two scripts still useful
[14:31] <Saviq> mterry, http://paste.ubuntu.com/7595031/ diff between working and broken unity8 env
[14:31] <ogra_> well, it cretaes a file .config/user-dirs.dirs
[14:31] <ogra_> /usr/share/upstart/sessions/xdg-dirs.conf
[14:31] <ogra_> but that file doesnt seem to get processed anymore
[14:32] <ogra_> (note that even though my name sticks in that upstart job, the code comes from cwayne, not sure what is supposed to process that file and export the vars)
[14:32] <mterry> ogra_, well that file isn't for DATA_DIRS
[14:33] <ogra_> oh, right, thats USER_DIRS
[14:34] <robru> sil2100, hey, mumble hangout time if you're not too busy
[14:34] <ogra_> Saviq, why do yu think that has any effect on the test though ?
[14:35] <Saviq> ogra_, because the test *does* prepend a dir there
[14:35] <Saviq> ogra_, but since it's not in env... it prepends to NULL
[14:35] <mterry> OK.  So what's the best way to set XDG_DATA_DIRS for the new world order?
[14:36] <sil2100> Oh
[14:36] <ogra_> sounds like something ubuntu-touch-session should do perhaps
[14:36] <ogra_> or do you need to use it in the greeter too ?
[14:36] <mterry> Ah, that's not a bad spot
[14:36] <sil2100> robru: coming soon
[14:36] <ogra_> then the greeter wrapper
[14:36] <mterry> ogra_, greeter uses ubuntu-touch-session too, but I don't think we need XDG_DATA_DIRS there anywhere
[14:37] <ogra_> so on desktop i have: XDG_DATA_DIRS=/usr/share/ubuntu:/usr/share/gnome:/usr/local/share/:/usr/share/
[14:38] <mterry> We probably don't need gnome there
[14:38] <ogra_> which means we could effectively just point to /usr/local/share/:/usr/share/ ?
[14:39] <ogra_> neither ubuntu i guess
[14:39] <mterry> Why not ubuntu?
[14:39] <mterry> And maybe we want ubuntu-touch in there, to simulate what gnome was doing there (using the xsession identifier)
[14:39] <Saviq> mterry, ogra_, we *can* default to it in the unity8 suite
[14:39] <ogra_> ah, i didnt notice it is actually existing on the phone
[14:39] <ogra_> i thought it comes from the ubuntu-session
[14:39] <Saviq> tedg, so, in any case, seems I completely misblamed you for the situation
[14:40] <mterry> ogra_, so /usr/local/share/:/usr/share/ would just be the default value anyway
[14:40] <Saviq> tedg, and sorry for that, I really don't understand how... I rebooted, worked, upgraded your stuff, died
[14:40] <ogra_> right, i 'm not sure if we need to use anything thats in /usr/share/ubuntu/settings/system/
[14:40] <mterry> ogra_, Saviq: so which thing is breaking due to the var not being defined?
[14:40] <Laney> that file also sets XDG_CONFIG_DIRS
[14:41] <ogra_> probably the settings app does
[14:41] <Saviq> mterry, unity8 ap dies
[14:41] <mterry> ogra_, but I don't think it gets there via XDG_DATA_DIRS
[14:41] <Saviq> mterry, and anything else that wants to append to XDG_DATA_DIRS
[14:41] <Laney> It might be easiest to just copy its logic to ease future migration pain
[14:41] <Saviq> mterry, 'cause it defaults to /usr/share or so, but not if you actually set it in env
[14:41] <Saviq> mterry, in which case only the ones in that var are used
[14:41] <mterry> Saviq, hrm.  Because they are appending poorly, but I suppose we could help them out by defining the var to its default for them
[14:41] <Saviq> without /usr/share fallback
[14:42] <Saviq> mterry, well, sure, appending to NULL
[14:42] <mterry> Saviq, but they should be providing the default when appending
[14:42] <Saviq> mterry, should (could) default to /usr/share:/usr/local/share
[14:42] <mterry> But no one does, because it's a pain
[14:42] <tedg> Saviq, No problem, there's a lot of change mixing in the pot right now. Hard to separate it out.
[14:42] <Saviq> mterry, sure
[14:42] <Saviq> tedg, indeed
[14:43] <mterry> OK, so something like this in ubuntu-touch-session:
[14:43] <mterry> if [ -z "$XDG_DATA_DIRS" ]; then
[14:43] <mterry>     export XDG_DATA_DIRS=/usr/share/ubuntu:/usr/local/share:/usr/share
[14:43] <mterry> fi
[14:43] <mterry> ?
[14:43] <ogra_> ++
[14:43] <ogra_> well, Laney mentioned XDG_CONFIG_DIRS
[14:43] <ogra_> XDG_CONFIG_DIRS=/etc/xdg/xdg-ubuntu:/usr/share/upstart/xdg:/etc/xdg
[14:43] <ogra_> thats what i have here
[14:43] <Laney> why don't you make it variable like it is currently?
[14:44] <tedg> ogra_, I'm worried about the upgrade testing not seeing autopilot-legacy. I guess that's what we get fixed with CI Airline doing images instead of PPAs though.
[14:44] <ogra_> tedg, yeah ... thats the plan
[14:45] <ogra_> tedg, in fact we want to have the emulator run the full test suite for every landing and give the uploader/commiter the resulting report
[14:45]  * tedg hopes we don't get a CI TSA along with that plan
[14:45] <ogra_> on such an image
[14:45] <tedg> That'd be awesome.
[14:46] <tedg> I imagine that's phase n+1 though
[14:46] <ogra_> yeah
[14:49] <alesage> fginther, that MP is ready for your review again, thanks
[14:50] <alesage> fginther, https://code.launchpad.net/~allanlesage/cupstream2distro-config/enable-webcred-coverage/+merge/219606
[14:55] <fginther> alesage, approved
[14:55] <alesage> fginther, 10Q
[14:57] <oSoMoN> sil2100, hey, I’d need landing request on line 33 to take precedence over the one on line 19, which is currently assigned silo 10 (it’s ok to reassign the silo)
[14:58] <sil2100> oSoMoN: hey! Let me look at that after the meeting
[14:58] <oSoMoN> thanks!
[15:12] <mterry> Sorry, got pulled into a hangout
[15:12] <mterry> Back on XDG_*_DIRS
[15:20] <mterry> Saviq, ogra_, Laney: https://code.launchpad.net/~mterry/ubuntu-touch-session/xdg-dirs/+merge/222213
[15:22] <Laney> mterry: how come the special treatment of 'ubuntu'?
[15:23] <mterry> Laney, my desktop session has it...  I don't think anything is put there though
[15:23] <mterry> Laney, I know you system-settings folk put files in that dir.  Do you need the dir in DATA_DIRS at all?
[15:24] <Laney> mterry: That's just a coincidence, it's not meant to be a data dir there
[15:24] <ogra_> Laney, mterry, there is stuff in it on the phone too
[15:24] <Laney> your desktop has it because of $DESKTOP_SESSION
[15:24] <ogra_> (not sure it is used at all though)
[15:24] <Laney> it's been a wishlist/low priority item for me to rename that directory to /usr/share/ubuntu-system-settings/
[15:25] <ogra_> /usr/share/ubuntu/settings/system/ has a lot of stuff
[15:25] <mterry> Laney, oh I was confused by the gnome presense in my XDG_DATA_DIRS, thought that was the DESKTOP_SESSION
[15:25] <mterry> Laney, but I guess that's from gnome-session?  which we still must use
[15:25] <mterry> Laney, ok, will drop from branch
[15:25] <Laney> right
[15:26] <Laney> we probably do want /usr/share/gnome on a gnome session
[15:26] <mterry> But gnome-session will do that itself
[15:26] <mterry> branch updated
[15:27] <Laney> fair enough
[15:28] <jhodapp> sil2100, any idea when you'll be able to land simple things like silo 18 again?
[15:28] <tedg> I'm a bit confused. This still lists autopilot: $ apt-cache rdepends gir1.2-upstart-app-launch-2
[15:28] <sil2100> jhodapp: hi! I'll land those in a moment :)
[15:28] <sil2100> jhodapp: we're just having a meeting right now
[15:28] <jhodapp> sil2100, excellent, that's great news
[15:28] <sil2100> We kicked a new image so we should be landable now
[15:28] <jhodapp> nice
[15:29] <robru> retoaded, hey, I'm getting some SMS messages... not sure what's up, my vanguard shift shouldn't start for 3.5 more hours, also I don't understand these messages at all
[15:30] <mterry> Saviq, so this was causing autopilot failures for unity8 or for everything?
[15:30] <ogra_> robru, it wants to make friends with you
[15:31] <robru> ogra_, yeah something about manta-01 being down. gosh I hope that comes back...
[15:31] <ogra_> (and probably didnt want its master to know)
[15:32] <ogra_> mterry, top approved
[15:32] <ogra_> (and thanks for the descriptive commit message :) )
[15:33] <mterry> ogra_, heh, I would normally just do the first sentence and leave the second just in the merge description.  I was swayed by your words
[15:34] <mterry> ogra_, so I guess we need a silo for this guy?
[15:34] <ogra_> yeah
[15:34] <ogra_> sil2100, ^^^
[15:36] <sil2100> o/
[15:36] <sil2100> Is there a landing prepared?
[15:37] <sil2100> ogra_, mterry: so, the ubuntu-touch-session change will fix our unity8 test failures?
[15:38] <ogra_> if i can belive Saviq it will :)
[15:38] <mterry> Yeah I think so
[15:38] <sil2100> Awesome
[15:38] <sil2100> mterry: you want to prepare a landing line for that?
[15:38] <mterry> working on it
[15:38] <Saviq> ogra_, I expect a few failures that crept in in the mean time
[15:39] <ogra_> Saviq, as long as its not all red we'Re fine i think
[15:39] <sil2100> oSoMoN: looking into your request now
[15:39] <Saviq> ogra_, +1
[15:40] <sil2100> oSoMoN: ok, so I flush silo 10 and assign a new one for 33 then
[15:40] <sil2100> oSoMoN: is that ok with you?
[15:40] <Saviq> ogra_, fyi bug #1325984 is what I expect to still kick us
[15:41] <mterry> sil2100, ready
[15:41] <sil2100> mterry: o/
[15:41] <mterry> Saviq, sil2100, ogra_: sorry about that.  My fault for not running the tests, but I didn't expect that change to affect them
[15:41] <Saviq> hum hum
[15:42] <ogra_> Saviq, qmlscene behaves very oddly if QT_SELECT (or however that is called) has the wrong value ...
[15:42] <Saviq> mterry, ogra_, wait for it, my phone just rebooted on start after applying that patch...
[15:42] <mterry> rebooted?
[15:42] <Saviq> mterry, yeah, failed to start
[15:42] <mterry> Ah
[15:42] <mterry> Huh
[15:43] <sil2100> ricmm: so... we have an urgent fix for our breakages in ubuntu-touch-session, so be sure to rebuild that one once you are finally free in landing papi
[15:44] <mterry> I'm not sure how we could screw up boot with this patc
[15:45] <ogra_> hmm, quoting issue ?
[15:45] <ogra_> but even then it shouldnt reboot
[15:45] <ogra_> the session/greeter should just not start
[15:46] <Saviq> mterry, ogra_, is fine now
[15:46] <ogra_> k
[15:46] <ogra_> heisenbug then
[15:46] <mterry> Building silo
[15:46] <Saviq> yeah, Schroedinger's cat
[15:49] <oSoMoN> sil2100, perfect
[15:50] <oSoMoN> sil2100, you can reuse silo 10 for the request on line 33 if that’s easier for you
[15:50] <sil2100> oSoMoN: I already freed up 10 and assigned a different silo - 008
[15:54] <oSoMoN> that works too, thanks!
[15:55] <imgbot> [15:55] <imgbot> [15:55] <robru> oSoMoN, silo 8, so work, much build! wow
[15:55] <sil2100> ;)
[15:56] <Chipaca> sil2100: silo 14 ready for landing \o/
[15:57] <sil2100> Chipaca: phew, you made it right before the e-mail where I announce TRAINCON-0 ;)
[15:57] <Chipaca> :(
[15:57] <brendand> messaging-app is taking a really long time to start
[15:58] <brendand> i installed silo 2 earlier
[15:59] <sil2100> Chipaca: ok, so this needs a packaging ACK from ogra_, but I already see he probably won't +1 it because of the changelog :(
[16:01] <Chipaca> is there a tool to build debian-style changelogs from bzr commits?
[16:03] <Chipaca> like git-debian-changelog but for bzr
[16:07] <robru> Chipaca, just import the bzr branch into git ;-)
[16:07] <robru> Chipaca, kidding, i'm sure that tool exists, but I don't know where to find it, sorry
[16:08] <Chipaca> bzr merge-upstream looks like it wants to do that but fails
[16:08] <Chipaca> maybe because i need to use that always for it to work
[16:09] <Chipaca> not sure
[16:17] <sil2100> Chipaca: ok, so as long as you remember to keep your changelogs/commit messages informative then I guess we can publish :)
[16:17] <sil2100> Let me just give it to ogra_
[16:17] <sil2100> ogra_: https://ci-train.ubuntu.com/job/landing-014-2-publish/25/artifact/packaging_changes_ubuntu-push_0.3+14.10.20140605-0ubuntu1.diff
[16:17] <sil2100> That's the packaging change, cmake for the win!
[16:22] <ogra_> Chipaca, thats the last time i'll let a "new upstream release" changelog through from you ...
[16:23] <ogra_> sil2100, ACK
[16:23] <sil2100> ogra_: thanks!
[16:31] <Chipaca> ogra_: ack. Thanks.
[16:33] <ogra_> sil2100, gallery-app exploded on 68
[16:33] <sil2100> How?
[16:33] <ogra_> dunno ... 15 failures
[16:34] <ogra_> Photo with image name /home/phablet/Pictures/sample04.jpg could not be found
[16:34] <sil2100> Well, let's fill in a bug and escalate! ;)
[16:35]  * sil2100 doesn't feel like digging into that instead of upstreams anymore today
[16:36] <ogra_> there are four that are not related to the missing photo
[16:36] <ogra_> the rest is though
[16:36] <ogra_> right
[16:36] <sil2100> I wonder if it's related to content-hub
[16:37] <sil2100> We had a landing with that today
[16:37] <ogra_> 68 isnt looking so great though
[16:37] <sil2100> kenvandine: ^
[16:37] <ogra_> smells like it will be worse than the former ones
[16:38] <pmcgowan> http://people.canonical.com/~ogra/touch-image-stats/68.changes doesnt show much
[16:38] <pmcgowan> applaunch stuff
[16:38]  * sil2100 generates commit logs
[16:38] <ogra_> pmcgowan, nope, there were autopilot-legacy fixes related to the upstart-app-launch renaming
[16:38] <ogra_> that were hoped to fix a bunch of issues
[16:38] <kenvandine> sil2100, ?
[16:39] <ogra_> and unity8 is still waiting for a landing
[16:39] <pmcgowan> but nothing to break gallery itself, and no content hub landed
[16:39] <sil2100> Actually, kenvandine's landing landed after 68
[16:39] <sil2100> tedg: could your click-app-confinement break anything?
[16:39] <sil2100> tedg: just trying to find anything that could have caused it
[16:39] <kenvandine> it was broken with that fix :)
[16:40] <ogra_> pmcgowan, the whole test suite is still untested against the new UAL renaming
[16:40] <pmcgowan> got it
[16:40] <kenvandine> whoops
[16:40] <sil2100> http://people.canonical.com/~lzemczak/landing-team/68.commitlog
[16:40] <kenvandine> s/with/without
[16:40] <ogra_> its all quite a mess
[16:41] <tedg> sil2100, Uhm, it would be confined instead of unconfined. But the gallery confinement should let it see Pictures.
[16:41] <ogra_> sadly we dont get proper logs until the whole test suite has run
[16:42] <tedg> sil2100, So the old version could have let a bug through, but I don't think it's changed.
[16:42] <sil2100> tedg: could you and kenvandine maybe look into that? Since with 68 only this basically landed, so I would suspect this to be a problem, or some other already existing issue
[16:42] <sil2100> Just want to make sure it's being investigated by people with experience :)
[16:42] <tedg> sil2100, Sure, let me flash 68
[16:42] <sil2100> Thanks!
[16:43]  * ogra_ would like to see some info about AP-legacy ... 
[16:43] <om26er> Dear people we are here for helping you get through these tough times of TRAINCON-0 - yours QA team. ;-)
[16:43] <ogra_> but without console log ...
[16:43]  * ogra_ hugs om26er 
[16:43]  * om26er hugs ogra_ back
[16:48] <sil2100> o/
[16:48] <sil2100> om26er: thanks :)
[16:49] <ogra_> sil2100, oh, if you are away, who will do silos tomorrow ?
[16:49] <sil2100> ogra_: Mirv  :)
[16:49]  * ogra_ has no clue how to assign them etc ...
[16:49] <ogra_> PHEW !
[16:50] <ogra_> thanks :)
[16:50] <sil2100> ogra_: I already asked him and he said he'll be around for silo management in our timezone
[16:50] <sil2100> And he's a veteran
[16:50] <ogra_> sil2100, calendar looks pretty bad too btw
[16:50]  * tedg needs photos to test with, everyone smile!
[16:50] <sil2100> Again all failed...
[16:50] <ogra_> i guess this will be much worse than 67
[16:51] <sil2100> ogra_: so it seems it wasn't autopilot py2's fault it seems?
[16:51] <ogra_> probably not
[16:51] <sil2100> Or wait
[16:51] <sil2100> hm
[16:51] <sil2100> Why do I see the same message again?
[16:52] <ogra_> 16:36:31.206 INFO _launcher:116 - Attempting to launch application 'com.ubuntu.calendar_calendar_0.4.296' with URIs '' via upstart-app-launch
[16:52] <ogra_> 16:36:44.219 ERROR proxies:410 - Introspect error on :1.55:/com/canonical/Autopilot/Introspection: dbus.exceptions.DBusException:
[16:52] <ogra_> org.freedesktop.DBus.Error.NoReply: Message did not receive a reply (timeout by message bus)
[16:52] <sil2100> Ah, ok, I didn't rename the text
[16:52] <sil2100> Maybe my branch was invalid
[16:52] <sil2100> tedg: did you see anything wrong with the debdiff I sent you?
[16:53] <tedg> sil2100, I didn't notice anything, do you have a branch that I can grep?
[16:54] <sil2100> tedg: sadly, since I didn't know which branch to use as I saw many direct-uploads in the archive
[16:54] <sil2100> So I wasn't sure if those got synced
[16:54] <sil2100> (so I used the sources from LP)
[16:56] <tedg> Okay, let me see if I can find this gallery issue.
[16:56] <tedg> Seems it can get the image, but not the thumbnail.
[16:56] <sil2100> oh
[17:07] <ogra_> :)
[17:08] <slangasek> sil2100: why was autopilot-legacy an issue for the landings?  Aren't all of the phone tests migrated to python3 already?
[17:09] <sil2100> slangasek: yeah, so...
[17:09] <sil2100> slangasek: from what we saw, some tests still used the old python2 autopilot, that's first
[17:09] <slangasek> sil2100: which tests are doing that?
[17:09] <sil2100> slangasek: the other problem was that unity8-autopilot depends on both python-autopilot and python3-autopilot, so by fetching the python2 AP dependency it was pulling in the old upstart-app-launch bits
[17:10] <sil2100> slangasek: let me look, one moment
[17:10] <slangasek> ah; is that unity8-autopilot for the desktop?
[17:11] <sil2100> No
[17:11] <sil2100> It's the touch unity8 autopilot package
[17:11] <slangasek> hmm, ok
[17:11] <sil2100> For instance, from the logs we saw that even ubuntuuitoolkit used the old AP
[17:12] <sil2100> slangasek: same for messaging-app I saw
[17:12] <sil2100> Not sure if that's wanted behavior or what... really didn't have time to do that investigation
[17:12] <ogra_> slangasek, unity8 pulls it in ... its not clear if they are used at all
[17:13] <ogra_> slangasek, the issue was the dependency that caused parts of the UAL renamed packages ot be reverted on the test image
[17:14] <ogra_> (i know that some are said to still use legacy ... but what mainly bit us was that AP-legacy brought uupstart-app-launch back)
[17:14] <slangasek> sil2100: right; this is reflected on https://blueprints.launchpad.net/ubuntu/+spec/core-1311-python3-roadmap as "debs needing porting to python3 but not blocking python2 removal from the image"
[17:15] <oSoMoN> om26er, hey, I’d need a QA signoff for landing request on line 33, since we’re in traincon0
[17:15] <slangasek> sil2100: (sorry, I should have remembered this)
[17:15] <oSoMoN> om26er, is that something you could do for me?
[17:15] <om26er> oSoMoN, yes, let me look
[17:16] <oSoMoN> awesome, thanks!
[17:16] <om26er> oSoMoN, does it require opening youtube.com on the device ?
[17:16] <om26er> asking since its blocked here
[17:16]  * sil2100 needs to finish off a script before going EOD
[17:16] <oSoMoN> om26er, to verify the fix, it does, I tested it myself here if it’s of any help
[17:17] <oSoMoN> (I suppose that’s not how a QA signoff works though)
[17:17] <om26er> ToyKeeper, do you think you can help in this case ?
[17:24] <ToyKeeper> oSoMoN: Can you add the relevant tests to the test plan?  It has no tests for the changes in the silo.
[17:26] <ToyKeeper> oSoMoN: It needs either a manual or automatic test to verify the correct behavior.
[17:27] <mterry> sil2100, do I need to do anything else for silo 002?
[17:29] <ogra_> so at least music anf filemanager got better
[17:29] <robru> mterry, QA signoff I guess
[17:30] <mterry> Oh right, traincon
[17:30] <robru> mterry, yep ;-)
[17:30] <mterry> :(
[17:30]  * mterry feels bad for contributing to traincon
[17:30] <oSoMoN> ToyKeeper, will do in a moment, thanks
[17:31] <robru> mterry, nah, the alternative is that your work would just bitrot forever in a silo
[17:31] <ToyKeeper> oSoMoN: It looks like only one of the four MPs has any tests, so the other three need them.
[17:32]  * mterry will offer sacrifices to the St. Christopher, patron saint of travelers (closest I could get to a train)
[17:47] <cjwatson> ToyKeeper: I need to land https://code.launchpad.net/~click-hackers/click/devel/+merge/222243 (getting our team's lander to do it for me), which I guess will need QA signoff.  Fixes a critical bug; checking that uninstall works is in click's test plan (https://wiki.ubuntu.com/Process/Merges/TestPlan/click).
[17:48] <ToyKeeper> om26er_: Can you help?
[17:48] <ToyKeeper> cjwatson: Is the change already in a silo somewhere?
[17:49] <cjwatson> Not yet, I've asked my team's lander to get it into the process
[17:49] <cjwatson> Need to go for dinner now, just thought it might save time to ask in advance
[17:49] <om26er_> ToyKeeper, cjwatson I can help with that
[17:49] <cjwatson> stgraber: ^-
[17:50] <oSoMoN> ToyKeeper, added test plan to the landing request
[17:53] <ToyKeeper> oSoMoN: Could you get it into the actual test plan?  It doesn't really matter unless it's documented for future testing (and preferably automated).
[17:53] <oSoMoN> ok, will do
[17:58] <oSoMoN> ToyKeeper, ok, I updated the test plan at https://wiki.ubuntu.com/Process/Merges/TestPlan/webbrowser-app
[18:00] <ToyKeeper> oSoMoN: Could someone use a site which doesn't require an account, instead of twitter?
[18:00] <ToyKeeper> Maybe tinypic or imgur or something?
[18:02] <ogra_> QA: could you guys treat silo 002 with priority ? that is supposed to get the image less red http://ci.ubuntu.com/smokeng/utopic/touch/mako/68:20140605.2:20140530/8424/ ...
[18:02] <ogra_> (for details talk to mterry but by feedback from tedg that seems ot be the roed out of traincon-0 )
[18:02] <ogra_> *to be the road
[18:02]  * ogra_ cant type anymore 
[18:05] <sil2100> Hey
[18:05] <sil2100> Shouldn't 002 go without a queue?
[18:05] <sil2100> It's an isolated fix right?
[18:06] <sil2100> ogra_, robru: remember that isolated bug-fix-only landings can go without QA sign-off
[18:06] <robru> oh, well in that case
[18:07] <robru> mterry, silo 2 published
[18:07] <mterry> yay
[18:07] <oSoMoN> ToyKeeper, yes, probably
[18:09] <robru> stgraber, you got silo 10, please build
[18:10] <ToyKeeper> ogra_: Yes, if it's related to fixing a blocker and it's an isolated change, it can bypass sign-off.
[18:10] <ogra_> mterry, ^^
[18:11] <oSoMoN> ToyKeeper, neither tinypic nor imgur will do, as they don’t seem to restrict the accepted mimetypes…
[18:11] <ToyKeeper> Traincon 0 only holds back the other changes, features, large branches, anything not related to fixing blockers.
[18:12] <ogra_> we need traincon training to be more trained ;)
[18:12] <oSoMoN> ToyKeeper, so unless you can suggest another login-free site, we’ll have to keep twitter on the test plan
[18:12] <ToyKeeper> oSoMoN: Okay, was just hoping for something which doesn't require an account, since that's a lot easier for random people and scripts to test.
[18:12] <ogra_> how is twitter login free ?
[18:13] <oSoMoN> ToyKeeper, agreed, if you come across such a site let me know so I can update the test plan
[18:13] <oSoMoN> ogra_, it’s not
[18:13] <ogra_> yeah
[18:13] <robru> oSoMoN, what kind of site? you can upload photos to imgur without logging in
[18:13] <oSoMoN> robru, yes, but they don’t seem to restrict the accepted mimetypes, which is what we’re interested in testing
[18:14] <robru> ah
[18:16] <robru> ok, well I'm off for lunch, happy to publish anything when i get back
[18:16] <ToyKeeper> I suppose I'll need to finally make a twitter account.
[18:19] <oSoMoN> ToyKeeper, found one: http://www.wufoo.com/html5/attributes/07-accept.html
[18:20] <oSoMoN> ToyKeeper, having a test twitter account around wouldn’t hurt though (I don’t use twitter myself but I have a test account for this sort of things)
[18:20] <oSoMoN> I’ll update the test plan
[18:21] <oSoMoN> done
[18:24] <ToyKeeper> ... testing
[18:25] <stgraber> cjwatson: things are building at https://launchpad.net/~ci-train-ppa-service/+archive/landing-010/+packages
[18:29] <fginther> mardy, alesage, the changes to collect coverage have been deployed. Future builds should now work.
[18:33] <alesage> fginther, thanks
[18:34] <alesage> fginther, when's your next vanguard btw? have to get a traunch of these ready for review
[18:34] <fginther> alesage, Tuesdays and Wednesday now
[18:35] <alesage> fginther, supah, will ask about next Tues.
[18:37] <ToyKeeper> Well.  Found a new bug in the browser...  not sure yet how wide the impact is.
[18:48] <ToyKeeper> Anyone know if this is a known issue in oxide?  http://toykeeper.net/tmp/browser-rendering-errors-while-scrolling.2.png
[18:48] <ogra_> never seen that
[18:48] <ToyKeeper> So far, I only see it in the uReadIt app.
[18:49] <ogra_> i have seen other funny things like the screen gettng squeezed for a split second etc
[18:49] <ogra_> nothing i can really put a hand on
[18:49] <ogra_> but font rendering was always very nice hhere
[18:50] <ToyKeeper> It's not just fonts.  Images behaved the same.
[18:50] <ToyKeeper> http://toykeeper.net/tmp/browser-rendering-errors-while-scrolling.png
[18:50] <sil2100> o/
[18:50] <ogra_> wow
[18:51] <ogra_> definitely never happened to me ... though i only use promoted images on the phone ...
[18:51] <ogra_> (i use my flo for the unstable stuff)
[18:51] <ToyKeeper> This is image 68, unmodified.  Was checking baseline behavior before testing a silo.
[18:53] <ogra_> i have 67 on my flo and the G+ app behaves just fine
[18:53]  * ogra_ upgrades to 68
[18:53] <ToyKeeper> I see this in uReadIt, and I don't see it in the xkcd webapp.  Haven't tried others yet.
[18:54] <ogra_> i'll try ureadit then ... probably an app bug though
[18:54] <ogra_> note that we have two browser frameworks installed ... could well be that ureadit still uses webkit
[18:56] <ogra_> mhall119, does ureadit use oxide or webkit ?
[18:57] <ToyKeeper> The slashdot app has no issues either.
[18:58] <ogra_> ureadit behaves fine on flo so far ...
[19:00] <ogra_> ToyKeeper, ha ! got it now as well on flo
[19:00] <mhall119> ogra_: oxide, but it's not great
[19:00] <ToyKeeper> mhall119: It's doing this: http://toykeeper.net/tmp/browser-rendering-errors-while-scrolling.2.png
[19:00] <ogra_> yeah, it seems to have issues i see in no other webapp
[19:01] <mhall119> ToyKeeper: all the time, yeah, it's really annoying
[19:01] <ToyKeeper> I was trying to figure out where the bug should go...  what component is responsible.
[19:01] <mhall119> ogra_: it's not a webapp, it's a QML app with an UbuntuWebView component inside a Page on a PageStack
[19:02] <ToyKeeper> mhall119: Known bug then?  Got a link?
[19:02] <mhall119> UbuntuWebView is responsible, I think it's part of the webbrowser-app project still, oSoMoN ^^
[19:02] <mhall119> ToyKeeper: I haven't filed it yet, no
[19:02] <mhall119> glad to know somebody besides me is still using uReadIt though :)
[19:03] <ToyKeeper> mhall119: It's in the webbrowser-app test plan: https://wiki.ubuntu.com/Process/Merges/TestPlan/webbrowser-app
[19:04] <oSoMoN> mhall119, ToyKeeper: the rendering issues would be an oxide bug, can you please file it there: https://bugs.launchpad.net/oxide/+filebug
[19:04] <ToyKeeper> oSoMoN: Yes, and that's where I was about to file it before I heard it might be webkit or something else.  :)
[19:05] <robru> cjohnston, hey vanguard buddy, is it time for my vanguard shift now?
[19:05] <mhall119> thanks ToyKeeper
[19:05] <oSoMoN> whether it’s webkit or oxide doing the rendering depends on whether version 0.1 or version 0.2 of Ubuntu.Components.Extras.Browser is being imported
[19:05] <mhall119> it's 0.2 now
[19:05] <oSoMoN> in the case of ureadit, it’s version 0.2, hence oxide
[19:06] <cjohnston> robru: yessir
[19:07] <robru> cjohnston, so, uh, hey, wanna give me a crash course in vanguarding? this is my first one
[19:08] <cjohnston> robru: do you know where the docs are and do you have pager duty setup?
[19:08] <tedg> K, so a reflash of 68 + the touch-session fix make clock/gallery/etc work for me in casual testing.
[19:09] <robru> cjohnston, I have pager duty set up sort of, it texted me a couple things at 8am this morning, I didn't understand them. also I don't have the doc handy
[19:09] <cjohnston> robru: so to understand the notifications you really need to go into PD..
[19:10] <ToyKeeper> mhall119, oSoMoN: https://bugs.launchpad.net/oxide/+bug/1326924
[19:10] <robru> cjohnston, yeah I was looking at those
[19:10] <cjohnston> if you look over the doc, that's probably a really good start...
[19:10] <robru> cjohnston, great, it's big ;-)
[19:11] <cjohnston> There are too many things that could happen to really give a 'crash course'... the doc is probably the best bet.. and then as you see stuff, investigate, read the doc, ask questions of others... etc... and please, since you are a fresh set of eyes, if you run into something that is missing from the docs, add it in
[19:12] <robru> cjohnston, oh, good call. ok will do
[19:13] <cjohnston> Someone could spend an hour telling you things, and then the way things work out you never see those issues.. heh
[19:13] <oSoMoN> ToyKeeper, thanks for the bug, how’s the testing going?
[19:15] <tedg> With silo 10 I can uninstall apps.
[19:16] <tedg> alecu, We should probably put quotes around the app name. I got "Uninstall YES or NO?" and then the buttons were Cancel/Uninstall. But "YES or NO" is the name of the app :-)
[19:18] <robru> tedg, brb, making an app called "Your mom"
[19:19] <robru> then another one called "the Internet"
[19:19] <ogra_> robru, FYI starting an image build so we get results for the ubuntu-touch-session change ...
[19:20] <robru> ogra_, excellent
[19:20] <tedg> Heh, good, good.
[19:20] <tedg> robru, Fill our app store with value! :-)
[19:20] <robru> tedg, I live to give! ;-)
[19:21] <ogra_> robru, err, or not ... it vanished
[19:21] <robru> ogra_, https://launchpad.net/ubuntu/+source/ubuntu-touch-session/0.108+14.10.20140605-0ubuntu1 what vanished?
[19:22] <ogra_> robru, weird ... rmadison didnt have it for a while
[19:22] <robru> ogra_, glitch in the matrix... looks fine to me even in rmadison. builds away!
[19:23] <ogra_> yeah, looks fine now
[19:23] <Saviq> ogra_, "<CI-SNCF> mterry, , saviq (landing-002): Migration: All packages are in destination. You can Merge and Clean now. (https://ci-train.ubuntu.com/job/check-publication-migration/19455/console)"
[19:23] <Saviq> oof
[19:24] <Saviq> ogra_, you scared me there!
[19:24] <ogra_> Saviq, yeah, i guess it uses rmadison
[19:24]  * Saviq has no idea who rmadison is, but feels for him/her
[19:24] <ogra_> Saviq, http://paste.ubuntu.com/7596770/
[19:25] <robru> ogra_, what does? nothing in citrain does (citrain polls lp for the package status, which has known limitations regarding when packages actually hit the archive)
[19:25] <ogra_> robru, how/where does it poll the package status ?
[19:25] <ogra_> rmadison is the only reliable source
[19:26] <ogra_> the LP UI is currently not in sync with the publisher ...
[19:27] <ogra_> not sure about lplib ... but i guess it uses the same DB
[19:27] <robru> ogra_, you're right, rmadison is the only reliable source, and citrain is emphatically not using that. it's caused us many problems.
[19:28] <ogra_> well, colin mentioned plans to improve the LP side
[19:28] <robru> that would be nice
[19:28] <ogra_> so we should hopefully see that becoming better soon
[19:29] <ogra_> image building btw ...
[19:29] <ogra_> bot should announce it soon
[19:30] <imgbot> [19:31] <ogra_> on that note ...
[19:31]  * ogra_ &
[19:44] <robru> OGRA AND.... OGRA AND... what?!?!
[19:48] <ToyKeeper> He daemonized.  He's daemonic now.
[19:49] <ToyKeeper> He's hiding in the background, lurking in the shadows.
[20:01] <robru> hide yo kids, hide yo wife
[20:09] <cjwatson> stgraber: thanks
[20:23] <om26er> cjwatson, regarding the click silo its status still says its not tested
[20:24] <om26er> it first needs to be tested by the dev team and then we (QA) test it
[20:28] <cjwatson> om26er: yeah I know, but I'm about to go out for the evening
[20:28] <cjwatson> om26er: so it'll be a few hours
[20:51] <tedg> om26er, cjwatson, not sure that I count for that approval, but it does work for me.
[20:56] <om26er> tedg, if you have tested you can formally change the status of 'Testing Done' in landing-010 tab ;)
[20:59] <tedg> om26er, I think it makes sense to wait for cjwatson at this point considering it won't get into an image anyway.
[21:00] <tedg> I really want to see how 69 turns out, hoping it'll focus our efforts.
[21:05] <tedg> Cool, music-app fixed in 68. http://ci.ubuntu.com/smokeng/utopic/touch/mako/68:20140605.2:20140530/8424/music_app/
[21:08] <tedg> Looks like the biggest regression in 68 was the clock app, which should be fixed in 69.
[21:12] <om26er> cjwatson, I gotta sleep now, in my absence ToyKeeper is your QA help
[21:12]  * om26er outy
[21:15] <imgbot> [21:15] <imgbot> [21:15] <tedg> Thanks imgbot!
[21:16] <robru> can't wait for the smokeng on this one
[21:16]  * tedg can't wait, downloads ;-)
[21:24]  * popey updates devices
[21:37] <ToyKeeper> Hmm, now my device isn't responding to screen input.  (image 68 still)
[21:38] <ToyKeeper> Well, not entirely true.  If the screen starts to dim from inactivity and I touch it, it'll go bright again.
[21:39] <ToyKeeper> Otherwise, no response to screen inpt.
[21:39] <ToyKeeper> input even.
[21:48] <tedg> 69 seems to behave well. Hoping the automated tests show the same.
[21:52] <robru> tedg, results are still coming in, but so far mediaplayer-app is showing improvements in the smokeng
[21:53] <tedg> robru, Nothing has failed yet, it's perfect!
[21:53] <tedg> :-)
[21:53] <robru> quick, RTM!
[21:54] <tedg> It does seem to be running warmer than I remember.
[21:54] <tedg> Not sure if it's just summer in Texas, or really an issue.
[21:55] <t1mp> robru: :)
[22:02] <fginther> balloons, the utopic tests have been re-enabled for the coreapps jenkins
[22:03] <balloons> fginther, all sorted then?
[22:04] <fginther> balloons, yep, I had to rebuild the test host, but it all appears to be working again
[22:04] <balloons> fginther, awesome thanks
[22:06] <ToyKeeper> Shouldn't we be at least a little bit worried about /sbin/init crashing before the user even has a chance to log in?
[22:06] <ToyKeeper> That has been happening for a while...
[22:09] <tedg> ToyKeeper, Not seeing any crash files for that, do you have one?
[22:10] <tedg> Oh, wait, I have one now.
[22:11] <ToyKeeper> Every time I flash, I get a /sbin/init crash before the device even gets to a welcome screen.
[22:11] <ToyKeeper> I also get a mediascanner crash soon afterward, but I think that's because I push a bunch of files to the device for testing after flashing.
[22:11] <ToyKeeper> (a bug, but one triggered by my own actions)
[22:12] <tedg> Uploading mine to daisy
[22:13] <tedg> Make sure the Upstart folks can get a retrace.
[22:20] <tedg> ToyKeeper, Seems to be this error: https://errors.ubuntu.com/problem/1b58afecf2814f03d3953923cfe361887edf91c0
[22:21] <tedg> ToyKeeper, So looks like we need an Upstart release to fix it, as it's already in trunk.
[22:22] <ToyKeeper> tedg: I can't view that page.
[22:24] <tedg> ToyKeeper, Hmm, that's odd. Anyway, the LP bug is bug 1222705.
[23:00] <veebers> tedg: you still around perchance?
[23:04] <veebers> tedg: nvm sorry, my issue has sorted itself out ;-)
[23:15] <cjwatson> ToyKeeper,tedg: Right, tested silo 10 in my device and it works fine for me, definitely fixes the uninstall issue; if QA signoff is necessary then could QA have a look?
[23:16] <cjwatson> That's bug 1326694
[23:17] <cjwatson> I don't think I have the ability to mark it as tested, or if I do then I don't know how.
[23:17] <ToyKeeper> cjwatson: Ah, it's built and in a silo now?  Sweet.
[23:18] <cjwatson> Yep
[23:19] <cjwatson> My testing is a bit hampered by the fact that the only way I have internet right now is via 3G tethering which I only know how to do via Android, so I can't simultaneously test and have internet on my laptop
[23:20] <cjwatson> A good part of https://wiki.ubuntu.com/Process/Merges/TestPlan/click from "Suggested Tests" down wasn't written by me and I can't vouch for how well it's likely to work; I'm not very convinced that creating more users is a good idea on a phone
[23:20] <cjwatson> Nor that it's interesting to try; that's what unit tests are for IMO
[23:20] <cjwatson> But meh, your call I guess
[23:24] <cjwatson> In fact I think a good chunk of the "Suggested Tests" are just can-never-work right now.  I'd suggest considering them a work in progress (which should go into automatic integration tests anyway) and skipping them for the moment
[23:25] <cjwatson> Dropped a note in the wiki explaining why
[23:26] <ToyKeeper> cjwatson: Thanks.  Part of what happens in traincon 0 is actual review of test plans by people who didn't write them, so I tend to complain a lot at people for that sort of thing.
[23:27] <ToyKeeper> I'm a bit slow today though, due to massive jet lag and being on call during the hours I should be asleep.
[23:30] <cjwatson> ToyKeeper: Right.  I'm off to bed shortly, because after midnight.  Hopefully it's not so bad that it can't be published; if you decide it's OK to sign off despite problems then I'd appreciate it if you could let a lander know so that they can publish it before I'm up again.
[23:31] <ToyKeeper> cjwatson: It'll probably be fine.
[23:32] <cjwatson> My only a priori concern was whether the broken desktop files would be automatically regenerated on session restart with the fixed code, but it turns out that (for me) they are.
[23:33] <cjwatson> I hadn't been going to go to special efforts to smooth the upgrade given that it's only in devel-proposed, but that saves me a mail to ubuntu-phone@ at last.
[23:33] <cjwatson> *least
[23:36] <robru> cjwatson, i gave you edit access to the spreadsheet in case you already didn't. you can mark testing: yes here: https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0AuDk72Lpx8U5dFVHQ3FuMDJGLUZCamJfSjYzbWh3Wnc&rm=full&pli=1#gid=28
[23:37] <cjwatson> robru: Ah, thanks a lot, done
[23:37] <robru> great