[02:31] <teward> where do we file bugs with the daily isos?
[02:31] <teward> [2015-02-08 21:24:29] <jsjgruber> Bug ; error dialog on boot of current daily iso
[02:31] <teward> [2015-02-08 21:27:48] <jsjgruber> User ubuntu .config file owned by root rather than by user ubuntu
[02:31] <teward> ^  from #ubuntu-bugs
[03:01] <ePierre> hey guys!
[03:02] <ePierre> Any news about Ubuntu 14.04.2? Wasn't it supposed to be released last Friday?
[03:19] <teward> ePierre: correct me if I'm wrong but the 19th is in 11 days?
[03:19] <teward> https://wiki.ubuntu.com/TrustyTahr/ReleaseSchedule
[03:19] <teward> (search for 14.04.2)
[03:40] <ePierre> teward, oh you're right! I saw a bunch of e-mails last week about the .2 version testing and thought it was due very soon :D Thanks for the reminder!
[14:40] <elopio> good morning
[15:05] <teward> there's no mirrors of the daily ISO builds are there?
[15:05] <teward> just the master cdimage servers?
[15:08] <balloons> teward, no official mirrors I know o
[15:08] <elfy> teward: not sure tbh - I've never seen mirrors
[15:08] <elfy> hi balloons :)
[15:09] <balloons> hey elfy
[15:11] <elfy> balloons: good weekend I trust
[15:12] <elfy> not broken anything
[15:12] <teward> balloons: elfy: I know the official mirrors https://launchpad.net/ubuntu/+cdmirrors don't have daily
[15:13] <balloons> no breaks.. needed a good bit of rest yesteday, tired!
[15:13] <teward> someone on Ask ubuntu is looking for a more local daily builds mirror near China, but meh
[15:13] <jpds> teward: That's releases.
[15:13] <teward> jpds: i'm aware - they said 'cd mirrors', but the question remains for daily mirrors
[15:13] <teward> and AFAIK there aren't
[15:13] <balloons> teward, ahh, excellent question. If they use zsync the diff should be small it should be ok
[15:13] <teward> (could probably repurpose one of my servers to do it, but i don't need em)
[15:13] <teward> balloons: assuming they have a copy - they still have to zsync the whole image down once
[15:13] <teward> or twice
[15:13] <teward> (depending on which images they pull)
[15:14] <balloons> teward, yes you do need a first image, and if the diff gets large then :-)
[15:17] <teward> balloons: indeed.  Y'know, it helps when people are specific and don't confuse things too - they think http://cdimage.ubuntu.com/precise/daily-live/ means precise still gets daily builds
[15:17] <teward> IIRC they don't, it's already 'stable'
[15:18] <elfy> balloons: Read vivid-desktop-amd64.iso. Target 27.1% complete.
[15:18] <elfy> the difference between utopic and vivid for me :)
[15:21] <balloons> teward, precise still gets new builds, since it's still supported afaik
[15:21] <balloons> it's just a rollup of the archive
[15:21] <elfy> balloons: and did you know that vivid ubuntu is still giving home/$USERNAME/.config owned by root
[15:22] <teward> balloons: but is that daily built or only around point-release dates
[15:22] <teward> because that's the confusion on the end user
[15:22] <teward> i understand if it's near a point release and testing, but the standard users don't.
[15:22] <teward> (being a power user gives good insights into things)
[16:47] <dobey> elopio: hey. did you get a chance to try my pay-ui adt-run-ap branch btw?
[16:48] <elopio> dobey: I gave it a try and got the same error you got.
[16:48] <elopio> couldn't dig futher than that on friday, sorry.
[16:49] <dobey> elopio: ok, i'm really confused on why it's not working :-/
[16:51] <elopio> dobey: can you run it with the null testbed? on your host machine?
[16:51] <elopio> that's what I was about to try, before things collapsed in the sanity suite.
[16:53] <dobey> elopio: --- null instead of qemu?
[16:53] <elopio> dobey: yes.
[16:53] <dobey> adt-run [11:53:35]: ERROR: unexpected error: "sh -ec awk '/^deb(-src)? .*(ubuntu.com|debian.org|ftpmaster)/ { if ($3 !~ /-/) { $3 = $3"-proposed"; print }}' /etc/apt/sources.list `ls /etc/apt/sources.list.d/*.list 2>/dev/null|| true`  > /etc/apt/sources.list.d/proposed.list" failed with stderr "sh: 1: cannot create /etc/apt/sources.list.d/proposed.list: Permission denied
[16:53] <dobey> no :)
[16:54] <elopio> right, that probably might even break your host.
[16:54] <dobey> from that i guess it wants to run it on my host machine and screw with my data
[16:55] <elopio> dobey: we can inspect the tree. Put a self.app.print_tree before the self.app.select_single(objectName='payui') that fails to you.
[16:56] <dobey> elopio:
[16:56] <dobey> err
[16:56] <dobey> elopio: ok
[16:56] <elopio> it takes as an argument the file where it will store the tree. If you pass a name inside the output dir of adt-run, it will be collected after the failure
[17:12] <dobey> elopio: where do those artifacts go though?
[17:12] <pitti> dobey: hey, how are you? you asked about where the test output goes -- that's -o <dir> or --output-dir <dir>
[17:12] <dobey> elopio: afaict, all the screenshots and such just get deleted when adt-run finishes
[17:12] <dobey> pitti: and where does it go by default?
[17:12] <pitti> dobey: if you want to poke around in the testbed after a failed test, use -s instead (or --shell-on-failure)
[17:13] <pitti> dobey: just to stdout/err
[17:13] <dobey> pitti: do artifacts not get copied to the host anywhere?
[17:13] <pitti> (and a temp dir, but that will be cleaned up)
[17:13] <pitti> dobey: they are, with --output-dir
[17:13] <dobey> would be nice if the temp dir was copied to /tmp on the host before the vm gets destroyed
[17:14] <dobey> by default
[18:31] <dobey> elopio: where are the docs for that print_tree() call?
[18:32] <elopio> dobey: https://developer.ubuntu.com/api/devel/ubuntu-14.10/python/autopilot/api/autopilot.introspection.html#autopilot.introspection.ProxyBase.print_tree
[18:48] <doug5> balloons, ping
[18:48] <balloons> doug5, pong
[18:49] <doug5> balloons, :) can you take a look at this when you have time? https://code.launchpad.net/~acerisara/ubuntu-calendar-app/WeekView-ScrollToToday
[18:49] <balloons> doug5, sure
[18:49] <doug5> balloons, merci
[18:51] <balloons> doug5, heh, why the change?
[18:52] <dobey> elopio: ok, i am not seeing any file in the output directory for the tree structure
[18:52] <dobey> elopio: i only see the autopilot.subunit file
[18:53] <doug5> balloons, https://code.launchpad.net/~pkunal-parmar/ubuntu-calendar-app/WeekView-ScrollToToday/+merge/243664
[18:54] <dobey> :-/
[18:54] <doug5> balloons, also, if there is something else to do...tests or autopilot, I'm available
[18:56] <elopio> dobey: I will give it a try.
[18:56] <balloons> doug5, awesome. Let's think
[18:57] <dobey> elopio: ok, thanks for helping with this. i really want to get it in and working
[18:58] <elopio> dobey: no problem. But I'm about to start our planning meeting, so I might be slow today.
[19:00] <dobey> elopio: ok, i'll look at other stuff in the meantime
[19:05] <balloons> doug5, I'm looking at https://bugs.launchpad.net/ubuntu-phone-coreapps/+bugs?field.tag=needs-autopilot-test&orderby=-datecreated&start=0. I know the calc reboot needs some stuff still for the scientific calculator they are adding
[19:11] <doug5> balloons, ok, I will take one of those :)
[19:12]  * balloons cleans up fix committed bugs
[19:12] <wxl> balloons: wha?
[19:12] <balloons> wxl, howdy
[19:12] <balloons> I was curious how the testing jam went! Any feedback?
[19:13] <wxl> oh yeah i need to write it up
[19:13] <wxl> i'm not sure what people usually do for time, but we had about 3 hours
[19:13] <balloons> wxl, if you wouldn't mind sending it along to the list as well, that would be awesome
[19:13] <wxl> most of the time was spent discussing testing and bug triage
[19:13] <wxl> it would have been wise to have only focused on one of the two
[19:13] <wxl> this being my first jam ever, i didn't know better, i guess
[19:14] <wxl> but there was a lot of good dialogue and with bdmurray around, the bug discussion went great
[19:14] <wxl> our local lug's head was there, so that was good as far as helping him grok the ubuntu way of doing things
[19:15] <wxl> so we enabled people to test but didn't necessarily get to much actual testing :(
[19:15] <wxl> but i would still consider that a success
[19:16] <wxl> i sure learned a lot for next time
[19:23] <balloons> wxl, excellent. I'd love a writeup and any feedback from myself / team would be wonderful. I'm glad you got to spend that much time together. Good stuff!
[19:24] <wxl> yes i'll make sure you have it today or tomorrow balloons
[19:24] <elopio> dobey: I have the tree file http://paste.ubuntu.com/10146794/
[19:25] <elopio> the pid is from a qfiledialog.
[19:25] <elopio> I think that's what happens when you launch qmlscene with the wrong path. It will ask you to choose a qml file to open.
[19:27] <dobey> elopio: ugh. how does one launch it with the wrong path?
[19:28] <elopio> yeah, looking at the subunit file I see that.
[19:28] <elopio> dobey: btw, you can use this to view the subunit screenshots: https://www.youtube.com/watch?v=jkLtbmQxXYc
[19:28] <elopio> 13:10:28.520 INFO _launcher:373 - Attempting to launch application '/usr/lib/x86_64-linux-gnu/qt5/bin/qmlscene' with arguments '/opt/click.ubuntu.com/.click/users/@all/com.canonical.payui/share/qml/payui/payui.qml --desktop_file_hint=/opt/click.ubuntu.com/.click/users/@all/com.canonical.payui/payui_payui.desktop' as a normal process
[19:28] <elopio> dobey: I guess it's that @all.
[19:28] <dobey> elopio: and what did you add exactly to get the file?
[19:29] <dobey> elopio: or it's embedded in the subunit file?
[19:29] <elopio> dobey: http://paste.ubuntu.com/10146868/
[19:31] <dobey> oh
[19:37] <doug5> balloons, can we try to push it on trunk?
[19:38] <dobey> elopio: why would the '@all' there be invalid? it's a valid file path :(
[19:39] <balloons> doug5, on your mp? it was going into mihir
[19:39] <balloons> 's right?
[19:40] <doug5> kunal actually
[19:40] <doug5> balloons, but yes, you're rught
[19:40] <doug5> *right
[19:42] <elopio> dobey: I didn't know that was valid. So there must be a different problem, like the file it's not there.
[19:43] <dobey> elopio: yeah, so i just tried locally by creating an @all symlink to another directory and loading a qml file via it, and it worked :-/
[19:43] <dobey> sigh
[19:55] <dobey> elopio: https://jenkins.qa.ubuntu.com/job/generic-click-autopilot-runner-mako/765/console <- this is apparently what happens when jenkins tries to run the tests on an actual device
[20:12]  * dobey wonders what is going on here
[20:43] <dobey> why does qmlscene complain about every letter in "-testability" not being a valid option?
[20:44] <dobey> does that mean qtdeclarative5-test-plugin is missing?
[20:46]  * dobey wonders if he should just propose this branch anyway
[20:48] <balloons> ahh yes, lovely bug
[21:13] <dobey> balloons: the -testability thing?
[21:13] <balloons> dobey, indeed
[21:13] <dobey> yeah, not fun
[21:14] <dobey> but also probably the least of my problems at the moment, with regards to autopilot tests :-/
[21:15] <elopio> dobey: the screenshot for jenkins looks a little different. It's not opening the file selection dialog.
[21:18] <dobey> i don't see a screenshot there
[21:19] <dobey> but unity8's complaining about being unable to connect to mir there
[21:19] <dobey> oh i guess i have to grab the subunit file to see the screenshots
[21:19] <dobey> and then use the trv thing to view them
[21:24] <dobey> hmm, and i can't seem to run it from the source tree
[21:29] <dobey> elopio: ok, well those screenshots make sense
[21:29] <dobey> elopio: since qmlscene fails to run, it just takes a screenshot of what's in the mir buffer anyway. in this case, it's just the dash
[21:29] <elopio> dobey: right. So the problem here is launching payui.
[21:30] <elopio> I'm not sure I'll be able to help you there. Ted might be more useful.
[21:30] <dobey> right
[21:30] <dobey> i don't think ted probably knows why qmlscene is complaining that -testability is an invalid set of arguments
[21:31] <dobey> elopio: besides, i've fixed the u-a-l issue already, by avoiding u-a-l all together, for the autopilot testing
[21:33] <dobey> elopio: look at the console output from qmlscene
[21:34] <dobey> not ideal :-/
[21:34] <dobey> but at least i'm not getting that locally
[21:34] <dobey> locally it runs but doesn't load the file for some reason
[21:34] <dobey> so i guess it thinks the file doesn't exist
[21:34] <elopio> dobey: yeah. You can ignore the testability part. That's ugly, but harmless. The mir is the one that I think shouldn't be there.
[21:35] <elopio> on the one running locally on my qemu, I don't see any output from qmlscene
[21:36] <dobey> right
[21:37] <dobey> because qmlscene runs ok there
[21:37] <dobey> it just thinks it has no file input and opens the file open dialog
[21:37] <dobey> i'm guessing when running on the actual device, the problem is that access to mir is confined
[21:38] <dobey> and thus we can't open the ui
[21:41] <dobey> adding a log message to make sure the file exists; maybe i got the path wrong but it should be correct
[21:44] <elopio> I'm going to get some lunch. I'll check back soon. dobey: please let me know if you find anything useful.
[21:44] <dobey> enjoy elopio
[21:45] <elopio> thanks. You too ^^
[21:47] <dobey> somehow lunch sounds much better than "wait for autopilot to fail"
[21:48] <balloons> elopio, how's the backlog on https://code.launchpad.net/ubuntu-manual-tests/+activereviews look?
[21:48] <balloons> elfy, sorry ^^
[21:58] <dobey> :-/
[22:18] <dobey> 17:15:42.167 ERROR testresult:44 - ERROR: unittest.loader.ModuleImportFailure.pay_ui.tests.test_pay_ui
[22:18] <dobey> wtf :(
[22:21] <dobey> ah, somehow an extra ) ended up in the code :-/
[22:29] <dobey> ah, bad reviewers
[22:38] <dobey> gah, ok so it thinks the file doesn't exist, which means 'click info' is lying or something :-/
[23:04] <dobey> elopio: ok, time for me to go. but it seems the path to the qml doesn't exist, so either click is lying about where it's installed, or something weird is going on there :-/
[23:44] <alesage> elopio, I'm seeing import failures presumably due to my being on utopic, what did we decide to do about those?
[23:45] <alesage> elopio, should I file a bug, do we wish to support utopic, etc.