[12:38] brendand, hi, could you please top approve https://code.launchpad.net/~canonical-platform-qa/ubuntu-system-tests/mtp_refactoring_merge_and_bugfix/+merge/265084 [12:40] cachio, yep was just waiting for rhuddie's ack [12:40] brendand, thanks [12:40] ah, yes i just approved that earlier [13:24] ubuntu-qa: since the change to use the OSK for text entry, we still can't get the ubuntu-system-settings tests to pass when running the full suite. Running the tests individually pass reliably though [13:25] i'm not sure it's directly related to the change in AP to use the OSK, but it seemed to coincide [13:25] kenvandine, have you got some failure logs? [13:26] i get a different number of failures each time i run the full suite, but when running test failed tests individually they pass [13:26] vivid and wily [13:26] rhuddie, sure [13:26] http://jenkins.qa.ubuntu.com/job/generic-deb-autopilot-runner-vivid-mako/2881 [13:26] that was a CI run for vivid [13:26] which included the fix we needed for dismissing the osk [13:27] and not all the failures actually enter text [13:28] rhuddie, i've hit a wall trying to figure it out... [13:28] kenvandine, just checking through the logs.... [13:28] rhuddie, thx [13:30] ValueError: Object 'None' does not have any recognised position attributes [13:30] in get_center_point [13:31] rhuddie, the tests that produce that traceback all pass individually [13:33] kenvandine, that is weird. the osk is being shown, so its like its not finding where the required key is for some reason. [13:33] rhuddie, what's driving me crazy is that it passes fine individually, so not the test [13:33] but something gets in a bad state or something [13:33] rhuddie, same thing in wily and vivid [13:34] mako and krillin [13:34] kenvandine, yeah... that really makes it tough [13:34] rhuddie, anyway... i have no idea what else to do to debug this [13:35] rhuddie, btw, i'm running the ubuntu_keyboard tests right now and i assume it's going to fail [13:35] watching the UI, I see the OSK get raised and no keys pressed [13:35] over an over [13:39] kenvandine, it looks like a bug in the autopilot helpers for ubuntu_keyboard, although why it is happening i have no idea. for some reason it is passing a None object instead of the key to press [13:39] rhuddie, should i file a bug against ubuntu-keyboard? [13:40] i think since it works individually it's unlikely to be the helper [13:40] the helper clearly works... === qwebirc846751 is now known as slickymasterWork [13:40] just some state issue [13:42] kenvandine, its difficult to say where the problem is exactly. but the call stack seems to indicate that key object coming from ubuntu_keyboard helpers is None for some reason. [13:42] rhuddie, i just pointed Elleo at that in the other channel === chihchun is now known as chihchun_afk === spineau is now known as spineau_afk [16:01] dkessel: hey, sorry i've only just seen your email, (had a bit of a backlog to work through). I'll reply later this evening, got to go sort the kids dinner now [16:03] o/ DanChapman [16:05] hey balloons :-) [16:05] My apologies to you as well for my slow responses, it's been a busy few weeks [16:07] DanChapman, no worries at all. Nice to hear from you :-) [17:46] ubuntu-qa, the latest autopilot landing is missing a depends, which is now breaking autopkgtests in wily [17:46] i proposed a fix at https://code.launchpad.net/~ken-vandine/autopilot/depends_for_gsettings/+merge/265580 [17:46] kenvandine: I'll take a look at that mp now, thanks for bringing it up [17:47] nuclearbob, happy to help, it's holding up my landings :-D [17:47] nuclearbob, here's the log showing why it's needed http://d-jenkins.ubuntu-ci:8080/view/Wily/view/AutoPkgTest/job/wily-adt-ubuntu-system-settings-online-accounts/68/ARCH=amd64,label=adt/console [17:48] kenvandine: yeah, the change makes sense to me, I'm going to look at what we need to get the release process going since I haven't done one in a while [17:48] nuclearbob, cool, thx! [17:48] I think we may need a sign-off from another team for autopilot packaging changes as well, since they potentially affect a lot of archive landings, looking into that [17:49] kenvandine, you're not alone FWIW https://bugs.launchpad.net/ubuntu/+source/autopilot/+bug/1477233 [17:49] Ubuntu bug 1477233 in autopilot (Ubuntu) "autopilot now depends on "/usr/bin/gsettings" and ubuntu-keyboard" [Undecided,New] [17:50] alesage, ah... i'll add the depends for the keyboard too :) [17:52] alesage, nuclearbob: pushed [17:52] kenvandine: thanks! [17:53] np [17:53] autopilot is holding up a bunch of packages right now :/ [17:54] we feel the urgency [18:07] nuclearbob, can you review the changes and prepare a release and veebers will top approve and land it [18:08] jibel: the changes look good to me, but I'm not familiar with the process for preparing a release. Is it just requesting a silo for the train, or is there more involved? I'd like to get familiar with it so I can do this in the future [18:09] i tried to get the release team to migrate settings for me to help get me moving on to another silo... but they said to just push you guys to land autopilot faster :) [18:09] nuclearbob, 2 hours won't change the face of the world but do the release with him and make sure it's documented so next time it won't block on him. [18:09] jibel: thanks, will do [18:11] kenvandine, understood. it'll be fixed ASAP but I prefer if veebers drives the landing so we don't waste more time. [18:11] sure [18:11] thanks guys!