[14:41] <mterry> Saviq: I'm not on my usual computer -- can you link me the CI bot's account that holds the silo merged bzr branches?
[14:42] <Saviq> mterry, https://code.launchpad.net/~ci-train-bot/
[14:43] <mterry> Saviq: cheers!
[14:43]  * mterry bookmarks
[14:43] <Saviq> mterry, not sure you saw - there's a greeter issue in 2272 - on boot, the first user's selected session is wrong
[14:44] <mterry> Saviq: fixed in my top branch -- looking for the test failure you noted now
[14:44] <Saviq> ack!
[14:44] <mterry> Doesn't appear in my top branch, so must be an integration issue
[14:44] <mterry> Hence why I need the merged bzr branch  :P
[14:46] <Saviq> :)
[14:49] <davmor2> mterry: what did you do to your other computer?
[14:50] <mterry> davmor2: it's fine, but I have both a xenial & a zesty computer -- finding myself on the xenial one more these days because of snap work and it isn't as well-used as the zesty one with all my bookmarks and such.  Should probably get a sync plugin for that
[14:51] <davmor2> mterry: probably :)
[14:52] <davmor2> mterry: although I'm all bored now was expecting a small explosion to be involved at least :(
[14:52] <mterry> Yeah I should punch that story up a bit
[14:53] <mterry> Turns out a zesty kernel bug overheated my motherboard and the CPU glue came off
[14:53] <mterry> So I'm on my xenial backup comp
[14:53] <davmor2> mterry: see way more exciting :)
[14:53] <mterry> :)
[14:54] <davmor2> of course now I'm trying to find the damn zesty bug I hate developer it's all or nothing with them ;)
[14:55] <mterry> I DON'T KNOW HOW TO MAKE YOU HAPPY DAVMOR2
[14:55] <davmor2> mterry: hahahaha
[14:55] <mterry> :)
[14:55] <davmor2> that'll do it the sound of defeat \o/
[14:56] <mterry> heh
[15:27] <Saviq> mterry, it didn't show up in CI either, so might need stress & co to reproduce
[15:27] <Saviq> but it happened at least twice in britney
[15:27] <mterry> Saviq: yeah haven't reproduced locally yet either  :(
[15:31] <mterry> Saviq: it's not 100% in britney, right?  Sounds like a flaky test that is exacerbated by brtiney
[15:32] <mterry> Looks like the test is trying to click on a name in the login list and ends up missing it, hitting the next name, according to the failure you posted
[18:41] <mterry> Can't reproduce the failure, even using stress-ng on my laptop   :(
[19:04] <Saviq> mterry, reproduced after a few seconds with stress and parallel
[19:04] <mterry> Saviq: oh nice!  I didn't use parallel, just ran the test over again in series
[19:04] <mterry> maybe that's my missing stress
[19:05] <Saviq> mterry, http://paste.ubuntu.com/23853449/
[19:05] <Saviq> not sure if there's a better way of feeding parallel
[19:11] <mterry> Saviq: unless you made your own changes, I'm guessing you ran into a test init issue that is separate from what britney would have seen -- I fixed the test problem in my own branch a little bit ago, but haven't yet reproduced britney's failure
[19:11] <mterry> The problem appaeared when you tried to run a specific test, not the whole suite
[19:34] <Saviq> mterry, right, lemme run the whole thing then
[19:54] <Saviq> mterry, got it to fail exactly how it did in britney
[19:55] <mterry> Saviq: yay!  just doing the same thing you passed before?
[19:55] <Saviq> mterry, yes
[19:55] <Saviq> http://pastebin.ubuntu.com/23853717/
[19:56] <mterry> Saviq: comforting...  I will continue to try to reproduce
[19:58] <Saviq> mterry, if you need, get me a more verbose version of the test and I can try and get you more data
[19:59] <Saviq> robert_ancell, hey, could you have a look at bug #1654365 please - Laney tried to fix dbus behaviour of the unity8 session, but for some reason the daemon dies straight away, leaving a bogus DBUS_SESSION_ADDRESS lying around
[20:00] <robert_ancell> Saviq, I'll have a look but I don't know a lot about the D-Bus daemons
[20:19] <mterry> The problem with a verbose version of the test is that it would affect the timing.
[20:34] <mterry> Saviq: instead of a verbose version, I think it makes sense to just try a fix and see how much better that makes it -- here's an attempt to settle things before the test starts: http://pastebin.ubuntu.com/23853952/ -- add that to end of of init() in tst_WideView.qml.  I still haven't been able to reproduce  :(
[21:08] <Saviq> mterry, that does seem to help, been running 10mins now without fail
[21:09] <mterry> Saviq: alright, they are harmless lines, might as well commit it and see what britney thinks
[21:09] <mterry> thx!
[21:10] <Saviq> mterry, yup
[21:11] <mterry> Saviq: done, rebuild at your leisure
[21:12] <Saviq> mterry, thanks