[00:01] <ianorlin> wxl bug 1434774 is annoying though
[00:01] <wxl> let's head over to #lubuntu-offtopic :)
[00:03] <Captonjamason> friends wanna play minecraft with me
[00:03] <Captonjamason> maybe in a bit
[12:46] <elfy> balloons infinity cyphermox - just so you are aware, testing today's dailies for us and Ubuntu - the fail to reboot after the remove media prompt is back
[12:48] <davmor2> infinity, cyphermox, jibel: there are huge issues on oem mode in latest dvd
[12:48] <davmor2> I'm going to start filing bugs
[13:07] <davmor2> https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1445474 , https://bugs.launchpad.net/ubuntu/+source/casper/+bug/1436715 , https://bugs.launchpad.net/ubuntu/+bug/1445479 , https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/1445481
[13:52] <elopio> good morning.
[13:56] <infinity> cyphermox: Did you see all of davmor2's bugs up there?
[13:58] <davmor2> infinity: I need to run a non oem install but I don't want to blow away the broken wifi till cyphermox has logs :)
[14:03] <jibel> pitti, running upgrade from utopic to vivid, I'll create other jobs once it passes
[14:03] <pitti> jibel: yay, merci beaucoup !
[14:04] <pitti> jibel: et comment on dit "yay" par français ? :-)
[14:04] <jibel> pitti, Youpi!
[14:04] <jibel> but it's a bit childish :)
[14:07] <balloons> jibel, interesting to know! :-)
[14:21] <cyphermox> infinity: I did
[14:21] <davmor2> jibel: oem fails hard
[14:23] <seb128> hey there
[14:23] <seb128> ubuntu-system-settings has some flacky tests, which I'm trying to look at
[14:24] <seb128> we use self.main_view.pointing_device.click_object() calls on listitems
[14:24] <seb128> but sometime it seems that the tap doesn't work properly
[14:24] <seb128> like onClicked: is not called for the item and the text/selection color in the UI is not changed
[14:24] <seb128> is that a known issue/any hint on how to debug?
[14:24] <seb128> kenvandine, ^
[14:25] <kenvandine> seb128, thx!
[14:25] <seb128> kenvandine, yw!
[14:25] <kenvandine> i'm wondering if missing that is causing other our general flakyness we've been experiencing lately
[14:31] <cyphermox> davmor2: I do get sound in the installer...
[14:32] <cyphermox> in qemu
[14:32] <cyphermox> I can try hardware as well
[14:32] <davmor2> cyphermox: I'm testing on hardware here
[14:32] <davmor2> cyphermox: HP g62 normally it works
[14:32] <cyphermox> ok, let me write this to a USB key and try again
[14:33] <cyphermox> sounds like it might not be the installer though
[14:41] <Captonjamason> hell
[14:41] <Captonjamason> hello
[14:43] <Captonjamason> can i get some help with LFS
[14:43] <Captonjamason> or would i move to a offtopic
[15:00] <fgimenez> elopio, didn't know about that flake8 bug, thanks :)
[15:04] <cyphermox> davmor2: and you had a pointer in OEM install?
[15:06] <cyphermox> AH! think I just noticed a pattern here
[15:07] <elopio> fgimenez: it's new.
[15:07] <davmor2> cyphermox: I have left the last run in the oem desktop incase you need logs and stuff for the bugs
[15:07] <elopio> but I have opened many. Now I will try to stick with this one, reopening projects if they regress.
[15:08] <cyphermox> davmor2: not sure what you mean there
[15:08] <elopio> fgimenez: that qakit project's trunk is still under thomi's profile.
[15:08] <elopio> not sure how to change that.
[15:08] <elopio> fgimenez: oh cool, you took the address book :) Thanks.
[15:09] <fgimenez> elopio, i feel in the flake8 mood this friday :)
[15:09] <elopio> fgimenez: the hard part with this is to get all the reviews, like cat herding.
[15:09] <elopio> fgimenez: when you are done with that, you can ask renatu in #ubuntu-touch for the review.
[15:09] <davmor2> cyphermox: I have a cursor
[15:10] <cyphermox> on real hardware?
[15:10] <davmor2> cyphermox: yes
[15:10] <cyphermox> I did not
[15:10] <davmor2> I'm moving it around the screen now
[15:10] <fgimenez> elopio, ok thx, about qakit, i see that in code -> configure code hosting it mentions thomi's branch
[15:10] <davmor2> cyphermox: tis another breakage then
[15:10] <cyphermox> also, reboot works in a live image, but fails in install-only
[15:10] <cyphermox> that's the pattern
[15:10] <elopio> fgimenez: yes, but if I move it, will I break your MP?
[15:10] <davmor2> cyphermox: it's getting worse not better dude stop touching it :)
[15:11] <cyphermox> also, I loathe systemd
[15:11] <fgimenez> elopio, don't know, there are other mps too
[15:12] <davmor2> cyphermox: no really did you every think it might hate you too?
[15:12] <cyphermox> I'm starting to
[15:12] <davmor2> cyphermox: hate it back it works for me
[15:13] <cyphermox> I wish it a fiery death
[15:13] <cyphermox> in the heat of a thousand suns
[15:14] <davmor2> cyphermox: don't forget the and it's ashes drop into the blackhole of oblivion
[15:14] <cyphermox> well, there wouldn't be any ashes.
[15:15] <cyphermox> stars tend to destroy most atoms :)
[15:15] <cyphermox> so, now on to the actual oem install
[15:20] <cyphermox> davmor2: I have sound and I have the oem icon
[15:20] <cyphermox> are you sure you were on 20150417?
[15:20] <cyphermox> 20150416 would have been very broken
[15:25] <elopio> oh crap, I missed the standup.
[15:25] <davmor2> cyphermox: http://people.canonical.com/~davmor2/oem-mode.png
[15:28] <cyphermox> what is this supposed to show?
[15:45] <davmor2> cyphermox: lack of oem icon, the fact I'm in oem desktop mode and that my davmor2 wifi connection isn't shown
[15:45] <cyphermox> ok
[15:46] <cyphermox> well, I don't know like this why it might have happened, things are looking fine here in a VM and I can't do an OEM install for now, for lack of hardware to do it on
[15:47] <davmor2> cyphermox: yes so this is after the install
[15:47] <cyphermox> right
[15:47] <davmor2> cyphermox: this isn't live session
[15:47] <cyphermox> but then what about the sound and reboot?
[15:47] <cyphermox> tbh those would be my priority for now
[15:48] <davmor2> cyphermox: reboot was an Issue still I had to hard power down the system
[15:48] <cyphermox> can you make sure you're running on vivid-desktop 2015-04-17
[15:48] <cyphermox> or let me know exactly which image you used
[15:48] <cyphermox> I can reproduce the reboot issue, it looks like it's only happening if you don't choose a live session
[15:48] <cyphermox> I need to dig in why, trying to convince systemd to log to a serial console right now
[15:50] <davmor2> cyphermox: woooo okay that is weird
[15:51] <davmor2> cyphermox: from the site 20e8bb0dac1df4506309bad377260d18 *vivid-desktop-amd64.iso on my server 8d57a7f4bb63531935f90b457b7f208d  iso/daily-live/current/vivid-desktop-amd64.iso  however it says there is nothing to update
[15:55] <davmor2> cyphermox: let me delete this image and update again and see what is happening
[16:02] <davmor2> cyphermox: okay now I got the right image so no idea what the hell went on with that
[16:03] <davmor2> cyphermox: I'll see what I can confirm
[16:03] <davmor2> elfy: check your md5sums it seems like zsync might of been playing up
[16:05] <elfy> the xubuntu images are right
[16:14] <cyphermox> rebooting again to try to convince plymouth and systemd to give me useful data.
[16:33] <elopio> dobey: ping. I'm following your hacking file in pay-ui, failing to create the chroot.
[16:33] <elopio> http://paste.ubuntu.com/10839516/
[16:33] <elopio> have you seen that?
[16:35] <kenvandine> ubuntu-qa, please see seb128's question from 2 hours ago, he forgot to include the vanguard ping
[16:35] <kenvandine> or i can repeat it if needed
[16:35] <dobey> elopio: i have not seen that, no
[16:41] <elopio> kenvandine: looking...
[16:41] <kenvandine> elopio, thx
[16:41] <elopio> kenvandine: do you have the trace of an error?
[16:41] <kenvandine> not the one he had specifically
[16:42] <kenvandine> but we've had lots of tests failing lately
[16:42] <kenvandine> flaky
[16:42] <kenvandine> and he did some debugging and managed to reproduce it locally on his n7
[16:42] <kenvandine> which i've never been able to do on my mako
[16:42] <elopio> I will look at the code to see if I find something obvious
[16:43] <elopio> kenvandine: any progress with the QML tests?
[16:43] <kenvandine> he made the header title change to "clicked" when it got onClicked in that test
[16:43] <kenvandine> and when it failed, that had never been changed
[16:43] <kenvandine> so it never got onClicked
[16:43] <kenvandine> elopio, no...
[16:44] <kenvandine> we need to get that on our backlog
[16:45] <kenvandine> elopio, so seb128's debugging pointed to sometimes click_object() never actually triggers the onClicked signal
[16:45] <elopio> kenvandine: yes, we need to clean up the autopilot tests in order to make them reliable. For that it's essential to turn many of them into lower level tests.
[16:45] <kenvandine> agreed
[16:45] <jibel> pitti, upgrade test passed with forcing upstart
[16:45] <elopio> kenvandine: click_object simulates an evdev event, which hasn't failed so far.
[16:45] <kenvandine> elopio, it had been reliable for quite a while
[16:45] <elopio> there could be a bug there, but I doubt it. It's more likely that the object we are trying to click is not ready
[16:46] <kenvandine> but in the past 2 weeks or so, we've been getting bit by this
[16:46] <elopio> like not enabled, not listening for the event, or moving.
[16:46] <kenvandine> could be
[16:46] <elopio> kenvandine: generally, that means a missing wait.
[16:46] <kenvandine> but... those tests haven't changed in months
[16:47] <kenvandine> https://jenkins.qa.ubuntu.com/job/generic-deb-autopilot-runner-vivid-mako/2021/testReport/junit/ubuntu_system_settings.tests.test_datetime/TimeDateTestCase/test_same_tz_selection/
[16:47] <kenvandine> elopio, ^^ that's one of them
[16:47] <kenvandine> but not the same as the run seb128 did verifying it never got the onClicked
[16:47] <elopio> kenvandine: thanks, that helps.
[16:48] <kenvandine> it's 2 of the tz_selection tests that seem the most common culprits
[16:48] <kenvandine> but, we've had random failures in others too
[16:48] <kenvandine> like in your branch yesterday
[16:48] <elopio> kenvandine: that's expected with these tests, something in the lower layers gets slower and then we find out that we were assuming things to be fast in some of the helpers.
[16:49] <elopio> the way to survive that is not to duplicate any code at all.
[16:49] <kenvandine> yeah, and hard to debug :/
[16:49] <elopio> then any breakage will require at most one change. Also adding log messages, to make it clear what the test was doing.
[16:50] <elopio> kenvandine: I can help you refactoring the test to make it clearer where it's failing. Probably I won't have more time today to debug it, but we can continue from there on monday.
[16:50] <pitti> jibel: \o/
[16:50] <kenvandine> elopio, thx, ping me or seb128
[16:50] <kenvandine> elopio, seb128 has a device he can repro it on, i don't
[16:51] <kenvandine> it never seems to fail on my device
[16:51] <kenvandine> just in CI
[16:51] <kenvandine> i think seb128 could only repro it on his n7, not mako
[16:51] <kenvandine> all frustrating :/
[16:51] <davmor2> cyphermox: 1 bug down I have volume
[16:52] <wxl> davmor2: elfy: were you guys talking about a bug where the system wouldn't reboot at the end of install?
[16:53] <elfy> wxl: yea
[16:53] <wxl> elfy: we've noticed that too. bug reports?
[16:54] <wxl> ianorlin: did you file a new bug for the system not rebooting at end of install?
[16:54] <elfy> bug 1436715 I guess
[16:54] <wxl> well yeah the fixed one :/
[16:55] <wxl> shouldn't be the exact same bug so much as a regression
[16:55] <ianorlin> wxl I haven't yet I realized it might be an sd card and havn't tested since then
[16:55] <wxl> ianorlin: yeah doesn't sound like it
[16:56] <elopio> kenvandine: ok, firt obvious thing that is wrong, you are calling one test from another test.
[16:57] <elopio> that's asking for trouble. The tests should be independent, and they should have only one action and one assertion. That goes against both :)
[16:58] <kenvandine> ugh... which test is that?
[16:58] <kenvandine> i've never really looked at the tz tests
[16:58] <elopio> test_same_tz_selection
[16:58] <elopio> kenvandine: second thing, checking that when you select the same tz doesn't show you a different page is something you should do on QML.
[16:59]  * kenvandine looks
[16:59] <elopio> that's a feature of your widgets, it's not a user goal. I make the distinction with a question like this: would a user open my app just to set the time zone twice?
[17:00] <kenvandine> yeah
[17:00] <elfy> wxl: not worried too much about it - people are aware of the issue
[17:00] <elopio> if yes, that's a user story and should be automated at a high level, with autopilot.
[17:00] <elopio> if no, that's something you should test at a lower level, maybe in cpp, qml or javascript, depending on the feature.
[17:01] <wxl> elfy: cyphermox: meaning we expect the fail to reboot to be fixed by release?
[17:01] <kenvandine> i wonder how many tests we have calling tests
[17:01] <cyphermox> wxl: we'll do everything possible to get it fixed before release
[17:02] <kenvandine> elopio, i know we have tons of tests that have more than one assertion
[17:02] <cyphermox> otherwise it will be in the release notes
[17:02] <elopio> kenvandine: no, only that one. http://paste.ubuntu.com/10839649/
[17:02] <kenvandine> excellent :)
[17:02] <wxl> cyphermox: well then maybe i should make a bug report for it just for the sake of being completist
[17:02] <elopio> kenvandine: I have like an hour I can spend on this. Let me see what I can get done in that time, and I'll ping you back.
[17:03] <kenvandine> elopio, i think the other 1 we see lots of failures in is test_tz_list_initially_empty
[17:03] <elopio> kenvandine: another clear candidate to be downgraded to QML tests.
[17:03] <kenvandine> elopio, oh, not that one
[17:03] <kenvandine> test_manual_tz_selection
[17:04] <kenvandine> that's the other one
[17:04] <kenvandine> so interesting that's the one called from the other
[17:04] <elopio> kenvandine: yes, I imagined as the error you showed me comes from that test.
[17:04] <cyphermox> wxl: sure
[17:05] <elopio> kenvandine: that test is interesting, because it would be good to keep it as a high level test, in autopilot.
[17:05] <kenvandine> test_manual_tz_selection is the test seb128 reproduced failing this morning
[17:05] <elopio> kenvandine: could be many things. Where do you get the list of timezones from?
[17:06] <elfy> wxl: if you do report - I'll edit my QA results then
[17:06] <wxl> elfy: we'll make sure to post here when we got a report
[17:07] <wxl> i'm trying to track down something more lubuntu specific
[17:07] <kenvandine> elopio, i really don't know, i know next to nothing about this whole plugin
[17:07] <kenvandine> Laney and seb128 did this one
[17:08] <kenvandine> -        self.assertThat(header.title, Eventually(Equals(_('Time & Date'))))
[17:08] <kenvandine> -        self.assertThat(self.tz_page.text, Equals('Europe/London'))
[17:08] <kenvandine> +        self.assertThat(self.tz_page.text, Eventually(Equals('Europe/London')))
[17:08] <kenvandine> elopio, how do you feel about that?
[17:08] <kenvandine> seems silly to check the page title there
[17:08] <kenvandine> and maybe using Eventually will make it wait for it to change
[17:08] <davmor2> cyphermox: pressing enter does nothing
[17:09] <davmor2> so that bug is still valid
[17:09] <wxl> davmor2: the symptoms are similar, but it's a regression, so it really should be a new bug
[17:09] <kenvandine> maybe they put the check for the title there as a way to make it wait
[17:10] <kenvandine> but perhaps that's racy
[17:10] <davmor2> cyphermox: on arriving at the oem desktop I see the icon I am not however connected to the wifi as expected, this time at least it is showing my ap \o/
[17:11] <cyphermox> oh, pressing enter is doing something alright
[17:11] <cyphermox> it's making me insane. :)
[17:11] <elfy> lol
[17:12]  * davmor2 presses enter again to see if turn cyphermox into a flying rage monkey
[17:12] <cyphermox> wxl: in your bug mention it's only happening for the non-live session
[17:12] <jibel> pitti, bug 1358203 is the bug that makes python import test fail
[17:12] <wxl> good point. we'll double check, cyphermox
[17:12] <cyphermox> davmor2: more like flying-animal guano-crazy ;D
[17:13] <cyphermox> wxl: that has been my discovery so far
[17:13] <cyphermox> fortunately pitti is right here
[17:13] <elfy> wxl: I would have looked at Lubuntu when I was checking others if the image was new
[17:13] <cyphermox> I'll need his help to figure this out, it's more systemd deep magic
[17:13] <pitti> jibel: ah, thanks
[17:14] <elfy> cyphermox: ohhh - so we blame pitti :p
[17:15]  * wxl throws cyphermox's monkeys at pitti 
[17:15] <davmor2> cyphermox: that brings us full circle to JamesTait's morning greeting of Good morning all; happy Friday, and happy Bat Appreciation Day! :-D  and my reply JamesTait: close enough right https://www.pinterest.com/pin/135319163775397306/
[17:18] <davmor2> damn it I'm still in OEM Configuration (temporary user) mode it's not been removed only the app
[17:19] <davmor2> I see the new user setup though
[17:19] <elfy> davmor2: oh - I tried that oem thing - new one on me - I saw that happen, you're not going mad
[17:21] <cyphermox> so, I'm not reproducingthe right bug here in a VM
[17:21] <cyphermox> it just complains that there is no splash to show
[17:21] <cyphermox> will need to reboot yet again
[17:23] <davmor2> elfy No I'm going flying-mammal guano-crazy along with cyphermox
[17:23] <elopio> kenvandine: I would prefer to understand the cause before putting a patch.
[17:23] <kenvandine> sure
[17:24] <elopio> and the screenshot of the failure shows me that the item was not clicked, so it would fail anyway.
[17:24] <elfy> davmor2: I'm grabbing some popcorn and a comfy seat to watch you both :p
[17:24] <elopio> probably, the wait needs to be while the query is being made.
[17:24] <elopio> kenvandine: if you are getting the list from a web service, it makes sense for it to fail sometimes, when the service is slow.
[17:25] <elopio> I'm digging a little in the code...
[17:26] <kenvandine> elopio, oh, i'm sure the data is local somewhere
[17:27] <elopio> we need a new idea then :)
[17:28] <kenvandine> i think it comes via dbus
[17:28] <kenvandine> org.freedesktop.timedate1
[17:29] <kenvandine> or maybe that's just for setting it
[17:30] <elopio> kenvandine: when I build the project, where is the SystemSettings qml module?
[17:30] <kenvandine> you mean in the builddir?
[17:30] <elopio> kenvandine: yes, which folder in there?
[17:31] <kenvandine> lib/SystemSettings/
[17:35] <elopio> kenvandine: http://paste.ubuntu.com/10839819/
[17:36] <elopio> what am I doing wrong there? I always have a hard time with the qml imports.
[17:36] <kenvandine> i've never tried running it that way :)
[17:36] <kenvandine> one sec
[17:37] <kenvandine> oh
[17:38] <kenvandine> the timedate plugin is in builddir/plugins/time-date
[17:40] <kenvandine> but it's not going to be able to load it like that
[17:41] <kenvandine> elopio, we have some magic in the plugin interface to load the private libs, not sure how
[17:41] <elopio> kenvandine: hum, we need that to test each qml in isolation.
[17:42] <kenvandine> they aren't public qml components, they are private
[17:42] <kenvandine> so the settings container does some magic to load them
[17:43] <cprofitt> hey all... trying to report this bug so it can get looked at before release.
[17:44] <cprofitt> when I empty trash I got the confirmation dialog and file window popping up.
[17:44] <elopio> kenvandine: I'm going for lunch. bbs.
[17:44] <cprofitt> wxl: suggested reporting against the file manager... anyone else want to weigh in?
[17:46] <elfy> cprofitt: can confirm at least, probably nautilus would be the right package
[17:47] <cprofitt> yeah, but only happens when I use the launcher
[17:50] <wxl> cprofitt: i'd just file against the file manager. we can always re-file if needed after we dig deeper
[17:50] <cprofitt> ok
[17:50] <wxl> well, not re-file; re-attribute?
[17:53] <elfy> wandering off now, wxl I'll check your tracker results later to see if you've filed
[17:53] <elfy> that new bug
[17:53] <cprofitt> ok
[17:53] <cprofitt> filing now
[17:53] <wxl> elfy: okie dokie
[17:53] <wxl> thx cprofitt :)
[17:54] <cprofitt> wxl elfy https://bugs.launchpad.net/ubuntu/+source/nautilus/+bug/1445595
[17:55] <wxl> cprofitt: elfy and i are probably two of the worst people to help push this forward as we are flavor leads. cyphermox or balloons might be of more help. you might want to see if anyone at #ubuntu+1 can help
[17:56] <elfy> thanks cprofitt - confirming that
[17:56] <elfy> bbl
[18:01] <ianorlin> although could that be with gvfs it might also effect lubuntu
[18:02] <kenvandine> elopio, i refactored out the call to the other test and made it more "wait" friendly, waiting for a CI run now to see if we're barking up the right tree
[18:02] <ianorlin> but doesn't effect pcmanfm
[18:02] <kenvandine> elopio, for seb128's case, it couldn't be waiting for the model, he was able to see the result filtered
[18:03] <kenvandine> but it listview item never go the onClicked
[18:03] <kenvandine> elopio, so could even be something in wait_select_listview_first
[18:06] <kenvandine> elopio, or wait_select_listview_first before the listview has a populated model
[18:06] <balloons> cprofitt, I'm not seeing empty trash in the launcher :-(
[18:07] <kenvandine> elopio, but i'm not seeing the exception raised from wait_select_listview_first
[18:07] <cprofitt> balloons: what do you mean - you are not seeing the behavior or you do not see the trash icon in the launcher?
[18:07] <balloons> cprofitt, ohh.. lol the trash icon, got it
[18:07] <balloons> I thought you meant on nautilus icon
[18:10] <cprofitt> I am trying to find the file that creates the action menu... I am guessing that is what is borked
[18:12] <cprofitt> balloons: any idea where that file is located?
[18:15] <ianorlin> bug 1445592 is the new one I filed for no restart at that final splash screen
[18:20] <balloons> cprofitt, I'm afraid not
[18:23] <cprofitt> wxl: I appreciate the help none the less
[18:23] <dobey> elopio: that error doesn't even make any sense to me, fwiw. why would "click chroot create" even care about the pay-ui trunk directory?
[18:23] <cprofitt> balloons: any chance you can triage that bug so it can get upstream to the deves?
[18:25] <elopio> dobey: I don't understand. It's like it umounts my /home/elopio.
[18:25] <wxl> elfy: cyphermox: balloons: see above for that restart now bug
[18:25] <wxl> cprofitt: balloons: interesting that elfy could confirm it
[18:25] <dobey> elopio: yeah, no idea
[18:25] <balloons> cprofitt, bdmurray tagged it and set importance already looks like
[18:34] <elopio> kenvandine: ok. I don't think I can get much done on the QML side without being able to launch the qml from qmlscene.
[18:35] <elopio> dobey: do you know who maintains click? is it mvo?
[18:36] <dobey> i think it might be now, yes
[18:39] <cprofitt> bdmurray: thanks for adding the tag.
[20:39] <elopio> dobey: do you still have the problem of qemu dying during the adt-run?
[20:39] <elopio> I've just gotten this: http://paste.ubuntu.com/10840692/
[20:40] <dobey> elopio: i always get qemu being killed at the end, when the tests are successful. i think that's just adt-run killing it
[20:41] <dobey> elopio: i don't get the weird python backtrace about unicode in testtools any more, when using my ppa with the new testtools, though
[20:41] <elopio> dobey: ah, maybe I'm not getting my results because I'm not collecting them :)
[20:42] <dobey> that could be :)
[20:43] <dobey> elopio: use --setup-commands "add-apt-repository ppa:dobey/testtools" immediately before the -U argument, and it'll install the new testtools to run the tests under, so you get all the proper attachments and no crazy crashes when the tests fail
[20:43] <elopio> dobey: cool.
[20:43] <elopio> I love that this now exports the xvfb display.