[00:55] elfy: kind of ot but do you have a link to the qa contest you did for stickers? === chihchun_afk is now known as chihchun [04:10] veebers: so, I don't fully understand. But to fall on the safe side, I can just make branches for those project to make everybody use the base from the toolkit. [04:16] elopio: I'm just trying something now which should hopefully clarify it for myself then I'll be able to explain it better [04:52] veebers: ok, so tomorrow I'll prepare the extra branches. [04:52] good thing we made this card. [05:43] wxl: http://xubuntu.org/news/help-the-community-with-testing-and-win-xubuntu-stickers/ [05:48] elfy: thx [05:48] yay just booted my raspberry 2 with snappy [05:48] squee [06:30] wxl: welcome === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === vrruiz_ is now known as rvr [10:44] vila: so adt-run -d --- ssh -d works for me with an LXC target, so this looks phone specific [10:45] vila: -U should never be used on a phone in general, but I suppose you are doing this on purpose? did you set the phone to r/w before? [10:45] ah, apparently you do as apt-get update worked [11:06] vila: ok, I give up on bug 1449446 for now, I need more information [11:06] bug 1449446 in autopkgtest (Ubuntu) "adt-virt-ssh -d breaks satisfying adt-satdep.deb dependencies" [Undecided,Incomplete] https://launchpad.net/bugs/1449446 [11:51] pitti: yup, writable to test silos with citrain device-upgrade [11:51] pitti: I haven't dig deeply to understand why I need -U though [11:56] pitti: the log file is fine but lp fails to display it, I just downloaded it to check. Apart from a couple of ^M ... which may enough to trick lp ? [12:06] pitti: and thanks for the shell advices ! I'll probably need more for some time ;) [12:42] vila: no, it's like 2 out of 3 messages are missing in the log, I'm not sure what happened there [12:51] pitti: doh [12:51] pitti: weirdly enough, I'm currently seeing an 'echo "Reboot #$X"' disappearing when X > 3 but still presents in the log ! [12:53] pitti: with qemu in that case [12:54] pitti: ha ha ! Comparing my buffer with the log, I got: [12:55] shell: tee: standard output: Resource temporarily unavailable [12:55] log: Reboot #3 [12:55] yeah, that indeed, exactly ;) [12:57] pitti: but anyway, when you say 2 out of 3, it is some big chunks (start/end) are missing or little chunks all over the place ? [12:57] vila: the latter; I really have no idea what happened there :/ [12:58] pitti: ok, I'll retry with a phone later and see, any hint on what should be there and isn't ? [12:58] vila: for example, "adt-run: DBG: testbed init" or "adt-run: DBG: $ vserver: adt-virt-ssh ..." [12:59] vila: oh wait! this log doesn't actually contain the debugging for adt-run [13:00] vila: so maybe it's just because the branch uses Depends: @, and that's somehow not satisfiable after the apt-get upgrade, or something strange like that [13:00] pitti: err, as in I forgot adt-run -d or it's not supposed to end up htere ? [13:01] vila: it is supposed to end up there (I ran it here, and it's all in the log [13:01] pitti: nope, it doesn't happen with the same command but without -d for adt-virt-ssh [13:02] hmm, unless something weird happened at X: adt-run --- adt-virt-ssh (no -d) X; adt-run --- ssh -d ; adt-run --- ssh [13:03] pitti: well, according to my own report, I didn't use adt-run -d /me blushes [13:04] only adt-virt-ssh -d, will redo [13:45] vanguard: hi guys, I need some help with writing an emulator to interact with on-screen notifications in ubuntu touch [14:12] elopio: is vanguard working? or should I use ubuntu-qa? [14:12] boiko: ubuntu-qa please. [14:12] ubuntu-qa: hi guys, I need some help with writing an emulator to interact with on-screen notifications in ubuntu touch [14:14] boiko, fwiw, this should get you on the right path [14:14] http://bazaar.launchpad.net/~unity-team/unity8/trunk/view/head:/tests/autopilot/unity8/shell/tests/test_notifications.py#L159 [14:20] fgimenez: ^ [14:22] Saviq, thx, elopio does http://bazaar.launchpad.net/~canonical-platform-qa/ubuntu-sanity-tests/trunk/view/head:/ubuntu_sanity_tests/helpers/unity8/settings_wizard.py contain also a good example? [14:22] fgimenez: of the helpers style, yes. But boiko needs to restart unity with testability if his tests are running on touch. [14:23] elopio, ok clear now :) [14:23] then wait for the notification to appear, and then swipe it away. This part is easy. I think it would be mostly to help him getting a nice base test class for unity integration. [14:43] elopio: so, some of the notifications are dialogs with multiple actions (snap decisions?), s not sure just swiping them away would work [14:46] boiko: I see. But the principle is the same, if the user in your tests needs to dismiss the notification, write a dismiss method with all the steps necessary. [14:47] elopio: yep, well, actually right now the dismissing of the notification is not part of the test, it is just to prevent the notification from getting in the way of the subsequent tests [14:47] boiko: fgimenez: it's worth analyzing that if the test is not testing the notifications, maybe we can stop the notification service or something like that. [14:48] but we'll need other tests that check the interaction with the notifications. [14:49] elopio: yep, FYI: the test I am working on is to check that dialer-app is doing the right thing on a valid USSD session, the dialog appearing is just a side effect [14:52] right, I guess that if there's no risk of the USSD session breaking the notifications, then you can split that test. Maybe even test the notification integration with lower level API tests. [14:54] elopio: still I need the dialog dismissed somehow at the end of the test :) [14:56] elopio: I was talking to tiagosh and he gave me the idea of killing the telephony-service-indicator process at the end of the test [14:56] elopio: that might be the easiest fix for now [14:57] boiko: I let you to discuss about it with fgimenez. Whatever solution you two come out with will be good for me. [14:57] elopio: thanks [15:00] boiko, i guess the notification doesn't have any control to interact with, right? [15:01] fgimenez: it has two buttons, "Ok" and "Cancel" but they both just dismiss the notification itself [15:02] fgimenez: in this particular case it is useless, but we have also tests for incoming calls that show the dialog to accept/reject the call [15:03] fgimenez: in that case it would be nice to be able to actually click the notification. Right now what we do is to trigger a dbus call to telephony-service-approver accepting/declining the call [15:05] boiko, you can create classes that expose methods for interacting with each kind of notification, for instance http://bazaar.launchpad.net/~canonical-platform-qa/ubuntu-sanity-tests/trunk/view/head:/ubuntu_sanity_tests/helpers/unity8/settings_wizard.py#L419 [15:05] fgimenez: the problem is: the notifications themselves are in unity (which is not running in testability mode by default) and dialer-app is in a different process [15:06] boiko, you have to restart it in testability mode first [15:07] boiko, then do your test [15:08] brendand_: ok [15:09] fgimenez: brendand_: thanks for the help, I will try to cook something in the next few days, let's see how it goes [15:15] boiko, you can ping the vanguard for review [15:15] fgimenez: nice! thanks [15:23] ubuntu-qa - anyone can confirm this sanity suite regression? unity8.shell.emulators.UnityEmulatorException: No scope found with id clickscope [15:23] brendand_, I've seen that one before [15:25] cking ping hey that rhymes :) [15:25] cking I'm wanting to persuade you to present some of you coverity findings during a UOS session next week, any interest? [15:25] rhuddie, intermittently? [15:26] rhuddie, it seems to be plaguing the whole suite now [15:26] rhuddie, unless i did something wrong maybe [15:26] brendand_, when I saw it, it was intermittent and very hard to reproduce. [15:27] brendand_, so that sounds different [15:27] alesage, don't you mean cppcheck findings? [15:27] cking exactly :) [15:27] cking, I'd like to do a "static analysis" topic [15:28] alesage, actually, I was aiming to just use these findings to try and get leverage and a coverty scan license [15:28] cking, no better forum to publicize this work :) [15:29] cking, but you have worked with the coverity scanner, yes? or am I misremembering [15:30] alesage, coverity scan is free and superior to cppcheck anyday, but it has big limitations to the number of times one can use it per week on large projects [15:30] cking, I'm very curious about the "onboarding" process for the free service [15:30] cking, I see [15:31] alesage, it is limited, one has to either own the project or be a significant author to sign a project up to it [15:31] cking do you get access to their web console, etc.? [15:31] alesage, yep, I get access to the webby front-end goodness [15:31] cking, anyway the session could literally be me asking you these questions :) and maybe us presenting a few interesting defects [15:32] i use it for all projects I package into debian [15:32] cking, doesn't have to be longer than 15 min [15:32] cking, balloons and I trying to sort some QA-relevant sessions [15:33] projects-team: I'm breaking the rule of 2 because these cards are boooring. [15:33] alesage, i'm more convinced we'll get better coverage if I blog about it, I get tens of K hits on my blog [15:33] I'm done, but waiting for three jenkins runs. [15:33] alesage, and I'm bogged down with some critical work that needs attendind to at the mo [15:34] cking, whell that's fair, can't stand in the way of your work [15:34] cking, thanks for considering [15:34] cking I'll have more questions at some point, ok? [15:35] alesage, ask anytime, I'm very happy to discuss it [15:35] however, I really think the perfect solution is to get a coverity scan licence for devs and the QA process [15:36] cking, alesage thanks for considering [15:36] cking, we'll have to discuss that further--would you be willing to give me a tour of the free process at some point? [15:38] alesage, sure, I can do that, it's very simple, here's a starter: https://scan.coverity.com/faq [15:39] alesage, sign yourself up, and I'll add you to a project so you can see what it looks like [15:39] cking, thanks [15:44] vila: are you ok with this one? https://code.launchpad.net/~canonical-platform-qa/qakit/flake8-fixes/+merge/256674 [15:44] elopio: I approved it long ago didn't I ? [15:45] elopio: fixing the key exposure will require changing it anyway [15:46] I can't top-approve it :/ [15:46] alesage: can you? [15:46] elopio, let me check [15:47] elopio: hmm, I can't either, NeedsFixing ;-) [15:52] elopio I can't either :/ [15:53] vila: oh, you figure the phone ran out of space and thus it couldn't create /tmp/ssh_askpass.XX any more? === chihchun is now known as chihchun_afk [16:01] elopio, vila, seeing the problem one sec [16:05] oSoMoN_: alex-abreu: returning to the issue killing the web container, is there something else you need from us? [16:05] elopio, not at this point, alex-abreu is debugging the issue [16:06] cool. The bug fix in autopilot is easy, but the release is painful so it will take some time. [16:24] thanks alesage. I suppose we don't have autolanding, right? [16:25] elopio, not sure if thomi set it up, assume no [16:25] elopio will look into [16:30] wxl: now I am awake enough to think about that - what you asked wasn;t at all offtopic for this channel :) [17:12] elfy: well, tangenital :) === chihchun_afk is now known as chihchun === om26er_ is now known as om26er [18:50] balloons: ping if you're still around, and don't mind a PM [19:07] teward, pong [19:07] balloons: mind a PM? [19:07] go for it