=== mapreri_ is now known as mapreri === ToyKeepe- is now known as ToyKeeper [04:58] Good morning [05:05] good morning [05:24] pitti, could you have a look at bug 933199 ? it a missing mkdir when the sandbox is created, unless it has to be created manually of course. [05:24] bug 933199 in apport (Ubuntu) "apport-retrace crashed with IOError in _search_contents(): [Errno 2] No such file or directory: u'~/.cache/apport/retrace/Ubuntu 12.04/Contents-i386.gz'" [Medium,Triaged] https://launchpad.net/bugs/933199 [05:27] jibel: looking [05:36] morning all [05:52] jibel: ok, finally got to this (sorry, some IRC diversions) [05:52] jibel: do you actually encounter this with current saucy? [05:53] pitti, yes, I do. [05:56] pitti, here is the trace on saucy http://paste.ubuntu.com/6174319/ [05:57] jibel: merci [06:06] jibel: so the workaround is indeed to manually run mkdir -p on that path [06:06] * pitti commits fix to trunk [06:45] pitti, thanks [08:08] thomi: seems autopilot run doesn't support @unittest.skipUnless? [08:08] is there another way to conditionally skip a test? [08:08] pitti: dunno, it supports testtools.skip* decorators [08:08] skipIf, skipUnless [08:09] thomi: ah, thanks [08:09] (*slightly grumble*..) [08:09] heh [08:11] ok, that works [08:13] * thomi -> EOD === vila is now known as vila-afk-biab [09:05] morning all === vrruiz_ is now known as rvr [10:02] Morning all [10:04] hey davmor2 === vila-afk-biab is now known as vila === _salem is now known as salem_ [14:48] good morrning. [14:48] morning elopio [14:50] morning elopio I do use this channel :P [14:51] * balloons notes davmor2 looks familar [14:52] morning folks.. morning davmor2 [14:53] elopio, do you think Autopilot feels slow launching and killing apps? I've been deep in the code and I think I have a nice optimization I can implement to make it a bit faster [14:54] cgoldberg: I had a bug where it took 15 seconds between launching and beeing able to interact with the application. [14:54] but after that, it seems normal. If you make it faster, I'm sure I [14:54] 'll like it mroe. [14:54] *more. [14:55] elopio, there is a 1 sec sleep inside launching, and another 1 sec sleep inside killing. I think I can get rid of those and still have it work correctly [14:55] but i need to write some tests for AutopilotTestCase first [14:57] cgoldberg: yes, please add the tests. With that at least we will know why we need the sleeps. [14:57] elopio, btw..unrelated... this script is awesome... you should add it to your .bashrc: http://iloveubuntu.net/beautify-your-terminal-screenfetch [14:57] https://plus.google.com/114546378907380458640/posts/1wzh9o93oci [14:59] cgoldberg, thomi recently did some work in that area of long-launch times [15:00] pitti, can I request you to get this MP in this week? https://code.launchpad.net/~dpniel/autopilot-gtk/autopilotgtkemulators/+merge/187673 [15:00] cgoldberg: :) I'll try it. [15:00] balloons, i'll talk to him when he gets on.. thanks. i'm gonna give my optimization a shot also. we need tests around AutopilotTestCase anyways.. there aren't any [15:01] cgoldberg: hey, I got the SSD. I just forgot to order a cable... I'll install it in the afternoon. [15:01] balloons: sorry, was a victim of TODO overflow.. I'll try really hard [15:01] elopio, you can probably use your existing sata cable that your old drive used [15:01] pitti, I know the feeling, but I'll be your reminder (at least I'll try) :-) [15:02] cgoldberg: I'd like to keep my old drive too. It's not completely broken. [15:02] ah [15:08] so DanChapman have a pleasant weekend mate? [15:09] balloons, Yeah was nice and relaxed thanks. How about you? [15:10] DanChapman, same. Had a lovely spontaneous sunday afternoon drive yesterday which was very relaxing [15:12] balloons: sounds lovely. :-) [15:14] DanChapman, it was. It's interesting driving through a place and taking a slightly different turn on a whim. Sometimes you discover big new things that were always right there [15:16] hi balloons DanChapman [15:17] hi elfy! [15:17] So many people around here! [15:17] 108 [15:17] even ... [15:18] or 106 [15:18] * DanChapman wishes he had a car again! [15:18] hey elfy o/ [15:18] DanChapman: I've not had one for almost 2 years - not something I miss much [15:23] having one car is handy for trips.. But I like it purely for pleasure, not as require transport [15:23] *required [15:23] I have access to one if it's required === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha === chihchun is now known as chihchun_afk === mhall119_ is now known as mhall119 === fginther is now known as fginther|lunch [16:56] ping ubuntu-qa - curios if there is a system level smoke test plan for the phone somewhere [16:57] no idea. [16:57] jfunk: I'm writing the test cases for publishing, reviewing, installing and updating click apps. [16:57] balloons: is this something you might know about? ^ [17:01] jfunk, system level? what would you be after specifically? [17:01] elopio, ohh really? didn't realize you were doing any click stuff [17:03] balloons: yes, I'll be working with mmcc automating the tests for the unity app preview, the U1 integration on system settings and the click update manager. [17:03] balloons, I mean a set of tests that briefly touches everything that a user touches [17:03] on the phone [17:04] a system wide tyre-kick [17:04] jfunk: that's a big set of tests :) [17:05] jfunk, ok, well there is nothing that can be considered definitive in that way. What you used to see on the tracker was probably the closest thing that has ever existed to such a suite [17:06] tracker? [17:06] jfunk: the test tracker for iso testing [17:06] qatracker, when the images were still managed in there [17:06] right [17:06] as far as testing and etc go.. For the moment, they are being handled outside of that === psivaa is now known as psivaa-afk === fginther|lunch is now known as fginther === wxl_ is now known as wxl === om26er_ is now known as om26er [18:57] good morning all [19:10] thomi good morning [19:11] morning alesage [19:21] fginther: some of the runs on that bug have all the tests failing. [19:21] I suspect that's because we couldn't open the application. So, I'll work on that test first. Make sure the app is visible before going on. [19:22] elopio, For each of those runs, I found a traceback that looked similar to a failed tab switch, but I didn't look beyond why that failed [19:24] thomi: is there a way to check if an application is on foreground? [19:28] fginther: yes, I've just checked that even if the app is on the background, it will say it's visible. [19:28] balloons, hello [19:29] I'm not sure who's to blame about that. I suppose Qt, and that's going to be a pain. [19:30] letozaf, hello [19:30] letozaf, how's docviewer? [19:31] mzanetti: ping, you around? [19:31] balloons, fine fixed that unstable state [19:31] elopio: hi [19:31] balloons, added two tests and integrated emulatro [19:31] sorry emulator [19:32] mzanetti: hey. I've just found that we are checking if the main view of the application is visible before starting the tests. But it's always True even when it's behind another window, thus not visible. [19:33] mzanetti: do you know if there's a property or a way to check if it's actually being shown? [19:33] balloons, yesterday I had a look at the stock ticker app, do you know two tests fail ? [19:33] elopio: yeah, QQuickView should have such a thing [19:33] letozaf, did you see my conversion in progress on stock ticker? [19:33] letozaf, https://code.launchpad.net/~nskaggs/stock-ticker-mobile-app/test-click-support/+merge/187950 [19:34] balloons, nope, gonne look now :p [19:34] gonna not gonne [19:34] mzanetti: I'm looking at the properties of the QQuickView, but I can't see any of them changing when I bring the app forward. [19:34] letozaf, it doesn't work, lol, and I can't figure out why.. there's something funny with it [19:35] balloons, :( [19:35] balloons, well if you cannot find the problem I doubt I will be able :P [19:36] balloons, but I can just take a look [19:36] letozaf, yes, go for it [19:36] balloons, ok [19:37] elopio: hmm... no... there doesn't seem to be such a thing. actually how should the application know that? [19:37] elopio: its the window manager that has that information [19:39] mzanetti: I was hoping for the application be able to know if it was actually being rendered. But yes, sounds complex. [19:39] elopio: well, it is rendered [19:39] elopio: but it can't know it the window manager paints something on top [19:39] balloons, :( I see docviewer fails too :'( [19:40] elopio: why is there something on top? [19:40] mzanetti: sometimes, because we don't have a desktop file so unity doesn't launch it properly. [19:40] other times, because we have the screen locked. [19:40] letozaf, you are missing a comma [19:41] in this case I'm debugging, I have no idea. I'm just trying to make sure no test continues if the application is not on top. [19:41] balloons, thanks, let me check [19:44] balloons, fixed, I really didn't see it [19:44] * letozaf needs to change glasses [19:47] balloons, I see you put a comma at the end of the "Depends" last line : ubuntu-ui-toolkit-autopilot, do we need a comma also on the last line ? [19:48] letozaf: it's good to have - means adding a new dependency only adds one diff line [19:48] but it's not required [19:48] thomi, thanks :) [19:48] * balloons finds it annoying, lol [19:49] but ty thomi for applying some reasoning [19:49] so letozaf put the comma in.. for diff's sake :-) [19:49] balloons, ok [19:50] balloons, done [19:55] balloons, OMG [19:56] thomi: can we get some help from mir or autopilot to tell if an app is actually being shown? [19:56] letozaf, ohh nice [19:56] letozaf, that's a weird archive issue it seems [19:58] balloons, how do you fix that ? [20:03] letozaf, well it's odd, but it's not a problem with your merge i don;'t think. Just need to wait it out [20:03] balloons, ok [20:03] the package will build out and be availible shortly [20:36] balloons, meh! cannot find what the problem is with your conversion in pogress on stock ticker, everything looks fine but it fails to get main_view visible property [20:37] letozaf, yep, indeed it does [20:37] I suspect the qml is to blae [20:38] *blame [20:42] balloons, going to bed it's getting late for me, sorry could not help much :p [20:42] letozaf, no worries. There are other apps and things to do [20:42] which don't require you to pound your head at a problem [20:42] I like letting things sit for a bit and coming back after doing something successfully [20:43] I'll figure it out I'm sure [20:43] balloons, ok [20:43] letozaf, ciao, sleep well [20:43] balloons, ciao [20:51] elopio, how do I mark an MP as a WiP? so nobody will review it yet? [20:51] cgoldberg: did you already make the MP? [20:51] elopio, yes.. just did [20:52] i can delete it and make another? [20:52] ok, then on top it will say Status: Need review. Click there and change to Work in process. [20:53] ah.. nice. thanks [20:53] if you are just making the proposal, there's a checkbox with the misleading label: "Needs review". So, uncheck it. [20:53] gotcha [20:55] elopio, wanna look at my WiP? I'm trying to make Autopilot stream results as tests are run. my biggest pet peeve with AP is how you get no progress during a test run in non-verbose mode [20:55] cgoldberg: sure. To learn, mostly, because I missed the streaming branches on SST. [20:55] my branch works.. even with verbose mode.. but i'm probably missing some cases, like using junitxml [20:56] elopio, this doesn't use real streaming via subunit... it just adds a unittest.TextTestRunner [20:56] https://code.launchpad.net/~coreygoldberg/autopilot/text-testrunner/+merge/188451 [20:56] cgoldberg: oh, right. [20:56] it's small. [20:57] yup [20:57] I suppose we need a way to select the runner. [20:58] elopio, everything *should* work same as before.. plus output as the tests run [20:58] cgoldberg: yes, I think it's an good improvement. [20:58] it's great if it works correctly [20:59] I was wondering why autopilot remained silent while running. Now I see. [20:59] yea.. it waited for test.run(result) to return [21:00] cgoldberg: one thing that I always find a pain in the ass is the verbosity. You have the verbosity of the runner, and the log level of SST/Autopilot [21:00] and as on SST, autopilot uses -v to set the log level. [21:00] right.. they are sorta 2 different things though [21:01] verbosity of the runner just tells it to print dots... or testcase names [21:08] cgoldberg: yes. I think a runner should receive to parameters: verbosity, and log level. [21:08] veebers: I found an easy way to reproduce the None _Backend error. [21:11] https://code.launchpad.net/~elopio/ubuntu-ui-toolkit/fix1233364-assert_visible/ [21:12] with that branch: autopilot run ubuntuuitoolkit.tests.test_base [21:18] elopio: awesome thanks, I'm just looking into that now [21:42] fginther: one good news is that the failure that looks like String(u'Tab3') != 'Tab2' is because of a change in the toolkit behavior. [21:43] elopio, yeah! one problem solved [21:44] fginther: almost. I'm going to try accessing the buttons by their index property, not just by their index on the list. [21:51] elopio: for me to run that branch do I need to do anything (i.e. build/install anything)? Or can I just branch and run autopilot? [21:52] veebers: qmake && make [21:52] elopio: awesome, cheers [21:52] veebers: hum, and maybe apt-get install build-dep ubuntu-ui-toolkit-autopilot [21:53] elopio: would that be the same for the gallery-app would you know? [21:53] veebers: I have never ran the gallery-app. [21:53] elopio: ack [21:53] generally, the apps don't need a make. [21:54] elopio: right ok. It seems running ap tests for gallery-app needs something extra (or I'm doing something dumb) [21:54] veebers: what's the error? [21:55] elopio: ugh lost it now (building u-u-t). but using your branch will work for what I'm doing [22:00] EOD all, have a good one! [22:14] veebers: will this give me always the latest application I launched? [22:14] self.process_manager.get_running_applications()[0] [22:16] elopio: hmm, good question. let me refresh my memory [22:16] elopio: also note, this will only currently work on the deskop [22:16] desktop [22:16] really? [22:16] there isn't an interface to the process management on devices yet [22:16] that's sad. [22:16] yeah it's in the works [22:16] agreed [22:17] so, we basically have no way to tell if the application launched is visible and focused. [22:17] fginther ^ [22:17] * fginther head -> desk [22:18] veebers: who [22:18] veebers: who's doing that work? and how do we increase its priority? :) [22:19] elopio: there has been some discussion recently (https://bugs.launchpad.net/unity-mir/+bug/1232054) there isn't any one person on this right now [22:19] Ubuntu bug 1232054 in unity-mir "[mir] Need to expose geometry for autopilot consumption" [Critical,In progress] [22:21] critical, in progress sounds good. [22:22] elopio: Not sure if it helps you right now, if you're testcase inherits from unity8testcase then you can get the current focused application id [22:22] the application lifecycle tests in unity8 do this [22:22] veebers, "your" :) [22:22] veebers: no, that won't help. [22:22] knome: d'oh, thanks [22:23] if unity provides an API to get the focused application, that would help. [22:23] but that would make the tests work only on unity. [22:25] elopio: correct. The autopilot process management stuff will be the best bet, unfortunately I don't have an eta for it right now [22:30] ok, I'm blocked until the SDK team wakes up, so I'm going for lunch now. [22:30] bbl. [22:32] EOD.. later all o/ === salem_ is now known as _salem