=== chihchun_afk is now known as chihchun [02:04] thomi: that last MR has been approved [02:05] thanks veebers [03:28] Good morning [03:45] jibel: FYI, we are getting I/O errors during adt on amd64 again, so far all of them on wazn [04:02] jibel: I restarted these jobs and temporarily took wazn-adt offline [04:09] veebers: when you get a chance, super-easy MP for you: https://code.launchpad.net/~thomir/autopilot/trunk-add-help/+merge/187968 [04:11] Good morning [04:11] pitti, I will look [04:11] jibel: bonjour -- wow, early for you [04:13] pitti, heh, I'm trying new working hours to finish earlier in the afternoon for other activities :) [04:13] I tried the "can't sleep any more" this morning :/ [04:16] thomi: hm, why didn't that get a "PS Jenkins" review? [04:16] pitti: I only just created it - if we waited it would [04:17] thomi: reviewed, there's a typo [04:17] pitti: the CI system polls for MPs - from memory every 15 minutes, so on everage, it takes 7.5 minutes before a MP is picked up [04:17] pitti, did you receive notifications of the failures? I didn't [04:17] ah [04:17] jibel: yes, tons [04:17] To: ubuntu-testing-notifications@lists.ubuntu.com, jean-baptiste.lallement@canonical.com, martin.pitt@ubuntu.com [04:17] pitti: good catch, thanks [04:17] jibel: ^ spam? [04:17] thomi: looking now [04:17] oh, do you still want me to if pitti is? [04:18] also, morning pitti and jibel o/ [04:18] hey veebers, how are you? [04:18] hm [04:18] veebers: I just pushed a fix for the typo, perhaps you want to approve it? [04:18] thomi: can do [04:18] good morning veebers [04:18] pitti: good thanks [04:19] thomi: done [04:19] jibel: ok, u-drivers-common finished, that was the last one [04:19] hey jibel, I wonder if you could take a look at this when you get a chance? https://bugs.launchpad.net/autopilot/+bug/1229932 [04:19] Ubuntu bug 1229932 in Autopilot "errors running an autopilot test with Xephyr" [Undecided,New] [04:20] pitti, did any test you re-ran failed with IO error again? [04:20] jibel: no, they all succeeded [04:20] also pitti, there's this packaging problem in autopilot - I'm not sure what the issue is exactly, I wonder if you could take a look if you get a chance please? https://bugs.launchpad.net/autopilot/+bug/1227797 [04:20] Ubuntu bug 1227797 in Autopilot " Upgrading python-autopilot to 1.4 with the ppa:autopilot/ppa doesn't upgrade the libraries " [Undecided,Confirmed] [04:20] jibel: except for the first few which still got assigned to wazn (I retried the first few without disabling wazn) [04:21] ah, right [04:21] thomi: oh, I thought we added Breaks: for the older versions [04:21] pitti: we did, and I think that might be part of the problem [04:21] ah, just the other way around [04:22] python-autopilot is missing Breaks [04:23] pitti: it should break older versions of libautopilot-*? [04:23] thomi: I followed up to the bug [04:24] thomi: can do an MP if you are ok with that proposal [04:25] pitti: thanks - I can do the MP, and get you to review maybe [04:25] thomi: WFM [04:25] :) [04:25] thomi: did another followup [04:27] pitti: I don't understand your last comment... [04:27] pitti, there is also this error which does't make sense to me: qemu-system-x86_64: -drive file=[...].img,if=virtio: could not open disk image /dev/shm/adt/[...].img: No such file or directory [04:27] thomi: if we lift the unversioned recommends to versioned depends, there is no need for the Breaks: any more [04:27] it doesn't hurt, but it's redundant [04:28] pitti: gotchya [04:28] pitti: is there a de factor standard for the order of the Depends, Recommends, Breaks etc. clauses in debian/control stanzas? [04:29] thomi: usually in the order you stated, i. e. from "strong plus" to "strong minus", but it doesn't matter that much [04:29] cool [04:30] depends, recommends, suggests, enhances, breaks, conflicts, replaces [04:30] we so much need Likes:, GoesWellWith:, and MaybeIfIamInTheMood: ! [04:31] heh [04:31] InstallOnlyIfThePlanetsAlign [04:32] who needs determinism in their packaging system anyway? === chihchun is now known as chihchun_afk [04:44] ahhh, gmail suddenly decided to send all the jenkins notifications to spam, how to I tell it not to do that! [04:45] pitti: care to review those changes? 3 MPs are: https://code.launchpad.net/~thomir/autopilot/trunk-fix-deps/+merge/187970 and https://code.launchpad.net/~thomir/autopilot-gtk/trunk-fix-deps/+merge/187971 and https://code.launchpad.net/~thomir/autopilot-qt/trunk-fix-deps/+merge/187972 [04:50] thomi: done [04:51] pitti: thanks! [04:52] thomi: btw, do the gtk tests work for you ATM? they seem to fail here [04:52] pitti: yeah, they work for me, but you do need AP 1.4 installed [04:52] I do [04:52] which creates a small bootstrapping problem :) [04:52] oh [04:52] how do they fail? [04:53] interactively they take ages and all fail with dbus timeout [04:53] * pitti runs them again and gets a log [04:54] hmm, I closed the terminal window, but I just built the package for the last MP [04:54] pitti: although mterry reported a bug similar to this, but I haven't had a chance to look into it yet [04:55] thomi: http://paste.ubuntu.com/6161510/ [04:55] pitti: https://bugs.launchpad.net/autopilot/+bug/1224256 [04:55] Ubuntu bug 1224256 in Autopilot "autopilot-gtk hangs when launching an app on saucy" [Undecided,New] [04:55] thomi: might also be something which I screwed up on my system, of course [04:55] I use the exp PPA [04:55] Package: python-autopilot [04:55] Version: 1.4+14.04.20130917ubuntu.unity.next-1experimental340~ubuntu13.10.1 [04:55] hmmm [04:55] 1.4+14.04 ? [04:56] pitti: talk to didrocks about that :) [04:56] we'll need a bigger version for t then, but I guess we can call it 1.4.0 [04:56] I never understand those crazy versino numbers [04:56] *version [04:57] pitti: the log looks good... but it does make me think that StateNotFoundError should derive from AssertionError rather than RuntimeError [04:57] happy to take counsel on that though [04:57] how does it look good? [04:58] well, let's see what the PS jenkins bot says about the recent MP [04:58] pitti: oh, I see a few failures now [04:58] pitti, 14.04 is the version for the head branch because 13.10 is now considered 'stable' by the daily release system [04:58] hmmm [04:59] thomi: mterry's bug certainly is for 1.3, though? [05:00] pitti: yes, true [05:02] pitti, I didn't find anything from system logs on wazn. I switched it back online, and reprovisioning a new base VM (and told gmail not to move notifications to spam so I can monitor failures) [05:03] jibel: thanks; the I/O errors are only within the test VM, usually from dpkg (but I guess only because that just touches a lot of files) [05:04] pitti, right, but sometimes there are IO errors because of some failure on the host system [05:05] jibel: it's weird that it only affects 64 bit; sounds like the saucy guest 64 bit kernel doesn't work with the raring host kernel through kvm [05:05] I haven't yet seen I/O errors on i386 [05:07] but there is something going on on the host itself, I got this 'No such file or directory' error when starting qemu again [05:29] morning all [05:32] thomi: so https://jenkins.qa.ubuntu.com/job/autopilot-gtk-saucy-amd64-ci/26/console fail with the same StateNotFound errors [05:32] pitti: ahhh yes, I know why [05:32] thomi: seems something changed in ap in that regard? it still worked at least two weeks ago at the sprint [05:33] pitti: yes, select_single no longer reutrns None if nothing was found [05:33] now it raises an exception [05:33] oh, I remember that [05:33] I thought we already adjusted -gtk/-qt for that [05:33] I can get into it as soon as I finish this MP [05:33] but perhaps it was just "we need to remember to do it" [05:33] pitti: the libraries themselves didn't need any changes, but the libautopilot-gtk autopilot suite obviously does :) [05:34] ah, right [05:35] yes, I make quite a few assertions that objects of a certain kind aren't found [05:35] so these now need to turn into assertRaises [05:35] yeah [05:41] pitti: how well do you know python-mock? [05:41] thomi: I used it once or twice in apport's test suite, so "basic leve' [05:41] l [05:42] ok [05:42] nvm then :) [05:50] pitti: could you cast your eye over this please? https://code.launchpad.net/~thomir/autopilot/1.3-add-large-select-warning/+merge/187979 [05:50] if you approve I'll push a MP for trunk as well [05:53] thomi: done [05:57] pitti: same thing, but for trunk: https://code.launchpad.net/~thomir/autopilot/trunk-add-large-select-warning/+merge/187980 [05:58] done [05:58] cheers [06:05] pitti: are you happy if I put the fixes for the libautopilot-gtk autopilot tests in the same MP as the debian/control changes? [06:06] thomi: sure, WFM and keeps overhead low [06:08] cool - I think I've fixed them now. I realise that the issue I had earlier was my version of python-autopilot was earlier than when we introduced that change [06:08] so, doing a dist-upgrade now, will confirm that the tests WFM, and re-push [06:09] thomi: just setting $PYTHONPATH ought to work, too? that's how I used to test -gtk changes against ap trunk [06:09] (and calling ~/upstream/autopilot/bin/autopilot) [06:09] pitti: hmm, yes, good point :) [06:13] it turns out one of the hacks in the tests is a perfect place to use the new wait_select_single as well :) [06:14] bzr st [06:14] nice [06:14] * pitti puts thomi's mouse back into the right place for focusing [06:14] pitti: all pushed - care to re-review? https://code.launchpad.net/~thomir/autopilot-gtk/trunk-fix-deps/+merge/187971 [06:15] thanks for the mouse move, btw [06:15] thomi, hey. When accessing object properties does autopilot always return unicode for string values? I'm trying to test the language list for ubiquity are all unicode and not squares and stuff but running type against each one it returns a and using get_properties on the object the string looks like "string(u'lang')". So not sure how I can actually assert it is indeed a unicode string [06:15] thomi: does the more complicated assertThat(... raises()) give you anything over the standard assertRaises() ? [06:16] pitti: probably not, but we settled on using testtools matchers in autopilot itself, so I kind of use it everywhere now [06:16] DanChapman: hmm, I'll need to think about that, one second [06:16] * thomi scans the bowels of autopilots introspection module [06:17] thomi thanks :-) [06:17] DanChapman: autopilot --version says '1.3.XXX' right? [06:17] thomi: hmkay; just takes 4 lines where one would be enough [06:17] thomi yeah :-) [06:18] pitti: well, I coulda made it one line, but I wanted to make it pep8 compliant. The assertRaises version was already too long :P [06:19] thomi: thanks, re-approved [06:21] DanChapman: so.. this is an area that autopilot needs more tests in. [06:21] DanChapman: however... [06:22] DanChapman: if your data is being shipped over the wire in a dbus.String type, then it's being shipped as unicode (see: http://dbus.freedesktop.org/doc/dbus-python/api/dbus.String-class.html_ [06:22] err http://dbus.freedesktop.org/doc/dbus-python/api/dbus.String-class.html [06:22] inside autopilot, we'll store all string data as unicode internally (if we don't somewhere, then that's a bug that will get fixed in 1.4 most likely - please let me know if you see this) [06:23] DanChapman: if you're trying to work out if the string contains unicode characters or not... [06:24] DanChapman: something like this might help you out: http://stackoverflow.com/questions/196345/how-to-check-if-a-string-in-python-is-in-ascii [06:25] but ultimately, autopilot should 1) transmit all text content as unicode. 2) present all text data to the test author as unicode. [06:26] DanChapman: I hope that helps - I have to sign off for the evening, but feel free to send me an email if you have more problems. I'll PM you my address [06:27] thomi, yes thats great thanks :-) i'll have a read through those links and see how I get on [06:27] nw [06:28] * thomi -> EOW === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === DanChapman_ is now known as DanChapman === Ursinha is now known as Ursinha-afk [10:12] Morning all === psivaa is now known as psivaa-afk === mzanetti is now known as mzanetti|lunch === chihchun is now known as chihchun_afk === psivaa-afk is now known as psivaa === mzanetti|lunch is now known as mzanetti === _salem is now known as salem_ === iahmad is now known as iahmad|afk [14:15] good morning. [14:16] morning elopio [14:16] hello DanChapman. [14:16] hey om26er, are you here? [14:16] elopio, I am [14:17] om26er: did you write an emulator for the slider? [14:18] elopio, not yet. Is that needed now? I waited because I though it won't go in given the time of the cycle we are at [14:18] if its needed in the near time I can get it done before monday [14:18] om26er: I remember somebody write something, I just don't remember who. [14:19] elopio, I wrote the inital proof-of-concept I talked to you about it as well [14:19] https://bugs.launchpad.net/bugs/1228061 [14:19] Ubuntu bug 1228061 in ubuntu-ui-toolkit (Ubuntu) "[touch]Provide a sane solution to unlock the screen on touch devices during autopilot test run" [Medium,New] [14:19] om26er: yes, I know about that. [14:20] elopio, atleast I think its the best solution out there. does not restart unity8 all the time and many other benefits. I'll experiment with it today and tomorrow to put it in a sane place so that apps will use it [14:21] best as in what I could think of :p [14:22] om26er: it's a nice solution. Please try to make some self tests for it. [14:23] the self test will only run on the phones [14:23] ofcourse [14:38] om26er: yes. For now. [16:30] anybody up for doing a small MP/review for autopilot? (only documentation) [16:30] elopio maybe? ^^ [16:34] cgoldberg, I would, but I'm drowning in my own MP's today -- lots of them [16:35] balloons, thanks.. no prob.. ping me if you free up. It's a very small review to do [16:35] cgoldberg, if you can't get anyone else, well :-) [16:36] I think elopio will be around after lunch :-) [16:36] he needs a break from his tickets [16:36] cool [16:36] cgoldberg: give it to me. [16:36] I'm going for a walk, and then get back to it. [16:36] elopio, <3 :-p === Ursinha-afk is now known as Ursinha [17:52] elopio, https://code.launchpad.net/~coreygoldberg/autopilot/doc-update [17:58] om26er, ping [17:59] balloons, yo [18:09] cgoldberg: just one typo. I left my approval with that in the comment. [18:09] elopio, great thanks [18:10] cgoldberg: thanks to you. [18:10] I'm going to have lunch. [18:56] balloons, Hello! [18:57] letozaf, hello [18:58] balloons, I need your help ! :P if you got time [18:58] letozaf, go for it.. I'm working on the music app stuff you victor and I spoke about [18:58] I hope it all works [18:58] balloons, cool, can't wait to see it :) [18:59] well in theory, nothing should change, hah :-) [18:59] balloons, :) [18:59] balloons, I am trying to integrate emulator in ubuntu docviewer but there is something wrong, I've checked with what you did for instance with sudoku app [19:00] balloons, but it is not working, can you take a look to see what I'm doing wrong ? [19:00] balloons, https://code.launchpad.net/~carla-sella/sudoku-app/emulator-integration [19:00] balloons, to me everything looks fine, but it's not I get AttributeError: Class 'MainView' has no attribute 'get_object'. [19:01] balloons, but cannot find where I'm wrong :( [19:01] balloons, just run the test_open_text_file [19:02] letozaf, ohh, I can tell you just from the error I think.. there is no get_object method [19:02] convert it to a select_single [19:02] balloons, but it's in the emulators.py file [19:02] balloons, and it is select single [19:03] balloons, I more or less looked how you did in sudoku app :P but still does not work [19:03] letozaf, ohh, I think you sent the wrong link [19:04] that's for sudoku app not docviewer [19:04] this it? http://bazaar.launchpad.net/~carla-sella/ubuntu-rssreader-app/Emulator-integration/revision/23 [19:04] balloons, :( [19:04] it's under rssreader, oops :-) [19:05] balloons, :'( let me push it again [19:05] letozaf, before you do, add the emulators.py file [19:05] check your bzr status [19:06] balloons, https://code.launchpad.net/~carla-sella/ubuntu-docviewer-app/Emulator-integration [19:08] letozaf, got it [19:08] balloons, fiew, what a mess I was making :p [19:08] balloons, just the test_open_text_file [19:08] yes, I was looking at that one [19:09] I would still convert all the get_objjects [19:09] but you are corec, they exist [19:09] balloons, this thing has been driving me nuts [19:10] letozaf, if it makes you feel better I hit a brick wall on the stock ticker app in a similar way [19:10] this all looks good :-) [19:10] balloons, well yes it makes me feel better :p [19:10] letozaf, I would just convert it to self.app.select_single [19:11] sorry err self.main_view.select_single [19:11] let me try real quick and see [19:11] balloons, thanks :) [19:12] yep works fine [19:12] I'd drop the get_objects functions in the emulator [19:13] so it'll be blank it seems for now :-) [19:13] textArea = self.main_view.select_single("TextArea", objectName = "textAreaMain") [19:13] make sense, convert the rest and you should be good to go :-) [19:13] balloons, thanks a lot ! [19:13] if you'd like, add the click support to it also [19:13] balloons, I will carry on now [19:14] did you see how that was done in sudoku? [19:14] balloons, yes [19:14] kk, go for it :-) [19:14] * balloons high fives letozaf [19:14] * letozaf high fives back === salem_ is now known as _salem [20:26] balloons: I found the slider emulator: http://bazaar.launchpad.net/~hiroshidi/ubuntu-terminal-app/autopilot-header-and-settings/view/head:/tests/autopilot/ubuntu_terminal_app/emulators/main_window.py [20:26] I'll try to copy and test it over the weekend. [20:27] the hard part is that I would like the API to receive the slider value, not the position. So I'll have to find a way to convert that. [20:28] elopio, ohh. sorry, that was e [20:28] *me [20:29] :) [21:21] balloons, ping [21:21] om26er, pong [21:21] balloons, you are using logger.debug in a proposed branch can you tell me what do I need to import for it ? [21:21] it says [21:21] NameError: global name 'logger' is not defined [21:22] import logging [21:23] aah I though it was something autopilot specific [21:26] om26er, you also have to create the logger object you are using [21:26] logger = logging.getLogger(__name__) [21:26] logger.debug('foo') [21:27] cgoldberg, I think that's what I was looking for. thanks [21:27] om26er, .. and import logging like ballons said [21:27] yeah [21:28] http://unity.ubuntu.com/autopilot/tutorial/advanced_autopilot.html#test-logging [21:28] ^^ in our freshly themed new docs [21:31] :-) [21:51] hi