| balloons | o/ | 16:26 |
|---|---|---|
| gang66 | Hello balloons | 16:27 |
| nik90 | hey | 16:27 |
| balloons | hey guys | 16:32 |
| balloons | so it's time to solve your autopilot issues! | 16:32 |
| nik90 | :) | 16:37 |
| balloons | just finishing up the last little details on https://core-apps-jenkins.ubuntu.com/ | 16:38 |
| nik90 | ooh nice | 16:39 |
| gang66 | ./ubuntu_clock_app/fixture_setup.py: 'sudo setprop custom.location.testing {}'.format(test), | 16:40 |
| gang66 | ./ubuntu_clock_app/fixture_setup.py: 'sudo restart ubuntu-location-service && ' | 16:40 |
| balloons | nik90, yea you can see my email. Login and you can build away | 16:40 |
| gang66 | Why do we need these sudo there? | 16:40 |
| balloons | you can't restart the location service without it. | 16:40 |
| balloons | it's more a bigger question of why we need to restart location service | 16:41 |
| nik90 | I don't remember adding that line either | 16:42 |
| nik90 | Could it be to show the location-trust-store prompt and press the "ok" button? | 16:43 |
| nik90 | at one point clock-app was failing because of it, and we fixed it once the trust-store became testable | 16:43 |
| balloons | right. If there's anymore of those, it's not a testable service, we should pushback on making sure it is testable. But I think we've rooted all those out | 16:44 |
| nik90 | balloons: I am admiring the jenkins service at the moment. Its so cool that I can provide it with branch link, framework to see if it passes or not! Superb! No more having to do empty commits in the Mp just to kick jenkins to rebuild and test. | 16:46 |
| nik90 | wait is jenkins running it on the krillin device *and* a local machine? Or just krillin atm? | 16:47 |
| balloons | nik90, jenkins builds on a machine, but the tests run on the krillin | 16:47 |
| * nik90 leaves university. bbs | 16:57 | |
| balloons | so, do the tests run for you guys? | 18:45 |
| balloons | as in, does the app launch and try to do something/ | 18:46 |
| gang66 | nope it is just crashing | 18:52 |
| nik90 | balloons: no it crashes after opening the app. Pretty much at MainView12 point | 18:53 |
| balloons | I'm trying again with a fresh branch | 18:54 |
| balloons | I'm not even getting that far, imho | 18:54 |
| gang66 | https://core-apps-jenkins.ubuntu.com/job/adt-krillin/15/artifact/logs/autopilot-stderr/*view*/ | 18:55 |
| gang66 | autopilot.exceptions.StateNotFoundError: Object not found with name 'MainView'. | 18:55 |
| gang66 | I'm curious why it cannot find MainView | 18:55 |
| balloons | there we go, finally it launches | 18:57 |
| balloons | ok, I'm on the same page as you now :-) | 18:57 |
| balloons | gang66, the issue with mainview is because of versioned qml classes and autopilot | 18:59 |
| balloons | indeed, mainview12 is the mainview, fun times | 19:00 |
| balloons | here's the bug https://bugs.launchpad.net/autopilot-qt/+bug/1341671 | 19:03 |
| ubot5 | Ubuntu bug 1341671 in Autopilot Qt Support "Versioned QML classes are not recognized by their public type name" [High,Confirmed] | 19:03 |
| gang66 | How we could workaround it? | 19:05 |
| balloons | I need to check and see if some code landed in autopilot 1.5 | 19:07 |
| balloons | the fix is really in autopilot itsel | 19:07 |
| balloons | indeed, it landed in 1.5, sweet | 19:08 |
| nik90 | oh | 19:09 |
| nik90 | what are we using in vivid jenkins at the moment? Autopilot 1.4 or 1.5? | 19:09 |
| balloons | vivid jenkins does make it hard | 19:10 |
| balloons | mmm.. but it's using the overlay, so we should be fine | 19:10 |
| balloons | yep, it should be using 1.5.1+15.04.20150908-0ubuntu1 | 19:11 |
| balloons | ok, so the fix is what's described in here https://code.launchpad.net/~canonical-platform-qa/autopilot/fix-cpo-having-different-name-1337004/+merge/262047/comments/668227 | 19:12 |
| balloons | nik90, you about? | 19:48 |
| nik90 | balloons: yes | 19:48 |
| balloons | so autopilot is complaining about finding multiple mainview matches | 19:48 |
| balloons | and indeed I see several mainviews amongst the qml files. I assume this hasn't changed from before | 19:49 |
| nik90 | multiple mainviews in the qml files? | 19:49 |
| balloons | app/upstreamcomponents/PageWithBottomEdge.qml: MainView { | 19:49 |
| balloons | app/ubuntu-clock-app.qml:MainView { | 19:49 |
| * nik90 checks | 19:49 | |
| nik90 | balloons: app/upstreamcomponents/PageWithBottomEdge.qml: MainView { doesn't exist | 19:50 |
| nik90 | it is a comment | 19:50 |
| nik90 | looking at http://bazaar.launchpad.net/~ubuntu-clock-dev/ubuntu-clock-app/trunk/view/head:/app/upstreamcomponents/PageWithBottomEdge.qml | 19:50 |
| balloons | ahh indeed ;-) | 19:50 |
| nik90 | ;) | 19:51 |
| balloons | Nevertheless, AP isn't happy. ValueError: More than one custom proxy class matches this object: Matching classes are: <class 'ubuntu_clock_app.PageWithBottomEdge'>,<class 'ubuntu_clock_app.ClockPage'>,<class 'ubuntu_clock_app.MainView'> | 19:51 |
| balloons | so I'll sort it | 19:52 |
| nik90 | ok, I am installing autopilot 1.5 from the PPA on my vivid machine now | 19:53 |
| nik90 | hmm I am actually in wily | 19:53 |
| balloons | you forgot you upgraded? | 19:53 |
| nik90 | doesn't 15.10 get autopilot 1.5 without PPAs? | 19:53 |
| balloons | yep | 19:53 |
| nik90 | its my development machine. So I kind of forgot about it :P | 19:53 |
| * nik90 updates | 19:54 | |
| balloons | I actually dropped one of my machines back to trusty for fun. So I have no vivid | 19:54 |
| nik90 | oh...I realize how easy it is when developing on the latest development release. | 19:55 |
| balloons | yea, autopilot is very unhappy | 20:04 |
| balloons | it can't even traverse the tree :-( | 20:04 |
| balloons | nik90, so I'd like to set an objectName for the MainView as one way to stop AP from being so confused. However, the app won't run if I set one | 20:08 |
| nik90 | hmm, the app doesn't run? | 20:09 |
| nik90 | let me try setting it | 20:09 |
| balloons | yes please | 20:09 |
| nik90 | balloons: in ubuntu-clock-appq.qml, we already have the objectName of mainView set to 'clock' | 20:11 |
| nik90 | http://bazaar.launchpad.net/~ubuntu-clock-dev/ubuntu-clock-app/trunk/view/head:/app/ubuntu-clock-app.qml#L33 | 20:12 |
| balloons | lol, ok let me try again | 20:12 |
| balloons | I tried the fix out on music, it works | 20:12 |
| balloons | so it should fix clock too | 20:12 |
| ahayzen | ...did someone mention music ? ;-) | 20:13 |
| balloons | ahayzen, so let me give you a little diff | 20:13 |
| nik90 | ;D | 20:13 |
| balloons | ahayzen, http://paste.ubuntu.com/12323350/ | 20:14 |
| ahayzen | heh the MainView vs MainView12 thing | 20:14 |
| balloons | ahayzen, that little method is a 'nice way' of working around the greater bug | 20:15 |
| ahayzen | looks it :-) | 20:15 |
| balloons | it also lets you set the query name for anything you wish, so you can call a local class 'andrewisawesome' and have it point at anything in the tree | 20:15 |
| ahayzen | sweet :-) | 20:16 |
| balloons | nik90, clock however isn't so nice.. It really is confused. I think we're hitting https://bugs.launchpad.net/autopilot/+bug/1350532 too | 20:17 |
| ubot5 | Ubuntu bug 1350532 in Autopilot "validate_dbus_object can cause more than one class in the cpo cache" [High,Confirmed] | 20:17 |
| nik90 | :/ | 20:18 |
| balloons | in a moment veebers will be here | 20:19 |
| balloons | then we'll get out the pitchforks | 20:19 |
| balloons | ahah, ok, so I get why now | 20:23 |
| balloons | woot, success | 20:24 |
| nik90 | wat did you do? | 20:25 |
| balloons | I did the simple fix, but I realized it was our own class declarations that was causing AP to go nuts | 20:25 |
| balloons | clockpage and pagewithbottomedge were both pulling in mainview also | 20:25 |
| nik90 | oh | 20:26 |
| balloons | i thought it was in the qml, but lol, nope it's in our helpers | 20:26 |
| balloons | ok, let me commit this | 20:26 |
| nik90 | so does it run now? | 20:26 |
| balloons | yea, it tries to do things now | 20:26 |
| nik90 | cool | 20:26 |
| balloons | lp:~nskaggs/ubuntu-clock-app/fix-ap-mainview | 20:27 |
| balloons | you still get my debugging info if you pull that now, lol | 20:28 |
| * nik90 test | 20:28 | |
| balloons | rev 380 is more or less a clean base to fix things | 20:29 |
| ahayzen | balloons, the new coreapps jenkins looks awesome btw :-) | 20:32 |
| balloons | I'm glad you like it. Some more goodies and tweaks to do, but it's ready to rock. | 20:33 |
| balloons | we've been delayed on getting the MP functionality working | 20:34 |
| balloons | I still have my MP ready and waiting on your project | 20:34 |
| balloons | nik90, so are you going to attempt to actual correct the tests now or ? I don't want us both working on the same thing, heh | 20:35 |
| nik90 | balloons: I am going to be real busy with uni for the next 1-2 weeks. But its something that me and gang66 are putting emphasis on for this release. | 20:36 |
| balloons | ok, I'll keep going and see if I can't get at least one test running and propose something | 20:36 |
| balloons | thanks for the help in deciphering this | 20:36 |
| nik90 | balloons: oke, but beware the whole navigation has changed..so its going to be lot of change to get that first test passing. | 20:37 |
| nik90 | (code wise) | 20:37 |
| balloons | righto | 20:37 |
| nik90 | balloons: and thnx a lot :) | 20:37 |
| veebers | barry: Hey if you're around, I believe this bug should be considered fix released? https://bugs.launchpad.net/autopilot/+bug/1488175 | 21:47 |
| ubot5 | Ubuntu bug 1488175 in Autopilot "FTBFS on Wily" [Undecided,In progress] | 21:47 |
| veebers | barry: unless there are other issues? | 21:47 |
| barry | veebers: i *think* it's due to testtools bugs, but this is mostly for tracking. if it's messing you up, the bug can be closed | 21:50 |
| veebers | barry: ah right, this is failing to build in the 3.5 ppa right? | 21:51 |
| barry | veebers: yep | 21:51 |
| barry | veebers: i suspect it would also fail if we rebuilt testtools in the archive | 21:51 |
| veebers | barry: ah ok, that's not good :-\ Is there anything I can do to help? Or is it waiting on something else | 21:56 |
| barry | veebers: upstream testtools isn't compatible w/ the version of twisted we have in the archive. there's a patch that *almost* gets us there, but i'm chasing down one last failure. i'll update the ap bug once it all gets straightened out. thanks for your help tho, much appreciated! | 21:57 |
| veebers | barry: nw, I'll be interested to see how it goes | 21:58 |
| barry | veebers: me too :) | 21:58 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!