| === ajmitch_ is now known as ajmitch | ||
| elopio | rhuddie, https://wiki.ubuntu.com/Process/Merges/TestPlan/ | 00:50 |
|---|---|---|
| === TheLordOfTime is now known as teward | ||
| elfy | morning DanChapman | 07:07 |
| DanChapman | morning elfy :-) | 07:09 |
| === zequence_ is now known as zequence | ||
| === lool- is now known as lool | ||
| === vrruiz_ is now known as rvr | ||
| === sfeole` is now known as sfeole | ||
| === fader_ is now known as fader | ||
| === charles_ is now known as charles | ||
| balloons | DanChapman: afternoon to you. Any luck/thoughts on ubiquity? | 16:02 |
| senan | DanChapman, hi | 16:08 |
| DanChapman | balloons, good morning to you sir. Well it's turning out to be a real PITA. See just trying to catch the exception is probably a rather bad idea as it could end up hiding other bugs. | 16:11 |
| DanChapman | xnox is there a way we could stop/disable the U1 sign in page for the ubiquity tests? | 16:11 |
| DanChapman | senan hey there sorry missed you yesterday. I have a snippet somewhere which solves the menu item problem (let me see if I can find it) | 16:13 |
| DanChapman | xnox_, ^^ | 16:16 |
| senan | DanChapman, Ok dan | 16:16 |
| xnox_ | DanChapman: yes, there is. | 16:34 |
| xnox_ | DanChapman: at the moment ubiquity has a critical bug and does not start, so i'm fixing that at the moment. | 16:34 |
| balloons | xnox_: DanChapman ok, if we can test without going through u1 screen (to avoid the crash for now), that would help the tests get integrated. I'll note evan also looked into why we aren't getting a good retrace on the crash; so there's some renewed hope in figuring it out | 17:11 |
| balloons | by integrated, I mean become part of the proposed->current migration :-) CI wants them green to land | 17:11 |
| DanChapman | xnox_, ok great, is it something I can sort out (if relatively trivial) as you sound rather busy? :-D | 17:11 |
| senan | DanChapman | 17:32 |
| senan | DanChapman, hi | 17:33 |
| senan | DanChapman, everything seems to be ok now.. I just resumitted the MP,can you please review it | 17:33 |
| elopio | om26er, http://bazaar.launchpad.net/~ubuntuone-hackers/unity-scope-click/trunk/view/head:/autopilot/unityclickscope/fake_services.py | 18:05 |
| balloons | DanChapman: can you reproduce the ubquity bug on your machine locally? | 18:16 |
| DanChapman | balloons, I achieved it twice for the first time yesterday. I havn't tried again today. I'll have a go now | 18:16 |
| balloons | ahh, jibel is mentioning if you have something (like a testcase) that makes it easy to reproduce he could have a look | 18:17 |
| balloons | he was also thinking about mocking out the tab to be blank during the tests | 18:18 |
| jtaylor | does jenkins store the installed packages for adt tests somewhere? | 18:32 |
| jtaylor | e.g. an archive of dpkg -l? | 18:32 |
| pitti | jtaylor: ATM adt-run doesn't do that, but I think it would be an useful thing to have | 19:31 |
| jtaylor | very useful | 19:31 |
| pitti | jtaylor: if you need it for a particular test, you can of course just run dpkg -l in your test and fish it out of the log | 19:31 |
| jtaylor | its not so easy to figure out what changed that caused something to fail | 19:31 |
| pitti | jtaylor: (I mean, if you are blocked by it) | 19:31 |
| jtaylor | I right now was able to get the list via a chroot I didn'T update yet | 19:32 |
| pitti | jtaylor: you can do something like "dpkg -l > $ADT_ARTIFACTS/packages.txt" in your test | 19:32 |
| jtaylor | is that in the spec? | 19:32 |
| jtaylor | ah it isn | 19:33 |
| jtaylor | need to update my runner then :) | 19:33 |
| pitti | jtaylor: it's a relatively recent addition | 19:33 |
| jtaylor | but the lsit of packages would be something sueful to have by default | 19:33 |
| DanChapman | balloons, . It's near on impossible to reproduce reliably with a simple testcase. From the autopilot log i got from one i produced locally i've found this http://paste.ubuntu.com/6972651/ it's raising a dbus.noreply exception when trying to select the main window but we do get the reply some 21 seconds later though. | 19:44 |
| DanChapman | thomi could you take a look ^^ could it be a bug between autopilot and dbus?? | 19:46 |
| DanChapman | here's a link to the full log http://ubuntuone.com/03EkHZPWWQgjYM7d9jbFVB | 19:48 |
| balloons | jibel: also of interest ^^ | 19:48 |
| jibel | the dbus exception is a consequence of the crash | 19:56 |
| jibel | i asked veebers to look at the crash report on errors.u.c | 19:57 |
| DanChapman | jibel, ok, just so I can understand this better. If the crash report on errors.u.c is saying it occurs in GtkNode::MatchStringProperty and the dbus exception is a consequence of this, how come it still eventually get's the reply 20 odd seconds after the crash? | 20:19 |
| Letozaf_ | balloons, hi | 20:36 |
| jibel | DanChapman, the query to dbus occurs a few seconds before the application crash, dbus doesn't know the state of the application, then the application crash, dbus cannot receive a reply and the query times out after 25s or so | 20:46 |
| balloons | Letozaf_: hello :-) | 20:54 |
| DanChapman | jibel, thanks :-D will a simple testcase using the testrunner be sufficient. I can't reproduce actually on my box but with the runner i can | 21:00 |
| DanChapman | jibel, lp:~dpniel/+junk/force_segfault_runner and run with "/run-autopilot-test --sdl --test tests ~/iso/path" and as usual use the config to set no shutdown | 21:08 |
| elopio | cgoldberg, https://jenkins.qa.ubuntu.com/job/webapps-demo-saucy-armhf-ci/19/console | 21:17 |
| elopio | cgoldberg, http://bazaar.launchpad.net/~elopio/webapps-demo/autopilot-tests-fixes/view/head:/debian/control | 21:19 |
| === _salem` is now known as _salem | ||
| jibel | balloons, if we set UBIQUITY_NO_SSO in the environment it should skip the 'Ubuntu One' page, I'll give it a try | 23:58 |
| balloons | jibel: wahoo! | 23:58 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!