| === chihchun_afk is now known as chihchun | ||
| === chihchun is now known as chihchun_afk | ||
| rhuddie | Hi pitti | 10:10 |
|---|---|---|
| pitti | hey rhuddie, how are you? | 10:11 |
| rhuddie | pitti, I'm great thanks, and you? | 10:11 |
| pitti | quite well too, thanks; slowly winning over NetworkManager's configuration syntax :) | 10:11 |
| rhuddie | oh, well, I hope I don't cause you more problems :) | 10:11 |
| rhuddie | I've found an issue using autopkgtest with latest phone images | 10:12 |
| rhuddie | the test runs get stuck at "Waiting for desktop to boot" | 10:12 |
| rhuddie | see here: https://anonscm.debian.org/git/autopkgtest/autopkgtest.git/tree/ssh-setup/adb#n219 | 10:13 |
| rhuddie | it looks that the dbus interface is changed | 10:13 |
| rhuddie | The required IsActive property is removed and replaced with com.canonical.Unity.FocusInfo, where isPidFocused method takes required pid and returns bool | 10:14 |
| pitti | urgh | 10:14 |
| rhuddie | yes :) | 10:15 |
| pitti | how unfriendly | 10:15 |
| rhuddie | yes, unfortunately I think this will affect all the automated phone testing | 10:16 |
| pitti | rhuddie: do you have replacement command? | 10:16 |
| pitti | not sure how the emulator works these days, if it can still boot phone images I can use that | 10:17 |
| rhuddie | pitti, this one seems to work, but I'm not sure that it will do exactly as the old one was: gdbus call --timeout 5 --session --dest com.canonical.Unity.FocusInfo --object-path / --method com.canonical.Unity.FocusInfo.isPidFocused $(pidof unity8) | 10:18 |
| pitti | rhuddie: but how is that specific to the greeter? | 10:20 |
| pitti | wouldn't unity8 also be focussed when unlocked? | 10:20 |
| pitti | oh, that doesn't matter actually | 10:20 |
| pitti | we just need some indication of "unity is ready" | 10:20 |
| rhuddie | pitti, I'm not sure of that. I can give it a try and run locally to see if it waits appropriately | 10:21 |
| rhuddie | pitti, the good news is that the command did wait for unity8 to be in running state before proceeding | 10:33 |
| rhuddie | pitti, the bad news is that the HideGreeter method is also removed, so it can't unlock the greeter | 10:34 |
| pitti | rhuddie: hmm; it seems that needs to be put back then | 10:35 |
| pitti | we must have some noninteractive way to get rid of the gretter.. | 10:35 |
| pitti | how come that this could even land? I thought we ran tests on MPs | 10:35 |
| rhuddie | yes, clearly some hole in the process | 10:36 |
| davmor2 | pitti, rhuddie: it was a required landing for a security fix for the osk in xmir apps as it could cross contaminate other x apps apparently | 10:39 |
| pitti | err | 10:40 |
| pitti | that sounds wrong | 10:40 |
| davmor2 | sorry I mean other xapp could see key presses | 10:40 |
| pitti | if I can send out arbitrary D-Bus commands, I already 0wn the box, so what's the point of dropping that API? | 10:40 |
| davmor2 | pitti: it was trigging osk on xmir not the app inside xmir so it needed locking down | 10:41 |
| davmor2 | pitti: so if you had 2 xapps open both of them would see the key presses, now it is done on the active pid which is obviously what is causing this issue :( | 10:43 |
| pitti | davmor2: but how is that related to removing a d-bus method? | 10:43 |
| davmor2 | pitti: I assume it is how they are triggering the osk correctly with the locked down dbus call saviq/mzannetti can possibly help you more with detail | 10:44 |
| pitti | davmor2: well, SEP :) | 10:45 |
| pitti | but thanks for the heads-up | 10:45 |
| davmor2 | pitti: https://trello.com/c/XFw0bn8O/3386-1595-ubuntu-landing-046-ota12-qtmir-elleo I don't think there is much more info in here to help you | 10:46 |
| === chihchun_afk is now known as chihchun | ||
| rhuddie | pitti, I've raised a bug about the missing dbus methods, it seems it wasn't an intended change: https://bugs.launchpad.net/canonical-devices-system-image/+bug/1599145 | 13:00 |
| ubot5 | Launchpad bug 1599145 in Canonical System Image "com.canonical.UnityGreeter dbus methods removed" [Critical,Triaged] | 13:00 |
| pitti | rhuddie: oh, thanks! good catch | 13:22 |
| rhuddie | pitti, no prob. best thing is there should be no change needed in autopkgtest :) | 13:24 |
| flocculant | wxl: likely your images will implode on install with bug 1599174 , currently it installs but hasn't built *today* yet | 14:32 |
| ubot5 | bug 1599174 in ubiquity (Ubuntu) "dpkg seg fault warning during install" [Undecided,Confirmed] https://launchpad.net/bugs/1599174 | 14:32 |
| === chihchun is now known as chihchun_afk | ||
| flocculant | tsimonq2: what does huh mean? easier to actually ask something ;) | 16:11 |
| tsimonq2 | flocculant: well you said Lubuntu wasn't building, I had to take a second look :) | 16:12 |
| tsimonq2 | or hadn't built yet | 16:12 |
| tsimonq2 | I was just wondering if something was off | 16:12 |
| flocculant | aah | 16:27 |
| flocculant | it hadn't built yet - so installed - suspect it will fail to reboot when it has built today :) | 16:28 |
| tsimonq2 | flocculant: ...why? | 16:32 |
| flocculant | nvm - I left a note for wxl to read | 16:34 |
| flocculant | wxl: as suspected :) | 18:49 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!