| elopio | veebers: I got the output for debian packages working. The click packages are getting stuck during app launch. I'll figure it out when I get back from the gym. And I'll add the rest of the apps too. | 00:05 |
|---|---|---|
| elopio | veebers: do you want me to send you the directories with the results? | 00:06 |
| elopio | veebers: I need to go now. I'll take care of the nfss script also when I get back. | 00:13 |
| elopio | bbl | 00:13 |
| veebers | elopio: sorry was at lunch. Following up now | 00:35 |
| veebers | elopio: fyi, looking at the nfss script right night, will have something to show you a little later on | 01:04 |
| veebers | (if you can please review those MPs from earlier so we can burn those cards) | 01:04 |
| elopio | ok, lets get this done | 03:31 |
| === roadmr is now known as roadmr_afk | ||
| veebers | elopio: you talking to me? :-) | 03:31 |
| === roadmr_afk is now known as roadmr | ||
| === enhydra` is now known as enhydra | ||
| === dkessel_ is now known as dkessel | ||
| === chihchun_afk is now known as chihchun | ||
| === roadmr is now known as roadmr_afk | ||
| === roadmr_afk is now known as roadmr | ||
| === roadmr is now known as roadmr_afk | ||
| === chihchun is now known as chihchun_afk | ||
| === pitti` is now known as pitti | ||
| === chihchun_afk is now known as chihchun | ||
| === roadmr_afk is now known as roadmr | ||
| === roadmr is now known as roadmr_afk | ||
| === roadmr_afk is now known as roadmr | ||
| === ara is now known as Guest56698 | ||
| === qwebirc88404 is now known as slickymasterWork | ||
| === chihchun is now known as chihchun_afk | ||
| === chihchun_afk is now known as chihchun | ||
| balloons | round 2! | 13:20 |
| DanChapman | balloons: hey! how's it going? Are you running any sessions tomorrow at 16:00 UTC? dpm kindly added testing topics to the etherpad so would be great to have you there to bounce stuff off if your available :-D | 13:23 |
| DanChapman | balloons: dekko session that is | 13:23 |
| balloons | DanChapman, hehe indeed. I know you know all about AP :-) | 13:23 |
| balloons | DanChapman, I don't see any conflicts at that time, happy to join | 13:24 |
| DanChapman | balloons: awesome thanks!! looking forward to it :-D . Although my AP skills are probably a bit rusty now ;-p | 13:27 |
| rhuddie | elopio, hey. when you are around could you give me a hand with getting health-check tests onto new project branch? | 14:39 |
| elopio | good morning | 14:43 |
| elopio | rhuddie: I'm here. | 14:43 |
| rhuddie | elopio, great | 14:44 |
| rhuddie | elopio, I registered a new project to put the health-check tests on, but I'm having probs actually pushing code to it | 14:44 |
| rhuddie | elopio, https://launchpad.net/ubuntu-health-check-tests | 14:44 |
| elopio | rhuddie: like this? | 14:46 |
| elopio | bzr push --use-existing-dir lp:ubuntu-health-check-tests | 14:46 |
| rhuddie | elopio, yes, I did try that, but I get this error: bzr: ERROR: Permission denied: "+branch/ubuntu-health-check-tests/trunk/" | 14:47 |
| elopio | um, that sounds familiar | 14:48 |
| rhuddie | elopio, but I've also seen this one too: http://pastebin.ubuntu.com/8988200/ | 14:48 |
| rhuddie | elopio, that one was where I tried to push from a branch of lp:ubuntu-test-cases/touch (to get the initial code populated before merging my changes) | 14:50 |
| rhuddie | elopio, I wasn't sure whether I should start with a branch of ubuntu-test-cases/touch, push that, and then try to merge my changes once that was done. | 14:51 |
| elopio | rhuddie: try pushing to lp:~ubuntu-test-case-dev/ubuntu-health-check-tests/touch | 14:51 |
| rhuddie | elopio, I'm fairly sure I tried that too, but let me double check | 14:51 |
| elopio | rhuddie: sorry, not touch. | 14:51 |
| elopio | trunk | 14:51 |
| rhuddie | yes, trunk | 14:52 |
| elopio | rhuddie: sorry, I don't know. But why are you making an independent project instead of merging it into ubuntu-test-cases/touch ? | 14:53 |
| rhuddie | elopio, that was the decision from yesterday's meeting | 14:53 |
| elopio | rhuddie: for all the project, or just for this one? I guess I missed that part. | 14:54 |
| rhuddie | elopio, just for the health-check one | 14:54 |
| elopio | rhuddie: got it. Maybe ask vila? | 14:55 |
| rhuddie | elopio, ok. | 14:55 |
| elopio | balloons: my sprint closing meeting is at the same time as the QML testing session. | 15:03 |
| elopio | balloons: was there something specific you wanted to ask me for that session? or just open discussion? | 15:04 |
| balloons | elopio, just open discussion | 15:09 |
| balloons | elopio, no worries if you can't make it | 15:10 |
| elopio | balloons: I think you know | 15:10 |
| elopio | my mind. | 15:11 |
| elopio | I won't make it, but feel free to ping me if needed. I don't think thomi will bother for a couple of minutes. | 15:11 |
| alesage | pitti, adt-clarification: I'm reading from man adt-run that exit 4 signifies at least one test has failed--just want to verify (question from veebers): should we all be using 4 to exit? (i.e. is this picked up in the infrastructure, informative, etc.) | 15:26 |
| pitti | alesage: well, if you want to; but I think the usual "non-zero is bad" should suffice in most cases; just adt-run differs between "some failures", "no tests", "some skipped", "bad testbed", etc. | 15:27 |
| pitti | alesage: e. g. unittest or autopilot already behave differently | 15:28 |
| alesage | pitti ok yes I thought I was just being more specific using 4 | 15:28 |
| alesage | pitti also I'd appreciate another review of that (our sprint is coming to a close ;) ) https://code.launchpad.net/~canonical-platform-qa/lrt/lrt-to-dep8/+merge/241353 | 15:28 |
| pitti | alesage: oh nice, flight-mode just works as user? | 16:01 |
| alesage | pitti let me verify, appears yes :) | 16:01 |
| alesage | (either this or the indicator crashed :) ) | 16:02 |
| pitti | alesage: ah yes, it' sjust a d-bus call | 16:02 |
| pitti | phablet@ubuntu-phablet:~$ /usr/share/urfkill/scripts/flight-mode 1 | 16:02 |
| pitti | dbus.exceptions.DBusException: org.freedesktop.URfkill.Daemon.Error.General: not authorized | 16:02 |
| pitti | :( | 16:02 |
| pitti | it might be limited to a local session | 16:03 |
| pitti | alesage: so this one is tricky; the script might be able to tell the network-indicator to do the request (as that's running in the unity session) | 16:04 |
| pitti | right, see /usr/share/polkit-1/actions/org.freedesktop.urfkill.policy | 16:05 |
| alesage | egads | 16:05 |
| alesage | we need a robotic hand to go in and re-enable instead :) | 16:05 |
| pitti | alesage: also, adt-run isn't using phablet-shell (although it's fairly close -- still ssh); it might actually behave differently there | 16:06 |
| alesage | pitti interesting | 16:07 |
| pitti | alesage: followed up to the branch with that summary | 16:10 |
| alesage | pitti thank you | 16:11 |
| alesage | (sorry in a mtg) | 16:11 |
| elopio | pitti: so, can we do reboots now? | 16:13 |
| pitti | elopio: yes, sir! | 16:13 |
| pitti | elopio: I figured that out for rhuddie, didn't know that you need it too | 16:14 |
| elopio | cool, just in time as I might need it now. | 16:14 |
| elopio | pitti: I didn't know either. Maybe the utah script I was looking at was doing a magic reboot | 16:14 |
| pitti | elopio: that is, reboots in qemu have worked for a long time, but I figure you mean touch :) you'll need 3.7.1git2 from vivid | 16:14 |
| * elopio updates | 16:14 | |
| pitti | elopio: hm, I didn't actually manage to get reboot working in UTAH | 16:15 |
| pitti | elopio: that was the original plan, to do systemd testing in UTAH, but I remember debugging that with psivaa and we couldn't get it to work | 16:15 |
| pitti | so it became an autopkgtest instead | 16:15 |
| pitti | (in retrospect I'm glad, as it's also making kernel testing and the like much better) | 16:15 |
| elopio | pitti: tests look much better this way. | 16:15 |
| rhuddie | pitti, I tested the reboots earlier on touch, it was working perfectly, thanks | 16:15 |
| pitti | \o/ | 16:16 |
| elopio | I'm puzzled as how this tests ever worked. | 16:16 |
| pitti | elopio: well, we didn't actually run a lot of them | 16:16 |
| pitti | Max had a look at the utah error back then, but it was too hard to quick-fix, so we just gave up | 16:17 |
| pitti | anyway, history | 16:17 |
| elopio | and we have a data file \o/ | 16:35 |
| elopio | rhuddie: did you figure out the bzr problem? | 16:36 |
| elopio | ping nuclearbob: | 16:37 |
| rhuddie | elopio, somehow! | 16:37 |
| rhuddie | elopio, https://code.launchpad.net/~ubuntu-test-case-dev/ubuntu-health-check-tests/trunk | 16:37 |
| elopio | nuclearbob: I need some help with this: http://bazaar.launchpad.net/~veebers/+junk/app-startup-nfss-gen/view/head:/scripts/app-startup-nfss-datafile-generator.py | 16:37 |
| elopio | rhuddie: cool :) | 16:38 |
| elopio | alesage: what about you? are you missing something? | 16:38 |
| alesage | elopio in a minute, thanks--just writing an e-mail | 16:38 |
| === chihchun is now known as chihchun_afk | ||
| elopio | pitti: adt-run requires us to put a debian/changelog file. I found that weird. Is there a reason? | 16:45 |
| pitti | elopio: oh -- I guess it uses that to determine the package version that you are trying to test, as we usually want this as an artifact | 16:47 |
| pitti | elopio: but this can easily be rectified, you'd then just won't get such a version in the output if it's not there | 16:47 |
| pitti | elopio: mind filing a but about it? | 16:47 |
| elopio | pitti: sure. What is the name of the project? | 16:47 |
| pitti | elopio: autopkgtest package (not project) | 16:48 |
| balloons | oO interesting | 16:49 |
| === chihchun_afk is now known as chihchun | ||
| elopio | pitti: I have the right version, but I got: | 16:51 |
| elopio | /sbin/autopkgtest-reboot: 3: /sbin/autopkgtest-reboot: cannot create /run/autopkgtest-reboot-mark: Permission denied | 16:51 |
| pitti | elopio: ah -- missing root | 16:54 |
| pitti | elopio: --- ssh -s adb -- -p s3kr1t | 16:54 |
| pitti | elopio: you are apparetnly not using 0000 as password :) (that's the one it tries by itself) | 16:54 |
| elopio | pitti: I'm passing it: --- ssh -s adb -P 1234 | 16:55 |
| elopio | should I add -- -p too? | 16:55 |
| pitti | elopio: it's ssh -s adb -- -p 1234 | 17:00 |
| pitti | elopio: the -- is important -- the -p is an argument for "adb", not to "ssh" | 17:00 |
| pitti | (and yes, the CLI is complicated -- you are dealing with three different programs :/) | 17:00 |
| pitti | adt-run, the virt runner (ssh), and the setup script for ssh (adb) | 17:01 |
| elopio | pitti: it is complex indeed. But I got the same error | 17:06 |
| elopio | $ adt-run -B --built-tree=tests/app-startup --setup-commands=tests/app-startup/debian/tests/setup --output-dir=output --- ssh -s adb -- -p 1234 | 17:06 |
| pitti | elopio: can I have the full log, please? | 17:07 |
| pitti | elopio: does it say anywhere something like "I have no root privileges"? | 17:08 |
| elopio | pitti: http://paste.ubuntu.com/8990168/ | 17:08 |
| elopio | oh, sorry, that one went with both -P and -p. It's the same with the line I pasted before. | 17:09 |
| elopio | I don't see any other errors. | 17:09 |
| pitti | elopio: ah! | 17:09 |
| pitti | elopio: that's output from the test itself | 17:09 |
| pitti | elopio: so, your test isn't running as root, thus it can't reboot | 17:09 |
| pitti | elopio: Restrictions: needs-root | 17:10 |
| elopio | pitti: oh, but then I get back to the ubuntu-app-launch error of yesterday. | 17:10 |
| pitti | elopio: argh -- I hope we won't need a suid root reboot helper :/ | 17:11 |
| pitti | elopio: when exactly do you need to reboot? | 17:11 |
| pitti | elopio: i. e. would it be possible to have something like: | 17:11 |
| pitti | Tests: autopilot | 17:12 |
| pitti | [...] | 17:12 |
| pitti | Tests: some-rooty-bits-which-reboot | 17:12 |
| pitti | Restrictions: needs-root | 17:12 |
| pitti | Tests: some-more-user-stuff | 17:12 |
| pitti | [...] | 17:12 |
| pitti | elopio: or asked differently, why do you need a complete reboot, as opposed to a mere session restart? | 17:14 |
| pitti | if your tests run as user, they shouln't be able to break anything on the machine which you couldn't revert as user? | 17:14 |
| elopio | pitti: ted said: <ted> ricmm, We reset the environment variables to set them before unity starts. | 17:15 |
| elopio | <ted> ricmm, So you can restart the session, but makes as much sense to reboot. | 17:15 |
| elopio | pitti: how do I restart the session? | 17:15 |
| pitti | elopio: in the simplest case, just kill -9 -1 :) but I'm sure that there's some more elaborate dbus-y way | 17:16 |
| elopio | pitti: that seems to do the trick allright. | 17:16 |
| pitti | elopio: so if you need a full reboot, you can still do the Tests: split from above | 17:17 |
| pitti | elopio: i. e. do preparation as root, then run the acual test as user | 17:17 |
| elopio | after I killed the session, I can't adb into the device anymore. I might have to try the other one. | 17:19 |
| pitti | elopio: ah right, you are killing adbd, it runs as user now | 17:20 |
| pitti | elopio: so that (dbus session shutdown), or test preparation as root, then the test as user? | 17:20 |
| elopio | pitti: test preparation as root. | 17:21 |
| elopio | that has the added benefit that I won't have to pass the setup script to adt. | 17:21 |
| pitti | right | 17:21 |
| alesage | o of course your desktop doesn't have flight mode, so you can't d-bus introspect to figure out which action to use >:( | 17:40 |
| alesage | need d-feet for touch | 17:41 |
| elopio | alesage: https://code.launchpad.net/~canonical-platform-qa/ubuntu-test-cases/dep8-app-startup/+merge/241613 | 17:43 |
| elopio | in case you have some time for a review. | 17:43 |
| alesage | elopio ok will switch to that | 17:43 |
| pitti | alesage: gdbus introspect works quite well on the phone | 17:43 |
| alesage | pitti ok was wondering which to use thx | 17:43 |
| pitti | alesage: in theory you can apt-get install d-feet on the device, and then use ssh -X to display it on your workstation; but it's pulling a ton of packages :/ | 17:44 |
| alesage | pitti whoa actual x-forwarding :) , maybe I'll actually get to the point of wanting that soon | 17:45 |
| alesage | that'd make a fun blog post | 17:45 |
| pitti | alesage: that bit works just fine; but it's needing gtk, libx11 and all that | 17:45 |
| pitti | (on the phone) | 17:46 |
| alesage | right would probably stomp what limited space I have | 17:46 |
| elopio | thanks alesage. | 17:49 |
| alesage | elopio, ping | 17:58 |
| === chihchun is now known as chihchun_afk | ||
| Letozaf_ | balloons, got time for filemanager ? | 21:19 |
| dkessel | balloons, knome - i am going to bed. just wanted to share a wip screenshot of my qa tracker branch: https://imgur.com/0dp7gsk . | 22:05 |
| dkessel | maybe you have some comments. but all in all, i am working to build what you guys gave me as a screenshot... | 22:05 |
| knome | looks good | 22:07 |
| knome | an good night! | 22:07 |
| knome | and too | 22:07 |
| balloons | dkessel, wow, that's pretty sweet already | 22:17 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!