=== Guest34025 is now known as `Chef [05:40] Hello [06:01] Good morning [06:44] good morning [07:07] bonjour didrocks ! as-tu passé un bon we? [07:14] pitti: bon week-end, merci, et toi ? [07:15] didrocks: pareil, c'était calme [07:15] il a plu :( [07:16] pitti: yeah, didn't stop raining here as well, so can't do a lot :( === jhenke_ is now known as jhenke [07:40] good morning desktopers [07:42] re seb128! [07:45] re :-) [07:55] good morning! Happy Monday :) [07:56] hey larsu, happy monday! [07:56] had a good w.e? [07:56] hi seb128! I did thanks. How was yours? [07:57] quite good, thanks! relaxed and had a fondue dinner with friends on saturday and played tennis yesterday [07:58] hey larsu, how are you? [07:58] didrocks: good thanks! how are you? [07:58] seb128: yummie [07:59] :-) [07:59] larsu: I'm great, thanks! [08:04] pitti: testing multiple autopktests at the same time for Make, medium ones (using another branch, which sources /etc/environment to docker config) and large! [08:04] at least, they all shows up correctly in the UI :) [08:22] pitti: ah, it seems though that it's reinstalling the ubuntu-make package in the first testbed :/ [08:22] pitti: so, I need to remove that from Depends and move the install (+ and keep optional uninstall if not system tests) in the setup script [08:28] hello world! [08:29] hey alexarnaud [08:31] good morning alexarnaud [08:36] didrocks: ah yes, every test will reinstall it if it's in Depends: [08:37] didrocks: did you already get some logs of the large tests? [08:37] didrocks: (sorry, was out for a bit for doctor) [08:38] pitti: ah, I understood the contrary on Friday when we say it was dummy and just comparing with previous tests test bed requirements [08:38] pitti: large and medium tests are still running [08:38] didrocks: ah, sorry for the misunderstanding [08:38] pitti: but medium doesn't find the docker index it seems [08:38] large are running (one failure for now) [08:38] didrocks: I meant that determining whether to *reset* the testbed doesn't check the actual package list [08:38] ok [08:38] didrocks: but dependencies will be installed always [08:39] I'll thus move apt install ubuntu-make in setup [08:39] (and remove from the dep) [08:39] pitti: so, it seems that sourcing /etc/environment didn't work for docker [08:39] or that the proxy doesn't allow this [08:39] didrocks: what do you expect in /etc/environment? [08:39] oh, proxy setting [08:39] yep [08:40] didrocks: proxy should allow accessing docker.io [08:40] I needed before to export http_proxy though [08:40] I don't know if docker config file doesn't expect "export" keyword [08:59] goooood morning all [09:01] hey willcooke [09:01] had a good w.e? [09:01] good morning willcooke [09:03] hello pals! [09:05] good thanks seb128 [09:06] hey Laney! [09:06] how is the u.k today? [09:07] * Sweet5hark managed to not read any email during vacation. First (smaller) inbox now has 714 new mails. [09:08] Wondering what the big one will bring ... [09:08] hey Sweet5hark, welcome back! did you have good holidays? [09:09] good morning Laney [09:09] hey Sweet5hark [09:09] w/b Sweet5hark [09:10] hey seb128 didrocks [09:10] seb128, didrocks, willcooke: morning all! [09:10] getting colder! [09:10] and Sweet5hark! [09:10] still around 6-7°C here [09:11] weather says 3° atm [09:12] seb128: holidays were great. Was at tenerife for a week. 25C and wonderful weather to fight off a winter depression. As a bonus spain is celebrating christmas until 5th january. So Santa Claus and stuff under palm trees ;) [09:12] Laney: heya. [09:13] Sweet5hark, nice :-) [09:13] you should have stayed and worked from there for a bit :-) [09:16] seb128: yeah, maybe. It was an incredibly cheap (for what I got) all-inclusive-travel booked on a short notice. The only downside was no flexibility on the dates and such. [09:16] I see [09:16] well at least if the week make you good it's great ;-) [09:16] made* [09:22] ... and the other inbox has 2589 new mails. [09:23] hey Sweet5hark, welcome back and "gesundes Neues!" [09:24] hey Laney [09:28] Sweet5hark, watched a few CCC videos over the weekend, lots of interesting stuff as always :) [09:32] willcooke: oh yes. [09:34] hey pitti! [09:34] you well? [09:35] willcooke: i was introduction a new visitor to the madness, so found less time to spend in picking up the usual conversations, 0-days and #vncroulette and stuff like that ... [09:37] willcooke: if you havent seen it yet, https://media.ccc.de/v/32c3-7483-computational_meta-psychology was a lot of fun -- esp. since someone mentioned the speaker having a cute resemble [09:37] thanks, I'll take a look :) [09:37] resemblance to Sheldon Cooper. [09:38] pitti: Frohes Neues! ;) [09:38] Sweet5hark, ha! Yes indeed [09:49] Laney, did you say you would reply on the desktop list about the calendar app choice? we sort of let the discussion pending but did the work, might be nice to wrap with a "that's what we did and why"? [09:54] yeah I should [09:55] pitti: I guess you need to kill the medium test for ubuntu-make-master [09:56] thanks [09:56] didrocks: done [09:56] jibel: hey, is there a new jenknis for the pending -> current tests now? [09:56] jenkins [09:56] didrocks: it would have timed out eventually, too [09:56] pitti: I will try a manual export for now and see if this work… [09:56] (to see if the issue is that it's just not exporting the env var) [09:57] Laney: should be https://platform-qa-jenkins.ubuntu.com/ now [09:57] pitti: good knowledge! [09:58] assuming it loads [09:59] Laney: it's VPN only [09:59] on .u.c? [09:59] Laney: I think all the Jenkaas instances are ATM [09:59] bleh [10:00] pitti: quite some large tests failure: https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-trusty-ubuntu-desktop-ubuntu-make/trusty/amd64/u/ubuntu-make-master/20160111_092137@/log.gz [10:01] sounds like starting apps in the session isn't working [10:01] * didrocks wonders why, working locally, should try with a trusty cloud image though [10:01] didrocks: ah, unity doesn't run? [10:01] sounds like it [10:01] only the non visual frameworks have passed [10:02] didrocks: do you reboot the VM or restart lightdm after installation? i. e. does that session actually run? [10:02] pitti: I do reboot it after setting up autologin, unsure if the session actually run [10:02] didrocks: also, the test runs a ssh session -- how does that do things in the unity session? [10:02] jibel: https://platform-qa-jenkins.ubuntu.com/job/mark-pending-current/8/console <- do you know if the keys got changed or something? [10:02] pitti: running dbus-launch and export DISPLAY [10:02] didrocks: but that's not sufficient [10:02] pitti: that's exactly what's passing on wily [10:02] the launcher is just a warning [10:03] hm, but I'm quite sure you also need to grab the real sessions' dbus [10:03] it's not what's making the thing failing [10:03] let me try first locally with a trusty cloud image [10:03] and see how this goes [10:03] you are using another upstart instance (or rather, none at all as ssh isn't a desktop session), different gvfs, bamf, can't talk to unity etc. [10:03] ok [10:04] pitti: yeah, not talking directly to unity isn't the issue making it fails here (but I'll change that at some point) [10:05] it's just not adding the dekstop file to the launcher [10:05] didrocks: so my best guess would be that the desktop session doesn't start up on trusty (LLVM pipe etc.), would that cause these errors? [10:06] pitti: yeah, and that seems to match the "it seems that only visual frameworks (needing really a X session)" are failing [10:07] at least, it means that download and extraction worked well [10:07] pitti: hum adt-buildvm-ubuntu-cloud -r trusty [10:07] Could not find setup-testbed script [10:07] * didrocks looks at the wifi [10:07] wiki [10:08] didrocks: ah, already fixed in git, sorry about that [10:08] didrocks: workaround: sudo chmod +x /usr/share/autopkgtest/setup-commands/setup-testbed [10:08] thx! [10:08] * didrocks reruns [10:13] pitti: chmod didn't fix it [10:13] didrocks: might be my fix https://anonscm.debian.org/cgit/autopkgtest/autopkgtest.git/commit/?id=0d7f5a9a516fc0005821fc136f1fb9dfe164672e [10:14] good job you have that 1gbps internet :) [10:15] Laney: yep :p [10:15] didrocks: argh; let me try here; 3.19 I suppose? [10:16] pitti: yeah [10:16] Laney, hey, yes the keys have been updated last week. There is a ticket open with IS. I'll see with nuclearbob when he's online today. [10:16] didrocks: 1 min to download cloud image [10:16] pitti: did you fix, rerunning [10:17] didrocks: ^ qoui ? [10:19] pitti: I applied https://anonscm.debian.org/cgit/autopkgtest/autopkgtest.git/commit/?id=0d7f5a9a516fc0005821fc136f1fb9dfe164672e and it works now [10:19] jibel: ok, might be worth poking IS assuming it's quick for them to do [10:20] didrocks: ah, so that was the other fix [10:20] jibel: is it right that all of these instances have to be vpn only ATM, BTW? [10:20] * Laney runs mark-current manually for now [10:20] didrocks: I thought about https://anonscm.debian.org/cgit/autopkgtest/autopkgtest.git/commit/?id=98e219ebfa93 [10:20] didrocks: but I guess you might have run into that next :) [10:21] yeah ;) [10:22] ok, image built, running now large test from this cloud image [10:23] didrocks: I hope you have acng [10:29] internal compiler error triggered by something on nux... i hate those errors :/ [10:43] Laney, it's fixed, it was an invalid path. Although there is still an 'Invalid Address' error in jenkins' logs but the copy works. [10:44] jibel: nice, thanks! [10:45] email notifications are misconfigured but only IS can fix it, it's relatively harmless,n we just don't receive notification on failure but since you ping us it's fine ;). I'll report it. [10:46] I did notice that on the Jenkaas wiki page there's a YES/NO for if it should be accessible to the public so it seems like this is supposed to work [10:48] pitti: hum, going for a run, but I don't find the ppa arguments in my adt-run [10:48] (which I need for deps on trusty) [10:49] didrocks: you mean how to tell adt-run to add an use a PPA? [10:50] didrocks: http://paste.ubuntu.com/14468336/ [10:51] pitti: ah, thanks! [10:51] didrocks: and yes, it's horrible :/ but I can't rely on add-apt-repository in the testbed [10:51] yeah [10:51] thx! [10:51] * didrocks really gone now [10:51] didrocks: bon cours ! [11:40] my dual-screen is not working =) [11:40] and i am sad =( [11:41] main screen is 1920x1080 the other is 1024x768 [11:41] * xnox ponders if kernel people broke it. [11:41] xnox: hit it, it works for gibbs [11:41] davmor2, did that. [11:42] xnox: damn it I'm all out of suggestions then I'm afraid ;) Maybe check your leads, and reboot [11:44] try old kernel then you can know to blame them [11:45] blaming kernel team is the best kind of blame :) [11:46] Laney: no blaming Saviq and ogra_ is the best blame cause they always redirect it to the right person almost instantly :) === hikiko is now known as hikiko|ln [12:59] Laney, hi, could you grab gettext 0.19.7-2ubuntu1 when you have a moment -- https://launchpad.net/~ricotz/+archive/ubuntu/staging/+sourcepub/5951902/+listing-archive-extra [13:03] ricotz: not right now but maybe later, would be good if you could file a sponsoring bug [13:03] thanks for working ont hat [13:06] pitti: hum, I reallyl can't seem to be able to add the ppa: http://paste.ubuntu.com/14468917/ [13:06] pitti: my apt get install in setup has --force-yes, I wonder if it's the dep requirements from debian/test/control failing [13:07] didrocks: ah, you didn't import the GPG keys [13:07] (as those are the build-deps) [13:07] didrocks: you can add [trusted=yes] to the deb sources, then gpg won't be needed [13:07] ah ok, doing then :) [13:07] yeah, but let me add it rather [13:07] thx! === hikiko|ln is now known as hikiko [13:26] hum [13:26] adt-run [14:25:02]: test pep8: [----------------------- [13:26] ssh: connect to host 127.0.0.1 port 22: Connection refused [13:26] qemu-system-x86_64: terminating on signal 15 from pid 28313 [13:26] adt-run [14:25:03]: ERROR: testbed failure: testbed auxverb failed with exit code 255 [13:27] pitti: ^ ssh is always installed though one? (and I used -s, so it should have stopped and offer me to connect) [13:27] s/one/no/ === rickspencer3_ is now known as rickspencer3 [13:37] didrocks: images normally have openssh-server, yes [13:38] that output is weird thus, I relaunched to see if that was transiant [13:44] pitti: ok, it seems that the image creation failed somewhere… no openssh installed by default [13:45] is there any way to download the one used by autopkgtests? I wonder if adt-vmbuilder isn't broken [13:47] * didrocks relaunches "adt-buildvm-ubuntu-cloud -r trusty" for the 5th time today [13:48] good afternoon all :) [13:48] hey andyrock [13:48] too late to use "morning" :) [13:48] hey andyrock [13:56] pitti: yeah, can be cloud-init or adt-buildvm-ubuntu-cloud, but the new image is broken as well. I'm adding openssh manually for now [13:57] good morning, desktop-ers! [13:58] and happy monday [13:58] hey desrt! [13:59] good morning, didrocks [14:05] interesting study in reliability and redundancy from this morning: a bridge in northern ontario just went out, cutting the trans-canada highway in half [14:05] alternate route? there is none. [14:06] what used to be a 2 minutes trip to the other side of town is now impossible. [14:07] north? we don't really have roads up there.... south? do you really feel like going all the way around one of the great lakes and through the USA? 19 hours... [14:08] the very definition of 'single point of failure' [14:09] hey desrt, how are you? had a good w.e? [14:09] yes. i did. [14:10] time with family, and saw an interesting play. [14:10] nice [14:10] and took mascha to the airport... === vrruiz_ is now known as rvr [15:05] pitti: for now, all large tests are passing here with my cloud image (not fully done yet) [15:05] larsu, hey, any news of the nautilus menubar? [15:06] didrocks: did you have to change anything? i. e. was that "I fixed it", or "I cannot reproduce the session startup failure"? [15:06] pitti: that was unfortunately the second case :( [15:07] the only thing I did as adt-buildvm-ubuntu-cloud doesn't create image with ssh in is to add it manually before running the tests [15:07] (cloud-init regression?) [15:07] seb128: desrt didn't like my patch-gtk approach (and she's right). Fixing that right now [15:07] didrocks: hm, waaaait [15:07] larsu, ok [15:08] didrocks: that might be a bug in setup-testbed, the purge list might be overzealous [15:08] pitti: |o| [15:08] pitti: I WANT MY SSH! :-) [15:08] * didrocks hugs pitti [15:08] didrocks: I tested it with sid and xenial, but maybe on trusty something that I purge is a dependency of openssh-server? [15:08] larsu, other topic, gnome-logs lists logs from decembre 16-17 as most recent for me, but journalctl has logs from jan 08-11 in -b 0 ... do you know what might be going on? [15:08] pitti: yeah, do you want me to try on xenial here? [15:08] to confirm it's trusty-image related [15:09] didrocks: you sohuld see "purging openssh" in the adt-buildvm-ubuntu-cloud output, do you still have it in backscroll? [15:09] pitti: hum, let me dig into my terminals, otherwise, I can relaun it for the 6th time :p [15:10] didrocks: note, if you are experimenting with this, download the trusty image into /tmp/, do "ln -s . current", "ln -s . xenial", and adt-buildvm-ubuntu-cloud --cloud-image-url file:///tmp/ [15:10] didrocks: as this isn't want you really want to debug (you can leave the openssh fix to me), just comment out the big purging from /usr/share/autopkgtest/setup-commands/setup-testbed [15:11] pitti: trying without purging then [15:11] seb128: does it list the latest boot in the boot menu? (dropdown from the title bar) [15:12] larsu, no, most recent is dec 16-17 [15:12] seb128: weird! [15:12] indeed ... [15:13] where does it get the info to populate that list from? [15:13] the journal api [15:13] same as journalctl [15:14] hm works for me, but I guess that's not what you want to hear :D [15:14] is there a dbus call I can do manually or something? [15:14] no, it's a c api [15:14] sd_journal_* [15:14] did you reboot since then [15:14] ? [15:14] yes [15:14] I reboot daily or most days [15:15] journalctl shows log messages old-to-new [15:15] and gnome-logs new-to-old [15:15] which gnome-logs are you using? [15:15] the xenial one [15:16] 3.18.1 [15:16] from when is the last message you do see in logs? [15:16] déc. 17 2015 11:58 [15:16] is the most recent entry [15:16] which is a "journal stopped" [15:17] oh [15:17] you can stop the journal? [15:17] but then, why would jorunalctl still show stuff? [15:17] it's done on shutdown [15:18] well, that was a shutdown log [15:18] ah ok [15:18] I guess no errors/warnings on stderr? [15:18] no [15:18] * larsu doesn't even know how to start debugging this [15:18] (gnome-logs:1234): Gtk-WARNING **: GtkListBox with a model will ignore sort and filter functions [15:18] only [15:18] but I doubt it's it [15:18] yeah [15:19] oh! [15:19] there was a bug where it used the first 5 boots [15:19] in the menu [15:19] instead of the last 5 [15:20] that change to only fetch 5 seems to not be in gnome-3-18 [15:22] I have more than 5 boots though and it woeks [15:22] larsu, don't worry I though maybe there was a directory or dbus api I could check, seems it requires proper debugging [15:22] I might have a look later [15:22] indeed [15:25] pitti: confirming the overzealous purge (without it, I have openssh on trusty [15:25] didrocks: ack; I'll fix this in a moment [15:25] pitti: still no failure (for now) in large tests, but still running [15:30] didrocks: ok, could then be a difference in QEMU's emulated graphics cards? (trusty vs. xenial) [15:31] pitti: ah, you mean without any dri support? [15:31] that could be why unity doesn't start… [15:31] didrocks: we could make it use xserver-xorg-video-dummy, that usually works a bit better [15:32] didrocks: is it easy enough to construct a fast autopkgtest which just sets up the session? [15:32] didrocks: for debugging this right on the cloud without having to wait for the full test? [15:32] pitti: we can change the TESTS parameters! [15:32] didrocks: or, doesn't even need to be a test, just a set of the shell commands which I can paste into a newly created instance [15:33] pitti: let me run one which should fail on the infra [15:33] didrocks: and then I reboot the instance and reconfigure X until unity starts [15:34] pitti: ok, let me proceed a run (once I get one short tests which should fail) [15:39] pitti: just submitted the eclipse one, which should be a short download, let's see [15:40] didrocks: ah, I purge python2.7-minimal, bad me -- this is still required for older releases [15:41] pitti: that's what happen when you demonstrate so much your hate for python2 ;) [15:43] seb128: VT keymap fix landed [15:44] pitti, I saw the upload this morning, thanks! [15:44] seb128: wow, you did? can I use your time machine, too? [15:44] pitti, sorry, it was after lunch ;-) [15:45] assuming [15:45] * Add debian/keyboard-configuration.links to (statically) enable [15:45] console-setup.service. It is not being pulled in by plymouth any more, and [15:45] is the fix [15:46] seb128: I wonder why I didn't get it though [15:46] didrocks: argh, and perl too [15:46] seb128: yes, and yay my sentences for being [15:46] ... cut off at the end :) [15:46] didrocks, yeah, I don't understand either, what is your systemctl status console-setup.service ? [15:47] lol [15:47] pitti: you reimplemented computer-janitor? [15:47] :) [15:47] seb128: Active: inactive (dead) [15:47] k [15:47] still azerty, even on a tty [15:47] so I don't know what sets the layout on vts for you [15:47] didrocks: it's a constant battle against the ever-fattening cloud images, but indeed purging interpreters is dangerous [15:47] didrocks: did it ever run? [15:47] didrocks: it didn't have RemainAfterExit=, I added that too now [15:48] but you should see if it ran in systemctl status [15:48] pitti: I should have some output, you mean? [15:48] $ sudo journalctl -u console-setup.service -b [15:48] -- No entries -- [15:49] didrocks: no, it should say inactive since Jan 9 blabla [15:49] didrocks: if console-setup.service never ran for you, then something else called loadkeys [15:49] didrocks taught his system that french and azerty are the true defaults [15:50] didrocks: setup-testbed fixed in git, thanks for spotting [15:50] didrocks: I want to give the "get correct apt-get source version" a day of real-life testing, then I'll upload 3.19.1 [15:50] seb128: the "right" way [15:50] :) [15:50] * pitti will break it again if the keybaord layout is "azerty", i. e. unusable [15:50] :-) [15:51] pitti: excellent! [15:51] pitti: http://autopkgtest.ubuntu.com/running.shtml#pkg-ubuntu-make-master should be one test running (and failing to execute eclipse) [15:59] medium PASS [15:59] didrocks: expected? ^ [15:59] didrocks: oh, it's the large ones, isn't it [16:02] pitti: large will pass, it's custom as we only run one tests [16:02] one or specific tests :) [16:02] hum, large is running though (and probably all tests) [16:02] argh [16:03] TESTS=es<…> [16:03] missing t [16:03] didn't match the start, so only the whole string [16:03] and in that case, the fallback I've made is to run the matching category [16:03] pitti: issued another run [16:26] pitti: the custom test is now starting [16:26] didrocks: EOD now; in case I get up much earlier than you, could you send me a pointer to which 5 shell commands I need to run in a trusty VM to make it do what your test does? [16:26] wow, timing :) [16:26] heh :) [16:26] sure, will do, once I confirm it's failing :) [16:26] (should be one command) [16:27] didrocks: I'll check again in 20 mins, just starting dinner prep [16:27] pitti: ok, it seems it already ended, let me check the status [16:28] yep, it failed the same way [16:28] * didrocks sends an email [16:36] hikiko: hi [16:37] i'm currently working on the possibility to add a base atk (a11y) support to compiz [16:37] and i've seen some gobjects in compiz, rather in python for ccsm. [16:38] i was wondering if addig some gobjects to compiz is acceptable? [16:38] hikiko: didrocks well, i'm looking for a way to not do that, cause i think it's not desirable. [16:40] ksamak: I really think that's hikiko's area (also TheMuso looked at that in the past IIRC), so yeah, let's see what they tell :) [16:41] ok. thx [17:00] yak SHAVED. [17:07] didrocks: hm, still running? [17:07] didrocks: ok, that's from your "ests" [17:08] pitti: yeah, that's the large one, the other one is linked in my email [17:10] didrocks: ok, I'll prepare an image with ubuntu-desktop installed this evening, then I can do this faster tomorrow morning [17:10] didrocks: thanks for the mail [17:10] yw! [17:10] thanks for looking into this :) === Jin-Baba is now known as JinBaba [17:12] * pitti waves good night [17:13] good night pitti [17:23] night pitti [17:32] Laney, what yak did you shave? [17:32] webkit? [17:43] * didrocks waves good evening and good night [18:04] Hello guys [18:04] i am new to Ubuntu [18:04] how to install Phonetic Input method for Gurmukhi Punjabi ? [18:05] It was available for Ubuntu 10 and 11 [18:05] but not 14 [18:11] seb128: virt-manager was busted [18:11] and now ubiquity doesn't work, whyyyyyyyyyy [18:13] one for tomorrow [18:13] laters === ajmitch_ is now known as ajmitch [19:47] hi folks! i read about gnome-software landing in 16.04 but so far all i find is darkxst's ppa with 3.18 (and some outdated libapt prevents me from testing that). is it even decided that gnome-software will be used by default and if so, when will it be testable? [19:50] ochosi, hey, robert_ancell is working on it amongst others. He should be online in a few mins. [19:51] perfect, thanks willcooke! [19:51] my general guess is that xubuntu will follow ubuntu's lead, but i'd still prefer to test before we take an (educated) decision [19:51] makes sense [19:52] of course, Ubuntu Software Centre will still be in the archives, but it's rather abandoned [19:54] yeah, we thought that might be the case, so at least there'll be a fallback [19:55] there was a (very brief) discussion about going back to synaptic, but since that one hasn't become more userfriendly since we dropped it from default ~4yrs ago, that was a fairly easy decision to make [19:56] ochosi: does xfce not have a default package manager at all? [19:57] nope [19:57] never had one [19:58] seb128: could you subscribe desktop-bugs to fonts-noto-cjk? [19:59]