[00:20] ToyKeeper: you on ? [00:20] guess not... [00:40] kgunn__: Hi. [00:41] kgunn__: I haven't needed to do AP tests yet since the password bits landed, but I think it *might* be working in the UITK test plan script. [00:41] It uses a trick to run basic commands, but I don't see the trick used for phablet-test-run. [00:41] "trick" meaning this: adb shell 'echo foo | sudo -S whatever' [01:12] root@ubuntu-phablet:~# ./foo.sh [01:12] bash: ./foo.sh: /bin/bash: bad interpreter: Permission denied [01:12] ... what. [01:12] (the same script runs fine if I log in via adb shell) [01:13] apparmor is blocking it [01:39] ToyKeeper: hey...stepped away for a bit, thanks for following up...yeah, i think there used to be a hole that got plugged :P [01:39] i ran unity8 ap last week...but alas, no joy tonight === Laney is now known as Guest45203 [02:04] robru: you on? i could use a reconfig on silo11 if you are [02:05] kgunn__: sure [02:05] === trainguards: IMAGE 243 building (started: 20140916 02:05) === [02:05] thanks man! [02:05] kgunn__: you're welcome! === Ursinha-afk is now known as Ursinha [02:23] So the UAL branch works for me on Utopic. [02:24] If someone was sitting around and wanted to play with something, I'd love it if they ran silo 13 [02:24] (on utopic) [02:24] Nothing specific, just make sure it works. [02:38] oh, bot is on a spree today [02:49] Ursinha: looks normal to me? I guess there was a flurry of landing activity [02:56] Ursinha: I'm on a landing spree [02:56] but calling in for the night now ;-) [02:57] sergiusens: :) [03:05] === trainguards: RTM IMAGE 44 building (started: 20140916 03:05) === [03:45] === trainguards: IMAGE 243 DONE (finished: 20140916 03:45) === [03:45] === changelog: http://people.canonical.com/~ogra/touch-image-stats/243.changes === [04:15] === trainguards: RTM IMAGE 44 DONE (finished: 20140916 04:15) === [04:15] === changelog: http://people.canonical.com/~ogra/touch-image-stats/rtm/44.changes === [05:06] kenvandine, ralsina_: silo rtm-011 doesn't appear to behave according to the description in its MP. [05:07] Specifically, gdbus generated a notification immediately even with auto-download enabled (comments say that should only happen without auto-download). And then it didn't auto-download the update. [05:08] I'm not sure if this is due to me having previously turned off auto-download so that I could get the silo installed on an older image without it updating too soon. Even racing to the settings after flashing, it had already found the newer image and downloaded about 5% before I could pause it and change the setting. [05:47] good morning [05:56] sil... isn't here. D'oh. [05:59] Spreadsheet row 50 seems to have an issue... its silo info disappeared, so the automatic triggers aren't working. It should point to rtm silo 011, which is approved but not moving forward. [06:00] looking [06:02] ToyKeeper: fixed. thanks for reporting. the spreadsheet loses those required id:s almost daily for some landings for unknown reasons (either google to blame or our scripts). [06:02] Thanks! [06:26] tvoss: you probably won't mind if I rebuild the qtubuntu-sensors manually now that platform-api finished which is pre-requirement [06:46] Mirv, yup [06:46] Mirv, I actually need to rebuild both [06:47] Mirv, forgot to push a revision :/ === tvoss is now known as tvoss|test === tvoss|test is now known as tvoss [07:33] hmmm, is the latest devel-proposed not booting? [07:33] sil2100, ping [07:34] tvoss: pong [07:53] psivaa_, do you know any reason why i'd get an IndexError when running phablet-click-test-setup? [07:53] brendand: not off the top of my head [07:53] IndexError? [07:54] sil2100: did the image last night get kicked after the input landing? :) [07:54] good morning [07:54] psivaa_, yeah no-one seems to know anything about it [07:54] yay [07:54] i couldnt rfeproduce the dash hang [07:55] * asac tries another time [07:55] brendand: psivaa_: I got too, but I then simply used the autopilot tests I had checkout on Friday... [07:55] maybe some python update causing it? [08:00] Mirv, i think it's somewhat related to the developer mode changes [08:00] Mirv, CI must do something that we don't do locally [08:00] Mirv, probably setting the password, but even that doesn't seem to work here === Guest45203 is now known as Laney [08:06] psivaa_, what's the line you use to put phablet in sudoers? [08:09] asac: I instructed robru to do that, but anyway the 3 o-clock one would normally have the issue fixed already [08:09] Soooo [08:10] Once Dave is up and does promotion testing, we'll be good for promotion [08:14] hmmm, the modem fails to attach, although I have a very sensible signal strength in my office [08:16] brendand: sorry, fighting with adb protocol fault issue with krillin smoke [08:16] sil2100: o/ (only one arm for now) [08:16] hehe [08:17] brendand: "adb shell 'echo ubuntuci |sudo -S bash -c '\''echo phablet ALL=\(ALL\) NOPASSWD: ALL > /etc/sudoers.d/phablet && chmod 600 /etc/sudoers.d/phablet'\'''" [08:18] asac: yeah ;) Since we don't know if there's no other promotion blocker here or there [08:21] morning all [08:22] davmor2: morning! [08:22] davmor2: could you give the latest krillin ubuntu-rtm image a spin? [08:22] no [08:22] Meanie [08:22] sil2100: I've decided it's safer not to [08:22] hah [08:22] ;) [08:23] i couldnt reproduce the input hang anymore with the instructions from the bug ... think thats safe [08:24] davmor2: anyway, 44 I think is the image we would like to check [08:24] ogra_: do you know why the number got bumped so much? [08:24] sil2100: yeah will do once everything is ticking along [08:24] sil2100, nope [08:25] ogra_: I know #41 had the new device tarball, but what about 42, hmm [08:25] sil2100, 43 was built by me when your input fix landed [08:25] and 44 is the auto build [08:26] not sure about 42 [08:26] sil2100: 42 was possibly the tarball build for nm [08:26] sil2100: blame lool [08:26] Time to get dressed [08:27] sil2100, 42 is definitely some external tarball ... not sure which though [08:27] bzoltan, how do you get phablet-click-test-setup to work with the uitk from the silo installed? [08:27] sil2100, and given the device version hasnt changed it must be the custom tarball [08:27] bzoltan, for me it fails and it seem to be related to looking for the new uitk version in the Release pocket [08:29] yesterday I triggered a new build around 7pm UTC and published a new image for utopic around 9pm UTC [08:31] lool: ACK, thanks [08:32] lool, right, but we are talking about RTM builds, i dont think you touched these, did you ? [08:33] brendand: phablet-click-test-setup --distribution=ubuntu-rtm --series=14.09 and I use the test plan scrip -> http://paste.ubuntu.com/8356325/ [08:33] ogra_: I did not [08:33] yeah [08:34] bzoltan, why dont you submit that to phablet-tools ?:) [08:34] (well, probably a bit more generic) [08:35] ogra_: It is part of the UITK project http://bazaar.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/trunk/view/head:/tests/uitk_test_plan.sh [08:36] bzoltan, right, but it looks liek a good way to generalize testing [08:36] ogra_: it is as generic as it can get .. it flashes Utopic or RTM, sets up whatever PPA or silo and runs the tests [08:36] ogra_: it works for me [08:37] ogra_: it is a simple shell script I keep tuning as the tools and the images change ... [08:38] bzoltan, the problem is that it wants to get the ppa version of uitk but it is looking in the archive [08:38] ogra_: more details here -> https://wiki.ubuntu.com/Process/Merges/TestPlan/ui-toolkit As I have understand from ToyKeeper, it is not perfect tool for all environment [08:39] brendand: for the UITK testing I do not care what the phablet-click-test-setup is doing with the UITK, because the first thing I do is to delete the crap it downloads :) [08:40] brendand: Ohh.. one important detail. You can not use the phablet-click-test-setup after you have added a PPA to the sources [08:40] bzoltan, so how do you test it then? [08:40] brendand: first you should do the phablet-click-test-setup on the stock image and after all the click tests are done you are free to add PPA [08:40] bzoltan, that's ugly [08:40] oh well [08:41] brendand: It is not a beauty contest :) [08:41] bzoltan, and then after i need to delete the autopilot/ubuntuuitoolkit directory? [08:41] brendand: that is how it works... I am validating the release candidate from the Silo or from the PPA, so I need the UITK tests from there, not the one what the phablet-click-test-setup pulls [08:43] brendand: check the function I use to provision the device -> http://pastebin.ubuntu.com/8356384/ [08:44] brendand: phablet-click-test-setup fails after adding a UITK silo... it works for any other silo, just not UITK. You can work around it by doing it pre-silo or by editing p-c-t-s to catch the error and not die. The other AP suites install fine, just not UITK. [08:44] brendand: each line is important .. no kidding. Each line, each sleep was hours of strugle [08:45] ToyKeeper: No, the phablet-click-test-setup will fail for all cases when the Silo has the package what the phablet-click-test-setup tries to download. That is how it works. [08:46] Perhaps, I just haven't ever run into the issue on anything else. [08:46] ToyKeeper: so if you have a silo with the unity8 and you add that silo to the image first then the phablet-click-test-setup will fail, because it tries to download the unity8 and get confused with the version numbers. [08:47] In any case, editing p-c-t-s allows it to finish, and as an added bonus, it fails on UITK so you won't need to manually delete the files for that part. [08:47] ToyKeeper: it is a known issue. Could not even be considered as a bug, because the phablet-click-test-setup works like that. It checks what version is available and tries to pull the one it finds from the archive .. but if N+1 version is in the PPA and N is in the archive than it can not download the N+1 fromthe archive. [08:48] brendand: This bit of p-c-t-s allows it to continue after a failed package: http://paste.ubuntu.com/8356416/ [08:49] ToyKeeper: you have to delete the ~phablet/autopilot/ubuntuuitoolkit ... it is important to make sure that the tests are not using the archive UITK tests when you validate a release candidate what most likely has newer tests. [08:49] bzoltan: Yes, and when p-c-t-s fails it has a nice side effect of not installing the tests which would normally need to be deleted. [08:49] ToyKeeper: brendand: I would suggest to use the script from the UITK project. It works out of box without any need to hack the phablet-click-test-setup [08:50] ToyKeeper: if the phablet-click-test-setup fails on UITK then it does not download anything else either. The solution is simpple.. first you do phablet-click-test-setup and then add the PPA. No need to apply hacks. [08:51] bzoltan: That's what my last link works around. [08:52] ToyKeeper: and that is not necessary if you do the provisioning in the right order. [08:52] Mirv: your cat let me down :'( [08:52] That's why I specified in the test plan to set up click tests before adding the silo. [08:53] This particular test plan requires provisioning a little differently than every other test plan, so I tried to document the process even though that info would normally be redundant. [08:54] But since brendand already ran into the click test issue, I was trying to provide info on ways to proceed. [08:54] ToyKeeper: That is what the script does automatically [08:54] p-c-t-s should really handle that better anyway [08:55] ToyKeeper: I would suggest to remove the last section from the UITK test plan, because it is partially redundant and could be confusing. I do not suggest manual provisioning to anybody, because that is error prone and could lead to mysteriously failing tests. [08:55] In any case, I was on my way to bed, just heard a ping and saw a question unanswered which I had already stubbed my toe on and found solutions. [08:58] I'm okay with removing the first section of the UITK test plan page, since nobody actually uses it and the list is just going to get more and more out of date over time. [08:59] davmor2: that was indeed a bit worrying so I had to go looking. I found him being lazy on the balcony: http://people.canonical.com/~tjyrinki/thecat.JPG [09:00] Mirv: Yay Cat \o/ world needs MOAR pictures of cats :) [09:00] ToyKeeper: I am talking about the last section. To keep the rest of the Test Plan up to date is my responsibility. [09:10] I forgot or don't know if I have to Ready? before getting a silo [09:10] oh [09:10] that answers that [09:14] Mirv: when is your MOTU meeting? [09:16] Mirv: that cat picture is cute, MOAR indeed! [09:27] sil2100: next Monday [09:29] psivaa_, could you hack /etc/init/android-tools-adbd.conf on one of the devices and change the "stop on" line there to "stop on stopped lighdm" ? [09:30] and then check if it still behaves weird === ondra| is now known as ondra [09:30] ogra_: will try that. we are unable to log in to the server that's hosting the devices yet. [09:31] uh, oh, why is that ? [09:32] there is some disk space warning in nagios. dont have more information on that. that probably needs a reboot [09:33] oh [09:33] probably thats our root cause then [09:43] davmor2: anything worrying so far? [09:48] sil2100: I've only scratched the surface it'll be much latter before I can tell you that [09:54] psivaa_, ha ! ... there is an adb.reboot() call missing in phablet-config [09:57] psivaa_, this should fix the world: http://paste.ubuntu.com/8356768/ [09:58] ogra_: ohh, great. could try that in the lab once we reboot the server [09:58] psivaa_, so disregard the former thing, that one line in phablet-config will do it [09:58] ogra_: ack [09:59] adb really doesnt get along well when the device reboots itself ... needs to happen via adb (else adb wait-for-device immediately returns because the shutdown takes a few seconds) [10:06] sil2100, promotion is going to happen today right? [10:08] psivaa_, once you can test: https://code.launchpad.net/~ogra/phablet-tools/fix-phablet-config-reboot/+merge/234792 [10:08] brendand: depending if davmor2 finds anything or not [10:09] sil2100, well how bad does it have to be :) [10:10] brendand: depends on how rigorous davmor2 feels today! [10:24] Hi. I was told line 14 has not landed because some mrs are not approved. Anyone know what those are (the landing sheet does not say and all of them should have landed to respective trunks already)? [10:26] brendand: it's me and I'm feeling particularly vindictive ;) [10:26] * davmor2 awaits the tears from sil2100 [10:26] sil2100, davmor2 - we really need a promotion for beta testing [10:27] brendand: I'm working on it dude so far so good, it's not perfect but I think it is the best image we have had for a long while :) [10:27] I suspect we should be promotable, I was just a bit worried yesterday when davmor2 mentioned those network issues [10:28] Which ogra_ said had on krillin from the start [10:28] But still [10:28] sil2100, well, i have three APs in my house ... obviously there is a long standing bug with such wifi setups [10:29] so my bug might rather be that one than davmor2's [10:51] brendand, davmor2, i would appreciate if one of you need to run phablet-config writable-image today, if he could test https://code.launchpad.net/~ogra/phablet-tools/fix-phablet-config-reboot/+merge/234792 alongside [10:53] ogra_: with the change a job goes past 'phablet config writable-image' ok [10:54] yay [10:54] and you see no more error spam either i guess [10:56] * sil2100 off to prepare lunch [10:56] o/ [11:00] trainguards: hi. we've actually approved the MRs for silo 17 now! (https://ci-train.ubuntu.com/job/ubuntu-landing-017-2-publish/build?delay=0sec) [11:12] sergiusens, once you are around, could you do a quick top approval of https://code.launchpad.net/~ogra/phablet-tools/fix-phablet-config-reboot/+merge/234792 ? [11:13] brendand: hi. apparently the stuff in rtm silo 14 is ready to be tested now === MacSlow is now known as MacSlow|lunch [11:28] ogra_, mzanetti, sil2100: Grrrrrrr phone just locked up opening an album in the music player from the scopes whole system is unresponsive :( check the crash reports now [11:28] gah [11:29] aw men [11:29] unity 8 crash [11:29] hmm... [11:29] but not the input thing, right? [11:29] mzanetti: no this is the whole system dead so no swipe actions either [11:31] mzanetti: system went to sleep and came back fixed so I assume the lock up was apport hogging the system why it did the report [11:31] mhm [11:31] davmor2: yeah, if you could get some useful trace, that's be great [11:32] ogra_: how did this work before? wasn't there an adb.reboot() there already? [11:34] sergiusens, the new code calls a dbus interface that touches the file on the device and just calls reboot ... but with the time it takes upstart to shut down the wait-for-device call immediately returns [11:35] that dbus code never had a reboot call on the PC side since i thougth upstarts reboot would be sufficient (which it isnt) [11:36] davmor2: check if it's reproducible [11:38] mzanetti: https://errors.ubuntu.com/?period=day&pkg_arch=armhf number 96 [11:39] sergiusens: btw. rejoice! We'll be doing a trial run soon with doing binary copies instead of source copies for synces between distros - this means no longer any worries with the ugly ~rtm version numbers o/ [11:40] If things will not be broken for a week we'll switch for reals [11:40] pete-woods: publish done [11:40] sil2100: really? Nice; most of my things are golang based, so take nothing to build, or are basically the same package in either archive ;-) [11:40] sil2100: no now it is working fine so no idea possibly sd card maybe [11:40] pete-woods, it hasn't been marked as tested though [11:40] davmor2: hmm. I'm confused by this [11:41] davmor2: why does it redirect me to gallery-app? [11:41] ogra_: ok, that backstory helps :-) [11:41] mzanetti: no idea [11:41] :) [11:42] brendand: oh, it has been tested. I thought you would be the one to tick the tested box, though. [11:42] ev, pitti: ^ https://errors.ubuntu.com/?period=day&pkg_arch=armhf why does 96 go to gallery? crash happened in unity 8 from scope opening music player [11:43] Mirv: thanks! [11:43] pete-woods, i tick the QA sign off box - you tick the Testing pass box === MacSlow|lunch is now known as MacSlow [11:44] brendand: okay, sure, will tick it then [11:51] sil2100, ^^^ [11:53] asac, sil2100 ... what do we do with devel/devel-proposed if we promote ... i assume there are still people using the devel channel that didnt get an update since 35 images [11:54] ogra_, hey will phablet-shell work again at some point? [11:54] ogra_, the shell provided by adb is pretty terrible [11:54] brendand, how does it not work ? it works forr me [11:55] ogra_: dunno ... not sure if devel-proposeed has anything terrible or if its on par wrt to quality [11:55] anyway lunch ... will be back in 30 or so [11:55] asac, i dont know either, but people being on devel shouldnt be left behind i think [11:56] brendand, i know there are a few issues i planned to work on now, but nothing that would make it not work at all [11:56] (so please tell me if there are worse things) [11:57] ogra_, ok weird - a couple of times i tried it and got some kind of key error, but it looks fine now [11:57] yeah, i plan to work on the key handling today [11:57] but that shouldnt block its only a little weird :) [12:00] ogra_: you don't have time for that you need to bring me desert remember ;) [12:00] haha [12:00] true [12:01] davmor2, bring you desert? [12:01] davmor2, i don't think ogra_ has magical powers [12:02] brendand: it's a joke from last night [12:02] yeah, you missed that one [12:02] davmor2, oh you mean dessert? [12:03] yeah, not 500 trucks full of sand :) [12:03] brendand: haha yeah sorry [12:03] davmor2, http://www.freelancewritinggigs.com/wp-content/uploads/2014/03/desert_or_dessert-700x466.jpg [12:04] haha [12:05] brendand: allow me to throw leaving somebody behind aswell, desert :P [12:05] brendand: English sucks [12:08] sil2100: hey, can you help me for a minute when you have time? [12:08] sil2100: I don't really understand why silo 11 fails to build [12:10] mzanetti: which silo is that? 11 ubuntu or ubuntu-rtm? [12:10] ubuntu [12:10] sil2100: ^ [12:10] sil2100: seems the gles-sync is messing up [12:10] sil2100, poke again ... a silo for line 73 please :) [12:11] ogra_: ACK ;) One moment === cprov changed the topic of #ubuntu-ci-eng to: Train support: trainguards | Vanguard: cprov | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: Latest flash update hung all makos on s-jenkins. [12:13] mzanetti: ok, so to properly use this 'trick' gles sync, you first need to build the qtmir base package, check which version it got and then modify the -gles merge to include the same version number as from the non-gles package [12:14] mzanetti: so let me just do a build of qtmir now, then you should modify the merge and build the -gles part explicitly [12:14] sil2100: I think the -gles branch should already be ok [12:14] at least it holds today's date in the version etc [12:15] it's just that it can't find it because of the build order then [12:15] Ah, ok, so we just need to build qtmir first, then just qtmir-gles [12:15] Since it needs to be uploaded already into the PPA [12:16] Even if you order the merges correctly, CI Train only uploads packages after the sources are prepared [12:16] mzanetti: ok, let me do that, ok? [12:17] ah, I see [12:17] sil2100: ok, sure, go ahead [12:18] sil2100: sorry to ask those noob questions... Saviq kinda does it for the team usually so the -gles and -rtm bits are new to me [12:18] mzanetti: no worries ;) It's a bit hackish so it's not entirely intuitive [12:23] ogra_: o/ [12:24] ogra_: sadly! We haf no silozzz [12:27] sil2100, sigh ... so we wont have smoke tests either ... [12:27] ogra_: let me try freeing one up [12:28] thanks ... i'd really like to have that in before the next test run [12:30] ogra_: we might have one free soon [12:30] k, thanks [12:30] pete-woods: ping [12:31] sil2100: hi [12:32] pete-woods: asking just in case - is ubuntu silo 14 still a test silo? [12:32] sil2100: not any more. I pulled out the unity8 MR so it stopped conflicting [12:36] sil2100, I was given a silo but AFAIK sergiusens already requested one for his change in udm, can you remove it etc...? [12:36] sil2100, he has the following => http://people.canonical.com/~platform/citrain_dashboard/#?distro=ubuntu-rtm&q=sergiusens [12:37] mandel: ah, so you want the other one, the one for you, removed? [12:37] sil2100, yes please, mea culpa, I did not know segio had one === mandel is now known as mandel|lunch === Ursinha is now known as Ursinha-afk [12:49] ogra_: anyway, as you can see you have a silo [12:49] \o/ [12:49] * ogra_ hugs sil2100 [13:02] sil2100: can you come to pot file merging meeting now? [13:02] sil2100: https://plus.google.com/hangouts/_/canonical.com/pot-file?authuser=0 [13:05] bfiller: be right there [13:09] pete-woods, if you know where satoris is tell him that for rtm silos we need the testing to be done on rtm [13:09] bregma: will do [13:09] pete-woods, this time i'll continue with the sign-off (before eod), but it should be done that way in future [13:09] brendand: will do [13:09] brendand: thanks! [13:10] pete-woods, also if he's still working remind him that landers/testers should be in this channel! [13:10] brendand: I have told him at least once today [13:10] he's here [13:10] just with lower case s :) [13:11] pete-woods, ah yes - i see now [13:12] bzoltan, silo 9 signed off [13:18] brendand: I'm right here. ;-) [13:19] And I have done that testing on rtm, the spreadsheet only lists mako for some reason. [13:19] err, what ? [13:19] sil2100, dont i have to click on publish anymore nowadays ? [13:19] uh? [13:19] What happened? [13:20] ogra_: o/ [13:20] satoris, you said #243. that's a utopic image number [13:20] satoris: the thumbnailer crashes landing? there's a line below it for rtm [13:20] brendand: https://docs.google.com/a/canonical.com/spreadsheets/d/17FjOXaT-OsTnuSPnyv7kzEfubJadTaEjDO1gSAjRNCM/edit#gid=0 [13:20] sil2100, i marked testing done and the silo automatically published it seems (unless Mirv clicked something ... which his above recation indicates) [13:20] ogra_: I opened the dashboard for the first time in some while, and there was a silo marked as ready, desktop only, ready to be published. maybe you mean that? :) [13:21] ;) [13:21] Mirv, heh, yeah [13:21] That's image 29 with silo 14. [13:21] good timing I see [13:21] Mirv is just fast ;) [13:21] perfect timing :) [13:21] Mirv, thanks then [13:21] brendand: plus I retested it an hour ago or so with image 44. [13:22] satoris, ok you should put that in the Testing pass column [13:22] satoris, at least for the RTM silo [13:22] brendand: I would have done it ages ago but don't have edit rights. [13:22] satoris, oh - you have them now though? [13:23] brendand: nope. Won't let me edit. [13:23] eh, i just got a passphrase entry dialog instead of pin entry [13:23] sil2100, can you add satoris to the spreadsheet? [13:23] brendand: sure [13:24] sil2100, in fact why isn't it just set to 'everyone at canonical with the link can edit'? [13:25] ogra_, did someone already confirm your phablet-config fix? [13:25] brendand, landed in rpoposed already [13:25] ogra_, cool [13:29] brendand: updated. [13:30] ogra_, immediately after running writable-image, then if i do phablet-shell 'ssh_exchange_identification: read: Connection reset by peer' [13:30] ogra_, that's what i was talking about [13:31] brendand, is that rtm ? [13:31] ogra_, yeah [13:31] brendand, silo rtm 006 ;) [13:32] you are to fast :) [13:32] ogra_, or you're too slow? [13:32] the fix in that silo makes sure the session is fully up before providing adb to you [13:33] no, i mean your phablet-shell is to early (for the adbd with the fix missing) [13:33] so you enter adb shell without having everything up and running [13:34] 006 only starts adbd after the session is up, dbus is running etc ... by then the ssh_exchange_identification bit should be available too [13:46] ogra_, oh so if i just wait longer it will work? [13:46] it should [13:46] just wait ... say ... 20sec [13:50] davmor2: ping [13:50] davmor2: any news? [13:51] just finishing off looking goodish so far if we ignore a couple of the newer issues [13:52] om26er: hey [13:52] sil2100, Hi! [13:52] om26er: today is utopic day, right? [13:52] :) [13:52] sil2100, it is but I am yet to start on it. I was investigating the dasboard failures [13:52] ACK [13:52] sil2100, though I am starting on the image testing and will send the email [13:53] om26er: anyway, if you get some results just give me a sign, since if it looks pretty good then we might also promote it === greyback__ is now known as greyback [13:58] \o/ [13:58] sil2100: https://docs.google.com/a/canonical.com/spreadsheets/d/1Mw46QHRVqmaf_NSL2L-jNcEYNlQzbcsJyaXvhjSgSSY/edit#gid=713699479 looking pretty good :) [13:58] * sil2100 prepares his "SHIP IT" stamp [13:59] davmor2: I like the looks of it [13:59] * ogra_ has no clue how to promote RTM :P [13:59] ogra_: use the force luke [13:59] asac: looks like we'll promote something today IF ogra_ figures out RTM promotion ;p [13:59] yeah, i'll manage somehow :) [13:59] asac: for utopic we'll wait for om26er to do his testing [14:00] sil2100: do it we can pick up the shrapnel after :) [14:00] brendand: ^ [14:00] ogra_: can you promote #44 with all your might and powa? [14:00] i'll try, no promises :P [14:00] sil2100, thats mako and krillin only, right ? [14:00] SHIIIIP IT! [14:00] mwahaha [14:00] or do we want other arches in RTM [14:01] sil2100: ok cool. when is that? tomorrow? [14:01] omer? [14:02] wonder if he found anything critical last few days he tested [14:02] asac: he's starting testing now IIRC [14:02] cool [14:04] asac, sil2100 krillin 44 and mako 41 promoted ... what other arches do we want ? [14:04] emulator ? [14:05] oh, wait. emulator wont work anyway until rtm silo 006 has landed [14:07] guess not then [14:07] ogra_: do we have flo or manta on RTM builds? [14:07] yes [14:07] but totally untested [14:07] Those are probably broken but we always promoted those as well, right? ;) [14:07] Would be good for a base-line! [14:07] manta will also need rtm silo 006 [14:08] ogra_: looks good :) [14:08] thx [14:08] :) [14:08] the rtm promotion i mean [14:08] yeah [14:08] (i didnt expect you to refer to my silo :) ) [14:09] hehe [14:09] \o/ [14:09] wheee ! any my new router HW arrived [14:10] so exciting ! [14:10] I think we need to modify the spreadsheet to indicate better what is promoted [14:12] well done! [14:12] ogra_: sil2100 \o/ for promotion! [14:12] john-mcaleely: jibel: ^ :) [14:13] asac, awesome! flashing 14.09 #1 [14:13] hehe [14:13] nice number [14:13] sil2100, the initial setup wizard is crashing [14:13] will be remembered [14:13] asac, sil2100 \o/ [14:14] om26er: at the end? [14:14] or at which stage? [14:14] asac, at the start [14:14] (and i assume its on devel?) [14:14] davmor2: nice column of 'pass'! [14:14] asac, utopic-proposed (yes) [14:14] hmm [14:14] * asac tries [14:15] * asac removes wizard-run file and reboots [14:15] make devel-propsed [14:16] mako [14:16] om26er: not here [14:17] i finished wizard [14:17] now i have boot animation [14:17] and now the device is running [14:17] asac, I rebooted and now my device started as well [14:17] om26er: but wizard worked for me [14:17] i didnt do a full wipe [14:17] just removed the wizard-has-run file [14:17] and rebooted and wizard worked [14:18] asac: remember om26er is testing on mako [14:18] yeah i am too [14:18] mako devel-proposed is what we talk about [14:18] krillin is on rtm-proposed [14:19] i like devel-proposed [14:19] osmtouch is working kind of [14:19] and the rest also [14:19] om26er: maybe finish more of the promotion testing there [14:19] and we can get back later to this if thats the only regression candidate [14:20] * sil2100 has a tear in his eye [14:20] :P [14:20] Our first RTM promotion === Ursinha-afk is now known as Ursinha === gatox_ultra is now known as gatox [14:20] asac, ok [14:20] *sob* [14:20] sil2100: yes!! [14:20] sil2100: let go for a few minutes [14:20] all feelings allowed [14:20] lol [14:20] I... I need to mark it in my calendar! Will celebrate it every year ;p [14:20] you can create a facebook fake account in honor of this event [14:21] that has birthday today [14:21] hehe [14:21] hah ;) [14:21] Anyway, thanks everyone! [14:23] brendand: \o/ Thank you! [14:24] bzoltan: published already a looong time ago :) [14:24] ogra_: since I cried so much now I can't attend our landing meeting, could you lead it? ;) [14:25] ;) [14:25] could I have a rtm silo for line 28 please? [14:25] But seriously, I won't be able to attend as it's Tuesday today [14:27] cyphermox_: o/ [14:28] sil2100: hey :) === cyphermox_ is now known as cyphermox [14:28] cyphermox: assigning, let's see how it goes [14:28] Mirv: like an hour ago? :D [14:28] sil2100: there's a job for MOTU at https://ci-train.ubuntu.com/job/ubuntu-landing-014-2-publish/ [14:28] sil2100: thanks [14:28] cyphermox: ah, ok, no silos available, but we should have some soon [14:29] Mirv: looking [14:29] sure [14:29] bzoltan: at _least_! [14:29] sil2100: thanks again [14:31] sil2100: we've a bit of a problem with the lack of rtm silos, we've been at full 20 for some days now, with things being queued without a silo [14:31] sil2100: I kept a note "please get your rtm silos tested" for one day at the top of the spreadsheet :) [14:31] not sure if it helped [14:31] ;) [14:32] cjwatson: I wonder if you could help getting https://launchpad.net/ubuntu-rtm/+source/ttf-ancient-fonts to release pocket? [14:32] sil2100, indeed [14:32] Mirv: there's one thing I want to double check regarding that landing [14:33] sil2100: ok [14:34] pete-woods: hey! [14:36] pete-woods: just want to make sure of something - I noticed that the RegistryObject constructor has some parameters modified === cprov changed the topic of #ubuntu-ci-eng to: Train support: trainguards | Vanguard: cihelp | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: Latest flash update hung all makos on s-jenkins. [14:37] pete-woods: since libunity-scopes-dev is a dependency of unity-scope-click, I just want to make sure it's not used there [14:38] sil2100: don't worry, it's an internal only symbol. sorry we didn't mention this in the changelog [14:38] sil2100: I know it's unusual (I don't like that we do it), but all our internal symbols are exported, so this happens nearly every release of scopes [14:39] this discussion also happens with nearly every release of scopes :) [14:44] anyone knows why this is failing? http://jenkins.qa.ubuntu.com/job/generic-deb-autopilot-runner-mako/4616 [14:44] no vanguard? === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [14:51] Mirv: That's been copied in a very strange way using CI Train, which is why it's broken. How about I just copy the binaries directly? [14:52] Mirv: Please don't use CI Train's mode where it rewrites the upstream version with ~rtm for packages whose upstream isn't managed by CI Train ... [14:54] ogra_: can you elaborate on the emulator problems with RTM and why we don't promote there? [14:54] slangasek: ^ [14:54] Mirv: Should be fixed shortly. [14:55] sil2100, slangasek adbd starts to early in the emulator and gets our of sync with teh gadget emulation, the rtm silo 006 landing has a fix for this [14:56] (which fixes it in general by starting adbd only when lightdm is up) [14:56] ogra_: why should that be a blocker for promoting the emulator image? I think it's more important to have the emulator in sync (or nearly so) with the other channels, otherwise surely it's useless for validating changes locally? [14:57] otoh maybe people should always be using the -proposed channel for that === mandel|lunch is now known as mandel [15:00] sil2100: ^ check cjwatson's comments on syncing new packages from utopic [15:00] cjwatson: ok then, thanks! [15:00] that will clean one rtm silo too [15:01] Mirv: I just did this FWIW: copy-package --from=ubuntu --from-suite=utopic --to=ubuntu-rtm --to-suite=14.09-proposed -b ttf-ancient-fonts [15:01] slangasek, what do you do with an emulator you can not log in to in any way and dont have a terminal app for ? [15:02] Much simpler and more correct than messing about with silos [15:03] ogra_: so this is a recent regression on the emulator? Because it was working for me as of version 30. If you mean the image is truly unusable (which wasn't what I understood from the above), then sure, no value in promoting it [15:03] slangasek, i would have happily approved, but not without working adb ... and QA was to busy to do the regression testing on rtm 006 yet [15:04] cjwatson: sil2100 probably used silo since that workarounds the access right problem (I did try similar copy-package before to see whether I've copy rights to the rtm distro) [15:04] ogra_: in the future, do we expect to be promoting the emulator in sync with the devices? right now it's not an issue because there's nothing for the emulator channel at all, but later I worry about what having these channels out of sync will mean for validation [15:05] Mirv: but sil2100 has the necessary access for this copy, since ttf-ancient-fonts is in universe. [15:05] cjwatson: ah, right, he has. and I should hopefully have next Monday. [15:05] slangasek, we definitely want to, yes [15:05] ok [15:06] and i would have preferred to for this rouond too [15:06] * slangasek nods [15:06] cjwatson: sorry about that, CI Train did't have an option of not rewriting the version back then [15:06] cjwatson: should have just done a binary copy for that indeed [15:06] sil2100: I don't think CI Train adds any value at all when copying new packages :) [15:07] cjwatson: CI Train will do binary copies soon as well! [15:08] sil2100: But what's the point in using a silo for this? There is zero risk of regression for a new package [15:08] sil2100: It just fills up silos better used for other things. [15:09] For this one it was pointless indeed, sorry about that ;) As we discussed before, this should be a quick binary copy straight away [15:10] greyback, another unity8 lockup it seems [15:11] greyback, backtrace looks different though [15:11] greyback, did you figure out where to get those debug symbols? [15:18] brendand: hey yes, http://ddebs.ubuntu.com/ubuntu-rtm should do it [15:18] greyback, shoot me that wiki link again [15:18] brendand: https://wiki.ubuntu.com/DebuggingProgramCrash#Debug_Symbol_Packages [15:19] brendand: in step 1, replace the URL with the one I gave you [15:20] cjwatson: it was simply an old request [15:20] greyback, and $(lsb_release -cs)-updates with 14.09, right? [15:21] brendand: probably. I'm not 100% [15:21] greyback, what was the package then? [15:22] brendand: that depends on what the backtrace was showing you. Can you pastebin it to me? === elopio_ is now known as elopio [15:24] ogra_: I'll be driving out soon, when I'm back I'll write the e-mail and such :) [15:24] oki [15:27] greyback, hmm. i can't access either of the pastebins [15:27] bregma: http://pastebin.ca/ maybe? [15:28] brendand: ^^ [15:28] bregma: unping, sorry [15:28] 'sokay, it happens a lot here [15:29] greyback, at last: http://pastebin.com/REPjL5Pq [15:29] bregma: get that brendand bloke to change his name to notbregma end of issue ;) [15:30] brendand: interesting. I'd like to get symbols for libusermetricsoutput , qtcore & qtdbus. So install qtbase5-dbg, libusermetricsoutput1-dbgsym please [15:34] greyback, agh more version issues. might be quicker if i just get the matching debs for each === gatox is now known as gatox_lunch [15:35] brendand: apt install libusermetricsoutput1-dbgsym= ? [15:36] greyback, yeah - of course, sorry [15:36] no worries, I just fight these fires a lot :) [15:38] brendand: while you're at it, please install libc6-dbg too [15:40] brendand: and when you have those things installed, in gdb please run "t a a bt" - which will show backgrace of all threads. QProcess::WaitForFinished is waiting for another thread to complete [15:40] greyback: that looks like the external translation process (the gettext binary) for libusermetrics is acting weird [15:41] pete-woods: ah this is your code, no? :) [15:41] greyback: yes [15:42] greyback: we do translations in an external process because they could potentially come from untrusted sources [15:42] pete-woods: okay. If that process dies/is blocked, I guess that blocks the main thread? [15:43] greyback: I thought it was happening on a non-main thread, but I could be mistaken [15:43] (sounds like I am, if you are investigating a freeze of some kind) [15:44] pete-woods: me too. This stuff I don't know too well. Can you work with brendand to get a good backtrace - and log a bug about it? [15:44] pete-woods: yeah brendan found unity8 froze [15:45] greyback, ah too late. it's finally come back [15:45] it'll probably come back after some timeout [15:45] right [15:45] the fail should be logged somwhere tho [15:45] unity8.log? [15:45] yes [15:46] brendand: pastebin output of ~/.cache/upstart/unity8.log please so we can have a look [15:46] pete-woods: we should not be blocking for any dbus traffic, it should all be async really. [15:47] greyback: yes. I thought I had the whole thing running on its own thread [15:47] pity, a bigger bt would have shown us where [15:50] greyback, http://paste.ubuntu.com/8358720/ [15:51] pete-woods: QProcess: Destroyed while process ("gettext") is still running. [15:51] yep [15:51] that's the one [15:51] it'd be nice to know why he's freezing [15:55] pete-woods: brendand https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1370127 [15:55] Launchpad bug 1370127 in unity8 (Ubuntu) "unity8 UI freeze for some time - external translations process/gettext a clue" [Undecided,New] [15:56] greyback, the case i saw it in was opening the camera from the content hub when creating a new MMS [15:57] brendand: do add that info to the bug, every little helps [15:57] * greyback doesn't have a SIM with working MMS [15:58] I think just opening the camera could trigger it [15:58] well, taking a picture [15:58] am I supposed to drop aborted landings from the spreadsheet? [15:58] silo has been returned already [16:00] no hangout for me again :( [16:01] sil2100, Hi! the image looks good to me. Things are working and didn't see any major new problem. Though there were small issues discovered. [16:03] balloons: joining the landing hangout? [16:03] working onit [16:04] haircut popey? [16:04] yes! [16:05] me likey! [16:05] hehe === Ursinha is now known as Ursinha-afk [16:22] kenvandine: around? need a core dev ack on some packaging changes https://ci-train.ubuntu.com/job/ubuntu-landing-014-2-publish/26/ [16:22] robru, sure [16:22] pete-woods: if I assign you that silo are you around to actually use it? [16:24] robru, ack from me [16:24] kenvandine: thanks [16:25] robru: I think it's fair to wait until tomorrow morning [16:31] sergiusens, one more peek? https://code.launchpad.net/~nskaggs/phablet-tools/remove-python2-support/+merge/233754 === gatox_lunch is now known as gatox [16:36] sil2100, ping [16:43] jibel, is there any known problems with otto right now? [16:44] https://jenkins.qa.ubuntu.com/job/autopilot-testrunner-otto-utopic/2942/testReport/junit/ubuntu_system_settings.tests.test_about/AboutTestCase/test_settings_show_correct_version_of_the_os/ [16:44] robru: http://people.canonical.com/~platform/citrain_dashboard/#?distro=ubuntu-rtm&q= this says mir which was in the silo so I now don't know what the hell is going on with it [16:44] jibel, ^^ looks like just general brokeness [16:44] robru: it was mir when I tested it [16:45] davmor2: yeah but the package you tested mysteriously disappeared and I had to rebuild it. so please give it a quick re-test to make sure nothing exploded [16:46] robru: will do [16:50] robru, ping [16:50] tvoss: pong [16:58] cihelp, could you look at kenvandine's issue with autopilot-testrunner-otto-utopic jobs? [16:58] fginther, ^ [17:02] jibel, fginther: we've had several failures like this today, looks like a general issue to me... thanks :) === Ursinha-afk is now known as Ursinha [17:45] tvoss: I get Cannot add PPA: 'ppa:ci-train-ppa-service/landing-002'. [17:45] Please check that the PPA name or format is correct. [17:45] tvoss: when I try to add the silo 002 ppa [17:46] nik90, just try citrain device-install 2 [17:46] from your host [17:46] tvoss: I don't have citrain package or script on my host [17:47] * nik90 installs phablet-tools-citrain [17:47] nik90, yup,was just about to say that :) [17:47] never knew about this package..pretty cool [17:49] hmm how does it know if it rtm silo or normal silo? [17:53] davmor2: hey can you try "citrain device-upgrade 2" [17:53] davmor2: I get http://paste.ubuntu.com/8359438/ [17:54] so not sure if it installed the silo or not [17:54] Hi, is it safe to do OTA updates on mako as "Latest flash update hung all makos on s-jenkins." ? [17:55] i think thats a few days old and nobody cleaned it up [17:55] ogra_, ah thanks, i'll try updating and see what happens :) [18:03] nik90: add-apt-repository is non working for the rtm archive [18:03] sergiusens: ah ok [18:03] best go to the ppa in a browser and copy the sources list line and add the gpg key if you want [18:03] I did that and I got the error that I pasted above [18:03] which is why I used citrain [18:04] " Cannot add PPA: 'ppa:ci-train-ppa-service/landing-002'. Please check that the PPA name or format is correct. " [18:11] nik90: tvoss: yeah the citrain tool bitrotted a bit recently, with the switch to non-root adb and also rtm, it stopped working and I haven't had time to really figure out a fix [18:11] also I was hoping that add-apt-repository would get fixed for rtm, but I can see now that's not anybody's priority [18:12] robru, so we are back to manual installation of silos? [18:12] robru: well I am testing a non-rtm silo and a manual add-apt-repository doesnt seem to work as well [18:13] tvoss: yep [18:13] nik90: yeah that's because rootless adb broke the script. the script assumes you are root when you 'adb shell' [18:13] which is no longer true [18:13] ah ok === gatox_ultra is now known as gatox [18:18] sil2100: did you get my email? === tvoss is now known as tvoss|test === tvoss|test is now known as tvoss [18:30] nik90: have a look at /etc/apt/sources.list.d/ [18:30] nik90: that is where the silo is likely to be added [18:48] robru: which one? [18:48] sil2100: the reply to your reply about the watch-ppa issue [18:48] robru: I saw it now, I was out for practice as per most Tuesdays ;) [18:49] sil2100: ah ok, wasn't sure if you were around as you said you'd work on that in the email ;-) [18:49] robru: so, give me some moments to finish my look into this watch-ppa - if I have to EOD before it works you will have to be the hero and save the train [18:49] Maybe not save the train, but, yeah, make it better ;p [18:49] sil2100: heh, ok [18:50] sil2100: I'm just about to land a minor branch, touching only packageinppamanager.py, shouldn't conflict with any work you're doing on watch-ppa [18:50] Feel free :) [18:51] ogra_, asac: so, as per Omer's testing, it seems mako and utopic look pretty promising [18:51] great === Ursinha is now known as Ursinha-afk [18:54] hi trainguards, i have line 60 ready if a silo is free [18:55] trainguards, are any rtm silos free? line 52 [18:55] dbarth_: you got 13 [18:55] for utopic [18:55] tedg: one sec [18:58] plars, can you help me on core apps jenkins? [18:59] plars, I promise it will be something easy as I know fginther is away :-) [18:59] balloons: I'll give it a shot, what's up? [19:00] plars, can you switch ubuntu-calculator-app back to normal landings via jenkins, instead of using ps-jenkins and running on a mako device? [19:00] tedg: ok you got rtm2, please build [19:00] plars, it was an experiment, but it's not working and we don't have the capability to debug atm [19:01] plars, I'll give you the job names, one sec [19:02] balloons: is this just a cupstream2distro-config change and redeploy? [19:02] balloons: looking at the history now [19:02] robru, Thank you! [19:02] plars, the normal ci job is http://91.189.93.70:8080/job/ubuntu-calculator-app-ci/, but you can see the autolanding job has migrated to http://s-jenkins.ubuntu-ci:8080/job/ubuntu-calculator-app-autolanding/ [19:03] balloons: I think I might see it... this is something that was done back in may? [19:03] plars, it has been that way for some time I think yes.. [19:03] balloons: ok, so I think reverting 1124 should do it [19:03] plars, link? [19:04] balloons: oh, I was just looking at it locally, hang on [19:04] ogra_, asac: so, as per Omer's testing, it seems mako and utopic look pretty promising - at least for makop [19:05] balloons: http://bazaar.launchpad.net/~cupstream2distro-maintainers/cupstream2distro-config/trunk/revision/1124 [19:05] ogra_, asac: maybe we could promote utopic as well? [19:05] plars, ahh right, I remember those cfg's.. yep [19:05] plars, I guess let me tweak and propose then [19:05] balloons: ok [19:06] robru: thx [19:06] balloons: I think I have it if you want me to just proposes it [19:07] *propose [19:07] plars, go for it then [19:09] balloons: https://code.launchpad.net/~pwlars/cupstream2distro-config/backout-r1124/+merge/234871 [19:10] plars, awesome.. please just confirm http://91.189.93.70:8080/job/ubuntu-calculator-app-autolanding/ looks good [19:10] heck, I guess I can quickly.. just run it [19:11] robru: retested still looks good qa signed off on silo 001 [19:11] plars, I see one parm wrong.. distros should just be utopic [19:11] balloons: I'll need to land this and redeploy the jobs before you'll see any effect there [19:12] balloons: that's not anything that came from my change, I haven't even deployed it yet and the distro was never touched [19:13] plars, yes I know.. just commenting about what parms would need to be updated [19:13] plars, I think your change is fine. I kicked off a job #100 on the old autolanding just to make sure [19:14] plars, I guess you are right though, the job needs to autogenerate before we can see what needs changed [19:15] balloons: ok, if you ack it then, I'll land it and regenerate the config [19:15] plars, yep.. acking now [19:16] tedg: you're welcome [19:17] dbarth_: you're welcome [19:17] ty plars ! [19:23] * davmor2 patiently awaits sil2100 's hell froze over email :) === Ursinha-afk is now known as Ursinha [19:31] fginther: https://code.launchpad.net/~robru/cupstream2distro/just---why/+merge/234864 merged with no comment from the autolanding job, is that normal? autolanding job only comments when there's a failure? [19:32] sil2100: ^ seems MPs are working now ;-) [19:35] davmor2: ;) [19:35] robru: oh, you mean automerging? :) [19:36] \o/ [19:36] sil2100: yeah [19:37] Yeaaah [19:40] balloons: ok, it just finished. want to give it a try and see if it works how you expect? [19:41] robru: can I get a silo for line 66 - just a sync to rtm [19:42] bfiller: sure, you got rtm5 [19:42] robru: thanks [19:44] dobey: not sure why you tested your utopic silo with an image that's 200+ revisions old. [19:45] robru: i tested it on rtm [19:45] dobey: how do you test a utopic silo on rtm? [19:46] also i can't even flash my mako with ubuntu-device-flash right now :( [19:47] dobey: sorry to hear that [19:47] 13:06 < dobey> 2014/09/16 13:06:21 Cannot push /home/dobey/.cache/ubuntuimages/pool/ubuntu-a7190e9e460a09f6756665cd2f3c09430be188ef2c46c79e23f21e9f4aaecc12.tar.xz.asc to device: free space on /cache/recovery is denied [19:48] robru: anyway it's only a packaging change to a Recommends. [19:49] dobey: not sure how to fix it [19:49] dobey: ok i'll publish if it's just minor but note I'm not thrilled to publish something essentially untested to utopic [19:50] well the diff in the PPA is a lie [19:51] but the change is literally - upstart, + upstart-bin, [19:52] robru: can I have a rtm silo for line 28 please? [19:53] sorry, line 26 now ;) [19:53] sil2100, do we have any final word ? [19:53] cyphermox: uh... no? you need to get the utopic silo published first [19:53] it was [19:54] cyphermox: nm was looking at 28 [19:54] ah, yeah [19:54] cyphermox: is this a sync from utopic then? [19:56] robru: not anyone's priority> well, you say that ... but it looks like support for "add-apt-repository ppa:user/distro/repository" (e.g. ppa:ci-train-ppa-service/ubuntu-rtm/landing-002) was uploaded to utopic last Thursday [19:56] robru: yes [19:57] cjwatson: I made it my itch to scratch, while at debconf. It just took a bit before it got reviewed by mvo [19:57] cjwatson, the question is ... did it make it into rtm :) [19:57] I noticed, thanks :) [19:57] I was actually surprised it wasn't already working when I tried to use it then [19:57] ogra_: I'm sure that would be easy enough to sort out with a binary copy ... [19:58] sure [19:58] just saying :) [19:58] oh, it's actually only in utopic-proposed as yet [19:58] oh [19:58] looks like mvo had a go at fixing autopkgtests and didn't quite manage it [19:59] I think it's only pyflakes ... [19:59] cjwatson: I still maintain that "ppa:user/repo" should default to the distro that the user is using. [20:00] robru: I strenuously disagree; let's work towards killing the two-arg forms [20:00] ambiguity FTL [20:00] cjwatson: I can probably fix it, it ought to be from my change [20:01] cjwatson: my citrain script was really simple before; now it has to go trying to detect what distro it's running on in order to request a ppa for the distro it's running on. there's no possible situation in which you'd want to add a ppa from a mismatched distro, so what you call "reducing ambiguity" I call "extra typing for no benefit" [20:01] robru: I beg to differ [20:02] I regularly want to add a utopic PPA to install on a rtm image ;) [20:02] well, its a shame that we cant easily adjust /etc/{os|lsb}-release [20:02] cyphermox: cjwatson: what's the easiest way to detect the distro that a script is running on? [20:02] cyphermox: it's not your failure, it's a missing test dependency [20:02] no idea [20:02] robru: maybe ask apt in some way [20:03] grep rtm /etc/apt/sources.list :P [20:03] yeah [20:03] I haven't played with rtm enough to have an intuition [20:04] cyphermox: I'm fixing it now [20:04] cjwatson: ok [20:04] cjwatson: I like how we're ruining PPA syntax for *all* PPA users for *all* of ubuntu for this one short-lived corner case. [20:05] robru: ah! [20:05] system-image-cli [20:05] robru, but it isnt ubuntu :) [20:05] it is ubuntu-rtm [20:05] robru: I continue to disagree [20:05] robru: it's not, ppa:robru/ppa will work [20:05] its a different distro [20:05] No matter how inflammatorily you want to put it ... [20:05] so will ppa:robru in fact for ppa:robru/ppa [20:05] ogra_: yeah, that's exactly my point [20:06] cyphermox: cjwatson just said he's working to abolish the two-arg form [20:06] Explicit is better than implicit, and maybe some day we'll fix the politics to have Debian PPAs too [20:06] robru: Way to misquote me! [20:06] ahah [20:06] I'm leaving this conversation now. Bye. [20:06] ogra_: people who are part of the ubuntu ecosystem but aren't involved in rtm are inconvenienced by having to now specify ppa:foo/ubuntu/bar when they don't care about rtm [20:06] robru: I think he just means educating people in being explicit with what PPA they want, for which distro [20:07] right [20:07] cjwatson: "robru: I strenuously disagree; let's work towards killing the two-arg forms" what did I misquote? [20:08] robru: like I said, you will still be able to use ppa:mathieu-tl/ppa or even ppa:mathieu-tl to mean the PPA named "ppa" for distro "ubuntu", for the foreseeable future, I think there's a few scrips in various places which depend on that [20:08] robru, it is really an improvement long term ... you might be able to install the latest cinnamon from a mint PPA, some funny debian stuff from a debian PPA etc [20:08] That's not "[I'm] working to abolish the two-arg form". [20:08] cjwatson: so "let's work" means "you should work"? [20:08] No, it's a position statement. [20:08] I'm still allowed to have opinions, right? [20:09] But like I say I'm going elsewhere now since I don't think this is productive right now. [20:09] cjwatson: come to think of it; maybe I should have made it a bit more easy to default to debian PPAs for software-properties installed on Debian, since that might be what's expected if/when debian PPAs exist. [20:09] ...burn that bridge... [20:11] I'm happy to resume it when we can talk a bit more calmly about reasons and trade-offs. (Perhaps over beer.) [20:11] cjwatson, we in qa can haz? [20:11] robru: I'm familiar with the software-properties code for this stuff, since I've touched it a few times. If you have clear requirements or expectations you could let me know [20:11] hmm, beer [20:12] brendand: I uploaded the fix for the autopkgtest, hopefully, so if I'm lucky it'll be in utopic soon [20:12] cjwatson, but we need it in RTM :) that's sort of the point [20:12] cyphermox: all I want is to update my citrain script to be able to add distro-appropriate PPAs without spending a thousand lines of code trying to determine what distro i'm on, or worse, asking the user, because that's not something worth trifling the user over. [20:13] robru: I think you can use system-image-cli for that [20:13] the channel line will include whether it's RTM or not pretty explicitly [20:13] and it will give you an idea whether you're on proposed or not [20:15] cjwatson, just to be clear - this change allows us to specify 'ppa:ci-train-ppa-service/ubuntu-rtm/landing-009' to add-apt-repository and have it work? [20:16] brendand, yes [20:16] ogra_, cool. i can't figure out why that might be bad [20:17] brendand, because you have to add /ubuntu/ to "normal" ppa lines too now [20:18] ogra_, oooh. why can't it just default if you don't specify it? [20:18] plars, just let me know when those jobs are updated [20:18] ogra_: you don't :) [20:18] balloons: see above, I did already [20:18] balloons: they are updated, give it a try [20:18] cyphermox, oh, i thought thats what rob is upset about [20:19] it is, afaik [20:19] but ppa:mathieu-tl/ppa does work and expands to ppa:mathieu-tl/ubuntu/ppa as it should [20:19] it's a bit hardcoded to ubuntu right now; so if you want a ubuntu-rtm PPA then you really do have to specify it [20:20] right [20:20] but it wasn't really possible to make it work for software-properties to default to ubutnu-rtm on rtm, because you can't use lsb-release/os-release to do that [20:20] yeah, which i still think is a shame [20:21] heh [20:21] it's not worth spending too much time on unless using ubuntu-rtm for an extended period of time [20:21] yeah [20:22] well, i doubt we will use it long or often, but we will occasionally [20:22] and there it shoudl work [20:27] ogra_: just needs fixing in {os,lsb}_release, and for the software-properties to look there [20:27] cyphermox, no, we cant change it ... it was changed and reverted already [20:27] ah? [20:27] why not? [20:27] ask infinity [20:28] i dont know the details but there was a discussion between pitti and him before pitti rolled it back [20:28] if infinity did try, then I believe it's not done for a good reason [20:28] right [20:29] Is there a 244 building? [20:29] ogra_, cyphermox - i believe the justification was related to the short-livedness of rtm [20:30] tedg, i dont think anyone triggered one [20:30] Oh, I thought they went just one after the other. [20:30] tedg, the cron build is still several hours away [20:30] they are ... 2:00 UTC and 3:00 UTC [20:30] rtm being the later one [20:30] Ah, so only once a day now. [20:31] ogra_, https://bugs.launchpad.net/ubuntu-rtm/+source/base-files/+bug/1362496 [20:31] Launchpad bug 1362496 in base-files (Ubuntu-rtm 14.09) "LSB release and /etc/os-release still say "Utopic", needs to be RTM" [High,Triaged] [20:31] Cool, thanks ogra_ [20:31] tedg, well, if we have stuff we want in a separate image we usually roll a manual one [20:31] and usually we have two ... but rarely on promotion days [20:32] cyphermox, see the bug above [21:13] robru, brendand, et al: Right. Following cocoa, here's a somewhat more considered version of my position on apt-add-repository etc. near which it'll hopefully be a bit easier to find consensus. [21:13] 1) Even though its main focus is Ubuntu, Launchpad is a webapp that serves multiple distributions, and many of the ways it does so are useful to us (syncing from Debian, multi-distro bug tracking, etc.). [21:13] 2) As a result of 1), we've generally found hardcoding Ubuntu in Launchpad to be a mistake. Making the distribution explicit may be slightly more typing but it makes the layout more regular and thus easier to understand, and simplifies the code. [21:14] 3) Unhardcoding Ubuntu from the PPA URL format (→ https://launchpad.net/~OWNER/+archive/DISTRO/NAME) was necessary to support ubuntu-rtm, and generally made things internally clearer. Redirections are in place. [21:14] 4) Launchpad now has a canonical archive reference form which can be used by any client tool to look up archives. This has made it possible to greatly simplify a number of client tools (particularly obvious in copy-package). [21:14] 5) It's easier to work out what client tools that interface directly with Launchpad do when their primary visible syntax matches Launchpad's layout reasonably closely, and when they behave roughly the same way as each other. [21:14] 6) As a result of 4) and 5), the primary syntax of add-apt-repository should be ppa:OWNER/DISTRO/NAME, since that matches the archive reference form most closely while still using the ppa: syntax people are familiar with. [21:14] 7) As cyphermox points out, it's sometimes necessary to refer to a distribution other than the running one, even for PPAs. The fully-explicit form should thus always be available to users. [21:14] 8) That said, I don't object to abbreviated forms that are resolved client-side when they can be handled reasonably straightforwardly. I just don't want them to be the only available forms, since that has always proven to require refactoring later. [21:14] (But this is a position, supported as best I can with rationale, not a list of instructions.) [21:16] Hopefully that's a bit clearer. I was having difficulty responding directly earlier without sitting back and writing it all out in detail in a text editor ... [21:19] cjwatson: i don't disagree that the explicit form should always be available, I just want ppa:OWNER/NAME to do what I expect when I'm eg writing a tool that needs to support both ubuntu and ubuntu-rtm but doesn't particularly care which it happens to be running on [21:20] Right, I'm fine with that as long as the explicit's available. The problem in the past has been that people write the short forms and never get round to making the full versions available. :-) [21:20] cjwatson: eg if you're on rtm and you type ppa:OWNER/NAME you're unlikely to want ubuntu, it should default to rtm, and then if you really do want ubuntu you can specify that [21:20] It'd be a bit of a change for ppa:OWNER/NAME to resolve to DISTRO != ubuntu, and it wouldn't (couldn't) match the Launchpad URL redirections. [21:20] But it's probably justifiable. [21:22] (As long as the abbreviated forms are always resolved immediately in context, rather than being potentially passed around among different distribution contexts and rendering them ambiguous again. But hopefully that wouldn't be the natural implementation.) [21:24] it's not that big a change, I just couldn't think of a way to do it properly with what we had available in ubuntu-rtm, that would work elsewhere too [21:25] ie. guessing that we're on rtm is complicated to do, except maybe using system-image-cli, which isn't available in ubuntu or elsewhere [21:25] so; seems to me like the right way to do it would be for base-files to say ubuntu-rtm on rtm [21:26] that said, https://bugs.launchpad.net/ubuntu-rtm/+source/base-files/+bug/1362496 was brought to my attention :) [21:26] Launchpad bug 1362496 in base-files (Ubuntu-rtm 14.09) "LSB release and /etc/os-release still say "Utopic", needs to be RTM" [High,Triaged] [21:53] fginther, you about for a moment? [21:53] balloons, I do have a moment [21:54] fginther, so I had plars revert autolanding on devices for calc. I'm not sure what's broken, but way too much going on to figure it out as it's not vital imho. Anyways, the autolanding job needs to use python3 for tests. [21:55] fginther, I also noticed stacks/phablet/ubuntu-touch-coreapps.cfg doesn't reflect python3 for many of the ones we've converted.. I suspect we should push an mp for them so the changes won't be generated over? [21:55] balloons, yeah, i have a branch I was working on for those changes [21:55] * fginther failed on that one [21:56] fginther, no worries.. :-) So I guess a couple things.. Could you toggle calc autolanding to use python3 for now? the branch can come after [21:56] Trying to land https://code.launchpad.net/~rpadovani/ubuntu-calculator-app/1357983/+merge/232630 [23:07] balloons, the job has been updated [23:07] balloons, sorry for the delay [23:10] fginther, ty