[02:09] === trainguards: IMAGE 272 building (started: 20141007 02:10) === [03:10] === trainguards: RTM IMAGE 88 building (started: 20141007 03:10) === [03:59] === trainguards: IMAGE 272 DONE (finished: 20141007 04:00) === [04:00] === changelog: http://people.canonical.com/~ogra/touch-image-stats/272.changes === [04:19] === trainguards: RTM IMAGE 88 DONE (finished: 20141007 04:20) === [04:20] === changelog: http://people.canonical.com/~ogra/touch-image-stats/rtm/88.changes === [05:16] ogra_: should we just sync ubuntu-touch-session? [05:16] ogra_: I have a landing that I want to do for rtm that includes ubuntu-touch-session and tp-ofono [05:17] but latest u-t-session includes your priority changes [05:17] not sure if you want qa to sign that off first or not [05:17] ogra_: so in case you decide to land that, mind just syncing latest ubuntu-touch-session and tp-ofono? [05:18] * rsalveti bed time [07:04] Mirv: as urfkill was tested and will hopefully land pretty soon, can I get a silo for line 36 to start building/testing? [07:05] Mirv: also, was wondering what's the state for line 6 as it says aborted but at the same time that it tested well [07:12] thostr_: it seems urfkill (line 6, silo 011) has regressed in the sense that it no longer builds for arm64, powerpc and ppc64el [07:12] cyphermox: ^ [07:13] and therefore the build job never finishes [07:13] if that's by plan, cyphermox should contact archive admins to get this architectures change accepted and uploaded [07:14] added a comment [07:14] thostr_: assigned line 36 ignoring conflicts [07:15] thostr_: that urfkill problem only affects utopic naturally, not rtm [07:15] Mirv: thanks [07:22] good morning [07:22] trainguards, could someone reconfigure silo 1 for me? [07:24] tvoss: done, it removed dbus-cpp from the silo [07:24] Mirv, yup, AlbertA landed that part tonight [07:24] and morning [07:42] tvoss: want a reconfigure for that ^ ? [07:42] sil2100, nope, Mirv already reconfigured the silo [07:43] ACK [07:43] sil2100, thanks for asking, and good morning :) [07:59] don't tell me silo 09 is also busted... [07:59] * brendand waits [08:01] brendand: what's up? [08:01] sil2100, same thing as 20 [08:02] uh? [08:02] hmmm [08:02] seems like a mess of dependencies [08:02] mzanetti, Saviq: ^ ? [08:02] mzanetti, Saviq: do you have any problems after installing rtm silo 009? [08:02] this time fallback adb doesn't work - maybe that got merged [08:02] ogra_, ^? [08:03] sil2100, not that I know of, kgunn tested it yesterday and gave it a +1 [08:03] lemme try it out [08:03] sil2100, ... [08:03] The following packages will be REMOVED: [08:03] ubuntu-system-settings-wizard ubuntu-touch ubuntu-touch-session [08:03] !!!! [08:04] uh? [08:04] Looks like a killer feature to me [08:04] With emphasis on 'killer' [08:06] I wonder which dep causes this to happen [08:09] tvoss: it looks like it didn't remove the package it said it removed, so removing dbus-cpp manually from landing-001 [08:10] Mirv, thank you [08:14] Saviq, yeah looks like kgunn tested it [08:15] trainguards, i'd like a silo for line 70 [08:16] brendand, same thing happened to me with the citrain tool [08:17] brendand, but apt worked fine [08:20] brendand, I'm reflashing now and will investigate step by step [08:21] ogra_: ubuntu-touch-session is already in rtm/002 [08:21] Some of the binaries in that PPA shouldn't be installed on a device, so you probably need to supply an explicit list of packages to phablet-config writable-image. [08:21] Sadly citrain doesn't offer that interface. [08:22] Mirv, well, it is already in utopic ... and rsalveti asked me to take care of his landing too (which i was planning to do afterwards) [08:22] Specifically e.g. qtmir-desktop but perhaps others [08:22] cjwatson, it does dist-upgrade, surely that should be ok? [08:22] brendand: What does? [08:22] cjwatson, the citrain tool [08:22] brendand: Only for host-upgrade. [08:22] ogra_: alright, 002 does not seem landing immediately [08:23] cjwatson, no for device-upgrade too [08:23] ogra_: rtm-022 for you [08:23] Mirv, thanks [08:23] adb shell SUDO_ASKPASS=/tmp/askpass.sh sudo -A apt-get dist-upgrade --yes [08:23] brendand: Oh, right, misreading ... [08:23] brendand: Well in that case need a less incomplete log [08:24] cjwatson, http://paste.ubuntu.com/8513034/ [08:25] The following package was automatically installed and is no longer required: [08:25] lightdm [08:25] wow ! [08:25] ogra_: Because ubuntu-touch is being forced out [08:25] Probably [08:25] yeah [08:25] most likely [08:26] Actually ubuntu-touch-session, but anyway [08:30] sil2100, brendand, ok, so the way the citrain tool is doing device-upgrade is the problem [08:30] sil2100, brendand, unity8-private has a new dependency on qml-module-connectivity [08:31] Oh yeah so it doesn't work because of disabling the primary archive during dist-upgrade [08:31] but the citrain tool removes sources.list during the dist-upgrade, meaning that package is nowhere to be found [08:31] yup [08:31] which breaks havoc [08:32] so writable-image, install qml-module-connectivity, then citrain [08:32] yeah, that'd work [08:33] qml-module-ubuntu-connectivity to be exact [08:33] brendand, sil2100 U [08:33] ↑ [08:33] Saviq: thanks! [08:35] Saviq, is that thing telling you e.g if you have data enable/are online? [08:35] seb128, yes [08:35] great [08:35] seb128, http://developer.ubuntu.com/api/qml/sdk-14.10/Ubuntu.Connectivity.index/ [08:35] is the api documented somewhere? [08:35] Saviq, thanks [08:51] Saviq, cjwatson - Mirv provided a more correct way of upgrading from a specific ppa so i will file a bug on phablet-tools to update citrain to use that [08:51] brendand, coolz, thanks [08:51] Saviq, in the meantime, if the install steps for a silo varies from just running citrain then you have to remember to provide those steps [08:51] Saviq, pass that on to your team [08:51] Saviq, we in QA don't read minds :) [08:51] we have many talents but not that [08:52] brendand, well, if I knew before, I just went manual I'm afraid [08:52] brendand, just hire someone then ! [08:52] brendand, but I'll be keeping a closer look now [08:52] ogra_, yes a mindreader would be useful [09:03] trainguards: note, 87 and 88 are ready but ubuntu/landing-014 is blocking as it still has a package in the unapproved queue [09:03] trainguards, i seem to have a weird gpg key issue in rtm-022 https://ci-train.ubuntu.com/job/ubuntu-rtm-landing-022-1-build/11/console [09:03] gpgv: Signature made Tue 07 Oct 2014 08:35:27 AM UTC using RSA key ID B879A3E9 [09:03] gpgv: Can't check signature: public key not found [09:03] dpkg-source: warning: failed to verify signature on ./ubuntu-touch-session_0.108+14.10.20141007-0ubuntu1.dsc [09:04] (that was a source package sync from utopic) [09:04] ogra_: That's not the error, it's just out-of-order output [09:04] I'm not sure exactly what the error is but it's not that :) [09:04] hmm [09:05] it complains about the changelog being empty a little higher up [09:05] Why not just copy the source package? [09:05] which is weird as well for a synced package [09:05] Rather than doing a weird citrain thing which furtles about with the package [09:05] cjwatson, i thought that is what was done when the silo was set up [09:05] Evidently not! [09:06] Do you even need to rebuild the source here? [09:07] theoretically not (its even only a 1char change in an upstart job) [09:07] Maybe it should have been "sync:69" [09:07] but right, there are no packages in there at all [09:07] it was originally [09:07] not sure who changed it [09:07] Personally I would just copy-package -b rather than attempting to understand this [09:13] GRRR °! [09:14] i hate that i always have to click twice in henkins (and that it forgets my selection over that) [09:14] * ogra_ wanted a watch only build ... indeed the second click swallowed the check mark [09:14] ogra_: familiar feeling [09:15] well, re-building wont do harm i guess [09:17] pfft [09:20] brendand, given jibel tested the ubuntu-touch-session change yesterday already, do you think it is ok to call it an isolated bugfix so i can land it after afunctional test of the package (after all iits only a one char change) ? [09:21] ogra_, what did i say yesterday :) [09:21] dunno .... [09:21] * ogra_ looks for the scrollback [09:21] :) [09:22] ogra_, if there was ever a better example of a one-character change having a big impact, this is it [09:22] ogra_, can we at least give it a brief smoke test? [09:23] brendand, i'll install it on my production phone anyway [09:23] which should be a "long smoke test" :) [09:23] tell me if you want to do additional QA signoff too ... [09:24] (i just want to know if i need to set the checkmark :) ) [09:24] ogra_, yes i do [09:24] thanks === alan_g is now known as alan_g|afk [09:55] Saviq, is there a way i can watch for apps being suspended (according to the lifecycle management model)? [09:56] brendand, ps ax | grep binary [09:56] brendand, if status is S, it's running, if T it's suspended === psivaa changed the topic of #ubuntu-ci-eng to: Need a silo? Ping train support: trainguards | Need help with something else? Ping vanguard: psivaa | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: robru is on vacation [10:19] cihelp could someone help with this merge which appears to be barfing on a packaging issue? https://code.launchpad.net/~dpniel/ubuntu-weather-app/timezonePlugin/+merge/235389 [10:20] popey: looking [10:20] thanks [10:28] ogra_, is it possible to exercise a measure of control over what's in the next image? [10:29] ogra_, can we be sure that only stuff in silos that we sign off will land? [10:30] brendand, you can check it with rmadison ... [10:32] brendand, the silo "publish" button should theoretically be your control [10:32] we have a way to block stuff in proposed ... but that woulld just dupliacate that functionality [10:36] brendand: that's basically why the publish step is manual [10:36] brendand: that we only publish what we want to land in the image - so if there's something you don't want to land in an image yet but is ready to publish, just give us a sign [10:49] can I get a silo for line 37? [10:51] popey: the upstream tarball that's available for this version 'ubuntu-weather-app_1.8.3ubuntu1' has deps issue. i dont know how to update that. will have to wait for fginther [10:52] psivaa: ok, thanks. [10:54] ted: Wellark: thostr_: please deal between yourself whether the new wellark landing goes in first or ted's Indicator Greeter Modes. both have silos and both have unity8. [11:01] Mirv: which one should i use for silo testing? ubuntu-touch/ubuntu-rtm/14.09-proposed or ubuntu-touch/ubuntu-rtm/14.09 ? [11:01] /me anxiously awaits rtm silo 9 to land so I can have pull-to-refresh [11:02] Wellark, proposed indeed ... [11:02] Wellark, the other oone is days old usually [11:02] Mirv: ted is not around, so he has no saying [11:02] just land my stuff first ;) [11:03] Wellark: yeah, I don't see him objecting to that ;) [11:04] could I get a silo for line 89? [11:04] oh, sorry [11:04] I have! [11:05] what!? [11:05] Mirv: what is that --^ [11:06] Wellark: something is not in sync, looking. [11:06] sil2100, psivaa - so u-s-s autopilot tests are ok here [11:06] when we have utopic + rtm + everyone wanting to land unity8 changes, it gets kind of complicated [11:07] brendand: hmm [11:07] Mirv: thanks! [11:07] brendand: was that from a fresh install? [11:07] psivaa, yeah [11:07] psivaa, utopic [11:08] brendand: ok, and if it's 272 then it's going to be tricky to narrow that down [11:08] tvoss, i just noticed something watching the oom scores while testing the session score fix here ... http://paste.ubuntu.com/8513826/ ... seems the oxide renderers never get bumped, even if the webapps are in the background [11:08] brendand: could be something that happened before is causing this [11:09] tvoss, in case of that paste dekko is the foreground app [11:09] and while the webapp-containers get bumped up, their backends never do [11:09] Wellark: cool, just one silo was missing the clean that became possible to do an hour ago. please rebuild now. [11:10] also, lots of indicator landings ongoing too [11:10] Mirv: what do I have to do to get a decent service around here.. [11:10] fine. I will rebuild it! [11:10] ;) [11:11] Mirv: is it enough to rebuild i-network alone? [11:12] Wellark: I think not, the unity8 didn't reach the PPA either yet [11:16] ev: should I be worried that apport is crashing? [11:17] sil2100: the crashes so far http://paste.ubuntu.com/8513883/ [11:17] Holy crap [11:18] davmor2: you were opening and closing a lot of apps, right? [11:18] sil2100: nope haven't got that far yet this is just running through the standard tests [11:18] WTH [11:19] davmor2: ok, that doesn't look to good, but just to be sure - did we check this during last promotion testing? [11:19] brendand: ^ ? [11:19] sil2100, hmm, i think i saw that too on some settle-before top outputs [11:19] (apport running) [11:20] sil2100: probably not I saw the settings app crash so wanted to check if there was a crash for it and that is what I saw [11:20] Most of them seem to be similar to what we see in smoketesting [11:20] sil2100: it might be a healthy thing to do in future [11:20] sil2100: yeah most likely [11:53] sil2100, hey, do you remember if :native qualifier was fixed somehow in -ci jobs? http://s-jenkins.ubuntu-ci:8080/job/unity8-utopic-amd64-ci/1610/console [11:54] Saviq: sadly I don't know [11:54] ok, fginther will when he's around [11:55] hello [11:55] we have a lot of (random?) failing tests in CI and autolanding for UITK since a few weeks [11:55] is this something that's happening in other projects as well, or only UITK? [11:56] sil2100, any updates on bug #1359667? or the staging idea we talked about (did I not file a bug about this?) [11:56] bug 1359667 in CI Train [cu2d] "There should be a hook mechanism available" [Undecided,Confirmed] https://launchpad.net/bugs/1359667 [12:07] pstolowski: I can probably remove your line 96 since I added the rtm silo straight beneath the utopic one (which I like for easier seeing the both) [12:10] Mirv, sure, thanks [12:11] ogra_: I wonder if you're interested in acking the package changes at https://ci-train.ubuntu.com/job/ubuntu-landing-018-2-publish/lastSuccessfulBuild/artifact/packaging_changes_ubuntu-app-launch_0.4+14.10.20141006-0ubuntu1.diff ? barry` tried to publish it yesterday to utopic. [12:12] mzanetti, there is a bug not fixed by rtm silo 009 that is mentioned as being fixed [12:12] mzanetti, https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1282460 [12:12] Ubuntu bug 1282460 in unity-scope-click (Ubuntu) "Missing rating info in app preview header" [High,Triaged] [12:12] Mirv, isnt barry code-dev ? [12:13] Mirv, ack [12:13] ogra_: yes he is, but he didn't launch it again. [12:13] ogra_: thanks. === psivaa changed the topic of #ubuntu-ci-eng to: Need a silo? Ping train support: trainguards | Need help with something else? Ping vanguard: cihelp | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: robru is on vacation [12:14] barry`: so if you see the "UNSTABLE" / yellow publish job results, you can review the debian/ changes and then use the ACK_PACKAGING option when running the publish job again [12:34] davmor2: apport itself? Yes, that's worth noting. [12:36] Mirv: land lines 36 and 37 quick before ted gets online! [12:40] Mirv: I think barry` should know that! I guess..! [12:40] Maybe he just forgot..? [12:42] ev: yeap it looks like 3 of the bugs haven't been able to upload so I'm assuming they ran out of memory trying to file it at a guess _usr_share_apport_recoverable_problem.32011.crash [12:42] _usr_share_apport_recoverable_problem.32011.upload [12:43] sil2100: can you check 016 utopic error please? [12:43] * Mirv needs to run [12:43] Mirv: ACK [12:43] o/ [12:44] davmor2: problems in recoverable_problem aren't critical. This is just an API for reporting problems the application was able to recover from. [12:44] but it's still important to fix [12:44] obvs [12:44] Mirv: ok, this needs to be rebuilt... [12:44] ev: :) [12:44] davmor2: do you know the errors.u.c URL for it? [12:45] ev: doesn't say uploaded so I don't know [12:45] thostr_: so, sadly! Silo ubuntu/016 needs a rebuild [12:45] thostr_: the indicator-network package there is no longer up-to-date, as 2 hours ago a new indicator-network landed to the archive [12:46] So it needs to be rebuilt [12:46] davmor2: 'ave a look at your log, see if you can find the crash ID [12:46] sil2100: ok, will trigger one [12:46] thostr_: this also means, sadly, another re-test ;/ [12:46] ev: and previous error reports still doesn't work [12:46] sil2100: no worries [12:47] Thanks! [12:47] sil2100: which one indicator network silo did you land [12:48] Saviq, is mzanetti around today? [12:48] brendand, yeah, I think he's lunching atm [12:50] umm.. what happened? [12:50] sil2100, thostr_ ? [12:52] Wellark: indicator reordering landed for utopic [12:52] ted, I know you are here somewhere! [12:52] thostr_: ok :) [12:53] thostr_: I didn't land it, but I think Mirv landed it === barry` is now known as barry [12:54] sil2100: ok, once silo 16 has built i-network again, please land it [12:54] it's ready [12:54] Wellark, Good morning! :-) [13:04] ted, aren't bug #1223850, bug #1323654 fixed? [13:04] bug 1223850 in url-dispatcher (Ubuntu Saucy) "please support http and https URLs" [Undecided,New] https://launchpad.net/bugs/1223850 [13:04] bug 1323654 in url-dispatcher (Ubuntu) "error launching calendar/g+ from gmail webapp" [High,New] https://launchpad.net/bugs/1323654 [13:05] Saviq, Yeah, I think so. [13:05] Saviq, We're still getting a few recoverable errors on the second one, but I think it's just non-upgraded devices. They're not in the newer versions. [13:05] Saviq, I think that what I've learned there is that the image deployment process is *really* slow at getting updates out. [13:06] ;) [13:06] Wellark: did you re-test? [13:06] thostr_: ^ ? [13:06] ted, one for you though: bug #v [13:06] bug #1378350 [13:06] bug 1378350 in url-dispatcher (Ubuntu) "Need to allow multiple apps to handle the same url patterns" [Undecided,New] https://launchpad.net/bugs/1378350 [13:07] sil2100: Wellark: this is still building... [13:08] Saviq, I think we have a bug for that, let me look. It's a 15.04 issue, needs design, but on the proposed list for DC. [13:08] ted, ah, didn't see one in the 20 or so bugs that are there for url-dispatcher{, (ubuntu)} [13:09] brendand: yes, I am [13:09] * Saviq dogfooded the phone for 3 weeks with my wife, got a slew of bugs to file everywhere [13:09] Saviq: At least some jobs have "hook_source: lp:~canonical-ci-engineering/+junk/native-strip-hook", which looks relevant [13:09] mzanetti, did you get my earlier ping? [13:09] Saviq, Hmm, yeah. I guess we didn't have a bug to track it. Yours gets to be it! :-) [13:09] \o/ [13:10] brendand: yeah... sorry... I just ignored this channel given that it pings me every 5 minutes [13:10] cjwatson, indeed [13:10] mzanetti, fix your IRC client [13:10] Saviq: my client is working fine [13:10] mzanetti, I don't get pings from here at all, wonder why [13:10] Saviq: not unity8, but I guess it's easy to add [13:10] cjwatson, yeah, should be [13:11] cjwatson, we'll need to add it to all the projects that have the direct g++ 4.9 dep to allow them to cross compile [13:11] i.e. bug #1353855 [13:11] bug 1353855 in unity8 (Ubuntu) "Explicit g++ 4.9 dependency breaks cross-building" [High,In progress] https://launchpad.net/bugs/1353855 [13:12] Saviq: Why do you need :native on that? I would have thought you normally wanted the compiler for the host architecture except in some special cases. [13:13] cjwatson, because apt can't resolve the deps for cross-building, try `apt-get build-dep -aarmhf unity8` [13:13] ted, my mind just thought about squashing content-hub and url-dispatcher together... urls are content after all ;) [13:13] Saviq: stripping :native won't help; you don't have any :native build-deps in unity8 [13:13] cjwatson, not yet, there's a branch [13:14] oh I see [13:14] https://code.launchpad.net/~dobey/unity8/fix-cross/+merge/234818 [13:14] * cjwatson peers [13:14] Saviq: that looks precisely backwards [13:14] Saviq, I don't think that's off the table, not sure if we want something simpler for URLs though. [13:14] cjwatson, I know... but it works... [13:14] cjwatson, I think it just tricks apt into resolving [13:14] Saviq, I'm happy maintaining less projects :-) [13:14] Saviq: I'll follow up, I don't approve [13:15] Saviq, And as a bonus, it gives work to kenvandine. [13:15] cjwatson, what *I* would like is for all the projects that have the g++-4.9 dep Just Work™ for cross-building [13:15] * kenvandine ignores ted, as usual :) [13:15] Saviq: the proper fix for this isn't to hack about in individual projects; Debian has been working on a better standard fix [13:15] came up at debconf [13:15] oh good [13:17] cjwatson: is there a better actual fix that can be done now? [13:17] let me follow up in the MP [13:17] you can probably do this but you really need to be aware of the downsides and they need to be explicitly commented in the control files [13:17] because they are not obvious [13:17] it's what we're doing in other projects (so i don't know why pbuilder is failing there) [13:21] so maybe "don't approve" is a bit strong, but anyway, IRC margin too narrow for this conversation, will comment in more detail [13:30] brendand: do you have time now? [13:31] sil2100, for what? lrt - i'm here [13:31] hi trainguards! I'd like to add a branch to an existing silo. Is it just a matter of adding the new MP to the right cell, and clicking on "rebuild" ? [13:32] Or is there any other step I should follow? [13:37] ok, just clicked on "Reconfigure silo". Maybe I should have waited till the previous packages were built. === jhodapp_ is now known as jhodapp === Ursinha-afk is now known as Ursinha [13:38] and the build seems to be chugging along.... trainguards nevermind the above ^ [13:39] alecu: hi. sorry, i missed your message, but glad you figured it out! :) [13:39] no problem, thanks anyway :-) [13:42] sil2100: jussi did a new testing round on silo 16. all green [13:42] I had a meeting, just got back [13:42] Wellark: aye! [13:43] alecu: oh ;) [13:43] Wellark: oh, and I see someone just published your silo! [13:43] Mirv: it was YOU, wasn't it?! [13:43] :) [13:44] well, he had to make up for the careless act of publishing the indicator reordering too soon :) [13:47] popey, https://code.launchpad.net/~dpniel/ubuntu-weather-app/timezonePlugin/+merge/235389 is building successfully now [13:47] thanks fginther [13:52] sil2100, what is the bot telling me there ? [13:53] (rtm 022) [13:53] ogra_: usually that means that the MPs you want to release are not in the top 'Approved' state [13:53] that was a sync from utopic ... not sure why it thinks anything could be unapproved [13:53] Ah, ok, [13:53] Someone left the merges in there [13:53] oh. yeah [13:54] ogra_: so, someone made the landing and didn't clean the merges list [13:54] sil2100, right, i dont know who changed it from syncing from the utopic silo to syncing from the archive either [13:54] Let me reconfigure and republish [13:54] i removed the url [13:55] (someone definitely shuffled the spreadsheet data over night) [13:55] The spreasheet needs to DIE [13:55] ++ [13:58] ogra_, barry: published o/ [13:58] :) [13:58] thx !! [13:58] sil2100: rock on! [13:58] now on to rsalveti's ubuntu-touch-session landing [13:59] barry: remember to clean the MP list for sync: silos in case you see people leaving those [13:59] barry: sync: silos shouldn't usually have any merges in them (even though it's possible to land one thing with some new merges) [13:59] i wonder if i can build already or if i have to wait til my package is in the archive [13:59] sil2100: i guess the best way to scan for those is via the dashboard? [14:00] barry: yeah, but you can also just look at the spreadsheet [14:00] ted: where does one approve the setuid changes? [14:00] ted: and hi! [14:00] I approve the packaging, sarnold approved the MP [14:01] so if it is good enough to mention it here, consider it mentioned. otherwise, please advise [14:01] jdstrand, Cool, yeah, it seems the packaging has been ack'd. [14:01] sil2100: hmm, it's not clear to me what criteria in the spreadsheet is easily scanned to know which syncs can be cleaned [14:01] jdstrand, Not sure if that was you or not. [14:01] it was not [14:01] jdstrand, But that's what we needed. [14:01] ok, well, double review then :) [14:02] ted: was this the landing from yesterday? [14:02] Heh, not a bad thing to double review :-) [14:02] barry, Yes [14:02] ubuntu silo 18 [14:02] ted: cool, i'll ack the pack [14:03] oh, looks like it's already been published [14:03] barry, Yeah, I think it's in flight. [14:03] ted: at least this whole process isn't racy [14:03] barry, I think the European gnomes must have gotten to it while we were sleeping :-) [14:04] ted: the lesson i learn from that is that i should go back to sleep! :) [14:09] sil2100, hmm, i made a mess in silo rtm002 ... [14:09] 002? [14:09] (i accidentially copied telepathy-ofono binaries where i shouldnt have) [14:09] Oh [14:10] after deleting the package i cant copy the source anymore [14:10] Where did you want to copy it to? [14:10] from utopic to rtm 002 [14:10] uh [14:10] Yeah, that's the common problem with LP PPA's [14:11] sil2100: this issue where the app you open looks to of crashed is getting old really quickly :( [14:11] You would need to reupload with a bumped version [14:11] davmor2: what do you mean? [14:11] sigh [14:12] you could probably use a different PPA [14:12] I updated the Trello card with ToyKeeper's comments, but I can't seem to remove the "blocked" tag to put it back on review. [14:12] Is this a permissions thing or am I missing how to do that? [14:13] sil2100: you use the phone for a bit open an app and the whole system freezes eventually it looks like it restarts something and then you have a working system again [14:13] Laney, oh, they dont use the same archive, yeah [14:13] sil2100, ^^ could you move that landing to another PPA ? [14:14] ogra_: hm, could be doable, but it'll take a few moments [14:17] Saviq: ping [14:19] nik90, pong [14:19] Saviq: I saw earlier that you were adding Ubuntu.Connectivity as a dependency of Unity8, which silo were you referring to? [14:19] Saviq: I just want to make sure that the qml-connectivity-plugin is present in the images by default before I start using it in my app [14:20] elopio: hey! Give me a ping once you're around [14:20] trainguards, I have a package for SRU in the UNAPPROVED queue, and it ended up without a bug number in the automatically-created changelog (for unknown reasons) ... do I just rebuild the package again even though it's already published to the UNAPPROVED queue? [14:21] argh [14:21] hmm [14:21] bregma: somoene from the SRU team pointed this out? [14:22] sil2100, yes, although I noticed the bug number was not in the changelog earlier (then forgot to follow up) [14:22] my bad for that [14:22] nik90, weelll, you should only use it if the framework ensures it's there [14:22] nik90, which I don't think is the case at the moment [14:22] bregma: so yeah, it can be rebuilt and republished in the same silo - just wondering what caused the bug number not to be registered [14:23] Wellark, do you know if Ubuntu.Connectivity will be added to a framework? [14:23] Saviq: framework as in the ubuntu-sdk-14.10 framework? [14:23] nik90, yeah, see `click framework list` [14:23] sil2100, OK, and yes I'm wondering that too: if a rebuild doesn't fix it I'll add an explicit changelog [14:24] nik90, there isn't a 14.10 no-dev one yet, but it will be there before release [14:24] Saviq: I am using the 14.10 framework for the app eitherway [14:24] Saviq: but it seems like the connectivity plugin isn't in the devel framework as well [14:24] nik90, well, sure, but that framework doesn't yet provide the connectivity module [14:25] bregma: ok, just give me a sign if the rebuild doesn't help, then I might look into it also from the train side [14:25] nik90, yeah, exactly, it would need adding there, I'm not sure who decides what gets there... [14:25] Saviq: ah I think I know who :-) [14:25] sil2100, sure [14:25] ogra_: ping ^^ [14:25] Saviq: no idea [14:25] Saviq: we have frameworks? [14:25] * Wellark hides [14:26] Saviq: that would probably be a question for the sdk team [14:29] sil2100, what's the right way to get a trivial change in utopic and rtm? [14:29] seb128: both at once? How trivial the change is? Usually the recommended way is landing it to utopic first and then syncing it up to an RTM silo and releasing [14:30] sil2100, it's adding a one line X-Ubuntu-Use-Langpack: yes to content-hub's debian/control [14:31] seb128: I suppose you could land it for utopic through a silo (since it's a CI Train managed project), and then - if QA agrees, we could sync it to an RTM silo and let it through without QA sign-off [14:32] sil2100, who do I ask if QA agrees? [14:32] sil2100, I need to first land to utopic and then redo a landing for rtm silo? [14:32] seb128: best consult davmor2 or brendand, they're the current QA guards :) [14:32] seb128: if you prepare a landing for utopic, we'll prepare a landing for RTM for you [14:32] I though we had facilities to get it uploaded to rtm as well as utopic? [14:32] sil2100, ok, thanks [14:32] let me do that then [14:33] But you'll have to test it anyway [14:33] barry, Can I get a silo for line 42 please? [14:33] Yeah, CI Train handles all the package copies so it's not much manual work === fginther changed the topic of #ubuntu-ci-eng to: Need a silo? Ping train support: trainguards | Need help with something else? Ping vanguard: fginther | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: robru is on vacation [14:34] sil2100, a simple rebuild did not fix the changelog ... it's in landing-012 if you want to take a look, but I'm going to prepare an explicit changelog MP so do not tarry too long [14:34] sil2100, not sure how to test "launchpad should import translations" before it lands in the launchpad pocket [14:35] sil2100, it's not like I can locally fake launchpad [14:35] I guess all we need is just installing it on the image and checking if content-hub still works, I guess [14:35] bregma: sorry for that ;/ [14:35] bregma: I wonder what's up [14:37] if (lander == "bregma"): [14:38] sil2100, so seems rsalveti says we can just copy over to the archive, dont assign a silo for the given up rtm-002 [14:38] Aaaaa [14:38] AaAAAaaa [14:38] ;) [14:39] But wait! [14:39] Without QA SIGN-OFF?! [14:39] sil2100, thats what i was just told [14:39] well, do we want qa sign off for every minor change? [14:39] and bug fix? [14:39] if so, fine with ppa [14:39] but I thought that minor fixes could go in directly [14:39] rsalveti: what kind of a change it is? [14:39] rsalveti, well, i just waited a whole day for my one char change :P [14:40] sil2100: disabling pulse autospawn, as we have an upstart job for it [14:40] If it's really minor we can let it through, but I prefer when QA guys are aware of the things that get copied over without their attention [14:40] brendand, davmor2: ^ [14:41] and landing via a silo will make then know what landed? [14:41] ted: looks like it was already done for you? [14:41] that's the same thing as looking for the rtm 14.09 changes mailing list [14:42] ogra_: let's just do it via silo 27 then and everybody will be happy it seems [14:43] barry, Hmm, I don't seem to have one for line 42, the rtm one? [14:43] rsalveti: it's not that, I want QA to say if it's ok to let it through without QA sign-off [14:43] rsalveti: usually we consult with them if they're happy to see the change going through without QA sign-off [14:43] sil2100: oh, so we need an ack for every change from QA basically? [14:43] brendand, rsalveti: let me get this straight, You are disabling pulse autostart because there is a pulse respawn upstart job? Why was this never used before? [14:44] davmor2: it was used, this is to fix a race === alan_g_ is now known as alan_g [14:44] that sometimes tp-ofono triggers upstart outside of the upstart job === alan_g is now known as alan_g|tea [14:44] we have a bug number for it, just a sec [14:44] ted, is it safe to override the dupe? https://ci-train.ubuntu.com/job/prepare-silo/2597/console [14:44] davmor2: bug https://bugs.launchpad.net/ubuntu/+source/telepathy-ofono/+bug/1375249 [14:44] Ubuntu bug 1375249 in telepathy-ofono (Ubuntu) "telepathy-ofono tries to open /proc/1728/fd/ and is DENIED access by apparmor" [High,Fix released] [14:45] barry, Yeah, it'll have to wait on rtm5 landing to land, but we can test it before. [14:45] ted: +1 [14:45] barry, Also can I get a silo for line 98 please? [14:45] ted: yep [14:45] barry, It'll conflict with silo 13, but that's okay. === alan_g|tea is now known as alan_g [14:53] Wellark: is the package name for connectivity called qml-module-connectivity? [14:53] Wellark: I can get it added to the ubuntu-seeds which would make it into the images [14:57] nik90, make an MP against lp:~ubuntu-core-dev/ubuntu-seeds/ubuntu-touch.utopic/ [14:57] ogra_: cool will do so [14:57] rsalveti: yeah so it is effectively just turning off the one in favour of the other then right? [14:58] davmor2: we always had the upstart job, and that works most of the times [14:58] davmor2: the problem is when tp-ofono starts before pulse, and that was triggering pulse to auto-start [14:59] davmor2: causing a few apparmor denials [14:59] so I just made the tp-ofono -> pulse calls to not request pulse to auto start [14:59] and also disabled that option in pulse's config [14:59] as we don't want apps starting pulse, because we have an upstart job taking care of that [15:00] rsalveti: right okay, that doesn't sound like a bad thing. and I'm assuming it will become blatantly apparently very quickly if sound doesn't work and then we all point the finger at you ;) [15:00] davmor2: yup :-) [15:01] sil2100: I'd say let it through if it is a redundant system lose it but keep hold of a revert just incase [15:09] nik90: yeah, I think it was [15:09] or ubuntu-connectivity [15:09] can't remember [15:10] Wellark: found it, its qml-module-ubuntu-connectivity [15:11] davmor2: will test latest rtm with https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu-rtm/landing-027 at least :-) [15:12] ogra_: https://code.launchpad.net/~nik90/ubuntu-seeds/add-qml-connectivity/+merge/237442 [15:13] nik90, thx [15:18] sil2100: nope! we've barry now who wakes up 3h earlier than robert did :) [15:18] crazy ! === alan_g_ is now known as alan_g [15:21] thostr_: I think your indicator-network rtm-026 can be forced to build, as the text is "A version (0.5.1+14.10.20140829~rtm-0ubuntu1) is available at the destination archive for that component but is not in the destination branch which is still at 0.5.1+14.10.20141006-0ubuntu1." ie this is not a problem otherwise but the fact that rtm will get possibly multiple newer versions at the same time [15:22] Mirv: just rebuilding using the force flag? [15:22] thostr_: I'm trying that right now, https://ci-train.ubuntu.com/job/ubuntu-rtm-landing-026-1-build/8/console [15:22] yes, I see [15:22] thanks === gatox is now known as gatox_lunch [15:23] thostr_: oh, hmm, aborting. actually, those MP:s shouldn't be relisted there. [15:23] thostr_: so it needs to be that sync:16 only. reconfiguring. [15:24] thostr_: and now it's building without the force. [15:24] or actually, just syncing. [15:24] Mirv: ok [15:25] Mirv: so once i went through we can start testing [15:25] thostr_: yes so once https://ci-train.ubuntu.com/job/ubuntu-rtm-landing-026-1-build/9/console is finished you can start testing [15:25] it waits for the packages to be properly published in the PPA. [15:26] uuuh [15:26] sil2100, that rtm 026 above looks bad [15:26] ugh [15:26] ogra_: sil2100: no, no, that was the _previous_ build run :) [15:26] lol [15:26] k [15:26] phew [15:26] ;) [15:26] and that beautiful error means "someone aborted the build job" [15:27] Ah, yeah, now after Robert's changes that's how it looks [15:30] brendand: hey, any problem with silo9 anymore ? [15:30] i saw irc chat between you and saviq earlier, but i see it listed as still under test [15:31] per ello.com/b/AE3swczu/silo-testing [15:31] kgunn, it needs a bit more testing [15:31] kgunn, ello - isn't that a new social network? [15:32] brendand: sorry meant trello.com/b/AE3swczu/silo-testing [15:32] :) [15:32] brendand: and thanks...we're just getting backed up again...any eta ? [15:33] kgunn, it should be signed off before eod [15:33] camako: ^ [15:33] ack.. that will unblock silo 20 [15:34] passengers: i'll be afk for a bit during lunch [15:38] kgunn, i just checked the test plans, it's fine i covered everything. done [15:38] brendand: \o/ [15:38] kgunn, i'll try and see to it that 20 gets tested next [15:38] camako: you owe brendand man-hugs [15:38] brendand, silo 20 is not ready yet.... [15:39] :-) [15:39] camako, i thought it would be fine once 9 landed? [15:39] brendand, once it lands , we'll have to rebuild qtmir and do some more testing.. [15:40] camako, ah so it needs to be rebuilt [15:40] camako, ok [15:40] kgunn, any other silo you want to prioritise then? [15:40] brendand, yea I flipped the spreadsheet "testing done" column to "No"... [15:40] ogra_, So the adb shell is in its own session now? [15:40] ogra_, Looking at the cgroup paths. [15:40] ted, ?? [15:41] ogra_, $ adb shell cat /proc/self/cgroup [15:41] ted, adbd is calling sudo -u phablet -i bash -ci [15:41] which is treated like a login shell [15:41] brendand: np, silo20 was it... [15:42] ogra_, K, trying to understand why starting apps from adb is broken now. I think that's it. [15:42] ted, (... and before it was just "bash -ci", i dont think anything changed due to teh sudo ) [15:43] (apart from the fact that sudo processes the pam stuff afaik [15:43] ) [15:43] ogra_, Yeah, I think the pam stuff is where logind connects in. [15:43] ogra_, Why didn't you run every test plan possible in the wiki before landing any change? ;-P [15:44] ted, if starting apps from adb would be broken we wouldnt have any AP results since two weeks [15:44] or even three [15:44] ted, i ran all tests from the testing wiki [15:44] like our daily smoke testing does too [15:44] Hmm, wonder how that works then. [15:45] ted, the last actual change that could have affected anything here landed on sept. 11th [15:46] ah, well, i'm lying, 13th actually [15:46] Hmm, okay, perhaps it's just clock. Calculator works. [15:46] ted, are you sure you are up to date with the host side ? [15:46] like ... on the latest adb [15:46] (not that this would change much though) [15:47] ogra_, Seems so. [15:47] ted, also ... do you properly use phablet-test-run ? [15:47] ogra_, I'm just adb shell, then calling ubuntu-app-launch. It works for calculator though. [15:47] (which is what our smoke testing infra does as well) [15:47] ogra_, I figured clock was as simple as it got. [15:48] ted, hmm, i remember elopio saying theer were some issues with AP tests that try to mangle $HOME [15:48] So, now, why doesn't clock work… [15:48] not sure which ones that were though [15:48] Yeah, this isn't the AP tests. Just launching the ap. [15:49] ted, ah, k ... just looking at the last successful rtm run of the clock tests from yesterady http://ci.ubuntu.com/smokeng/utopic/touch_stable/mako/75:20141007:20140929.1/10836/ubuntu_clock_app/ it seems to work fine [15:50] ted, while it doesnt in non-rtm images http://ci.ubuntu.com/smokeng/utopic/touch/mako/272:20141007:20140929.1/10831/ubuntu_clock_app/ [15:51] ted, both have the new adbd since a while ... [15:51] so i'm not sure if/how adbd would be to blame [15:52] Hmm, yeah. We won't be able to get the process list from adb, but I don't know that we need that. [15:53] well, happy to add any changes if you tell me which :) [15:53] * ted queues up his wish list [15:53] if you need logind, we might be able to hook somethin into /etc/profile.d for example [15:54] I think it might be a GSettings issue. [15:54] (which definitely gets processed to get you DBUS and UPSTART_SESSION variables in the shell) [15:54] if there is any additional var we need to set, we can do it there with shell snippets [15:54] But I don't understand how it works normally if that's the case. [15:57] cwayne: hey, so slangasek wanted me to chase up the custom image building disagreement (bug 1367332) [15:57] Error: Launchpad bug 1367332 could not be found [15:57] ted, well, adbd cant actually enter the running UI session ... it can only talk to it via dbus and upstart socket [15:57] ogra_, Sure, and I think that's partially the issue, it's ending up with a different environment. [15:57] ted, i.e. i dont think there was ever logind involvement [15:58] For instance ubuntu-app-triplet isn't working either, seems to not find the click db. [15:58] Which is odd as click list does. [15:58] ted, try with phablet-shell, just for comparison [15:59] cwayne: I think our position here is that our images ought to be buildable in one shot on our primary Launchpad-based infrastructure without having to fetch stuff from Jenkins. Could you point me to the code that builds the custom tarballs so that I can see what's involved here? I might have some suggestions [16:00] ogra_, phablet-shell doesn't work for me as my ssh config is too complex for it. Need to fix that, but haven't yet :-/ [16:00] cwayne: Oh, is it lp:~savilerow-team/savilerow/rtm-custom-tools ? [16:00] cjwatson: yep [16:00] ted, ah, bad [16:00] ted, ssh should be the same though [16:00] Yeah, I think so. [16:01] cwayne: I think the point of disagreement here is fundamentally that Jenkins is anything but standard as far as most of Ubuntu goes :-) [16:01] cjwatson: that's fair [16:01] Oh, so that's actually not a lot of code [16:01] nope, the most complex part is installing it in an armhf pbuilder [16:01] cwayne: So my suggestion was to borg this into livecd-rootfs [16:01] which isn't particularly complex :) [16:02] cjwatson: but isn't that how it already is done? [16:02] cwayne: That's how the rootfs is built ... [16:02] e.g. https://launchpad.net/~ubuntu-cdimage/+livefs/ubuntu/utopic/ubuntu-touch/+build/8424 [16:02] right [16:02] We could spit out a custom.tar.xz as another artifact there, and that way we always get rootfs and the baseline custom image changing in sync [16:03] quick question about packages in unapproved; is there a way to unlock a silo (like rtm 14) to make room for another silo for oSoMon? [16:03] cjwatson: but also important to note that the custom tarball will have more than just those custom apps, it'll also have lool's HERE stuff [16:03] That's proprietary, isn't it? [16:03] yes [16:03] So I'm curious whether the TB has signed off on having proprietary code as part of baseline Ubuntu images [16:03] sil2100, landing-012 has been rebuilt with the bug explicitly added to the changelog and needs a re-publish, if you would be so inclined [16:04] cjwatson: there's proprietary code from android in these images [16:04] Whether it's done in LP or Jenkins doesn't massively change that [16:04] bregma: o/ Let me deal with that [16:04] from multiverse [16:04] Sure, different kind of code though [16:04] Is HERE essentially hardware enablement? [16:04] no [16:04] Hardware enablement has traditionally been regarded differently [16:05] bregma: hm, that's an SRU for trusty, right? [16:06] sil2100, correct [16:06] davmor2: meeting! [16:09] lool, cjwatson, cwayne: I understood from the discussion that we were not planning to put the HERE code in the "default" custom tarball [16:09] slangasek: we are [16:09] slangasek: I mean, it's in right now [16:09] slangasek, there might be different "default" tarballs [16:09] however there are two versions; one for community images [16:09] lool: but it's currently not a default [16:09] and one for production [16:09] I'm talking about the community image [16:10] slangasek: it's in the utopic mako and krillin images [16:10] albeit it's opt-in at runtime [16:11] lool: what images? [16:11] slangasek: I understood it would be.. [16:11] it's not in the rootfs (which has no mako or krillin version) [16:12] slangasek: it's in the custom tarball [16:12] slangasek, its in the custom tarballs [16:12] the "default" custom tarballs :) [16:12] so we're currently providing a custom tarball on the default channel, which contains HERE? [16:12] yes [16:12] yes [16:13] ok; so yeah, it was my understanding that we were not going to do that for the default channel on the community images [16:14] but we currently do [16:15] ogra_, So here's the variables that are different between "adb shell env" and "adb shell initctl list-env" : http://pastebin.ubuntu.com/8515498/ [16:15] ogra_, Can we have adb shell just pull a bunch of those from the upstart session? [16:15] ogra_: yes, I understand that's the status quo; but that's buggy [16:15] ogra_, I think that /usr/share/ubuntu-touch is probably the issue, but the others probably break something else. [16:16] ted, you can put an /etc/profile.d snippet in place that transfers them i guess [16:16] slangasek: in the initial addition, I had suggested we'd use two channels, one with and one without a custom tarball [16:16] I remember a discussion with cwayne around; I'd have to dig into logs to see what pushed towards the current implementations, it might have been due to the fact that rtm + krillin would have a custom tarball by default [16:16] *around that [16:16] https://lists.ubuntu.com/archives/ubuntu-announce/2007-February/000098.html "The Ubuntu project reaffirms that it will not include closed-source components in its default configuration except where allowed by the existing exception for the provision of sufficient hardware support." [16:17] how the mighty have fallen [16:17] ted, not sure how you get access to these values when not inside the session though [16:17] ogra_, How do we get the dbus session address? [16:17] lool: so the problem then is that if we're relying on the custom tarball to ship the core apps, we can't quite do that [16:18] ogra_, The script that grabs it from the runtime dir? [16:18] ted, same way as always, by reading it from a file (now in /run, before in ~/.cache) [16:19] cjwatson: the original desire and the longer term plan is to ship this kind of software in a click package [16:19] slangasek: what if we did 2 custom tarballs, one built on cdimage with *just* core apps, and one built on jenkins with whatever bits we need [16:20] ted, whatever sets these vars would have to put them somewhere for us ...*if* we cant just attch somehow to the upstart session and pull them at runtime from there [16:21] ogra_, Well if we can't attach to the upstart session starting apps isn't going to happen regardless. [16:21] ogra_, They're hardcoded else where, just avoiding doing that twice. [16:22] ted, how does autopilot achieve that then ? === gatox_lunch is now known as gatox [16:22] we obviously have apps starting during tests [16:22] ogra_, I think they're starting in a different environment, just clock is the only one using a gsettings key that needs that set. [16:22] cwayne: yes, the right answer is to build separate custom tarballs, the default community one without HERE [16:22] and i highly doubt any of these vars you list have ever been available in adb [16:23] cwayne,slangasek: we may have to add a new package with the code involved here; it seems suboptimal to have to copy it around [16:23] cwayne: I'm fine with the HERE one being built in jenkins [16:23] cwayne,slangasek: though I don't know, maybe it doesn't change often enough to be worth caring [16:23] ogra_, Yes, but there's a libual change that moves more environment processing to the adb shell side of things (part of the performance work). [16:23] cjwatson: to share the custom tarball building code, you mean? [16:23] ogra_, And I think that's what's causing the diff. [16:23] ted, aha ! [16:23] * ogra_ finally understands the issue then [16:24] ted, well, then we somehow need to get the vars over the fence [16:25] but i'm not sure that will actualy be enough ... [16:25] i.e. if some service asks back at logind it will get the wrong info [16:26] slangasek: Yeah [16:26] ogra_, Yeah, I'm looking, but I'm curious if we need to move part of the session init script into somewhere more common. [16:26] ogra_, So both the unity session and the adb session just get the same vars. [16:27] right [16:28] ogra_, Heh, the touch session writes ~/.profile [16:28] ogra_, I guess we just need to add to that :-/ [16:29] ted, please talk to jdstrand before though ... [16:29] not sure we want to expose that stuff outside of XDG_RUNTIME_DIR [16:30] (in fact whatever writes to ~/.profile should probably write to a file in XDG_RUNTIME_DIR ... and ~/.profile should get a single line sourcing this file) [16:30] cjwatson: ok. I wasn't sure if this would live in livecd-rootfs, and jenkins could just call it via a different target / configuration? [16:30] Not sure that DESKTOP_SESSION is really privileged. [16:31] ogra_, Yes, definitely. But at the top of the file it admits that it's all a total hack :-) [16:31] trainguards, If I want to leave a branch out (i.e. not land) that is in a silo, do I just delete it from the spreadsheet, and click on 'Reconfigure Silo'? [16:31] top of which file ? [16:31] ~/.profile surely doesnt say that here [16:31] ogra_, /usr/bin/ubuntu-touch-session [16:32] ogra_, It's the one writing to ~/.profile [16:32] ted, well, mterry added that comment ... [16:32] ogra_, And setting up the XDG dirs. [16:32] Heh, mterry is right. [16:32] (a year after the file existed) [16:33] well, it is less hacky than the original one we had :) [16:33] *way* less hacky :) [16:34] i agree its a hack ... but the only way to not have hacks would be to re-write adbd ... to properly hook into pam, logind and to actually use a prtoper tty and termcap :P [16:35] sil2100, how do I request content-hub from utopic to get into rtm as well? [16:35] ogra_, I actually see it the other way, this script is a replacement for Xsession.d which was better :-) [16:36] seb128: taking care of that! Sorry, been in meetings [16:36] sil2100, thanks [16:36] Okay, lunch time. Going to try to fix after lunch. [16:36] ted, other way round ... UAL and lighdm are replacements for the initial MWC quantal phone session script :) [16:37] ted, let me know if you need me for anythin [16:37] g+ [16:38] seb128: do you know if your landing got a silo already? Since the spreadsheet looks a bit strange [16:38] slangasek: That's possible too if it's a separate script there; that said I don't know if PES require more agility than that [16:38] Ah [16:39] seb128: ok, I see [16:39] seb128: nvm! [16:39] cjwatson: "agility" in what sense? [16:39] sil2100, silo 25 [16:39] seb128: yeah, I saw that, but I saw it's migrating already so somehow felt it's not the right thing [16:39] slangasek: How frequently they need to make code changes in order to do things for specific customer images [16:39] sil2100, why not? [16:40] seb128: brain blockage, no idea why [16:40] seb128: anyway: https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu-rtm/landing-002 [16:40] sil2100, k ;-) [16:40] sil2100, thanks [16:40] cjwatson, slangasek the custom tarballs requires QA signoff in any case ... so agility is definitely ruled by QA person availablility anyway [16:40] seb128: here's your ubuntu-rtm silo with the binaries copied over by CI Train [16:40] sil2100, ok, so what do I need to do next to have that approved? [16:40] no matter how many code changes you make at what frequency [16:41] ogra_: I meant ones other than the default channels [16:41] seb128: you have an ubuntu-rtm phone? If yes, just install and make sure it works on it and set to 'Tested: Yes' and I'll publish it [16:42] well, handling that is still up in the air afaik [16:42] * ogra_ guesses you can make up whatever you want here atm [16:42] sil2100, thanks [16:42] with later adjustments [16:42] sil2100, (yes I've a krillin) [16:42] Excellent [16:45] sil2100: could you reconfigure line 27 for me please? still not working when I do it [16:45] too many MR's [16:46] bfiller: ok, sure [16:47] sil2100, If I want to leave a branch that is in a silo out (i.e. not land), do I just delete it from the spreadsheet, and click on 'Reconfigure Silo'? [16:47] camako: yes, but not sure if you won't need a trainguard to do the reconfigure for you [16:47] camako: try removing it and pressing reconfigure, if it doesn't work just ping me [16:48] sil2100, ok I have a build of another package going (not the one I want to remove), do I need to wait until the build is done? [16:48] ogra_: will the default ones with just core-apps require qa signoff? [16:49] camako: hm, is it building in the PPA already? [16:49] sil2100, yes... it's ok I'll just wait .. almost done.. [16:50] camako: if it has been uploaded to the PPA already, you can abort the build job, reconfigure and rebuild [16:50] sil2100: also I need silos for line 101 and 102 when you have a chance [16:52] bfiller: ok, only now the reconfigure passed [16:53] sil2100: thank you [16:56] cwayne, i ahve no idea [16:56] cwayne, i guess thats a QA decision [16:59] cwayne, but i imagine they would want some testing "in context" [17:00] sil2100, ping [17:01] tvoss: pong? [17:02] "advantage tvoss " [17:02] 15:0 [17:02] ted, we have a few of your silos on the backlog, do you have an order you want them done in? [17:03] ted, currently silo 005 is top: https://trello.com/b/AE3swczu [17:04] ralsina_, we see silo 14, you can keep track of it here: https://trello.com/b/AE3swczu [17:04] thanks brendand! [17:04] ralsina_, currently we have a bit of a backlog but it should get picked up tomorrow [17:05] brendand: one note: because the id of the apps doing the polling has changed, it's easier to test this starting with a clean device [17:05] sil2100, do you have 10 minutes to talk about something [17:05] brendand: that' s perfectly fine, no rush [17:05] ralsina_, don't worry we always do [17:05] cool then :-) === alan_g is now known as alan_g|EOD [17:11] awe: ^ [17:11] tell them not too if you don't want it to land yet [17:15] trainguards, please don't land ubuntu/landing-011 or ubuntu-rtm/landing-024 ( urfkill, indicator ); some re-testing due to last minute update to urfkill, and indicator-network requires re-test [17:15] I'm adding a comment to the spreadsheet, and will ping trainguards when all is set to publish [17:16] sil2100: ok, it looks like the unity8 unlock script is failing quite badly now [17:16] sil2100: it tells me that it's unlocked, but it's clearly not [17:17] plars: hm, do you know if the 'looping fix' from mterry landed already? [17:17] awe: ok, thanks! barry ^ [17:17] sil2100: still looking into it, and I'm in a meeting now, but this could be the source of the strange system-settings or dropping letters problems [17:17] sil2100, I tested content-hub from the rtm ppa on the current rtm device, works fine [17:17] seb128: ok, publishing [17:17] sil2100: I think it's been there for a few days now on utopic right? [17:17] sil2100, thanks [17:18] brendand: sadly, busy right now :< Can we connect tomorrow? [17:18] sil2100, ok [17:19] sil2100, added a comment to line 6; also applies to rtm silo as well [17:19] thanks! [17:25] sil2100, awe ack [17:25] ah, right forgot you were us-lander of the week [17:25] ;D [17:26] yeah ;) [17:41] sil2100, the reconfig seemed to work but building a package afterwards I got the following : [17:41] https://ci-train.ubuntu.com/job/ubuntu-rtm-landing-020-1-build/47/console [17:41] at the very end, abt signatures === Ursinha changed the topic of #ubuntu-ci-eng to: Need a silo? Ping train support: trainguards | Need help with something else? Ping vanguard: Ursinha | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: robru is on vacation [17:45] ogra_, I pushed a developer mode session for the sprint [17:45] balloons, awesome, thanks !! [17:46] I feel better about what we can do for RTM, and then we can step back and do it better for 2.0 [17:46] ogra_, thank you :-) [17:46] :) [17:49] sil2100: ok, I don't know that it's necessarily caused by that unlock script change that mterry landed [17:49] sil2100: more like it's not a complete fix [17:49] sil2100: there's still a race [17:49] plars, more racing! [17:49] plars, we retry 5 times, waiting 20s each time [17:49] mterry: ah, you're here now! great! [17:49] plars: oh, mterry what did you do! [17:49] plars, that's a crazy race [17:49] sil2100: I'm saying he didn't break it, not that he did :) [17:49] davmor2: so, any news on the bug? [17:50] Ah, ok ;) So you mean the race might be somewhere else? [17:50] mterry: so it seems there's a narrow window just before the greeter comes up, where running that dbus call succeeds, but the greeter isn't all the way up [17:50] mterry: and when it comes up fully, it's still locked [17:50] plars, oh that's annoying [17:51] mterry: I was able to reproduce in a really ugly way by hand, just doing adb reboot && adb wait-for-device && adb shell, then running 'gdbus call --session --dest com.canonical.UnityGreeter --object-path / --method com.canonical.UnityGreeter.HideGreeter; echo $?' for a bit until I got 0 [17:51] mterry: but at that point, the greeter came up fully, and was still locked [17:51] plars, yeah I could imagine that happening... === Ursinha is now known as Ursinha-afk [17:52] plars, will have to queue up the signals or delay starting the dbus service I guess [17:52] plars, is it at least better than it was before? [17:52] mterry: well, in ways it feels worse, because we had some very unusual test failures today [17:53] mterry: everything looked normal, but the tests seemed to be getting stuck in a loop waiting for things to show up that never did [17:53] mterry: when I reproduced by hand, I discovered the test was running but the screen wasn't unlocked [17:53] plars, worse... :( [17:54] sil2100, the reconfig seemed to work... but building a package afterwards, I am getting errors abt key not being certified and location-service (the one I deleted) being in the ppa... Can you please help? [17:57] plars, is there a new bug or shall I re-use the last one? [17:58] mterry: I think the last one still describes it pretty well as "racy" [17:58] mterry: if you prefer though, I can start a new one [17:59] plars, no it's fine [18:03] camako: which silo was that? Let me look at it in a moment [18:03] sil2100, rtm 20 [18:04] sil2100, thanks... you can see the error in the dashboard also === Ursinha-afk is now known as Ursinha [18:16] can someone assign me a silo for line 98? [18:16] * jdstrand added it to the spreadsheet a while ago, but it hasn't popped up yet [18:47] jdstrand, So do you have any issue with us putting the XDG variables in ~/.profile? [18:47] jdstrand, Seems like the sanest solution :-/ [18:48] There's a bunch of different cases here for config. [18:49] what are the official instructions how to enable a silo? [18:49] brendand-nexus5: ? [18:49] brendand: ? [18:49] ted: no... not per se. I'm curious what this is solving [18:50] jdstrand, adb shell getting the proper XDG paths. We add a few in the ubuntu-touch-session including things like the custom one. [18:50] asac: you should use the citrain tool [18:50] asac: from your laptop: citrain device-upgrade [18:50] happy to , but how? [18:50] jdstrand, But now that adb is more independent (a good thing) it's not coming from the same environment tree. [18:50] asac, citrain tool, but i'm not sure if that's actually written down somewhere [18:50] err, that's from memory, but I'm sure it's reasonably accurate [18:50] i dont know where that tool lives [18:50] thomi, that's correct [18:50] i dont have it installed [18:51] asac, phablet-tools-citrain [18:51] asac: the package is phablet-tools-citrain [18:51] ok, I'll be quiet now and let brendand do the tralking :D [18:51] *talking [18:51] asac, if you're on utopic it's in the archive [18:52] ted: so the idea is to put them in ~/.profile so that bash picks it up when adb is started? [18:52] jdstrand, Correct, we do that already with GRID_UNIT_PIX, QTWEBKIT_DPR and NATIVE_ORIENTATION. [18:52] jdstrand, Adding XDG_DATA_DIRS is the plan. [18:53] plars, I'm having reliable success with: adb reboot && adb wait-for-device && adb shell 'while ! gdbus call --session --dest com.canonical.UnityGreeter --object-path / --method com.canonical.UnityGreeter.HideGreeter 2>/dev/null 1>&2; do echo -n .; done' && echo -e '\nDid it work?' [18:53] interesting [18:53] plars, let me confirm I'm on latest code [18:53] ted: curious why not use /etc/profile.d/ ? [18:53] jdstrand, Seems like all of those should go to somewhere more generic, but don't want to build a system for this fix. [18:53] jdstrand, I don't think ubuntu-touch-session runs profile.d [18:53] Though, I guess it could run specific scripts out of there. [18:55] * asac runs citrain device-upgrade and watches phone reboot [18:55] ted: I'm not saying you have to. I personally don't have a problem with it, but I have a sneaking suspicion someone else might [18:55] mterry: right, so the dbus call will exit with a return code of 0, but did it really unlock? you can't trust the return code if you hit the race condition [18:56] mterry: I was on utopic latest on mako when I tested it [18:56] jdstrand, Well, ogra_ only said I had to ask you :-) [18:56] plars, that's what my script tries to test -- it loops until it gets a 0 and then asks if it actually worked -- it always did in my cases [18:57] plars, I'm not on latest utopic code, will try again [18:57] ted: I can't come up with a compelling reason not to [18:58] mterry: true, I wonder if something landed that opened a window for this to happen in [19:06] fginther: beat you to it after all :) https://code.launchpad.net/~cwayne18/phablet-tools/clickbuddy-with-sessions/+merge/237477 [19:06] wait wtf that diffs wrong as hell [19:07] brendand: got another bug fix to land, can you tell me if you're ok with calling this an isolated bug fix? https://code.launchpad.net/~jonas-drange/ubuntu-system-settings/cellular-slot2-online-status-not-persisting-1375832/+merge/237070 [19:08] plars, yup, that little script reproduces the problem with latest unity8 code. It doesn't seem due to my previous unlock-device script change, just coincidence. Will have to investigate why something changed [19:08] will add a comment to bug [19:11] rsalveti, Could you look at this MR? https://code.launchpad.net/~ted/ubuntu-touch-session/xdg-to-profiles/+merge/237476 [19:11] barry: hi! can I have an rtm silo for apparmor, line 98? [19:11] rsalveti, Would like to fix the clock tests. [19:11] jdstrand: sure! [19:12] ted: sure [19:12] \o/ [19:12] rsalveti, Thanks! [19:13] sil2100, I know you're busy.. I hope you haven't forgotten about me :-)... I 'm done testing the silo and all I need is proper reconfig before I can send it to QA signoff [19:13] mterry: thanks [19:17] camako: yeah, sorry, a lot on my head - let me look at it now [19:23] ted: do we have any specific bug opened for this MR? [19:23] just so we can add and get them closed automatically [19:23] rsalveti, Not that I know of. [19:23] ted: alright [19:24] rsalveti, I think that people probably saw test failures they couldn't figure out, and this is the cause, but I'm not sure of the bugs. [19:24] ted: yeah, makes sense [19:25] ted: happroved [19:25] rsalveti, Let's fix it and pretend they're crazy if they bring it up :-) [19:25] rsalveti, Cool, thanks. Do we need a silo for that? Or how is touch-session done? [19:25] ted: yeah, via silo [19:25] K, I'll get that one. [19:26] kenvandine, yeah sounds fine [19:26] brendand: thx! [19:28] barry, Can I get a silo for line 105 please? [19:30] ted: yeppers [19:30] barry, And 99 please [19:31] ted: please ack the dupe: https://ci-train.ubuntu.com/job/prepare-silo/2606/console [19:33] camako: should be ok now [19:33] sil2100, thanks [19:33] barry, Yes that's fine. [19:33] * ted got confused and was looking at the rtm silos [19:35] barry, Thanks! [19:43] davmor2: any other blockers found? [19:43] davmor2: how does the situation look like? [19:44] sil2100: so alarm went off 2 minutes late if the device isn't connected to the usb lead by the look of it [19:44] can I get a silo for line 90? [19:44] oh? [19:44] I remember seeing it off anyway before [19:45] * sil2100 remembers ogra_ having harps late into the meeting [19:45] sil2100: indeed and not the end of the world as it goes off [19:52] sil2100: https://docs.google.com/a/canonical.com/spreadsheets/d/1Mw46QHRVqmaf_NSL2L-jNcEYNlQzbcsJyaXvhjSgSSY/edit#gid=478828577 make of it what you will :) [19:52] ;) [19:54] sil2100: photos taken in camera still are not right in the scope but show correctly in backdrops now, alarm didn't go off at the right time but did go off, issues with the new media controller in the sound indicator and my issue with 3g === fginther changed the topic of #ubuntu-ci-eng to: Need a silo? Ping train support: trainguards | Need help with something else? Ping vanguard: fginther | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: robru is on vacation [19:58] sil2100: so number 1 is a partial fix but not perfect so we can whitelist that and the other three I think are already in last promoted so are safe-ish [20:00] Ok, I checked the document as well and I think we'll risk it today [20:00] ogra_: you still there? [20:00] sil2100: I concur jibel one image promotion coming your way \o/ [20:06] ogra_: oh come on we all know you are still online really you must be your nick is here ;) [20:19] ogra_: anyway, once you're around give us a sign! We need an image to promote! [20:19] I mean [20:19] We need to promote an image! [20:20] (order of words sometimes is important) [20:20] camako, silo 20 is currently 9th in our queue, we see it but the ETA is likely to be no earlier than EOD tomorrow at this point [20:21] camako, and that's being ambitious. i'll update you tomorrow, but in the meantime have a look at our trello board: https://trello.com/b/AE3swczu [20:21] barry, Can I get a silo for line 106 please? [20:21] camako, if you need to escalate it just let me know [20:21] barry, And can you publish silo ubuntu/19 ? [20:22] bfiller, silo 3 is 10th in our queue, so it might not get seen to until EOD tomorrow or even into thursday. if this is a problem let me know [20:22] ted, barry: handling that [20:22] bfiller, i'm sure you're aware of the trello board already? https://trello.com/b/AE3swczu [20:23] elopio: hi! [20:24] sil2100: thanks [20:24] brendand, ok I'm good with that [20:24] sil2100, Thanks! === Ursinha is now known as Ursinha-afk [20:58] ToyKeeper, rtm/9 should fix the clock issue if you want to look at that at the same time. [20:59] ted: Awesome. I'm going to try to get things in oldest-first order for now though. :) [21:01] ted: Any ideas how to fix the indicator-messages test plan? It doesn't really test anything other than that it's visible. [21:01] ted: Or, is that all covered in AP testing? [21:03] ToyKeeper, Well it needs love, the problem is too many projects and not enough devs, trying to add tests when we implement features. [21:03] Fair enough. [21:04] ted, preach it brother! [21:05] kenvandine, Heh, you're only like 1:5 ;-) [21:05] :) === Ursinha_ is now known as Ursinha [21:06] rsalveti: hey! :) [21:06] sil2100: hey [21:08] rsalveti: sorry to bother again, but it seems that poor ogra_ is not around anymore - do you have all the tools and power to promote images to the ubuntu-rtm/14.09 channel? [21:08] sil2100: haha, never did that [21:09] might be able to try, if you're fine with the risks [21:10] sil2100: http://paste.ubuntu.com/8517194/ [21:10] that's probably how it was done [21:11] sil2100: hello [21:16] rsalveti: hah, most probably! [21:17] rsalveti: if you're sure that's how it is then we would appreciate promoting the latest krillin image (#88), mako and maybe emulator as well [21:17] rsalveti: from what I know ogra_ was rewriting the image number always [21:17] rsalveti: so those should be image #4 I guess [21:17] ogra_: ^ [21:17] elopio: let me get back to you in a moment! [21:17] sil2100: problem is that I'm not really sure [21:18] sil2100: guess it might just be better to wait for ogra_ [21:18] rsalveti: ok then, ogra_ will pop up sooner or later I guess [21:19] rsalveti: anyway, thanks for looking into that [21:19] no worries, sorry for not being able to help [21:23] elopio: ok, so two things! [21:25] elopio: first thing: any news on the issue with gallery and calculator AP test failures on utopic? [21:32] sil2100: I haven't had time to patch them again. I'm doing manual testing. [21:32] the calculator has a clear bug reported. The gallery is more confusing, but probably it's the same cause of the tmp HOME. [21:39] sil2100, oh, just saw the mail [21:48] rsalveti, next time just see /home/ogra/rtm-promo.log ... i just run the commands in there as cdimage user on nusakan ... (no bzr or anything involved) [21:48] ogra_: alright, noted [21:53] ogra_: \o/ :) [21:53] sil2100, i assume someone boot checked the emulator, right ? [21:53] elopio: ok, then the last thing (although not really that important due to a lot of other stuff): [21:54] elopio: if by some accident you find some time today, coould you take a look at the autopilot infinite-hang problem in utopic images we had today? plars has some details there [21:54] ogra_: erm [21:54] sil2100: is that while running toolkit tests? [21:55] ogra_: is it possible to also 'promote' 14.09.es-proposed to 14.09.es [21:55] elopio: it seems random, once I heard it happened in ubuntu-system-settings tests, but I think plars said something about dropping letters [21:55] cwayne, which image and did QA sign that off ? [21:56] sil2100: plars was the ui hung? [21:56] ogra_: ok, so maybe let's wait with promoting the emulator [21:56] sil2100, hrm [21:56] robotfuel: it's hard to say, since we only saw it on the logs... [21:56] sil2100, it is in the middle of processing [21:56] ogra_: ...then nevermind ;) [21:56] not sure what happens if i ctrl-C that [21:57] ogra_: whichever corresponds to what we're promoting elsewhere, and no, it's the same exact custom tarball as krillin rtm, but sets the default language in the wizard to spanish [21:57] sil2100: unlikely to have time, because I have to scroll down to see all the silos in the queue. But tomorrow. [21:57] ogra_: it's just that we didn't check if the emulator boots, but I suppose *someone* would tell us if it was different [21:57] cwayne, well, i just promoted krillin 88 ... the es channel is far from that number :) [21:58] sil2100, sure, but we should really check that before next time [21:58] elopio: plars it might be this bug https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1377332 [21:59] Ubuntu bug 1377332 in unity8 (Ubuntu) "UI randomly freezes" [Undecided,Confirmed] [21:59] robotfuel: yes, that's likely to cause an infinite loop when dragging. [22:00] ogra_: ha, fair point [22:00] cwayne, i assume you want 11 [22:00] ogra_: yea, that's latest? [22:01] http://system-image.ubuntu.com/ubuntu-touch/ubuntu-rtm/14.09.es-proposed/krillin/ [22:01] looks like to me [22:02] cwayne, you got image #1 in 14.09.es [22:02] ogra_: thanks man [22:03] robotfuel, elopio: ok, this sounds like a possible suspect then [22:03] === Image 88 Promoted === [22:04] * popey blinks [22:04] (that is: krillin 88, mako 75 and both emulatorrs 71) [22:05] and introducing for the first time image ... krillin 11 as the famous 14.09.es #1 ! [22:05] lol [22:05] * ogra_ vanishes back into the night [22:05] :) [22:06] err [22:06] that meant indeed [22:06] === Image 4 Promoted === [22:06] sorry :P [22:07] \o/ [22:08] Damn, ok, I think this is the right time to EOD [22:08] o/ [22:24] robotfuel: elopio: we think we found another race with the unity8 screen unlock [22:24] that could be causing the failures we saw earlier === fginther changed the topic of #ubuntu-ci-eng to: Need a silo? Ping train support: trainguards | Need help with something else? Ping vanguard: cihelp | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: robru is on vacation