| boiko | robru: if I am not around by the time silos 31 and 6 are published, would you mind triggering a rebuild on silo 8? | 00:04 |
|---|---|---|
| boiko | robru: assuming you will still be there, of course :) | 00:04 |
| robru | boiko: I'm here for a couple more hours | 00:04 |
| boiko | robru: nice! I am going for dinner and then supermarket, I might check the status once I am back | 00:05 |
| robru | boiko: although i don't see any movement on proposed migration, not much i can do about that. | 00:05 |
| boiko | robru: that's fine, I guess I can continue with all of that tomorrow morning, thanks for the help anyway | 00:06 |
| boiko | see you! | 00:06 |
| robru | slangasek: are you able to check on proposed migration? Seems excuses hasn't updated in 7ish hours. | 00:06 |
| slangasek | robru: let's see what I can see | 00:07 |
| slangasek | robru: while I'm looking at this, do you have any insight into the question I asked kenvandine above, re: two binaries with the same version number but different builds? | 00:08 |
| robru | slangasek: first I've heard of it... Want me to dig in? | 00:09 |
| slangasek | robru: yes please - we really don't want packages being published to the stable overlay ppa that have the same version number as packages in the main archive if they're not identical packages | 00:10 |
| robru | slangasek: indeed they both look like manual uploads by kenvandine, nothing train-created. He must have just goofed on the version number. | 00:13 |
| slangasek | robru: ok. Should the train have its own guard against that? | 00:13 |
| robru | slangasek: not sure how that would work... you mean a publish-time check to make sure that the version number doesn't already exist in a different series? there's already checks to make sure that the same version doesn't already exist within the same series. | 00:14 |
| slangasek | robru: yes | 00:14 |
| robru | slangasek: is is possible to query lplib for a version number without specifying the series? seems like it would be tricky to check all possible series. | 00:15 |
| slangasek | robru: if you're already checking that it doesn't exist within the same series, it makes sense IMHO to check for that version in any series | 00:15 |
| slangasek | robru: hum I'm sure it's possible by looking at the publishing history, but I don't know offhand where in the API that lives. bdmurray might know | 00:15 |
| slangasek | robru: also note that you'd have to check both archives (stable-phone-overlay + Ubuntu) for a conflict in either direction, since LP itself doesn't enforce this | 00:16 |
| jamesh_ | robru: I only marked the silo as being tested. I don't have permission to upload to the archive directly | 00:16 |
| === jamesh_ is now known as jamesh | ||
| jamesh | I'm not sure what would have happened | 00:17 |
| robru | jamesh: something really goofed up in your silo then, because the package is already in wily when I tried to publish it today | 00:17 |
| slangasek | robru: the command that's causing p-m to fail is: http_proxy=http://squid.internal:3128/ wget --mirror -nH --cut-dirs=3 -np -R *index.html* -P /home/ubuntu-archive/mirror/phone-overlay http://ppa.launchpad.net/ci-train-ppa-service/stable-phone-overlay/ubuntu/dists/ | 00:17 |
| robru | oh there's a note from mirv I just noticed... | 00:18 |
| slangasek | don't know yet why | 00:18 |
| robru | slangasek: weird | 00:18 |
| bdmurray | robru: yes it is - https://api.launchpad.net/devel/ubuntu/+archive/primary/?ws.op=getPublishedBinaries&binary_name=systemd-shim&version=9-1bzr3&exact_match=true | 00:19 |
| jamesh | robru: there is a comment left on the spreadsheet row that looks like it explains it: "Mirv/20150608: wily released, vivid put back on hold. The line was erronously published before-hand since QA sign-off field was empty and the monkey that looked at the dashboard read just "wily" while it had "wily vivid"." | 00:19 |
| robru | jamesh: k, I'm muddling through a fix for this | 00:20 |
| robru | bdmurray: thanks | 00:20 |
| robru | bdmurray: that only shows wily though, I guess not vivid because it's just a ppa. hrm | 00:21 |
| jamesh | robru: I had left the QA sign-off field blank because that's what the docs said it should usually be set to. | 00:21 |
| robru | jamesh: well vivid needs qa, but qa approved it so I'm trying to publish now and everything's crazy | 00:21 |
| kenvandine | slangasek, ugh | 00:23 |
| slangasek | robru: the error is that http://ppa.launchpad.net/ci-train-ppa-service/stable-phone-overlay/ubuntu/dists/devel/main/i18n/ is a 404 despite being listed on http://ppa.launchpad.net/ci-train-ppa-service/stable-phone-overlay/ubuntu/dists/devel/main/, I think this might be related to the workarounds being implemented for translation support | 00:23 |
| kenvandine | i thought we did a sync with a rebuild or something | 00:24 |
| slangasek | wgrant: could http://ppa.launchpad.net/ci-train-ppa-service/stable-phone-overlay/ubuntu/dists/*/main/i18n/ please not 404? :) | 00:24 |
| robru | kenvandine: those were both manual uploads as far as I can see... nothing synced, you just uploaded it twice with different series. | 00:24 |
| slangasek | kenvandine, robru: is it possible this was a sync, but at a time when the "sync" behavior in the train was buggy? | 00:25 |
| kenvandine | robru, hmm, i know at least at one point we had it on the spreadsheet as a sync | 00:25 |
| slangasek | publication was 5/22 | 00:25 |
| kenvandine | but don't recall exactly what ended up happening | 00:25 |
| kenvandine | there was a long delay with qa for the overlay silo | 00:26 |
| kenvandine | which was where it went originally | 00:26 |
| robru | slangasek: I'll double check the uploader. | 00:26 |
| kenvandine | several days i think | 00:26 |
| kenvandine | it's possible that i manually screwed it up too... since there was such a delay, like we changed our mind about where to land it first or something | 00:27 |
| wgrant | slangasek: Hm, I removed that rewrite months ago, I thought. | 00:27 |
| kenvandine | i really can't recall, it was all stressful because it was a blocking bug | 00:27 |
| wgrant | slangasek: It was there to minimise IO contention due to apt being moronic. | 00:27 |
| wgrant | But let me see. | 00:27 |
| robru | kenvandine: sorry, looks like I was mistaken | 00:27 |
| robru | slangasek: kenvandine: https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-035/+packages?field.name_filter=systemd&field.status_filter=&field.series_filter= shows here ken uploaded the vivid version | 00:27 |
| slangasek | wgrant: ok. it definitely seems to be causing snakefruit's mirror script to error out now | 00:27 |
| robru | slangasek: kenvandine: then train uploaded wily version: https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-001/+packages?field.name_filter=systemd&field.status_filter=&field.series_filter= | 00:27 |
| kenvandine | yeah, i know the vivid upload was first | 00:27 |
| slangasek | wgrant: I don't specifically need it to not 404; it's also acceptable for it to not be shown in the index | 00:28 |
| kenvandine | but it didn't get published right away, and we did a sync with rebuild | 00:28 |
| kenvandine | to wily | 00:28 |
| slangasek | wgrant: but the current behavior, showing in the index but 404, makes wget cranky | 00:28 |
| robru | kenvandine: I'm not sure why the sync code didn't mangle the version though. it's supposed to do that & has tests for that as well | 00:28 |
| wgrant | slangasek: snakefruit doesn't actually use wget to mirror archives, does it... | 00:28 |
| slangasek | wgrant: it uses it to mirror the indices | 00:29 |
| slangasek | only for p-m | 00:29 |
| kenvandine | yeah, well it might be because this is a manual package, not one where there was a MP | 00:29 |
| wgrant | wat | 00:29 |
| kenvandine | so maybe that only happens when there's a branch involved? | 00:29 |
| kenvandine | robru, just throwing darts here :) | 00:29 |
| slangasek | wgrant: hey man, the launchpad team wrote this code | 00:29 |
| robru | kenvandine: no, the sync logic is a totally different code path from the branch logic. | 00:29 |
| kenvandine | ok | 00:30 |
| robru | slangasek: kenvandine: https://ci-train.ubuntu.com/job/ubuntu-landing-001-1-build/182/console here's the build log that created the sync | 00:30 |
| kenvandine | must be case you aren't testing then :) | 00:30 |
| wgrant | slangasek: Heh | 00:30 |
| slangasek | wgrant: oh. Also, it was written /today/, so it never worked :-P | 00:30 |
| slangasek | so... yeah. Disabling for now | 00:30 |
| wgrant | slangasek: Huh? | 00:31 |
| slangasek | wgrant: the commit that introduced this stable-phone-overlay mirroring code was introduced by cjwatson at Mon 2015-06-08 17:45:53 +0100 | 00:32 |
| wgrant | Oh | 00:32 |
| slangasek | which is suspiciously like the time p-m stopped working ;) | 00:32 |
| robru | slangasek: kenvandine: oh I see what's going on here. the sync code only mangles versions when syncing ubuntu->ubuntu-rtm. That code path wasn't updated when sil updated the sync code to sync within ubuntu. | 00:33 |
| kenvandine | ah | 00:33 |
| slangasek | mmk | 00:33 |
| kenvandine | that's not good | 00:33 |
| slangasek | robru: so it's a train bug - can you open a bug report to track this? | 00:33 |
| kenvandine | is systemd-shim the only package? | 00:33 |
| robru | slangasek: sure. what rule do we want to use for version mangling intra-ubuntu syncs? | 00:34 |
| slangasek | robru: should it be any different? I don't remember what rule was used for the overlay | 00:36 |
| robru | slangasek: would be easy enough to append ~vivid for a backwards sync, but that's probably not what we want for a vivid->wily sync wince foo~wily would be less than foo. | 00:36 |
| slangasek | robru: well also the version numbers should round-trip correctly | 00:37 |
| robru | slangasek: the problem is that manual packages like this don't have the usual train datestamp that mps use. | 00:37 |
| slangasek | ah | 00:37 |
| slangasek | in that case maybe we want to disallow syncs and make the uploader sort it out manually | 00:37 |
| robru | slangasek: I'm not sure how we'd detect "that case" programmatically though. in the sync logic it just downloads the source package, rebuilds and reuploads. it doesn't really "know" if the version number is one of ours or not | 00:39 |
| slangasek | robru: it certainly should be able to detect if the version number is ours or not | 00:39 |
| slangasek | robru: because if it's a train-generated version number, we should be able to do a round-trip transformation on it | 00:39 |
| slangasek | if that doesn't work, it wasn't ours | 00:39 |
| robru | slangasek: but the thing is that we support syncing packages that aren't train-managed packages. | 00:40 |
| slangasek | robru: why do we do that? | 00:40 |
| robru | slangasek: I dunno, there were lots of cases where random things needed to be synced from ubuntu to ubuntu-rtm that weren't necessarily train managed packages | 00:40 |
| wgrant | slangasek: Oh, the files are accessible, just not the indices. | 00:40 |
| wgrant | indices == directory listings | 00:41 |
| slangasek | robru: we should only support this if we can do it right | 00:41 |
| slangasek | wgrant: right, ok | 00:41 |
| robru | slangasek: the current version mangling code for syncs is very basic, it literally just inserts '~rtm' before the last hyphen if syncing to rtm, and rips that out if syncing to ubuntu. | 00:42 |
| robru | it doesn't consider the structure of the version string beyond that. | 00:42 |
| slangasek | right | 00:42 |
| robru | slangasek: I think the easiest way to keep the current feature set and support roundtripping would be to just s/~.*-/~series-/ and that way the right version tag would always be there. | 00:43 |
| slangasek | that's not going to DTRT, so I think we should disallow such syncs and take our time to think about how we can do it right | 00:43 |
| robru | ok | 00:43 |
| slangasek | robru: that doesn't round-trip in the case that the original package wasn't train-managed. | 00:43 |
| robru | true | 00:44 |
| wgrant | slangasek: Is it still interesting to you? Or can you use something sensible like debmirror? | 00:44 |
| slangasek | wgrant: what interests me at the moment is having p-m back online, so I've commented out the mirroring. I don't know what this particular change was for, it seems to be a WIP by cjwatson; if I can figure out a quick fix to the wget commandline that doesn't require me to think about it I will, otherwise I'll leave it for him | 00:47 |
| wgrant | slangasek: We can probably tighten the apache rule a bit without too much danger, but I'd prefer not to if the solution is just "don't use wget -r"... | 00:49 |
| slangasek | wgrant: yes, I don't think you need to make a serverside change for this | 00:49 |
| robru | slangasek: https://bugs.launchpad.net/cupstream2distro/+bug/1463218 | 00:50 |
| ubot5 | Launchpad bug 1463218 in CI Train [cu2d] "sync logic doesn't mangle versions correctly when syncing within ubuntu" [High,Triaged] | 00:50 |
| slangasek | robru: thanks | 00:50 |
| robru | slangasek: your'e welcome | 00:50 |
| robru | kenvandine: slangasek: anyway to fix the problem at hand, wouldn't a manual upload with a bumped version for wily fix it? | 00:52 |
| slangasek | robru: yes | 00:52 |
| robru | slangasek: should I prepare that in a silo? ;-) | 00:53 |
| slangasek | robru: sure, go ahead :) | 00:55 |
| robru | slangasek: k, on it | 00:55 |
| robru | kenvandine: slangasek: k, version bump building: https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-017/+packages | 00:59 |
| robru | kenvandine: when that build is done can you smoke-test it? | 01:00 |
| slangasek | robru: p-m run successfully | 01:01 |
| robru | slangasek: great! | 01:01 |
| kenvandine | robru, sorry, i'm about to sign off here and i'll be out tomorrow :/ | 01:01 |
| robru | bah | 01:02 |
| kenvandine | sorry man! | 01:02 |
| robru | slangasek: ok well version bumped package is building but I have no idea how to verify it doesn't explode. | 01:02 |
| kenvandine | robru, quick test... change the phone from network time to manual and see if it persists | 01:02 |
| kenvandine | or the opposite | 01:03 |
| kenvandine | in system-settings | 01:03 |
| kenvandine | if that works, you're good | 01:03 |
| robru | kenvandine: how do I flash a wily image on my phone? | 01:03 |
| kenvandine | there's a channel for it | 01:03 |
| robru | hmm | 01:03 |
| kenvandine | don't recall off hand | 01:03 |
| slangasek | ubuntu-touch/devel-proposed/ubuntu | 01:03 |
| slangasek | or if you need bq bits, ubuntu-touch/devel-proposed/krillin.en | 01:03 |
| kenvandine | i gotta jet though, gotta hit a store before it closes | 01:03 |
| robru | kenvandine: goodnight | 01:04 |
| kenvandine | you too | 01:04 |
| slangasek | wgrant, cjwatson: found an option to (uglily) make wget skip the unreadable i18n dirs, so we're good now | 01:04 |
| robru | slangasek: ok I'll check that tomorrow morning | 01:04 |
| robru | pah | 01:09 |
| robru | boiko: Hmmmmmmm, some problems with silo 8... It's set for vivid primary, aka an SRU, is that what you want? Not overlay PPA? Also it seems is a sync, it may suffer from a version bug we just discovered today, so pay attention to the resulting version numbers... | 01:34 |
| boiko | robru: well, actually kgunn is the one who needs this silo/sync, they need to land shell rotation | 02:58 |
| boiko | robru: I'll check with him tomorrow | 02:58 |
| Mirv | good morning | 04:46 |
| === chihchun_afk is now known as chihchun | ||
| seb128 | hum, does anyone know what's wrong in https://jenkins.qa.ubuntu.com/job/dialer-app-vivid-i386-ci/89/console | 07:10 |
| seb128 | oh, ignore that, "Depends: qtdeclarative5-ubuntu-addressbook0.1 which is a virtual package." | 07:10 |
| seb128 | I guess it needs silo 008 to land | 07:14 |
| cjwatson | slangasek: thanks. this whole thing was an emergency fix to allow us to have enough history to reconstruct a snapshot of the overlay in case we need to do hotfixes to it, so I'm glad you got it put back together - I was a bit scared when reading scrollback about people disabling it | 07:52 |
| cjwatson | slangasek: (I suspect that it worked well enough for apt-mirror-snapshot to be able to commit the result, and I didn't notice that wget was exiting non-zero) | 07:54 |
| cjwatson | wgrant: I'd probably have used debmirror if it were installed on snakefruit, but this was an OMG-need-to-have-this-working-in-the-next-half-hour fix | 07:55 |
| wgrant | cjwatson: Ahh, I see. | 07:55 |
| sil2100 | cjwatson: is that something related to snapshotting? | 07:57 |
| cjwatson | sil2100: yes | 07:58 |
| * sil2100 is apparently missing some backlog | 07:58 | |
| sil2100 | :) | 07:58 |
| cjwatson | sil2100: it's ok, it was an overnight fix | 07:58 |
| cjwatson | you can ignore | 07:58 |
| seb128 | hey Łukasz | 08:00 |
| seb128 | how is the langpack patching going? need help with that? | 08:01 |
| sil2100 | seb128: hey, just sitting down to it once again, do you know where I can find the list of strings that we wanted changed? | 08:07 |
| seb128 | I can do it if you want | 08:08 |
| seb128 | but otherwise the strings are the ones on | 08:08 |
| seb128 | https://translations.launchpad.net/dialer-app/trunk/+pots/dialer-app/sr/+translate?show=untranslated | 08:08 |
| seb128 | (just picked a random language with 4 untranslated strings, which are the new ones ;-) | 08:09 |
| seb128 | let me ask for a wily export and compare to the langpack | 08:09 |
| seb128 | sil2100, http://paste.ubuntu.com/11666800/ | 08:15 |
| seb128 | so yeah, no string removed in wily | 08:15 |
| sil2100 | Waiting for the es export to arrive on my e-mail | 08:15 |
| seb128 | so it's easy to copy over a wily export of the po | 08:15 |
| seb128 | and upload that | 08:15 |
| sil2100 | Ok, then even better - if you're not too busy then feel free to do the de one, I'll prepare the es | 08:17 |
| seb128 | k | 08:18 |
| seb128 | sil2100, how do I upload? dput directly to the overlay ppa or in a silo or...? | 08:18 |
| sil2100 | seb128: I would say let's dput directly | 08:18 |
| sil2100 | We don't have any QA control over translations anyway... | 08:19 |
| sil2100 | Was dialer the only obvious one missing? | 08:20 |
| seb128 | yes | 08:20 |
| seb128 | but that makes me unsure about the adressbook | 08:20 |
| seb128 | there were changes to import it from dialer/messaging and new ui bits like google import string | 08:21 |
| seb128 | but it might be that was before vivid release | 08:21 |
| seb128 | let me check/ask an export of that one as well | 08:21 |
| sil2100 | I suppose version number 1:15.04+20150608.1 would be fine? | 08:23 |
| seb128 | yes | 08:24 |
| sil2100 | Ok, source package done, I think fr was also on the list of critical languages, right? | 08:25 |
| sil2100 | Ah, but you probably did that one already ;p | 08:25 |
| seb128 | yeah | 08:26 |
| seb128 | k, no new strings in wily addressbook | 08:27 |
| seb128 | so good | 08:27 |
| sil2100 | Ok then, let's upload to the overlay PPA | 08:28 |
| sil2100 | Ok, so it seems we have two issues temporarily resolved | 08:30 |
| sil2100 | seb128: did you upload your langpacks? :) | 08:42 |
| seb128 | sil2100, yes, https://launchpadlibrarian.net/208639682/language-pack-touch-fr_1%3A15.04%2B20150608_1%3A15.04%2B20150608.1.diff.gz | 08:45 |
| seb128 | sil2100, you forgot to do the same for messaging-app in your -es upload | 08:47 |
| sil2100 | Yeah, just noticed that, re-uploading | 08:47 |
| seb128 | -de update, https://launchpadlibrarian.net/208639914/language-pack-touch-de_1%3A15.04%2B20150608_1%3A15.04%2B20150608.1.diff.gz | 08:48 |
| seb128 | sil2100, want me to do the same for the few other languages pat listed in https://bugs.launchpad.net/canonical-devices-system-image/+bug/1460144/comments/1 ? | 08:50 |
| ubot5 | Launchpad bug 1460144 in Canonical System Image "New swipe actions tutorial needs translations" [High,In progress] | 08:50 |
| seb128 | like catalan/italian/portuguese | 08:51 |
| sil2100 | seb128: it's not required, but if you're not super busy right now I suppose we could do that | 08:51 |
| seb128 | k, let me do those | 08:52 |
| sil2100 | To save your time I'll start from the other end and do portugese, just need to wait for LP to export those | 08:52 |
| sil2100 | ;) | 08:52 |
| seb128 | k | 08:52 |
| seb128 | starting with catalan | 08:52 |
| seb128 | to make dpm happy | 08:52 |
| dpm | ;-) | 08:52 |
| jgdx | cihelp: what's up with the ubuntu system settings job? Disabled? | 08:56 |
| jgdx | I can't find any reason as to why in emails, channel hist… :'( | 08:56 |
| psivaa | jgdx: i'm not sure why, it could be to let the other jobs to finish and may be for more slave chroot updates, i've enabled that job and kicked one jobs, for https://code.launchpad.net/~jonas-drange/ubuntu-system-settings/apn-prototype/+merge/258992 | 09:01 |
| jgdx | thanks psivaa. | 09:01 |
| psivaa | jgdx: np | 09:01 |
| seb128 | sil2100, k, done catalan and doing italian, then I think we should have enough | 09:03 |
| sil2100 | seb128: thanks! I finished portugese and uploaded :) I think afterwards we should indeed be good | 09:05 |
| seb128 | sil2100, verified with the fr langpack that it works | 09:07 |
| seb128 | the tutorial is translated on the dialer | 09:07 |
| seb128 | k, they are all built, it/pt still to publish then I guess you can kick an image build | 09:16 |
| jgdx | trainguards: you can clean out silo 13 (row 18) | 09:17 |
| jgdx | (job failed for me) | 09:19 |
| Mirv | jgdx: ok.. so cancel the whole landing? | 09:50 |
| Mirv | well, it's a manual upload so yes just add another line if needed | 09:51 |
| jgdx | Mirv, yep | 09:59 |
| sil2100 | jibel: what was the other bug that we needed for OTA-4? Was that the 3G->2G switching one? | 10:25 |
| davmor2 | sil2100: that exists in rtm too | 10:27 |
| Mirv | sil2100: I didn't publish 015 in the morning since currently I'm not keeping exact track what is supposed to be landed... is there exact list of OTA-4 bugs still? | 10:37 |
| sil2100 | davmor2: oh, ok, since I though BQ wanted that fixed badly | 10:37 |
| sil2100 | Mirv: anything can land, the landing gates are open :) | 10:37 |
| Mirv | sil2100: oh :D | 10:38 |
| sil2100 | We decided to open them up, we have a 'snapshot' PPA to deal with OTA-4 | 10:38 |
| davmor2 | sil2100: well I reproduced it yesterday on vivid 26 and rtm22 | 10:38 |
| Mirv | sil2100: ah I misread your e-mail. OTA4 still has things, but gates open anyway | 10:38 |
| cjwatson | sil2100: FYI https://code.launchpad.net/~cjwatson/launchpad/do-not-expire-phone-overlay/+merge/261438 - not on production yet but we should be able to deploy that before the world ends (or would have ended if you hadn't created the snapshot PPA) | 10:38 |
| sil2100 | cjwatson: \o/ | 10:55 |
| sil2100 | davmor2: ok, then we'll have to discuss with Pat if we're blocking OTA-4 on that or not | 10:56 |
| === MacSlow is now known as MacSlow|lunch | ||
| sil2100 | uh | 11:50 |
| sil2100 | Ok, I jump out now for lunch | 11:50 |
| jibel | sil2100, it was bug 1461593 | 11:58 |
| ubot5 | bug 1461593 in network-manager (Ubuntu RTM) "No data connection switching from 2G only to 3G" [Critical,In progress] https://launchpad.net/bugs/1461593 | 11:58 |
| rvr | popey: I don't understand this "Allow tapping repeat and sound ListItems to trigger action (affects vivid only) where as in the current clock app, one had to explicity click the checkbox button to enable it while now one can enable it by clicking on the listitem itself" | 11:58 |
| nik90 | rvr: I can explain that :) | 12:00 |
| === alan_g is now known as alan_g|lunch | ||
| rvr | nik90: Yes, please :) | 12:00 |
| nik90 | rvr: The Ubuntu SDK changed the default behaviour of listitems when moving from Utopic -> Vivid. | 12:00 |
| nik90 | In utopic, you could click on a listitem to enable a child checkbox.. while in vivid this default behaviour was removed. So one has to explicity press the tiny checkbox in the listitem to enable it | 12:01 |
| nik90 | this changes affects the clock app and so I adjusted the code to ensure that the old behavior is retained | 12:01 |
| nik90 | on in the page where you choose alarm sound, you can click on the listitem to choose that alarm sound rather than having to click on the checkbox | 12:02 |
| nik90 | rvr: does that make it clear? | 12:02 |
| rvr | Reading carefully | 12:02 |
| rvr | ahh | 12:02 |
| rvr | Old behaviour kept | 12:03 |
| nik90 | yup | 12:03 |
| rvr | Cool, because I compared stable and rc-proposed and didn't found any difference :) | 12:03 |
| rvr | Thanks | 12:03 |
| nik90 | yw | 12:04 |
| jibel | nik90, someone was mentioning this change in behaviour to select Wifi access point, do you know if there is a bug report for it? | 12:07 |
| nik90 | jibel: I was told that it is a intentional change in the Ubuntu SDK. | 12:07 |
| jibel | ah ok | 12:07 |
| nik90 | jibel: while testing, it seems some of the indicators may have been fixed to retain the old behavior. | 12:08 |
| popey | nik90: fyi I asked QA (so I guess rvr) to test the multi-fat click. | 12:16 |
| popey | I tested it on my vivid unity8 machine and it dies soon after starting, but I think that's not an app problem but the image is missing something gsettings related | 12:16 |
| popey | other apps die similarly | 12:16 |
| nik90 | popey: ah ok | 12:17 |
| popey | hoping the image seb128 and didrocks are working on fixes these things and we can test that soon | 12:17 |
| seb128 | popey, -didrocks, he moved to other work | 12:18 |
| popey | ah, okay. so it's your baby? | 12:19 |
| popey | super looking forward to testing it (much like everyone else) | 12:19 |
| seb128 | popey, I don't feel like strong ownership, but it landed on my shoulder it seems so trying to move it forward, getting to stand by itself :p I'm sure other are going to come and play around with it once it's working thoguh | 12:24 |
| popey | heh, okay | 12:24 |
| jibel | rvr, silo 3 passed verification? | 12:28 |
| seb128 | what are the landing rules now for vivid? | 12:29 |
| seb128 | it's open for landing, but does it mean any sort of changes or just selected bugfixes still? | 12:30 |
| pmcgowan | seb128, its open, there is not a hit list yet | 12:32 |
| rvr | jibel: Yesterday, yes | 12:34 |
| rvr | popey: I'm testing com.ubuntu.clock_3.3.272_multi.click | 12:34 |
| popey | rvr: good! :D | 12:35 |
| seb128 | pmcgowan, k, so basically we can land whatever we think is good? | 12:35 |
| rvr | popey: I had to reboot to start it, but otherwise, it is working | 12:35 |
| seb128 | pmcgowan, no fear that it derail stability for the next ota or such? | 12:35 |
| rvr | popey: Most of the changes have been verified | 12:35 |
| rvr | popey: I'm now finishing with manual tests | 12:35 |
| popey | \o/ | 12:36 |
| === _salem is now known as salem_ | ||
| jibel | rvr, can you update the landing spreadsheet? | 12:37 |
| rvr | jibel: Which part? | 12:38 |
| rvr | Ahhh | 12:38 |
| rvr | The spreadsheet | 12:38 |
| rvr | So can I land it now? | 12:38 |
| === pat__ is now known as pmcgowan | ||
| rvr | Done | 12:39 |
| jibel | rvr, you can sign it off at lease :) | 12:39 |
| jibel | least | 12:39 |
| pmcgowan | seb128, not sure you saw my reply earlier, but anything can land for now | 12:39 |
| rvr | jibel: sil2100 asked me yesterday to wait | 12:39 |
| jibel | np | 12:40 |
| sil2100 | Ah, right, I didn't report back to you since it was late | 12:40 |
| seb128 | pmcgowan, I saw, thanks, just pondering the implications it has on the next ota but I'm not complaining, I like that better than diverging from trunk and not being able to land fixes ;-) | 12:40 |
| rvr | sil2100: No problem, I forgot to ask you this morning | 12:40 |
| pmcgowan | seb128, ack | 12:41 |
| seb128 | pmcgowan, btw not sure if you saw but we updated manually the ca/de/es/fr/it/pt langpacks for the new dialer/messaging string in the overlay | 12:41 |
| seb128 | should be good now | 12:41 |
| sil2100 | pmcgowan: hey! So regarding bug LP: #1461593, davmor2 mentioned he was able to reproduce this on RTM | 12:41 |
| seb128 | I tested the new deb on my fr phone | 12:41 |
| ubot5 | Launchpad bug 1461593 in network-manager (Ubuntu RTM) "No data connection switching from 2G only to 3G" [Critical,In progress] https://launchpad.net/bugs/1461593 | 12:41 |
| pmcgowan | sil2100, I can believe it | 12:42 |
| pmcgowan | sil2100, awe has a fix coming today | 12:42 |
| rvr | seb128: Does image 28 have those langpacks? | 12:43 |
| seb128 | rvr, unsure | 12:43 |
| seb128 | where do we see what changes are on what image? | 12:43 |
| === chihchun is now known as chihchun_afk | ||
| rvr | seb128: Ok, checked. Nope. | 12:44 |
| seb128 | :-/ | 12:44 |
| ogra_ | seb128, at the bottom of http://people.canonical.com/~ogra/touch-image-stats/vivid/ | 12:46 |
| === tvoss is now known as tvoss|test | ||
| === tvoss|test is now known as tvoss | ||
| sil2100 | pmcgowan: so we wait with ota4 for that fix? | 12:46 |
| seb128 | ogra_, thanks | 12:46 |
| pmcgowan | sil2100, three fixes - the langpacks, the mms default and the 2g to 3g, yes | 12:47 |
| === marcusto_ is now known as marcustomlinson | ||
| sil2100 | pmcgowan: ok, I knew it was on our list 'blocking ota4', but since davmor2 said it was already on stable I was wondering if we still block on that | 12:48 |
| sil2100 | But ok | 12:48 |
| pmcgowan | sil2100, I think its just more frequent but not sure | 12:48 |
| pmcgowan | sil2100, essentially bq asked it be fixed | 12:49 |
| jibel | langpacks and mms as default landed this morning, waiting for a fix for 2g to 3g | 12:49 |
| seb128 | did we take the langpack infra issue somewhere for discussion? | 12:50 |
| seb128 | we patched around it manually for this time but it's a one time workaround for an easy case | 12:51 |
| sil2100 | seb128: the LP team will be discussing the possible options today on their meeting | 12:51 |
| seb128 | we need to not forget to get a proper resolution | 12:51 |
| seb128 | k | 12:51 |
| sil2100 | We're waiting for news from their side to start a formal discussion | 12:51 |
| seb128 | thanks | 12:51 |
| sil2100 | I already molested slangasek about all our issues yesterday :D | 12:51 |
| jibel | ToyKeeper, with the new thumbnailer that landed in silo 15, I've no thumbnail for music and video and a crash file for thumbernail-service. Did you notice that? | 12:54 |
| === alan_g|lunch is now known as alan_g | ||
| === tvoss is now known as tvoss|test | ||
| === tvoss|test is now known as tvoss | ||
| dobey | is nobody else seeing segfaults causing build failures on armhf/ppc in silos? | 13:01 |
| fginther | jgdx, psivaa, sorry about the ubuntu system settings job. I had an update running to fix the broken amd64 job, but it crashed leaving the job disabled. It should be back to working now | 13:01 |
| sil2100 | jibel: can you get anyone else to confirm that? Since we'll need to revert the package upload then | 13:03 |
| ogra_ | dobey, nah, we all just quickly look in the other direction when that happens | 13:04 |
| jgdx | fginther, thanks. Seems I'm getting passes. | 13:04 |
| jibel | ops-team: ^^ can anyone install the new thumbnailer from the overlay on the latest rc-proposed/bq-aquaris.en ? | 13:04 |
| jibel | and confirm if you see crashes or not and if thumbnails are correctly generated | 13:04 |
| jgdx | fginther, aah, you meant the silo? | 13:06 |
| fginther | jgdx, no, I was referring to this: http://s-jenkins.ubuntu-ci:8080/job/ubuntu-system-settings-ci/ | 13:07 |
| fginther | jgdx, The amd64 subjob was trying to collect test results that didn't exist. I was updating the jenkins config to not do that | 13:07 |
| jibel | sil2100, I removed previous thumbnails and it works. It looks like an upgrade problem from 1.3 to 2.0 | 13:07 |
| sil2100 | Oh, ok, hm | 13:08 |
| jibel | I'll downgrade the package and try to reproduce | 13:08 |
| sil2100 | Might be a problem for people uploading though | 13:08 |
| dobey | ogra_: appreciate the sarcasm, but combined with the "240 B/s" download speeds i'm getting from ports.u.c on my nexus 4, and my need to test new builds of this stuff on a device, i'm starting to get annoyed with things breaking :) | 13:08 |
| sil2100 | I mean, upgrading | 13:08 |
| jibel | right | 13:09 |
| davmor2 | jibel: I can have a look in a minute I'm nearly done with this click package | 13:13 |
| jibel | davmor2, I think you need to install OTA4, generate some thumbnails with current version then upgrade the thumbnailer | 13:17 |
| jibel | and reboot | 13:17 |
| rvr | popey: I found a problem | 13:17 |
| === MacSlow|lunch is now known as MacSlow | ||
| rvr | popey: Set an alarm (e.g. +2 minutes). Let it activate. Go to the alarms page, check that the alarm is outdated. Switch the checkbox to activate it. | 13:18 |
| rvr | popey: In my case, it crashes. | 13:18 |
| popey | on vivid? | 13:19 |
| rvr | popey: Yes | 13:19 |
| rvr | popey: I'm using rc-proposed 28 | 13:19 |
| popey | nik90: ^ | 13:19 |
| popey | lemme try to reproduce | 13:19 |
| * popey updates to 28 | 13:20 | |
| jibel | hm, this time the gallery crashed | 13:27 |
| jibel | davmor2, ^ please have a look, there is something wrong with this thumbnailer | 13:27 |
| davmor2 | jibel: will do about 5 minutes | 13:28 |
| seb128 | nik90, popey, rvr, the dbus.log has also quite "libecal-Message: e_cal_recur_generate_instances_of_rule(): bogus component, does not have DTSTART. Skipping..." added when creating a reminder, unsure what component's bug that is/if that's known | 13:28 |
| popey | rvr: nik90 reproduced on my vivid 28 krillin, and it doesn't happen on my retail krillin | 13:29 |
| popey | so yes, that's a regression | 13:29 |
| popey | yes, i see that too seb128 | 13:30 |
| seb128 | oh, also the alarm list checkbox widget toggle animate twice when focussed | 13:30 |
| seb128 | like if you open some indicator and close it | 13:31 |
| seb128 | or switch to something else and back | 13:31 |
| seb128 | checkbox->switch widget | 13:31 |
| seb128 | the on/off one | 13:31 |
| rvr | popey: Ok, moved the trello card to fail. | 13:31 |
| rvr | I find the checkbox hard to interact with. I often activate the listitem instead. | 13:32 |
| rvr | My fingers are big :) | 13:32 |
| davmor2 | popey: good news now reminders is good :) | 13:33 |
| davmor2 | popey: erm I mean notes | 13:33 |
| popey | :) | 13:33 |
| popey | rvr: nik90 is it this https://bugs.launchpad.net/ubuntu-clock-app/+bug/1413027 ? | 13:37 |
| ubot5 | Launchpad bug 1413027 in Ubuntu Clock App "One-time alarms cannot be re-enabled using the alarm switch after they have gone off once" [Medium,Fix released] | 13:37 |
| popey | perhaps not fully fixed | 13:37 |
| rvr | popey: Well, it crashes :-/ | 13:38 |
| popey | indeed. | 13:39 |
| popey | we'll see what nik90 says. I was just filing a bug and re-found that one | 13:39 |
| nik90 | rvr, popey: just got back and reading the backlog | 13:40 |
| nik90 | rvr: so I take it the only fail is the crash when activating a outdated alarm? | 13:42 |
| nik90 | rvr: the issue about the checkbox being hard to activate is a design that that I will bring to the attention of the design team. Right now clicking on the listitem will open the alarm and we cannot use that to enable/disable an alarm. | 13:43 |
| === chihchun_afk is now known as chihchun | ||
| nik90 | rvr: are you using ubuntu-touch/rc-proposed/ubuntu channel? | 13:49 |
| popey | nik90: ubuntu-touch/rc-proposed/bq-aquaris.en image 28 | 13:50 |
| popey | (that's what I tested on, which given rvr said 28, I assume same) | 13:51 |
| nik90 | popey: ok..I am first testing on n4, before moving to bq | 13:52 |
| nik90 | popey: hmm :/ crashes on N4 as well | 13:53 |
| nik90 | popey: although strangely when I reopened the clock app, the alarm that I tried to reenable is now activated..so it seems the activation works but it crashes few seconds later due to some reason | 13:55 |
| seb128 | doesn't crash here on the same image | 13:56 |
| nik90 | seb128: N4 or Bq? | 13:56 |
| seb128 | the switch widget double toggle when the app gets focussed though, but I guess that's a qml animation issue only | 13:56 |
| seb128 | nik90, bq | 13:56 |
| nik90 | my N4 image number seems to be r1 strangely despite me explicitly flashing rc-proposed..but either way for me it crashes 3-4 seconds later after enabling it | 13:57 |
| rvr | current build number: 28 | 13:57 |
| rvr | device name: krillin | 13:57 |
| rvr | channel: ubuntu-touch/rc-proposed/bq-aquaris.en | 13:57 |
| rvr | nik90: ^ | 13:58 |
| nik90 | rvr: the crash seems to be fluctuating .. I cannot reproduce it every time.. Do you see the same? | 13:59 |
| rvr | nik90: Let me check | 13:59 |
| nik90 | this is bad since it might require a patch in the sdk which is a big target to update on the phone :/ | 13:59 |
| davmor2 | jibel: confirmed I get music notes and and what looks strangely like a youtube logo on videos | 14:00 |
| ChrisTownsend | trainguards: Hello, could I get a couple of silos please? | 14:04 |
| Mirv | ChrisTownsend: sure | 14:05 |
| ChrisTownsend | Mirv: Thank you | 14:05 |
| Mirv | ChrisTownsend: 003 + 019 | 14:06 |
| ChrisTownsend | Mirv: Great, thanks again | 14:06 |
| popey | nik90: we should get a bug filed to track this as soon as we can. | 14:07 |
| nik90 | popey: I will file it when I am able to reproduce the crash again (to get logs)..I reproduced the crash once out of 7-8 tries | 14:08 |
| nik90 | (on N4) | 14:08 |
| popey | heh, okay | 14:08 |
| rvr | nik90: Right, not always | 14:10 |
| nik90 | popey, rvr : Do either of you have the app log when it crashed? | 14:12 |
| nik90 | I will try a couple more times | 14:12 |
| popey | nik90: http://paste.ubuntu.com/11672904/ | 14:14 |
| popey | nik90: http://paste.ubuntu.com/11672918/ dbus.log | 14:15 |
| nik90 | popey: I dont think the clock app log has the crash event in it..it all seems pretty normal to me..more the warnings are sdk stuff that most apps display. | 14:16 |
| popey | hmm | 14:16 |
| popey | probably because qmlscene just dies on its arse | 14:16 |
| popey | -rw-r----- 1 phablet whoopsie 14668426 Jun 9 14:29 _usr_lib_arm-linux-gnueabihf_qt5_bin_qmlscene.32011.crash | 14:17 |
| popey | it uploaded, lemme see if i can find the crash in errors.u.c | 14:17 |
| nik90 | the only thing that I cannot understand in the dbus log is libecal-Message: e_cal_recur_generate_instances_of_rule(): bogus component, does not have DTSTART. Skipping... | 14:17 |
| popey | hmm, i have 116 errors on errors.ubuntu.com :S | 14:18 |
| nik90 | ;) | 14:18 |
| popey | /usr/lib/arm-linux-gnueabihf/qt5/bin/qmlscene:11:Change:QVector:resize:pop:QAbstractItemModel::endMoveRows | 14:18 |
| popey | thats what it says under "problem" | 14:19 |
| popey | i think i see your crash there too | 14:19 |
| nik90 | popey: I am creating a bug now, can you upload that log to it then unless there is something personal in them? | 14:21 |
| popey | is a pdf of it any good? | 14:21 |
| popey | I see no other easy way to get it | 14:21 |
| popey | gimmie the bug number when you have it and I'll attach what I can | 14:22 |
| popey | but it looks like a qt/toolkit issue to me | 14:22 |
| nik90 | popey: https://bugs.launchpad.net/ubuntu-clock-app/+bug/1463430 | 14:25 |
| ubot5 | Launchpad bug 1463430 in Ubuntu Clock App "Clock app crashes when renabling an alarm" [Undecided,New] | 14:25 |
| popey | nik90: updated | 14:29 |
| * nik90 looks | 14:29 | |
| popey | I'd add a task for whatever package provides qmlscene / qt5 | 14:29 |
| nik90 | popey: I think that was qtubuntu if I recall | 14:30 |
| popey | looks plausible :) | 14:31 |
| popey | we should set priority high as this is blocking landing | 14:31 |
| nik90 | I set it high for clock app..dont have permissions for qtubuntu | 14:32 |
| nik90 | popey: I see qt5/qml/Ubuntu/Components/libUbuntuComponents.so at the very end of the pdf...otherwise I cannot make head or tail of the crash log | 14:36 |
| Mirv | nik90: libUbuntuComponents.so -> ubuntu-ui-toolkit | 14:36 |
| popey | me either | 14:37 |
| Mirv | (looking at https://errors.ubuntu.com/bucket/?id=/usr/lib/arm-linux-gnueabihf/qt5/bin/qmlscene%3A11%3AChange%3AQVector%3Aresize%3Apop%3AQAbstractItemModel%3A%3AendMoveRows ) | 14:37 |
| kgunn | cihelp hey there, josharenson pointing out that ci seems to be taking unusually long to run here | 14:38 |
| kgunn | https://code.launchpad.net/~josharenson/unity8/integrated_lightdm_refactor/+merge/260588 | 14:38 |
| kgunn | any ideas ? | 14:38 |
| Mirv | QtCore -> qtbase-opensource-src | 14:38 |
| popey | thanks mirv | 14:38 |
| nik90 | Mirv: thnx..I will add the necessary projects to the report | 14:39 |
| seb128 | having a debug bt would help there I guess | 14:41 |
| nik90 | zsombi: We seem to have a clock app crasher related to alarms which we detected while trying to push out a new clock app to the store https://bugs.launchpad.net/ubuntu-clock-app/+bug/1413027 | 14:44 |
| ubot5 | Launchpad bug 1413027 in Ubuntu Clock App "One-time alarms cannot be re-enabled using the alarm switch after they have gone off once" [Medium,Fix released] | 14:44 |
| nik90 | bah wrong bug link .. here is the correct one https://bugs.launchpad.net/ubuntu/+source/qtbase-opensource-src/+bug/1463430 | 14:44 |
| ubot5 | Launchpad bug 1463430 in Ubuntu Clock App "Clock app crashes when renabling an alarm" [High,Confirmed] | 14:44 |
| nik90 | zsombi: can you take a brief look at https://errors.ubuntu.com/bucket/?id=/usr/lib/arm-linux-gnueabihf/qt5/bin/qmlscene%3A11%3AChange%3AQVector%3Aresize%3Apop%3AQAbstractItemModel%3A%3AendMoveRows and see if the sdk is involved? | 14:45 |
| nik90 | zsombi: We're currently trying to figure out the project causing the crash which could be the clock-app, sdk, qtubuntu, qtbase-opensource-src | 14:46 |
| balloons | fginther, ping | 14:55 |
| josepht | kgunn: looking | 14:59 |
| nik90 | popey: I think I can confirm this is a non-clock issue since the code that enables/disables alarm in that alarm list page has not been touched since before r191 while the store version is at r192. | 14:59 |
| popey | nik90: ok. | 15:00 |
| fginther | balloons, one moment please | 15:20 |
| fginther | kgunn, the MP is running again. The machine used to run the unity-phablet-qmluitests-wily test has failed and I've had to disable that test | 15:25 |
| kgunn | fginther: ta | 15:26 |
| === chihchun is now known as chihchun_afk | ||
| pstolowski | sil2100, hey, i've just noticed in pmcgowan's email a mention of not detecting bad package dependencies. i've just flashed the phone with rc-proposed/bq-aquaris.en and after applying a silo on top of it i ended up with half-broken system (a mix of new & old version of *thumbnailer* packages installed) and had to manually upgrade broken packages; is it that issue? | 15:38 |
| pmcgowan | it could be | 15:41 |
| pmcgowan | jibel, rvr were the translations confirmed? | 15:51 |
| fginther | balloons, what's uP? | 15:52 |
| balloons | fginther, :-) Hey, so I was thinking about the core apps problem again and I'd like to try the idea of building the packages on vivid, but running them on an older version of ubuntu. I think this will work, as long as we grab the newer version of autopilot and the UITK helpers package. | 15:53 |
| balloons | technically the requirement for vivid comes from needing to build, and the builders don't require the test environment setup | 15:54 |
| fginther | balloons, yes, I think should work | 15:55 |
| balloons | fginther, so I was trying to check this locally but couldn't find the zip of artifacts from the clock app jobs for instance | 15:55 |
| fginther | balloons, you just need the debs produced by http://91.189.93.70:8080/job/ubuntu-clock-app-vivid-amd64-ci/ ? | 15:57 |
| balloons | fginther, sure.. Just to double check before I ask you to do the same | 15:58 |
| balloons | I was going to try utopic and trusty.. If things would work on trusty that would be ideal for running | 15:59 |
| rvr | pmcgowan: Not available in 28 | 15:59 |
| fginther | balloons, http://91.189.93.70:8080/job/ubuntu-clock-app-vivid-amd64-ci/ should be collecting them now | 16:00 |
| rvr | Hopefully in 29 | 16:00 |
| pmcgowan | rvr, ah ok | 16:01 |
| === tvoss is now known as tvoss|dinner | ||
| fginther | balloons, hows this? http://91.189.93.70:8080/job/ubuntu-clock-app-vivid-amd64-ci/9/ | 16:17 |
| balloons | grabbed'em.. let's try | 16:18 |
| === alan_g is now known as alan_g|EOD | ||
| robru | dbarth: https://ci-train.ubuntu.com/job/ubuntu-landing-002-2-publish/104/console need these merges approved | 17:21 |
| dbarth | robru: right, i was just checking that, but i spotted a potential regression meanwhile on one of the branches | 17:39 |
| robru | dbarth: ah ok | 17:40 |
| dbarth | robru: well, so i'll revert the test status for now while we check with alex/mardy | 17:45 |
| dbarth | ah, you beat me to that ;) | 17:45 |
| robru | dbarth: heh, yeah | 17:46 |
| === oSoMoN_ is now known as oSoMoN | ||
| === tvoss|dinner is now known as tvoss | ||
| === tvoss is now known as tvoss|dinner | ||
| * robru -> lunch | 18:26 | |
| ogra_ | pmcgowan, hmm, all my scopes have no online images anymore (apps is fine, video shows local vids, music shows local music but no thumbnails for online stuff) | 18:44 |
| ogra_ | (latest rc-proposed ... seems like a thumbnailer regression, did that land recently ?) | 18:45 |
| ogra_ | sil2100, ^^ | 18:45 |
| jibel | ogra_, which image? | 18:45 |
| ogra_ | latest meizu.en | 18:45 |
| ogra_ | (rc-proposed) | 18:45 |
| ogra_ | updated earlier today | 18:45 |
| jibel | ogra_, what is the version of the thumbnailer on this image? | 18:46 |
| ogra_ | one sec | 18:46 |
| jibel | ogra_, I think the new thumbnailer crashes when it's upgraded, but not on a fresh install | 18:46 |
| ogra_ | 1.3+15.04.20150312-0ubuntu1 | 18:46 |
| jibel | ogra_, not thumbnailer's fault. Latest version is 2.0 | 18:47 |
| ogra_ | ah | 18:48 |
| ogra_ | i'll wait for the next image then :) | 18:48 |
| jibel | ogra_, however there is a new mediascanner on this image | 18:49 |
| ogra_ | yeah, | 18:49 |
| ogra_ | just checked the changelog | 18:49 |
| ogra_ | hmm, it runs but there is no recent log | 18:50 |
| ogra_ | uh | 18:50 |
| ogra_ | download-manager fails to download a click update | 18:50 |
| ogra_ | i'm definitely online though ... browser and webapps work fine | 18:52 |
| * ogra_ reboots | 18:52 | |
| ogra_ | reboot helped ... i'll keep an eye on that | 18:53 |
| sil2100 | Yeah, we just landed thumbnailer today, so it can't be on latest proposed image | 18:58 |
| jibel | sil2100, BTW davmor2 confirmed that after upgrade he doesn't have thumbnail for albums and videos. It would probably be safe to revert to 1.3 until the problem is understood. What do you think? | 19:02 |
| sil2100 | jibel: agreed, I'll prepare a soft-revert shortly | 19:02 |
| sil2100 | Just want to be done with preparing dinner | 19:02 |
| robru | ATTENTION: taking the train offline for emergency migration from ps4 to ps4.5 | 19:32 |
| robru | PPAs are fine but no new jobs will be able to be run temporarily | 19:32 |
| awe | ;(- | 19:36 |
| * awe thinks he added his new line to the wrong spreadsheet | 19:36 | |
| awe | robru, is ps4.5 the version of a new spreadsheet instance? | 19:37 |
| robru | awe: no, spreadsheet is unrelated. I'm talking about jenkins. | 19:37 |
| awe | whew | 19:37 |
| awe | robru, I just created a request for a NM sync silo, per the same method we last used to binary sync from the phablet-team/telephony PPA | 19:38 |
| awe | sil2100 handled the sync link last time, but I think I got it right | 19:38 |
| robru | awe: ok | 19:38 |
| robru | will look in a sec | 19:39 |
| awe | thanks | 19:39 |
| robru | awe: fixed it | 19:42 |
| awe | what did I miss? | 19:43 |
| awe | did I put the comments in the wrong column? | 19:44 |
| robru | awe: it's supposed to start with "sync:" and the ppa definition needs to be ppa:team/ubuntu/name | 19:44 |
| awe | ah, thanks | 19:44 |
| robru | awe: also wait, why do you have an MP and a sync? that doesn't make sense. | 19:44 |
| awe | it's just there for reference sake | 19:45 |
| awe | as we record test results in our MPs for ofono, and been trying to do the same for NM | 19:45 |
| awe | I noted this in the comments | 19:45 |
| awe | If that's too confusing, I won't do that in the future or maybe just list the MP in the comments | 19:45 |
| robru | awe: ok well that's not going to work where it is. if the train sees an MP in the MP field it's going to build it and that's going to interfere with the sync. | 19:45 |
| awe | got it | 19:46 |
| robru | awe: but I notice the MP isn't merged, so maybe you really just want to build the MP? | 19:46 |
| awe | no. this is network-manager | 19:46 |
| robru | awe: I dunno much about nm ;-) | 19:47 |
| awe | yea, it's maintained with quilt patches, don't think the CI build code knows how to deal with it | 19:47 |
| awe | that's why it's a binary sync from the telephony PPA | 19:47 |
| awe | this is how sil2100 and I handled the last upload of NM | 19:47 |
| awe | I believe cyphermox used to do something similar... | 19:48 |
| awe | except he may have directly uploaded to a silo | 19:48 |
| robru | awe: ok no worries. train doesn't have any special quilt coverage, train just uses "bzr bd". you might want to check if nm can be built with "bzr bd" ;-) | 19:48 |
| awe | it can, but I was told train wouldn't work with NM | 19:48 |
| robru | awe: yeah cyphermox used to do direct uploads himself. | 19:48 |
| robru | awe: yeah we should probably figure out for sure why it can't and if it can't, fix it. I really dislike these special cases ;-) | 19:49 |
| robru | awe: but not today, too busy | 19:49 |
| awe | robru, agreed. I will have more NM updates on the way. I'll re-review the email trail on this, and follow-up with you later in the week | 19:50 |
| robru | awe: thanks | 19:51 |
| awe | but that said, do I need to take any action to trigger the sync ( ie. do I still need to click 'Build' )? The PPA is just now finishing it's publishing of the package | 19:52 |
| robru | awe: yeah I need to assign the silo and then you'd have to click build (being careful to check INCLUDE_BINARIES_IN_SYNC option). but we have to wait for the train to come back online, #is is migrating it. | 19:52 |
| awe | robru, got it. I have to leave for the gym shortly, so I'll check back when I return | 19:53 |
| robru | awe: sure | 19:54 |
| balloons | fginther, fyi on the earlier experiments. Everything went more or less fine, however it still needed the newer qt components to run sadly. So the tests fai | 19:58 |
| fginther | balloons, :-( | 20:02 |
| balloons | I think nik90 works around that on trust by using an lxc container, but perhaps that's getting too deep for us | 20:04 |
| nik90 | balloons: Indeed I do..however its not that easy and is a bit too extensive and too much for your use case imo. | 20:08 |
| nik90 | welcome to my world ;) | 20:08 |
| balloons | nik90, :p | 20:08 |
| robru | awe: ok, started the copy for you since you're the guniea pig for the brand new production instance of the train ;-) https://ci-train.ubuntu.com/job/ubuntu-landing-031-1-build/9/console | 20:24 |
| === robru changed the topic of #ubuntu-ci-eng to: Need a silo or CI Train support? ping trainguards | Need help with something else? ping cihelp | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: We just cut over to a brand new production deployment of ci-train, please report any issues you find to robru immediately | ||
| sil2100 | jibel: package reverted | 20:33 |
| robru | awe: looks good, please test that silo and then we can publish | 21:13 |
| robru | charles: I see the error in silo 22, just investigating | 21:57 |
| awe | thanks robru! | 22:04 |
| robru | awe: you're welcome! | 22:04 |
| robru | awe: charles wasn't so lucky though, looks like a firewall issue busted his silo | 22:04 |
| charles | :( | 22:04 |
| awe | good luck charles! | 22:04 |
| charles | thanks awe | 22:05 |
| robru | charles: I poked the #is vanguard about this, shouldn't take too long to fix. | 22:05 |
| charles | robru, thanks! | 22:05 |
| boiko | robru: hi, I added a new component to silo 8 (telephony-service), could you please reconfigure it? | 22:12 |
| robru | charles: you're welcome | 22:12 |
| boiko | robru: vivid silo 8 I mean | 22:12 |
| robru | boiko: I can, but there's a train issue preventing MP builds from working currently, so you won't be able to build right away | 22:12 |
| boiko | robru: well, that silo is a sync silo, is it also affected? | 22:13 |
| robru | boiko: yeah, sorry, source syncs are also affected. only binary syncs can work right now (dput is broken, so any source upload fails, including source syncs or MP builds) | 22:14 |
| robru | boiko: so don't build 43 yet either ;-) | 22:14 |
| boiko | robru: ok , got it! thanks :) | 22:14 |
| robru | boiko: you're welcome. will ping you guys once it's safe | 22:15 |
| boiko | sure thing | 22:15 |
| boiko | robru: sorry for the silo request flood, but we had a long queue of small-to-medium fixes that were pending for quite some time now | 22:23 |
| robru | boiko: haha, no worries. we have lots of silos now | 22:23 |
| boiko | great! :D | 22:23 |
| robru | boiko: yeah, 14 free right now. | 22:25 |
| boiko | robru: that's nice! it means I can take my time on testing those, good :) | 22:26 |
| robru | boiko: sure, but not too long ;-) | 22:27 |
| boiko | robru: :) | 22:27 |
| robru | boiko: ok try your builds now | 23:06 |
| robru | charles: I fiddled yours manually, should be working now, but let's keep an eye on that | 23:06 |
| robru | charles: hm, looks like there's a build failure on powerpc, was that expected? | 23:07 |
| charles | robru, not sure what the issue there is... will look into it tomorrow morning, I'm EODing | 23:22 |
| charles | robru, thanks for getting the silo unstuck though :-) | 23:22 |
| robru | charles: ok no worries, yeah the train seems to be working, as far as I can tell this failure is on you ;-) | 23:22 |
| robru | charles: good night! | 23:22 |
| charles | heh | 23:23 |
| charles | good night | 23:23 |
| robru | boiko: https://ci-train.ubuntu.com/job/ubuntu-landing-046-1-build/6/console I'm kicking off one of your builds just to test the train a bit. apologies if it wasn't ready | 23:27 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!