[00:14] Robru, I added the MP for Saviq to the spread sheet ---> https://code.launchpad.net/~mir-team/platform-api/rtm-14.09-compat-mir-0.8.1/+merge/246370 [00:14] robru ^^ [00:15] camako, ah, I see you unconfused this, glad [00:15] Saviq, yeah it builds [00:15] camako, you need to include this diff before your changelog change https://launchpadlibrarian.net/194801806/platform-api_2.5.0%2B14.10.20141015-0ubuntu1_2.5.0%2B14.10.20141015-0ubuntu2.diff.gz [00:15] camako, the train will complain otherwise [00:16] Saviq, you added platform-api to column G, now we are adding papi MP.. I dunno if it makes a difference [00:16] Saviq, ok I'll take that in too [00:16] camako, yeah, remove it from the sources column [00:17] Saviq, ok [00:19] robru, camako, because of the nature of the change I would not "revoke" the QA sign-off, just build it and publish straight away to unblock the migration [00:20] Saviq, I think I'm ok with that [00:20] one sec [00:20] Saviq, changelog updated [00:21] camako, thanks, ACK [00:35] Saviq, camako: sorry guys I gotta run to the doctor's, will be back in about 2 hours. Hopefully rsalveti is around to hit publish on that when it finishes building, otherwise I can do it later. [00:36] robru, nw, it's almost build, I'll just do a quick validation and set the spreadsheet accordingly [00:54] ok, robru, rsalveti, if you're around ↑, everything seems finally in order [00:54] please publish rtm/018 (only papi needs publishing, not sure how the job looks) [02:10] === IMAGE 65 building (started: 20150114-02:10) === === robru_ is now known as robru [02:38] Saviq: ok, sorry for the delay, republished [02:40] ToyKeeper, i saw your comment about the autopilot failures [02:40] you need python3-evdev installed [02:40] which i can't figure out why it isn't getting installed [02:40] python3-autopilot depends on it [02:40] for some reason when CI runs the AP tests, that package gets installed [02:41] but installing ubuntu-system-settings-autopilot doesn't... [02:41] i don't want to add a depends for it myself, since ubuntu-system-settings-autopilot depends on python3-autopilot, which depends on python3-evdev [02:56] ToyKeeper, i went ahead and added the depends and kicked off a rebuild [02:56] this way i know it doesn't bite us again [02:57] but if you like, you can just install that package then start the tests [02:57] the rebuild just adds the depends on the package [03:00] ToyKeeper, and silo 3 can land separately, just without it any settings test that mocks upower will fail [03:00] but they are failing now anyway [03:11] === IMAGE RTM 193 building (started: 20150114-03:10) === [03:18] ಠ_ಠ [03:21] Saviq: you still around? some kind of merge conflict cleaning your silo. [03:21] just looking into it now... [03:31] Saviq: nm, fixed it. all landed and merged now! [04:15] === IMAGE RTM 193 DONE (finished: 20150114-04:15) === [04:15] === changelog: http://people.canonical.com/~ogra/touch-image-stats/rtm/193.changes === [08:06] sil2100, good morning [08:07] sil2100, could you take a look at line 52, I adjusted needs qa to no, elaborated in the test plans column [08:07] Morning! [08:09] tvoss: I'll check that in a moment, but I might have to consult QA as well [08:21] pstolowski: hey! I would need you to get https://code.launchpad.net/~unity-team/unity-scope-click/fix-1390191-rtm/+merge/242214 approved by someone [08:22] tvoss: ok, so I'll have to talk about your silo with QA - normally I would let it in with this rationale, but since it's location-service QA might think there's additional risk assigned etc. [08:22] sil2100, hi! uhm, sure [08:25] dobey: ^ [08:25] (actually I see dobey was the lander) [08:28] sil2100, done [08:28] sil2100, dobey will be available in a few hours [08:32] sil2100: I am done with the rtm silo10, feel free to empty it and put back to the pool. Also the line 58 can be deleted from the sheet. Thank you a lot for the resources. It was a super valuable exercise. The Vivid edition of the UITK passes all tests on RTM. [08:34] bzoltan: that's good to know - does it also run without any user-visible problems? You dogfooded it? [08:34] pstolowski: thanks! [08:35] ogra_, davmor2, popey: I have a doctor's appointment in 10 minutes so I'll have to jump out, in theory I should make it in time for the meeting but this depends on the waiting queue === sil2100 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: sil2100 out for a doctors appointment, trainguard back in ~1h [08:38] sil2100, sure, let me know if I can help [08:47] sil2100, hey, rtm silo 012 passed QA but Omer must've forgot to push the button, what do we do in that case https://trello.com/c/E7oTClp2/571-ubuntu-rtm-landing-012-unity8-saviq ? [09:00] ah actually no, it's the broken silo [09:10] sil2100, ogra_ - is now a good time to push a *vivid* device tarball? [09:10] john-mcaleely, sure, go ahead [09:10] o/ [09:11] thanks [09:11] john-mcaleely: yeah, no plans for vivid this week, so +1 === sil2100 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: - [09:11] Saviq: broken silo..? [09:11] sil2100, yeah, see Timo's comment [09:11] I hope davmor2 will be along at some point today with news of the rtm device tarball :-) [09:11] Ah, this one [09:11] Yeah [09:12] Saviq: it will land soon then [09:12] sil2100, well, it's my fault as I accidentally built it and we tried to recover by bincopying to a different silo, but it never go to a coherent state [09:12] sil2100, thanks [09:12] ok, so vivid tarball pushed [09:12] john-mcaleely: no I decided I was bribed enough as a child [09:13] heh [09:13] I'm not sure I get that [09:13] sil2100, it seems like it won't do anything automagically, might even require manual copy to proposed [09:13] unless you find a way to get it to work... [09:16] geez .... test results for 192 and 193 look really bad [09:16] ~20 new crashers [09:17] Wow, 193 had mir, but this would mean it's unrelated? [09:17] seems unity-scopes_scoperunner crashes in every test [09:17] sil2100, yeah, started one image before [09:18] Looks like a scoperunner issue [09:18] sil2100: I blame ogra_ [09:19] ogra_: yeah... [09:19] sil2100, might be an apparmor one though, i cant open any log file so i cant tell [09:19] john-mcaleely: wasn't even now it might make more sense :) [09:19] * sil2100 logs in into the VPN [09:22] ogra_, health-check didn't run on 192 and 5 tests are failing. [09:24] jibel, yeah, thats not the issue, we have a new constant crasher (unity-scopes_scoperunner) [09:25] that worries me more than a few new crashes [09:25] right, could be bug 1357143 ? [09:25] err [09:25] bug 1357143 in unity-scopes-api (Ubuntu) "/usr/lib/arm-linux-gnueabihf/unity-scopes/scoperunner:6:__gnu_cxx::__verbose_terminate_handler:__cxxabiv1::__terminate:std::terminate:__cxxabiv1::__cxa_throw:unity::scopes::Variant::get_string" [Critical,Confirmed] https://launchpad.net/bugs/1357143 [09:25] failures [09:25] ogra_: looking at the console logs and comparing with earlier images, so far I don't see any additional apparmor denials [09:26] sil2100, yeah, smells like it could be the bug above [09:27] funnily on my own device ... while i have a unity-scopes_scoperunner.crash file, it is from the 9th ... so 5 days old [09:28] Maybe autopilot triggers it in a funny way [09:28] seems https://code.launchpad.net/~alecu/unity-scope-click/fix-empty-variant/+merge/246365 has the fix [09:28] alecu: ping [09:32] trainguards, I can has reconfigure on rtm silo 006 please, added qtmir there [09:37] hi there, any pending issues with ci ? [09:45] pstolowski: hey! [09:45] sil2100, hey === sil2100 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: ubuntu-rtm landing gates closed! [10:03] tvoss: hey, so we had a talk about your silo, and it will have to wait for next milestone - sadly one of the reasons is also that it's not yet approved by the product team [10:03] john-mcaleely, ^, could you take care of that issue, please? [10:16] john-mcaleely: can you throw me the link for the tarball again please [10:19] sil2100, any word on publishing rtm silo 12? [10:19] Saviq: trying this now :) [10:19] sil2100, ah, thanks :) [10:19] john-mcaleely: nevermind found it [10:21] Saviq: if you could make sure to document adding dependencies in commit-messages/changelog-entries I would be really happy ;) Anyway, it seems it's publishing! [10:22] sil2100, I'll try to remember, sorries [10:22] sil2100, can I have a reconf on silo 6 in the mean time [10:22] No worries, it's cosmetics really [10:22] Saviq: sure [10:23] sil2100, btw, https://ci-train.ubuntu.com/job/ubuntu-rtm-landing-012-2-publish/26/console seems to have a wrong path to the branch it pushed to [10:23] missing "ci-train-bot" [10:24] Saviq: this is really strange, actually the ci-train-bot LP account seems to be missing [10:24] Where did it push it to then..? [10:25] sil2100, it pushed fine [10:25] sil2100, just the log is wrong [10:25] sil2100, well, actually, maybe it pushed to ~/, which is the user's path then [10:25] Saviq: might be the case here then [10:26] Since CI Train prints exactly the string it did bzr push to [10:26] Didn't know you could do ~/something/something === marcusto_ is now known as marcustomlinson_ === marcustomlinson_ is now known as marcustomlinson [10:29] sil2100, see the last commit in https://code.launchpad.net/~indicator-applet-developers/qmenumodel/trunk - something's funky there as well [10:30] Saviq: yeah, noticed that in another trunk as well, the user config is somehow busted [10:30] The problem is that I don't have much power over the CI Train machine now after it got pushed to IS [10:30] But I'll re-try today to get someone looking at this [10:30] sil2100, nw, thanks [10:39] sil2100, sorries, one last resync of rtm silo 6, added a sync of qtubuntu from vivid :) [10:40] Saviq: sure ;) [10:40] s/qtubuntu/qmenumodel/ === cking_ is now known as cking [10:41] Done! [10:54] sil2100, how do I get tvoss gps testing silo approved? [10:54] sil2100, it's an urgent request from a customer [10:55] john-mcaleely, probably best to check with pmcgowan once he comes up [10:55] john-mcaleely: is this required for this milestone? [10:55] sil2100, yes. testing GPS needs to be done well before other dates coming soon :-) [10:56] It's a little bit latish you know ;) [10:56] *lateish [10:56] jibel: ^ [10:57] sil2100, my fault. I forgot to get the bug raised to the product team. Mostly because it was a feature request... [11:01] sil2100, is the silo creation closed? I'd need to create one for rtm quite urgent :-/ [11:02] sil2100, I though it could be done yesterday night [11:02] mandel: for testing you mean? [11:03] mandel: since for a silo to get it for this week's promotion it's already too late [11:03] sil2100, I guess it was promotion.. but well, if it is too late is too late, nothing we can really do I suppose [11:03] sil2100, I'll add it to the spreadsheet then [11:08] john-mcaleely, it's too late and not reviewed by the product team. You can escalate it to Pat. [11:20] jibel, do you know when the product team triage canonical-devices-system-image? It seems I filed it there four days ago [11:20] https://bugs.launchpad.net/ubuntu/+source/location-service/+bug/1408984/+activity [11:20] Launchpad bug 1408984 in location-service (Ubuntu) "No tool exists to test SNR &TTFF on a handset" [High,In progress] [11:20] pmcgowan, when you're online ^ please get this in this milestone [11:22] (the alternative is for our customer to wait at least a week, possibly two, to start this testing.) [11:24] john-mcaleely, idk when the review is. [11:29] hi trainguards, i have silo 007 vivid finally tested and good to land on vivid [11:29] I have also tested backports to 14.04 if I can get a sync to Trusty, to initiate the SRU process [11:30] dbarth: hey! Ok, let me do that [11:30] But let's first release to vivid [11:30] sure, that's the goal [11:30] upstream, then SRUs [11:31] dbarth: btw. you know why the build job was waiting for infinity? Do you remove some arches in this silo? [11:32] sil2100: it was a build wait because the chrome build fails on powerpc, etc. [11:32] same as oxide [11:33] That shouldn't have caused the build job to wait unless the package was previously built on those architectures [11:33] Which would then imply that you'll have trouble getting out of -proposed without archive admin intervention, and it would be best to have it analysed up-front [11:34] Yeah, looking at LP it seems that previously it had powerpc binaries [11:34] but: [11:34] unity-chromium-extension | 3.0.0+14.04.20140318-0ubuntu1 | vivid/universe | source, amd64, armhf, i386 [11:34] cjwatson: i think i stopped the job cause it was being stuck on those arches [11:34] No, wait [11:34] so IMO the build job is wrong [11:34] that one is a new piece of code [11:34] dbarth: You did, but it shouldn't have broken this way [11:34] using the newest chrome API [11:34] cjwatson: https://launchpad.net/ubuntu/+source/unity-chromium-extension [11:34] cjwatson: this page confuses me [11:34] ie, a replacement for the old NPAPI which has been deprecated [11:35] cjwatson: the version in -proposed seems to have powerpc binaries [11:35] cjwatson: while the one in the archive, LP says it's build-dep-waiting on something [11:35] sil2100: Yeah, clearly managed to build somehow, but that was in a utopic PPA so who knows [11:35] cjwatson: ah, scratch the last sentence of mine [11:35] cjwatson: it's actually dep-waiting on the 3 arches [11:36] do you have links to the individual builds that were on dep waits ? [11:36] And it was stuck in -proposed because those three arches had unsatisfiable binaries [11:36] dbarth: that's not necessary [11:36] ah right, so it couldn't have worked anyway [11:36] sil2100: so how about I just remove the binaries for those three arches from -proposed [11:36] they aren't going to work, and they're evidently now no longer built [11:36] +1 [11:36] cjwatson: sounds like a plan, thanks :) [11:37] it's disappointing that nobody followed this up for the 201 days that unity-chromium-extension has been stuck in -proposed [11:37] somebody dropped the ball in epic style there [11:37] 201 days? [11:37] oh, that was the initial landing attempts [11:37] and then we went into fixing the package-kit problem [11:37] given how easy this was to address, it should have been addressed ~200 days ago [11:38] and not ignored [11:39] 201 days?! Oh damn [11:39] I've removed those binaries now; I strongly suggest letting the 200-day-old upload migrate before dealing with the next one [11:39] sorry, i was not aware this package was still blocked there [11:39] I suspect somebody cleared a silo when they shouldn't have [11:40] i asked the initial silo to be unloaded, knowing it would take a bit to fix the other bug [11:40] yeah, but this is how things get lost [11:40] and then we went into 'phone focus' mode... ;) === tvoss is now known as tvoss|lunch [11:58] davmor2, any news on the device tarball? (I'm on hols today, and will be out this afternoon, if it needs pushing) [11:59] john-mcaleely: running still should be done in about 25-30 minutes [12:00] davmor2, cool, no problem [12:03] * sil2100 goes off to prepare dinner [12:03] john-mcaleely: once you get a +1 from davmor2, feel free to push it [12:04] sil2100, thanks! [12:05] must be quite some dinner if you start preparing it around noon :) [12:05] lol [12:10] jibel, looks like victorp got to it before pmcgowan . I believe #1408984 is now approved for this milestone [12:10] sil2100, ^ === MacSlow is now known as MacSlow|lunch [12:34] ogra_: it has to cook in the oven for over an hour ;) [12:34] :) [12:34] ogra_: oh, and I meant the 'dinner' dinner [12:34] You know, dinner can mean 'supper', but can also mean 'lunch' [12:34] Confusing [12:35] heh [12:52] sil2100, john-mcaleely: tarball complete [12:53] davmor2, yay. +1 ? [12:53] pete-woods: did you ever figure out the "(missing upload)" thing for the store? [12:53] john-mcaleely: yeap [12:53] davmor2, double yay [12:55] sil2100, ogra_ pushed, per your comment above. Coming to an rtm build near you soon :-) [12:55] thank you! [12:55] john-mcaleely, enjoy your holiday now ! [12:55] :) [12:55] o/ [12:55] :-) [12:59] So now we just need to wait for rvr to finish testing telephony-service [12:59] sil2100: I did, but I need boiko to confirm an issue [12:59] Oh [12:59] Either him or bfiller [13:00] john-mcaleely: now get outta here ;) [13:00] jibel, brendand, davmor2, rvr: did you notice any scoperunner crashes during todays usage [13:00] ? [13:00] _usr_bin_webapp-container.32011.crash [13:00] _usr_bin_webapp-container.32011.upload [13:00] _usr_bin_webapp-container.32011.uploaded [13:00] sil2100: ^ that's it [13:02] _usr_share_click_preinstalled_com.ubuntu.camera_3.0.0.412_camera-app.32011.crash [13:03] No scope crashes, though === alan_g is now known as alan_g|lunch [13:08] davmor2, rvr: great news [13:08] We'll anyway inform cwayne once he's up [13:08] sil2100: whatsup? [13:09] you broke it !! [13:09] :) [13:09] cwayne: hey! So, after the custom tarball upload yesterday we noticed scoperunner crashes for every autopilot test in the smoketesting suite [13:09] cwayne: it's always reproducible on the testing environment, but it seems to happen only during running autopilot tests [13:10] hmm, there were only changes to like 2 scopes.. shouldnt have affected any of the other ones [13:10] sil2100, that is wierd, nothing else landed on the actual scope running? [13:10] cwayne: it seems some scope is crashing, and since no scope-related thing got updated besides the custom tarball [13:11] victorp: no... nothing besides that [13:11] do you have some logs? [13:11] victorp: there were a lot of landings, but nothing to do with scopes [13:11] as i said before, apparmor changes landed too, but we dont seem to have any apparmor denials in the logs [13:12] iirc the apparmor changes were custome related [13:12] -e [13:12] it's happening for every single autopilot test? [13:12] victorp: http://rtm-dashboard.ci.ubuntu.com/smokeng/utopic/touch_stable/krillin/192:20150113:20150107-5ffb790/223/reminders/ <- here's for instance one of the tests that had the crash, you can access the crashfile there [13:12] cwayne, yup [13:12] cwayne: almost every, but in overall there's like 20 scoperunner crashes in the smoketesting runs [13:12] http://rtm-dashboard.ci.ubuntu.com/smokeng/utopic/touch_stable/krillin/192:20150113:20150107-5ffb790/223/ [13:13] sil2100, it looks like I cant see that :( [13:13] I guess I dont have vpn setup [13:20] sil2100, not today, but I've a scope runner crash from 2 day sago === kalikiana_ is now known as kalikiana [13:30] bzoltan: aargh, unapproved branches! :) Check merges from silo 10 [13:31] sil2100: ohh, I am sorry. So amateurish mistake ...fixed [13:37] sil2100, so you know why the status of rtm 004 is "ready to build" on the dashboard when the package is built in the ppa since yesterday? [13:38] seb128, an MP branch or is it a source package upload ? [13:38] ogra_, copy from a vivid source [13:38] (the latter need a watch only build) [13:39] then try a watch only build [13:39] well, it's built [13:39] https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu-rtm/landing-004 [13:39] the train doesnt watch the actual build for source packages [13:39] you need to tell it to ... [13:39] it only works fully automatic for MPs [13:44] ogra_, danke, that worked [13:44] :) [13:44] john-mcaleely, om26er is on silo19 === alan_g|lunch is now known as alan_g [13:45] john-mcaleely, I am actually waiting for tvoss|lunch to give me steps to verify the fix. [13:45] cwayne: no. it's still happening today === tvoss|lunch is now known as tvoss [13:46] om26er, please see the bug report linked on the spreadsheet. Also: the patch does not alter the service's behavior [13:46] jibel, ^ :) [13:47] ogra_: do you know what happened with vivid? [13:48] rsalveti, was something supposed to happen with it ? [13:48] ogra_: well, it failed to build [13:48] E: Unable to locate package libmirclient8driver-android [13:48] E: Unable to locate package libmirplatform4driver-android [13:48] still that hack? [13:49] sigh [13:49] yeah, needs build system changes [13:50] ogra_: right, let me bump that [13:52] actually, there was no mir upload for days, maybe it just got promoted [13:52] maybe component-mismatches? [13:53] oh, no, those are the old names [13:53] there was a mir upload to rtm yesterday ... it should have gone into vivid too [13:53] yeah, you need to fix livecd-rootfs [13:53] $ bzr grep libmir live-build [13:53] live-build/auto/config: add_package install ubuntu-minimal libmirclient8driver-android libmirplatform4driver-android ubuntu-touch [13:53] live-build/ubuntu-touch/hooks/80-enable-libhybris.chroot:if dpkg -l libmirclient8driver-android 2>/dev/null|grep -q ^ii; then [13:53] yup [13:53] live-build/ubuntu-touch/hooks/80-enable-libhybris.chroot:if dpkg -l libmirplatform4driver-android 2>/dev/null|grep -q ^ii; then [13:54] right, but the upload for vivid happend at 8th [13:54] is there a mir ABI bump checklist that this could go on? there are several things that seem to be forgotten nearly every time [13:54] rsalveti, if you touch it anyway, drop that ugly check in the .chroot script [13:54] ogra_: sure [13:55] yeah, we need a checklist [13:55] camako: ^ [13:55] rsalveti: I processed NBS binaries yesterday: https://launchpad.net/ubuntu/vivid/armhf/libmirclient8driver-android [13:55] the old not-built-from-source binary would have been hanging around in the archive until then [13:55] right, that explains then [13:57] yeah, package got renamed [13:57] that was a big change [13:57] -Package: libmirclient8driver-android [13:57] +Package: mir-client-platform-android [13:57] wonder why nobody from touch was involved in that landing [13:58] most likely because it was prepared last year ... lying around til after holidays and then picked up again and touch people were forgotten [14:00] ogra_: why dropping that from .chroot? [14:01] only the check [14:01] not the command it wraps, just that dpkg -l code ... it is pointless [14:02] needed for libmirplatform5driver-android at least [14:02] right [14:03] argh, will need to brb, family around, need to have lunch at the right time [14:03] will fix this once back [14:04] pstolowski, sil2100: eh? [14:06] dobey, ? [14:06] the apparmor changes were to apparmor-easyprof-ubuntu and were only to allow more access to the sdcard for the camera and gallery in reserved policy groups. there is nothing in there that would've caused scopes to fail === jdstrand_ is now known as jdstrand [14:07] there was a custom tarball landing that incorporated those changes to fast boot. I can't comment on what else was in there [14:07] pstolowski: you and sil2100 mentioned me earlier. about your branch i guess? [14:07] s/to fast/for fast/ [14:09] dobey: all is cool [14:09] dobey, yes, we needed top-approval [14:09] seb128: it looks ok to me now [14:09] jdstrand, right, perhaps a scope falls over *because* it can access the SD card now :) [14:09] dobey, the silo was tested alreadt by om26er [14:10] well, they had access before, the only change was being able to readdir /media// [14:10] sil2100, seems the store scope is broken :/ [14:10] yeah. oh. i was pretty sure it was top approved already. anyway, ok. :) [14:10] ogra_: for users as well? [14:10] * sil2100 tries [14:10] sil2100, try it [14:10] but that would only affect scopes that were confined and used these policy groups-- and I know of none [14:11] i only get a wqallpaper and bouncin bottom bar [14:12] ogra_: hmm, I'm running 193 here and the store scope works fine [14:12] Let me experiment a bit [14:13] sil2100, i think victorp sees it too [14:15] sil2100, after reboot it works [14:15] ogra_: really strange, looks fine here [14:15] i upgraded to 192 this morning and to 193 just now ... so it probably only happens after a while [14:17] sil2100, or perhaps a server issue ... [14:17] sil2100: Approved silo 16 [14:18] rvr: \o/ [14:18] ogra_: so, you anyway think it might be the store scope at fault? [14:19] sil2100, see the other channel ... seems there is a server issue === MacSlow|lunch is now known as MacSlow [14:27] sil2100, yeah, I needed to do a watch only rebuild [14:27] thanks [14:29] ogra_, jibel: so I think we can label the scoperunner crash as low impact for now, right? [14:29] yeah [14:46] Ok, I see telephony-service has migrated completely, nothing in the proposed migration is left [14:46] ogra_, jibel: what do you say for building the candidate image now? === charles_ is now known as charles [14:53] tvoss, it timedout for me: "Problem executing the CLI: Wait for fix timed out." [14:53] om26er, sure, that's fine [14:53] FWIW I am sitting near a window [14:54] om26er, it's gps only, might well happen [14:54] om26er, could you pastebin the complete output? [15:01] rsalveti, something wrong with Mir? [15:01] tvoss, it took a long time to scroll in the terminal to copy that text. ~15000 lines. http://paste.ubuntu.com/9748370/ [15:01] camako: we need some sort of checklist when landing mir, because we're currently hardcoding the package name at the script that builds the touch image [15:02] camako: latest update included a package rename and abi bumpa [15:02] *bump [15:02] which broke vivid, which I'm currently fixing [15:02] om26er, thanks [15:03] rsalveti, I see.. bummer... who owns the script? [15:04] camako: it's part of livecd-rootfs, so any core dev should be able to update that if needed [15:04] camako: but just ping either me or ogra_ and should be fine [15:04] until we fix this properly [15:06] rsalveti, I was under the impression that it was fixed. I'll talk to Chris (RAOF). [15:06] the issue is the update alternatives for the android/desktop backend [15:06] to fix this properly we basically need to remove mesa from touch [15:07] rsalveti, gotcha [15:07] ogra_: once you're around, maybe let's kick a new image if jibel has nothing against it [15:07] because we're installing both mir-client-platform-mesa and mir-client-platform-android [15:07] when we should only be installing mir-client-platform-android [15:07] but the dep chain ends up bringing both [15:08] rsalveti, ok I'll chase it [15:15] camako: so, with latest mir we don't need update-alternatives for the client library, right? [15:16] just for the server [15:16] platform actually [15:16] rsalveti yes only platform [15:16] camako: great === plars changed the topic of #ubuntu-ci-eng to: Need a silo or CI Train support? ping trainguards | Need help with something else? ping plars | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: ubuntu-rtm landing gates closed! [15:28] sil2100, waiting for silo 19 [15:28] jibel: oh, wait, so this got approved by Pat? [15:28] sil2100, it was a server issue, according to beuno [15:29] sil2100, no by victorp [15:29] Ah, ok [15:31] om26er: how's testing going? [15:33] sil2100, so far so good [15:33] jibel: in this case it seems we'll indeed end up with an image for the evening meeting [15:33] sil2100, a longer test is running which should take ~30 minutes I believe. [15:34] sil2100, yeah, the estimation this morning was without an additional silo to land today [15:34] tvoss, does the logs look good ? [15:34] tvoss, what causes this messsage: http://paste.ubuntu.com/9748909/ ? [15:35] om26er, it's on purpose, the mode is standalone gps, non assisted [15:42] davmor2, can you try to run one of the tests in bug 1408984 -- see comment#8 [15:42] bug 1408984 in location-service (Ubuntu) "No tool exists to test SNR &TTFF on a handset" [High,In progress] https://launchpad.net/bugs/1408984 [15:42] standalone gps just does not get a fix for me (which is not related to this change). [15:48] om26er: yeap flash has just finished need to get the silo installed [16:01] sil2100, sorry, only returned now ... do you have an image building already ? else i'll kick one now [16:01] ogra_: no no, scratch that [16:02] ogra_: we're still waiting for that location-service silo that product team enforced [16:02] It's in testing right now :) [16:02] oh, the one that wasnt approved this morning ? [16:05] No, we were waiting for Pat, but then victorp approved it :) [16:06] ah, k [16:08] * mterry really wants another image but sees the above [16:11] tvoss, jibel, om26er: so both the command in the description and the command in comment 8 but exit with a first fix time [16:11] _status_update(uint16_t, void*): status=4, context=0x1432d8 [16:11] Mean time to first fix in [ms]: 58745 [16:11] Variance in time to first fix in [ms]: 11578391457 [16:11] ubuntu-location-service start/running, process 9126 [16:12] davmor2, that basically means its doing what its supposed to, I believe. [16:12] om26er: I'd say it works then :) [16:12] and that. [16:13] om26er: is that all you need from me? [16:13] davmor2, yep [16:13] davmor2, thanks :) [16:13] om26er: no worries [16:14] sil2100, ^ [16:14] o/ [16:14] Thanks! [16:15] Publishing, once this migrates we kick a new image [16:15] I suppose the US QA team will only be able to start testing [16:16] sil2100, so new image in ~2h ? [16:22] trainguards can you reconfigure silo 21 for ubuntu-rtm ? [16:22] ^^ L55 [16:22] om26er: yeah... [16:22] alex-abreu: sure [16:22] sil2100, thank you [16:24] alex-abreu: done, remember that unity-webapps-qml is also in your rtm silo 13 [16:25] sil2100, thank you, yes [16:40] ogra_: ok, rmadison seems happy (britney as well) [16:40] ogra_: let's kick teh new image! [16:41] kicking [16:41] Thanks :) [16:41] * ogra_ looks for his shoes [16:41] Use your best kicking shoes [16:42] kicked [16:50] Today's meeting will be real quick [16:50] === IMAGE RTM 195 building (started: 20150114-16:50) === [16:52] * davmor2 passes ogra_ so hob nail boots to give it a proper kick [16:54] trainguards just need an rtm silo for line 83, so rickspencer can test a potential solution seems he is special and is able to repro [16:58] kgunn: k [16:59] kgunn: so it's ready for assignment? [17:00] sil2100, as usual, snappy meeting running over, i'll be late [17:00] sil2100: yep [17:01] ogra_: ok [17:13] * ogra_ pokes google with a pointy stick [17:35] What is the usual build time for a new image? [17:36] mterry: 1 to 2 hours as far as I know. [17:38] * mterry hugs robru [17:38] and goes to the gym while he waits [17:38] We should have a new image in ~1h [17:38] ;) [17:38] * robru blushes [17:39] trainguards ↑ please :) [17:39] robru, thanks for yesterday, and sorry about all the fuss [17:39] Saviq: no worries, that's my job ;-) [17:40] Saviq: just need you to approve https://code.launchpad.net/~mir-team/qtmir/gles-sync/+merge/246339 [17:40] robru, done [17:41] robru: btw. since I didn't have time to look into that today - but did you notice that when ci-train-bot commits to branches, the user seems to be b0rken? Consider for instance this branch: https://code.launchpad.net/~indicator-applet-developers/qmenumodel/trunk [17:41] robru: looks like some variable is not set properly [17:42] sil2100: yeah I fixed that a couple times already but somehow my fix doesn't seem to be sticking. not sure what keeps resetting that :-/ [17:42] sil2100: do you have an example from within the last week? [17:43] hmmm, I remember seeing that once yesterday as well, but can't remember the project it was in [17:43] Let me dig a little bit [17:43] sil2100: https://ci-train.ubuntu.com/job/cyphermox-test/527/console actually it seems fine currently [17:43] robru: maybe those are all just leftovers for now [17:44] sil2100: one thing about the staging instance is that it comes with a whole new charm, and I'm reasonably confident that the charm sets this correctly, so once we get a new production deployment it should be fixed permanently. just not sure how the existing charm in the existing deployment keeps breaking this. === alan_g is now known as alan_g|EOD [18:43] davmor2, so you happy to sign off the silo? === plars 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: ubuntu-rtm landing gates closed! === chihchun_afk is now known as chihchun [19:01] tvoss: 19 went ages ago [19:02] tvoss: was om26er to sign off [19:02] davmor2, cool,thx :) [19:03] ogra_: is everything alright with the build I'd of thought it would of appeared by now [19:10] davmor2, just checking [19:11] the rootfs seems to have built https://launchpad.net/~ubuntu-cdimage/+livefs/ubuntu-rtm/14.09/ubuntu-touch [19:12] ogra_: great so just taking its time then right? [19:12] the importer seems to run ... hmm [19:22] davmor2, there were issues with the importer that stephane fixed only 20min ago ... should pop out the image soon [19:23] ogra_: nice thanks for the update [19:23] * ogra_ vanishes again [19:23] *fnop* [19:25] === IMAGE 65 DONE (finished: 20150114-19:25) === [19:25] === changelog: http://people.canonical.com/~ogra/touch-image-stats/65.changes === [19:25] === IMAGE RTM 195 DONE (finished: 20150114-19:25) === [19:25] === changelog: http://people.canonical.com/~ogra/touch-image-stats/rtm/195.changes === [19:26] \o/ [19:32] ping cihelp: Hi! I'd like to have lp:pay-ui autopilot tests being run by jenkins, or some other bot [19:39] alecu, let me look into that [19:40] great === fginther changed the topic of #ubuntu-ci-eng to: Need a silo or CI Train support? ping trainguards | Need help with something else? ping fginther | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: ubuntu-rtm landing gates closed! [20:27] kenvandine: bit of an issue for me on mako and krillin on the latest update http://paste.ubuntu.com/9751834/ the updates page is blank when I rewind the revision 10 [20:28] Saviq: http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html#unity8 seems to be a problem with your unity8 migration from silo 18. [20:28] kenvandine: I'll write up a bug ToyKeeper has just confirmed it [20:32] Hi! mardy [20:33] davmor2, blank? or just spinning checking? [20:34] kenvandine: https://bugs.launchpad.net/ubuntu/+source/ubuntu-system-settings/+bug/1410968 see screenshot in bug [20:34] Launchpad bug 1410968 in ubuntu-system-settings (Ubuntu) "In image 195 krillin 163 mako system settings is not showing updates" [Undecided,New] [20:35] kenvandine: 3 qa see it on 4 devices [20:39] nt.qml:31 module "Ubuntu.Connectivity" is not installed^M [20:39] doh! [20:39] davmor2, looks like something was removed in that image [20:39] yes, it was dropped from the image [20:39] i guess settings really needs to depend on it [20:39] anyone know why it was dropped? [20:40] jfunk: ^ [20:41] kenvandine: system settings wasn't imported from vivid was it? if so it might be version incompatibility maybe? [20:41] davmor2: ack [20:41] davmor2, nope [20:41] http://people.canonical.com/~ogra/touch-image-stats/rtm/195.changes [20:41] the package was dropped [20:42] i'm adding the depends now [20:48] ogra_, slangasek: image is bad a user will not be able to update from it till kenvandine fix is back into the system is there a way we can knock the ota on the head for this image so users don't get it? [20:48] i wonder why it was removed, but we should have had a depends for it anyway [20:49] kenvandine: it hates you? Hate it back it works for me ;) [20:49] davmor2, oh... it also broke telegram [20:49] that package is part of the platform [20:49] click apps depend on it existing [20:50] click packages can't depend on it [20:50] me adding the depends is a good idea, but we have apps using it as well which can't do the same [20:50] kenvandine: man that sucks I wonder if it was a seed change, I know there was an issue with the build it took longer than normal :( [20:51] maybew [20:52] i'm not seeing a seed change [20:52] ogra_, ^^ thoughts? [20:56] kenvandine, it was removed in unity8 silo [20:57] i.e. the dependency was removed from unity8 which I guess was pulling it into the image. [20:58] Saviq, ^ [20:59] davmor2: I'm not in the loop, what image are you talking about? [21:01] image 195 dropped a package that is part of the platform [21:01] the connectivity api [21:01] slangasek: last rtm proposed image krillin 195 mako 163 krillin.es 144 emulator 157 (but I guess that one doesn't matter so much.) [21:01] broke settings and telegram [21:01] and probably other click apps too [21:01] davmor2, i have the dep change building in rtm silo 3 [21:01] davmor2: ok. So we should roll back to the previous image, for all devices? [21:02] slangasek: I think there is an ota tag that cna stop people upgrading to it. [21:02] I don't know anything about that [21:03] my understanding is that rollbacks on the channel are handled by re-publishing the previous image [21:03] slangasek: rollbacks work too [21:03] kenvandine: looks like unity8 is to blame ← Saviq it's your fault :) [21:04] i added the dep to settings [21:04] but it should probably be seeded as part of the framework [21:04] kenvandine: indeed [21:05] kenvandine, lets fix the seed [21:05] ogra_, can you seed it? [21:05] davmor2: I don't see an ubuntu-touch/ubuntu-rtm/14.09-proposed/krillin.es channel [21:05] i'd like to have the package depends in there too, since settings is a debian package [21:05] oh, it's 14.09.es-proposed, hum [21:05] i'm anal that way :) [21:06] slangasek: indeed thanks sorry I couldn't think of the exact channel off the top of my head :) [21:07] pmcgowan, i added an ubuntu-touch task too [21:09] kenvandine: only issue then is if there is a newer version and yours is updated and the seed isn't and it all fails because of it ;) [21:15] davmor2: rolled back to previous versions for all the devices you mentioned; should be publishing shortly [21:15] slangasek: thanks [21:15] jfunk: ^ [21:16] oh jeez, everything's my fault again? [21:17] who forgot to depend on the connectivity module? [21:17] Saviq, aren't you used to it? :-D [21:17] Saviq, me... and of course some click apps :) [21:17] Saviq, it's know to have broken settings and telegram [21:17] Saviq: not everything only critical breakages of the system that lie squarely at your upload :P [21:18] kenvandine, sounds like it should be part of the framework then, if apps use it? [21:18] Saviq, yeah... it should be seeded [21:18] Saviq: this is the conversation we were just having :) [21:18] Saviq, most importantly it broke updating :) [21:18] yeah I can see that [21:19] so very hard to fet a fix [21:19] kenvandine, yay :| [21:19] s/fet/get [21:19] Saviq: remind me to buy you a beer at the next sprint ;) [21:20] davmor2, I'll remember to grep through / whenever I drop a dependency next time :P [21:20] robru, "unexpected error: test dependencies are unsatisfiable" [21:21] night all thanks for the quick responses kenvandine and slangasek [21:22] Saviq: yeah i dunno, i was hoping you knew ;-) [21:22] alecu, we got an adt failure on -scope-click after a unity8 upload https://jenkins.qa.ubuntu.com/job/vivid-adt-unity-scope-click/lastBuild/ARCH=amd64,label=adt/ [21:22] alecu, apparently test deps are uninstallable, any idea? [21:23] * alecu looks [21:23] if only it said more [21:23] I wonder if proposed got broken, /me adds to chroot [21:24] no idea about this... [21:24] dobey: can you take a look at that log too? ^ [21:24] sure [21:25] IIUC this would mean that B-D for unity-scope-click would be uninstallable in proposed [21:25] there are a bunch of broken deps in that log: [21:26] eg: Broken unity-scope-click:amd64 Depends on account-plugin-ubuntuone [ amd64 ] < none -> 14.04+14.10.20140910 > ( universe/libs ) [21:26] Saviq: no, it's the binary packagses that aren't installing [21:26] Saviq: build-depends installed as it actually built [21:26] dobey, right [21:27] so seems like maybe archive is broken? [21:27] ie, bad timing? [21:27] robru, i'm thinking we need to add a line or two of code to citrain to remove dropped dependencies. we keep on missing these things in silo testing because citrain doesn't do that [21:27] dobey, could be [21:28] Saviq: it looks like ubuntu-system-settings-online-accounts might have borked something [21:28] it failed twice already https://jenkins.qa.ubuntu.com/job/vivid-adt-unity-scope-click/ [21:28] can anyone test silo 3 for me when it finishes building? just verify it pulls in connectivity module back in as a dep? [21:28] i'm going to need to leave in 20m or so [21:28] and it won't be done building by then [21:29] huh [21:29] Saviq: well, the scope hasn't changed in vivid in a month... [21:29] if not i can test it when get back, but didn't want to hold up the fix === fginther 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: ubuntu-rtm landing gates closed! [21:31] dobey, yeah, not saying it's the scope's fault, rather that it seems to be an actual issue, not intermittent [21:31] but why does adt-run install all those packages and then give up with "oh btw, that was a huge waste of time because we can't actually resolve deps" [21:32] hmm, i can't really tell what is going on there [21:32] yeah me neither, really [21:34] trying to build u-s-click with proposed here, let's see if I find any issues [21:39] well, it builds fine according to the log. the problem is installing the packages (i'm not entirely sure how adt does that exactly) [21:40] yeah, I want to try and install the built packages [21:41] but... http://pastebin.ubuntu.com/9752191/ ← dobey [21:41] no build :/ [21:41] huh [21:41] but why did it build in the adt run then? [21:42] yeah, good question [21:43] [----------] 6 tests from BootstrapTest (19 ms total) [21:43] in the adt log [21:43] so that code built and passed the tests :) [21:45] it's passing here too [21:45] (on my desktop, still utopic) [21:46] ah, smart, you have a dummy adt test there to just verify the package builds :P [21:46] alecu, yeah, no question there, really, just it failed to build for me for some reason [21:46] yes [21:47] interesting, I think DEB_BUILD_OPTIONS="parallel=5" caused this? it seems to be building fine now... ¿? [21:47] hmm [21:47] weird [21:47] brendand: sorry i didn't understand. What is the problem? You mean if a package drops a dep for something it still really needs? [21:49] * dobey tries adt-run locally on vivid [21:52] dobey, yeah, looks like unity-scope-click and -autopilot are uninstallable on proposed [21:52] trying to find out why [21:55] oh yay, circular dep [21:56] oh because we depend on unity8 and unity8 declares it depends on us? [21:57] i think unity8 depending on specific scopes is wrong [21:57] dobey, actually no, we only Recommend [21:57] oh [21:58] dobey, but yeah, can't install unity-scope-click from proposed [21:58] that wasn't added to unity8 today was it? [21:58] qtdeclarative5-ofono0.2 : Breaks: ubuntu-system-settings (< 0.3+15.04.20150114) but 0.3+15.04.20150112-0ubuntu1 is to be installed [21:58] oh, that's the problem [21:58] lovely [21:59] so the unity8/scope dep isn't the issue [21:59] no [21:59] that didn't change [21:59] oh hmm, but there is a new settings in proposed [21:59] so it might indeed still be just a timing problem [22:01] yeah, it just resolved itself [22:01] dobey, ↑ [22:01] now who can restart adt tests (that's still alive at this hour) [22:02] hmm [22:02] not sure if i can [22:03] I'm sure I can't [22:03] jibel, you around by any chance? could you restart http://d-jenkins.ubuntu-ci:8080/view/Vivid/view/AutoPkgTest/job/vivid-adt-unity-scope-click/85/ for us? the issue should be resolved now [22:04] thomi maybe? [22:04] he should be starting his thursday by now i think [22:04] soon, yeah [22:05] or robru perhaps? [22:05] robru, I don't think you can deal with adt can you? [22:06] if trainguards can't poke jenkins to re-run adt jobs, i think we should fix that [22:07] trueth [22:07] Saviq: dobey: I can't remember, let me check [22:07] robru, http://d-jenkins.ubuntu-ci:8080/view/Vivid/view/AutoPkgTest/job/vivid-adt-unity-scope-click/85/ [22:07] if you can (re)build this, then you can :) [22:07] Saviq: ah poop, I don't have the right VPN configured, gimme a few minutes to sort that [22:08] robru, there's only one VPN! [22:08] one vpn to rule them all, i thought [22:08] Saviq: yeah I don't have that one set up yet ;-) still hobbling along with the old one [22:19] Saviq: bah I'm on the VPN but the DNS isn't working. can you give me the IP for d-jenkins? [22:19] robru, 10.100.0.2 [22:24] bah! I can access jenkins by the IP but then SSO redirects to the domain so I can't complete the login without DNS working. I'll poke IS about this [22:28] Anyone have an idea where we are for the missing dep fix? [22:28] Saviq,dobey,robru: retrying now [22:28] Saviq: yeah sorry, I can probably normally retry that job for you but right now everything's busted. no #is vanguard, not sure how long it will take to resolve. [22:28] cjwatson: thanks [22:29] thanks cjwatson [22:29] the new VPN is much better, I suggest configuring it :) [22:29] ToyKeeper, it's building in silo 3 IIRC [22:31] cjwatson: yeah I was putting off because I'm busy. trying it now [22:31] Looks like 3 is built. [22:31] ToyKeeper, yeah, that's it, kenvandine had to go [22:32] ToyKeeper, so it needs someone to jump on it and confirm the fix [22:33] Saviq: I just installed silo 3 on the same 195 install where I first saw the issue... and it's not working. [22:34] I'll try a reflash to be sure, but so far it doesn't look good. [22:34] ToyKeeper, did it pull in qml-module-...? [22:36] Saviq: It's hard to tell. I already started reflashing, and our silo tools aren't verbose enough to show apt's output. :( [22:36] Saviq: I don't suppose you know the IP of the dns server that resolves *.ubuntu-ci do you? [22:36] robru, it should get in with the vpn config [22:36] Saviq: yeah I would've thought, but not working here [22:37] robru, like I don't have anything special set up, let me dig in syslog [22:37] robru, 10.172.192.1 [22:37] Saviq: dunno, this happened to me before, last time I set up the old VPN, had to do something special to make *.ubuntu-ci resolve. can't find the wiki page for that though [22:38] NetworkManager[1148]: Internal DNS: 10.172.192.1 [22:38] robru, do you have something similar in your syslog? [22:38] when connecting to the new VPN [22:39] Saviq: yeah I do, now that I configured it by hand ;-) [22:40] robru, there's a bunch of domains it resolves for, you might be missing stuff if you do by hand, are you using the NM GUI to configure the VPN? [22:41] Saviq: yes [22:44] robru, dunno, WFM :P [22:45] Saviq: With 197 + silo 3, I see lots of qml-module-* packages... but the settings/updates page is still completely blank. [22:46] I also see the welcome wizard checkbox for HERE has returned, so I think things have improved... but not all the way yet. [22:49] kenvandine: When you return, it doesn't appear that silo 3 is a complete fix. :( [22:50] ToyKeeper, I'm not in on the details [22:51] Saviq: In 195, the Updates page stopped working in System Settings... totally blank, and made OTA updates impossible. The theory is that it was caused by missing deps. [22:52] I think the deps are pulled in now, but Updates still doesn't work. I don't know why. [23:04] ToyKeeper, trying here, got the problem reproduced [23:04] installing silo [23:05] ToyKeeper, were you using `citrain` to install the silo? [23:05] ToyKeeper: do you know if mako 161 was broken in that respect too? [23:05] dobey, 163 was the first according to davmor2's email [23:06] Saviq: i just saw that e-mail, but my phone is also not showing an image update available, though i see 165 in the server tree. :-/ [23:06] dobey, I think that's because it was removed from the index [23:06] dobey, to not update people to the broken image [23:07] Saviq: right, but i don't want to update to 163. 162 and 165 are both in the archive though (this is actually on a hammerhead, but the image builds for it mirror the mako images) [23:08] dobey, there's no fixed image yet [23:08] dobey, so IIUC your only upgrade would be to a broken image [23:08] dobey, so that's prevented [23:08] oh [23:09] ToyKeeper, everything's fine, just citrain can't install the packages from silo 3 because they require a package that isn't in the silo, an "apt update" and "apt dist-upgrade" after citrain helps, though [23:09] Saviq: ok, I fixed DNS, I can totally rerun that job now ;-) [23:09] :) [23:09] robru, do you get the dns from vpn now or manual? [23:10] yay, it published [23:10] dave's mail said "it can be resolved from a pc to updated" which implied to me that a newer version fixed it, but if you have the broken image, you have to flash over usb to upgrade [23:10] Saviq: well I left my manual config in place, I should double-check that it works automaticaly now. the problem was that I had both 'dnsmasq' and 'dnsmasq-base' installed (because I have no idea what I'm doing), uninstalling dnsmasq made everything magicallystart working [23:10] dobey, yeah, just there's nothing non-broken to flash to, yet [23:11] :) [23:12] Saviq: yep, seems to be working even without manual dns config. great ;-) [23:22] oh well, past time to go anyway [23:26] ToyKeeper, as far as I can tell, silo 3 is good to go [23:47] Saviq: Sorry, had to relocate. Yes, I used citrain. Looks like you found the answer about mako already. [23:48] ToyKeeper, so yeah, citrain can't deal with silos like that because installing the silo requires access to the archive, which citrain disables temporarily (blame robru) [23:48] Saviq: Thanks for the info on out-of-silo packages. [23:48] Disabling the archive temporarily usually fixes a lot more issues than it causes, but... it's still error-prone. [23:49] Saviq: what? it does that on purpose because the whole point of it is that you're testing the silo, not random updates from distro. [23:49] robru, I know ;) [23:49] Saviq: it sounds like whatever dep you needed should have been in the silo [23:49] robru, wha? do you mean we should be uploading all deps (even if they're in archive already) to the silos? [23:50] Not if it didn't need a rebuild... [23:50] Saviq: not all deps, just ones where the archive version isn't new enough [23:50] I think the issue here is more general. Our whole silo infrastructure is bad at handling dep changes... and it's a really difficult issue to fix. [23:50] Saviq: or rather the image version [23:51] robru, the archive version *is* new enough, but this is a case of completely new deps [23:51] robru, that were not in the image before the silo [23:51] Saviq: oh, ok. yeah that wouldn't work ;-) [23:51] robru, no, I totally understand why it does that, we just need to think about improving that case, at the very least fail with a clear error message [23:52] Saviq: well in theory one day the ci airline will provide images built for every silo, and then you just install the image. that'd certainly resolve this particular issue, but who knows how far away that is. [23:52] robru, fginther does! [23:52] * Saviq hides [23:53] Saviq: BTW, which exact package(s) need to be grabbed from the archive? [23:53] Saviq: lemme put it this way... spreadsheet replacement just got deprioritized. so now we're waiting even longer just for the spreadsheet to get replaced. who knows how much longer after that to replace the whole train. [23:54] robru, oh, wasn't it almost ready back in DC? [23:55] ToyKeeper, libconnectivity-qt1 qml-module-ubuntu-connectivity [23:55] Saviq: yeah, the spreadsheet replacement is about 99% ready as far as I can see, but management decided dep8 was more important, so now spreadsheet replacement is on hold [23:55] Saviq: Hmm, "apt-get update ; apt-get upgrade" didn't mention either of those after installing silo 3. [23:56] ToyKeeper, `apt-cache policy qml-module-ubuntu-connectivity`? [23:57] qml-module-ubuntu-connectivity: [23:57] Installed: (none) [23:57] Candidate: (none) [23:57] Version table: [23:57] So... maybe citrain left my sources broken.