[03:25] robru: ping [03:25] michi: hey hey hey waddya say? [03:25] Hi [03:26] Hey, I have an absolutely trivial change in silo 29, ticket 1374 [03:26] There is no point in testing or QA. [03:26] Can we short-cut that? [03:27] michi: indeed that looks quite trivial. [03:27] It broke on yakkety [03:27] michi: not really my call though, you should ask qa to just rubberstamp it [03:27] Because of the xenial test in teh script. [03:27] OK [03:27] Thanks! [03:27] michi: you're welcome [03:27] michi: also that silo won't fix yakkety, only xenial+vivid [03:28] I know. [03:28] Someone requested a no-change rebuild of the cache, which is how I found out. [03:28] https://launchpad.net/ubuntu/+source/persistent-cache-cpp/1.0.4+16.04.20160117-0ubuntu2 [03:28] You can see the failure in the build logs there. [03:29] It’s not urgent for myself. [03:29] So, if it goes through the normal QA thing, that’s OK too. [03:29] I just thought we could save some people some additional work. [03:30] michi: well, if you target yakkety in the silo, it doesn't need qa. qa is only for duals. [03:30] michi: that broken rebuild is in yakkety [03:30] Yes. [03:31] I can rebuild for yakkety too. It’s just that, to get into trunk, it really needs to go through the dual-landing build. [03:31] Hmmm... [03:31] When I select yakkety in the train, it gives me a warning about an SRU. [03:31] Shall I ignore that? [03:32] michi: yeah sorry there's a branch that fixes that but it's blocked for other reasons. [03:32] NP [03:32] I should split that up [03:32] Building for yakkety now. [03:32] I guess, with that, I can just set QA to “not required”? [03:33] michi: it should switch automatically with the yakkety build [03:33] Sweet! [04:08] robru: What’s with the changelog problem? That version is indeed in the changelog. [04:09] michi: no it isn't. the manual upload at dest was never synced to trunk so when you made a branch and built it it doesn't include that changelog entry [04:09] Hmmm... [04:09] So, how do I fix this? Build again for vivid+xenial? [04:11] michi: http://bazaar.launchpad.net/~ci-train-bot/persistent-cache-cpp/persistent-cache-cpp-ubuntu-yakkety-landing-029/view/head:/debian/changelog [04:11] michi: what is there to fix? there's no harm in publishing a package that clobbers a no-change rebuild changelog entry, it's meaningless. [04:11] michi: just drop your debian/changelog changes and build it and publish it [04:12] Well, it won’t hurt to have the changelog entry. [04:13] michi: ok, then copy the debian/changelog from distro over the one in your branch and let the train generate the new entry for you. [04:13] I just uncommitted. [04:13] I guess that’s easier after all. [04:13] Thanks for your help! [04:14] michi: you're welcome! [04:15] michi: your MP still touches debian/changelog, that's probably wrong. [04:15] I just pushed with overwrite. [04:16] Sigh... [04:16] michi: I'm still seeing this: http://bazaar.launchpad.net/~michihenning/persistent-cache-cpp/merge-devel/revision/26?start_revid=26#debian/changelog [04:16] Me too :( [04:16] michi: just uncommit and push --overwrite [04:18] I forgot the revert after the uncommit. [04:18] OK, should be good now. [04:18] I’ll kick the build === chihchun is now known as chihchun_afk [04:22] looks good [05:25] michi pls [05:25] ? [05:25] I just don’t get it. [05:26] I can’t publish without the change because the upload at destination. [05:26] michi: what are you even trying to do with this changelog entry? it's so wrong in so many ways. [05:26] There was a change to the code. What’s wrong about having a changelog entry for that? [05:27] michi: because the train generates that for you. the one you wrote does not do what you are thinking it does. [05:27] So why is it bitching at me then? [05:27] About the unexpected upload? [05:27] michi: first of all, putting it as UNRELEASED means the train will munge your version number when it releases it. so it doesn't solve the issue of that version number missing from the changelog. [05:28] Yes, it then create a new version, right? [05:28] michi: second of all, if you fix UNRELEASED to 'yakkety', it's a lie, because your upload now does not correspond to that version number [05:29] So, previously, whenever we landed something on trunk via the train, we had an unreleased changelog entry. [05:29] michi: I guess you didn't look at this last time I linked it. this is what is in the silo: http://bazaar.launchpad.net/~ci-train-bot/persistent-cache-cpp/persistent-cache-cpp-ubuntu-yakkety-landing-029/view/head:/debian/changelog [05:29] That then got rewritten, adn things worked just fine. [05:29] Now I’m doing the same thing and it isn’t working. [05:30] michi: it is working, in the sense that it's doing exactly what you're telling it to do. you're just doing the wrong thing. [05:30] OK, I need to know what to do then. [05:30] Because, without the changelog change, it doesn’t work, and with it, it doesn’t work either. [05:30] michi: you need to take the changelog that is already in yakkety, copy that overtop of your debian/changelog (just totally clobber the one you wrote by hand) and commit and pusht hat. [05:30] michi: at no point have you yet attempted to include that changelog change. [05:30] Aha. [05:31] So, the problem is that there is still a xenial entry at the top, but for the yakkety build, it wants a yakkety one? [05:32] michi: no, the problem is that when you write UNRELEASED, the train clobbers your version number, so the fake changelog entry with the version number you are fabricating is just ignored. [05:32] michi: if you take the real changelog from yakkety, the train will generate a new entry on top of that and it'll be fine. [05:32] Ah, OK. [05:32] The error messages don’t really make that clear. [05:33] michi: this is standard distro syncing mechanics. any time there's a manual distro upload you are supposed to copy it manually from distro to trunk. [05:33] I’ll try that, thanks. [05:33] I see. [05:33] michi: I'm not really sure how it could be clearer but please do file a bug if you have suggestions. [05:33] So, this happens each time we start a new series. [05:33] michi: no, this happens each time there's a manual upload in the archive that didn't use the train. [05:34] I see. [05:34] michi: the fact that this is the first yakkety upload is a completely unrelated coincidence. [05:34] I didn’t eve know that there was a manual upload. [05:34] michi: yes you did. the no change rebuild that you saw and copied the version number of. that's a manual upload. [05:35] Aha. I live and learn :) [05:35] michi: the train only ever generates version numbers that end in 0ubuntu1. anything else (in this case 0ubuntu2) is a manual upload. [05:37] So, this is what the changelog looks like right now: http://bazaar.launchpad.net/~ci-train-bot/persistent-cache-cpp/persistent-cache-cpp-ubuntu-yakkety-landing-029/view/head:/debian/changelog [05:37] So I’ll use that verbatim [05:39] michi: no, what? that's the one IN the silo. you want the one from yakkety. [05:39] Looking for that... [05:39] Where the hell do I find that? [05:39] Extract it from the yakkety source package, I guess. [05:40] michi: yeah, 'apt-source persistent-cache-cpp' and then copy the changelog from that to your branch [05:40] Yep === chihchun_afk is now known as chihchun [05:40] apt-get source, excuse me [05:40] (1am here) [05:44] Just pulled that. [05:44] It looks the same as our current head without the incorrect changelog entry. [05:45] Top version is s (1.0.4+16.04.20160117-0ubuntu1) xenial [05:46] michi: i can't even [05:46] michi: uncommit what you did and apply this patch: http://launchpadlibrarian.net/257871596/persistent-cache-cpp_1.0.4+16.04.20160117-0ubuntu1_1.0.4+16.04.20160117-0ubuntu2.diff.gz [05:47] OK, will do, thanks! [05:47] you're welcome === Mirv changed the topic of #ubuntu-ci-eng to: Train trouble? ping trainguards | CI problems? Use JenkaaS: http://bit.ly/jenkins-docs | Train: http://bit.ly/1hGZsfS | QA Signoffs: http://bit.ly/1qMAKYd | Known issues: robru sprinting in Texas, public holiday in large portion of Europe [06:23] including me and eg France/Germany based people but apparently not sil2100 [07:06] yay === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [08:41] sil2100, hey, can you please recycle the regression for us https://requests.ci-train.ubuntu.com/static/britney/xenial/landing-071/excuses.html thank you [08:47] Saviq: done [09:24] davmor2, hey, can you mark https://trello.com/c/yhjTyyxt/3121-1078-ubuntu-landing-071-unity-api-unity8-unity-scopes-shell-pstolowski-cimi as unblocked? it's re-running the britney regression (known flaky test) so is good to test again [09:25] Saviq: Done [09:25] rvr, thanks! [09:28] rvr: thanks [11:34] ooh, found a nice unity crash. If you get a notification which has a Yes/No prompt - like "your credentials are outdate, do you want to fix" - press no, all is fine, press yes, unity8 dies [11:37] * popey waits for daisy to process the crash [11:45] https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1578601 [11:45] Ubuntu bug 1578601 in unity8 (Ubuntu) "Crash when choosing "yes" to some popups when locked" [Undecided,New] [11:59] popey: o_O [12:03] yeah [12:41] robru, how I can mark this silo as landed? https://requests.ci-train.ubuntu.com/#/ticket/1362 === _salem is now known as salem_ [13:28] * tedg notices sil2100's ust landing and is excited, as it will make people stop bitching to him about those lttng messages in their logs. [13:35] managed to reproduce https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1578601 again [13:35] Ubuntu bug 1578601 in unity8 (Ubuntu) "Crash when choosing "yes" to some popups when locked" [Undecided,New] [13:35] added screenshot [14:17] sil2100, hey, what is the correct way to mark this silo as landed? https://requests.ci-train.ubuntu.com/#/ticket/1362 [14:18] sil2100, it only contains the click package which was already published on store [14:25] kenvandine: ping [14:25] rvr, pong [14:26] kenvandine: Should System Settings be resizeable? [14:26] no [14:26] that was something that was recently changed [14:26] i think we set the max size [14:26] because the design isn't optimized for a wide screen, etc [14:27] kenvandine: I see it can only be maximized [14:27] for now [14:27] i think that'll change eventually [14:27] something in the shell needs to change to support the max stuff [14:28] Ok, so current behavior is a feature and not a bug. [14:28] yes [14:28] Great, wasn't sure [14:28] i think once the shell handles it, when you maximize it the window will grow to the max size [14:28] or maybe the maximize control won't show, not sure what the plan is :) [14:29] but that's all outside of settings [14:29] renatu: hey! [14:30] renatu: hm, there's no easy way, but let me try something [14:30] sil2100, i tried to publish it even thought there was no package or MR [14:30] it didn't like that :) [14:30] i had hoped that it would just mark it as landed [14:32] hah, no, that wouldn't work sadly ;) [14:43] ChrisTownsend: Silo 31 approved [14:48] renatu: sil2100 kenvandine: clicks are an abomination unto the train and I've abandoned the request in order to free the PPA. This should have never been assigned and there's no way currently to mark click tickets as landed. [14:50] robru, i expected as much [14:51] kenvandine: clicks are on my list to completely redesign the ux for but unfortunately it's low on the list, after git and parallelism [14:52] robru, i'd like to see publish for clicks upload to the store :) [14:54] kenvandine: file a bug i guess? Clicks are a totally unconsidered afterthought bolted on for no particular reason. Clicks would probably be better served by people creating trello cards directly, except that it's also my intention to eventually roll trello board into bileto one day too [14:55] But that's even further down the line. [14:58] rvr: Awesome, thanks! [14:59] ChrisTownsend: And congrats for your spotlight award :) [14:59] rvr: thanks:) [15:49] pmcgowan, robru, sil2100, just realized one big drawback of not landing into Y - bugs don't get auto-closed now on release :S [15:50] Saviq, ah that is unfortunate, we will start soonish [15:50] robru, is waiting for the go ahead [15:51] Indeed [15:52] If everyone's ready for dealing with tripple landings already then we can switch that on [15:52] I thought of maybe waiting at least till OTA-11 is done [15:53] Since triple landings will mean additional effort for everyone [15:53] (and I like to have some spare time in case the triple landing bits have some issues) [15:53] sil2100, maybe once we snapshot, sometime next week? [15:54] I suppose it should be safeish already, velocity should be low [15:54] pmcgowan: let's coordinate it tomorrow on the status meeting then [15:54] :) [15:54] Since I'd need to batch copy what we have in xenial-overlay to yakkety once we enable triples [15:58] trainguards: Hi! Do you have the powers to publish silo 31^^^ Or do I need to find someone else? [15:58] ChrisTownsend: let me take a look [15:58] sil2100: Thanks! [15:59] ChrisTownsend: yeah, manual package uploads need either a trainguard or a core-dev for publishing, on it now [15:59] sil2100: Ok, great, much appreciated! [16:00] yw! [16:09] pmcgowan: sil2100: I'm going to poke at this a little today, as the train is showing SRU warnings against yakkety landings, so I'll clean that up and make sure the trio landings are more ready to go than they already are while I'm at it [16:11] great [16:54] sil2100, would you be so kind and publish https://requests.ci-train.ubuntu.com/#/ticket/1078 [16:58] kenvandine, or you ↑↑ [16:59] since mterry's hiding behind network issues [16:59] sure [17:14] kenvandine, thanks! [17:15] saviqnp [17:15] lol :) === marcusto_ is now known as marcustomlinson [18:20] sil2100, jibel: fyi, the staging-proposed channels were created but the devel-proposed channels had not been moved to import from yakkety; doing this now === josepht` is now known as josepht === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [21:19] jgdx: I think this is ready to land, but the automated signoff failed. What's the story there? https://requests.ci-train.ubuntu.com/#/ticket/1314 === salem_ is now known as _salem