[07:35] <Mirv> nooo
[07:36] <Mirv> the bot could shout and wail every time a build fails
[07:38] <Mirv> I guess even smaller GCC5 updates may cause symbol changes
[07:39] <Mirv> oh, no, actually there's a reasonable explanation for this -gles build failure
[07:39] <Mirv> of course
[07:49] <doko> Mirv, where else besides destructors?
[07:59] <Mirv> doko: this was me not basing on what's already in -proposed (since I've a local bzr but the -gles package is not in LP bzr), so I missed the removing of symbols _ZTV20QTextObjectInterface@Base and _ZTVN29QWindowSystemInterfacePrivate10InputEventE@Base
[08:01] <Mirv> Debian has a good overview on what kind of symbols went missing from qtbase for example, in addition to those two http://anonscm.debian.org/cgit/pkg-kde/qt/qtbase.git/log/?h=experimental
[08:01] <Mirv> mostly destructors of course, but some others too
[08:11] <morphis> sil2100, Mirv: can one of you upload me another android pkg to silo 48?
[08:12] <Mirv> morphis: o/
[08:19] <Mirv> I'm still laughing at this https://bugs.launchpad.net/bileto/+bug/1481215
[08:23] <ogra_> sil2100, FYI the i386 images still have the dependency issue
[08:25] <sil2100> ogra_: ok, will look into that, thanks
[08:33] <sil2100> ogra_: first I need to deal with some uninstallables in wily-proposed though
[08:34] <ogra_> sil2100, well, the emulator is broken anyway, i guess there is no hurry :)
[08:34] <sil2100> That's also what I think ;)
[08:34] <sil2100> Still, it needs to be fixed
[08:34] <ogra_> indeed
[08:59] <greyback> sil2100: are we doing vivid+overlay only landings atm?
[09:00] <greyback> I've a dual silo which is ready to test, can the wily packages be just deleted, instead of rebuilding everything
[09:17] <dbarth> sil2100: hey; i'm trying to verify which build of oxide landed in the archive/wily; i don't have the silo on my radar anymore
[09:17] <dbarth> from the build logs, it still says gcc-4.9
[09:17] <dbarth> so what happened to the rebuild in silo 26 yesterday?
[09:28] <Mirv> dbarth: the rebuild didn't happen. now the rebuild is there and ready, I guess sil2100 could publish it.
[09:35] <dbarth> Mirv: hmm, but if i check the buildlogs, there are still mentions of gcc-4.9
[09:35] <dbarth> https://launchpadlibrarian.net/213714289/buildlog_ubuntu-wily-armhf.oxide-qt_1.8.4-0ubuntu2_BUILDING.txt.gz
[09:35] <dbarth> is that expected?
[09:38] <cjwatson> dbarth: That's just updating the base chroot, which still contains gcc-4.9.  Harmless and expected.
[09:39] <cjwatson> You can see it installing gcc-5 as well and upgrading gcc to 4:5.2.1-3ubuntu1.
[09:43] <dbarth> ah perfect, sorry for the false alarm
[09:54] <Mirv> dbarth: right.
[10:38] <greyback> trainguards: are we recommended to land vivid+overlay only at the moment?
[11:00] <Mirv> greyback: I guess the answer depends on whether you feel blocked by your package staying at wily-proposed for a little while still
[11:01] <greyback> Mirv: I wouldn't, as long as I can land some code in vivid+o
[11:01] <Mirv> if it can wait there and you don't immediately have the next landing to be done (so that it'd need to merged to trunk), it's ok to have vivid to vivid+o and wily to wily-proposed until all GCC5 fixes are done. also, we can manually merge+clean the silo if the next landing is needed
[11:02] <Mirv> greyback: so I'd personally continue doing dual landings, but sil2100 can comment on which kind of view he has
[11:03] <Mirv> of course, _testing_ on wily is also "challenging"
[11:03] <ogra_> we should just roll back to gcc-4.9 ... this transition is way to painnful :P
[11:04] <Mirv> ogra_: ok, uploading new gcc-defaults!
[11:04] <Mirv> ;)
[11:04] <ogra_> ;)
[11:04] <Mirv> GCC 2.95 was just fine
[11:05] <cjwatson> 2.96 FTW!
[11:05] <cjwatson> *shudder*
[11:06] <ogra_> haha
[11:07] <ogra_> cjwatson, oh, just FYI, the checking for the PPA sources.list in config didnt work either ... in the end i decided to just rip out the whole sources.list mangling for the android package (which is fine since this livecd-rootfs is PPA only anyway)
[11:09] <cjwatson> ok
[11:49]  * sil2100 off to eat lunch
[12:01] <rvr> Mirv: Approving silo 30
[12:04] <Mirv> rvr: thanks!
[12:05] <Mirv> sil2100: bzoltan_ just managed to remind me that we've a case of lost ubuntu-ui-toolkit from silo 013 in wily... it's not in any queue either
[12:06] <Mirv> sil2100: not Rejected, the rsync line looks correct, version number is higher than the current one. the -gles package did get in.
[12:12] <cjwatson> Mirv: LP logs suggest it didn't even try to copy
[12:12] <sil2100> Mirv: I don't see it as well... I think we need to copy-package
[12:12] <cjwatson> AFAICS anyway
[12:14] <Mirv> sil2100: yeah, I think so too but I thought it'd be interesting enough to mention. I can't copy myself so we can get eg. ogra_ to run ./copy-package --from=~ci-train-ppa-service/ubuntu/landing-013 --from-suite=wily --to=ubuntu --to-suite=wily-proposed -b ubuntu-ui-toolkit
[12:16] <sil2100> Mirv: same here, ogra_ do you have a moment for a copy-package?
[12:17] <cjwatson> https://ci-train.ubuntu.com/job/ubuntu-landing-013-2-publish/68/console seems to only have published to the overlay
[12:20] <sil2100> cjwatson: yes, the publish to the archive should have been done by snakefruit's copy2distro, but something didn't work
[12:20] <cjwatson> Oh, right, I was misreading the publish job
[12:21] <cjwatson> No evidence from LP access logs that it even tried the other copyPackage
[12:21] <cjwatson> sil2100: can I suggest that the publish job should cat the rsync packagelist to stdout in future, to make this easier to debug?
[12:22] <cjwatson> oh, it's an artifact isn't it
[12:22] <cjwatson> https://ci-train.ubuntu.com/job/ubuntu-landing-013-2-publish/68/artifact/packagelist_rsync_landing-013-wily/*view*/  so WTF really
[12:22] <sil2100> cjwatson: it's in artifacts, just look for rsync there
[12:22] <cjwatson> yeah, just getting old
[12:22] <sil2100> It looks okayish, especially that it published its -gles counterpart from the rsync fine
[12:23] <cjwatson> ah
[12:23] <cjwatson> ci-train-ppa-service/ubuntu/landing-013ReleasewilyProposedwilyubuntu-ui-toolkit1.3.1584+15.10.20150730-0ubuntu11.3.1549+15.10.20150710.1-0ubuntu1timo-jyrinkiubuntu
[12:23] <sil2100> So copy2distro parsed it and copied the -gles one, but something went wrong with the normal UITK - we don't have any logs from what I understand
[12:23] <cjwatson> er, with whitespace
[12:23] <cjwatson> distro_version_at_prepare_time will have been 1.3.1549+15.10.20150710.1-0ubuntu1
[12:23] <cjwatson> but look at https://launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/+publishinghistory
[12:23] <cjwatson> so it will have complained about a newer manual upload in the archive
[12:24] <sil2100> Ah, hm, but from what I remember it shouldn't have caused the copy to fail, at least I don't remember it being like that in the past
[12:24] <cjwatson> copy2distro fails in that case
[12:24] <cjwatson>                         'Manual upload of {} {} is newer than {}. '
[12:24] <cjwatson>                         'Not uploading {}.'.format(
[12:24] <cjwatson> etc.
[12:25] <sil2100> Damn, that script needs some love and fixing really
[12:25] <sil2100> geh
[12:25] <sil2100> I think we once got bitten by that before, as this really doesn't make much sense - all checks like these are done on the train side anyway
[12:25] <cjwatson> it maybe shortens race conditions
[12:25] <ogra_> do you still need me to copy ?
[12:25] <cjwatson> but there probably ought to be a way to force it
[12:26] <sil2100> Let me propose a fix for that but in the meantime we need a copy anyway
[12:26] <sil2100> ogra_: yes, if you can :)
[12:26] <cjwatson> anyway I think a forced manual copy is OK in this case, as long as it was built with gcc5
[12:26] <sil2100> cjwatson: I'll double confirm, but I'm pretty sure this silo is from the times when we already landed the new gcc in -proposed
[12:27] <sil2100> ogra_: could you wait with that copy?
[12:27] <ogra_> wow, that was in the last nanosecond :)
[12:28] <Mirv> sil2100: oh fruck, I checked "ok it built on Saturday" but I looked at a wrong month.
[12:28] <sil2100> ogra_: thanks ;) Since I think it might be built with the old gcc...
[12:28] <sil2100> SInce the last build was '7 days ago'
[12:28] <sil2100> Mirv: yeah, needs a rebuild
[12:28] <sil2100> bzoltan_: ^
[12:29] <sil2100> cjwatson: thanks for the help here :)
[12:29] <sil2100> ogra_: phew, don't copy anything, we need a rebuild
[12:29] <cjwatson> np
[12:29] <Mirv> sil2100: I guess I could remove the vivid from the silo since it was already copied, and then they'd sync up version numbers for the next release again?
[12:29]  * ogra_ presses n
[12:29] <ogra_> :)
[12:30] <sil2100> Dodged a bullet here
[12:31] <Mirv> hmm, maybe a cleaner to do a merge+clean and a more manual rebuild, to not diverge the version numbers to much
[12:33] <sil2100> +1
[12:38] <sil2100> cjwatson, Mirv: the good thing is that the -gles packages at least built against gcc-5 already
[12:38] <sil2100> cjwatson, Mirv: this also confused me and made me think we're ok
[12:39] <Mirv> sil2100: also since it's not in archives I can use the same version number, just via different silo
[12:40] <Mirv> thanks everyone for help
[13:02] <ogra_> LOL
[13:02] <ogra_> http://blog.darknedgy.net/technology/2015/08/05/0-androidinit/
[13:05] <pmcgowan> sil2100, ogra_ hey am I correct to assume the nexus 7 build should be up to date
[13:05] <ogra_> wily or vivid ? :)
[13:10] <pmcgowan> ogra_, sorry vivid
[13:11] <ogra_> yeah, should be all up to date
[13:11] <pmcgowan> thanks
[13:11] <pmcgowan> and thanks for fixing the android stuff
[13:11] <ogra_> yeah, turned out to take a lot more than planned :/
[13:17] <sil2100> pmcgowan: yeah, all should be ok, although I still didn't find the time to fix the i386 builds
[14:27] <bzoltan_> sil2100: I see Mirv has pushed my release candidate to the silo17 where it was re-build. Is there anything I should do? Do you want a new QA round? I doubt that any functionality could be effected... but it is not my call :)
[14:28] <bzoltan_> Hmm... I see the landing branch is landed on the UITK trunk. Nice... thanks folks
[14:32] <dobey> hmm
[14:33] <sil2100> Oh, Mirv rebuilt it already?
[14:36] <pstolowski> sil2100, hey, may i ask for purging of ppa 41? having some weird dependency issues there...
[14:36] <sil2100> pstolowski: you want another silo assigned instead?
[14:37] <pstolowski> sil2100, whatever works for you
[14:38] <kenvandine> rvr, great bug reports, thanks!
[14:38] <rvr> kenvandine: You're welcome
[14:39] <rvr> I'm finishing testing the silo
[14:39] <kenvandine> cool, thx
[14:40] <sil2100> pstolowski: ok, let me try that then
[14:46] <sil2100> pstolowski: reassigned to 27
[14:46] <pstolowski> sil2100, thanks!
[14:57] <rvr> kenvandine: Silo 32 approved
[14:57] <kenvandine> rvr, thx
[14:59] <sil2100> kenvandine: you publish your own silos, right? :)
[14:59] <kenvandine> sil2100, already done :)
[14:59] <kenvandine> thx
[15:49] <pstolowski> sil2100, hey, any idea why it failed with missing dependency https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-027/+build/7767198 when it's in the same silo: https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-027 ?
[15:51] <sil2100> pstolowski: it's building?
[15:51] <sil2100> pstolowski: looks like it was a temporary build wait until the binaries got published :)
[15:51] <sil2100> (if I'm looking at the right logs)
[15:52] <pstolowski> sil2100, i've just interrupted
[15:52] <sil2100> Interrupted what?
[15:52] <pstolowski> sil2100, so i was too impatient? ;)
[15:52] <pstolowski> sil2100, interrupted the build
[15:53] <sil2100> pstolowski: ah, on the train? Yeah, probably, sometimes it might take a bit longer for the PPA to notice a dep appearing :)
[15:53] <sil2100> A watch_only build should be fine
[15:53] <pstolowski> sil2100, uh, hmm
[15:55] <sil2100> pstolowski: everything ok now?
[15:56] <pstolowski> sil2100, i'm not sure yet, started watch only build
[15:58] <robru> sil2100: pstolowski: if a package is in "Dependency wait" state, in my experience it takes about 2 hours for the PPA to automatically retry the build. it's  not something you want to hold your breath for...
[16:00] <pstolowski> sil2100, robru i see. allright, i think it's going to be fine, one of the architectures just got the dependency
[16:02] <robru> pstolowski: if you notice the dependency is published (eg it's all green and doesn't say "Pending publication" on the ppa page), we can retry the other depwait builds to speed it up
[16:31]  * sil2100 AFK for a while
[18:02] <rvr> dobey: Silo 11 approved
[18:03] <dobey> rvr: great, thanks
[21:23] <veebers> robru: can you confirm for me that the autopilot release didn't get released? (due to me screwing up the qa acceptance)
[21:23] <robru> veebers: well i didn't publish it. lemme check
[21:24] <robru> veebers: right, victim of bad timing I'm afraid. alesage was making some mistakes and your change happened right around the same time. it didn't correspond with anything in the trello board so we assumed it was a mistake and set it back
[21:25] <veebers> robru: nw, the fact that it wasn't released clears something up. I'll get that sorted today
[21:25] <alesage> robru, veebers yep my fingers need to lose some weight
[21:25] <veebers> alesage: ^_^ lol
[21:25] <veebers> robru: thanks for confirming
[21:26] <robru> veebers: you're welcome. I'm happy to publish it when ready
[21:26] <veebers> robru: I might re-test it now, these issues that brendand had have made me paranoid that something has changed inbetween :-)