/srv/irclogs.ubuntu.com/2015/07/29/#ubuntu-release.txt

infinityLogan: No, but it seems someone else just did.00:16
Loganah, but I wanted the infinity stamp on it :(00:17
RAOFLogan: Trusty's done; would you kindly add the bug reference to vivid?00:17
Loganoh, did I not nominate it for Vivid? awks00:18
RAOFNo, you didn't include the bug in the changelog.00:18
LoganOH00:18
Logando I need to bump the version?00:18
infinityNo.00:18
RAOFNo.00:18
infinityRejected versions aren't "used up".00:18
LoganI figured00:18
LoganRAOF: ^ thanks!00:22
infinityRAOF: Probably poor form to accept mesa-lts-vivid/trusty, but not the matching mesa/vivid.06:02
RAOFinfinity: Hm. Didn't notice that one.06:04
infinityRAOF: For HWE backports, I tend to verify that the origin exists in the archive before accepting the backport.06:04
infinityRAOF: Since it's a bit backwards the other way. :)06:04
RAOF:)06:04
infinityRAOF: (pls fix, since you have context from the first review)06:05
RAOFDoing so now.06:05
jamespageif there is a sru team member around the horizon in trusty proposed addresses a regression introduced with the last update06:19
jamespagehttps://bugs.launchpad.net/ubuntu/+source/horizon/+bug/147641706:19
ubot93Launchpad bug 1476417 in horizon (Ubuntu Trusty) "[SRU] FloatingIpManager in neutron.py missing is_supported method" [Critical,Triaged]06:19
jamespageI'd like to get that out asap as horizon is quite broken in trusty right now06:20
infinityjamespage: Looking.06:20
jamespageinfinity, thankyou06:20
infinityjamespage: Accepted.  If it's critical, feel free to turn around a quick verification with some indication of reasonable regression testing, and we can fasttrack the promotion to updates.06:22
jamespageinfinity, on that now06:23
infinityjamespage: Well, ideally wait for the binaries to spit out. ;)06:23
jamespageindeed06:23
jamespageinfinity, (or any other sru team member): I've worked through the horizon webui as both an end-user and a admin user - fix for bug 1476417 tests out OK and causes no regressions that I can see08:10
ubot93bug 1476417 in horizon (Ubuntu Trusty) "[SRU] FloatingIpManager in neutron.py missing is_supported method" [Critical,Fix committed] https://launchpad.net/bugs/147641708:10
jamespageif we can get that released today much appreciated08:10
Laneyflexiondotorg: You dropped ubuntu-mate-wallpapers-common but stuff depends on it08:23
Laneydidn't notice that, partially my bad08:23
tseliotcan an admin approve nvidia-graphics-drivers-352 and nvidia-graphics-drivers-352-updates in wily NEW, please?08:34
tseliotthey belong in "restricted"08:34
infinitytseliot: I'll be looking at those later, but first you need to fix your nvidia-* trusty SRUs for me.08:41
infinitytseliot: Note the comment on the bug, you got your conflicts/replaces wrong (the old packages provide a different virtual package than the one you replace in the new ones)08:42
tseliotinfinity: ok, let me check that08:42
tseliotinfinity: what bug report did you use for that comment?08:44
infinitytseliot: The one mentioned in the SRUs?08:48
infinitytseliot: https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-346/+bug/146570608:49
ubot93Launchpad bug 1465706 in nvidia-settings (Ubuntu Trusty) "SRU: New upstream releases of nvidia for 14.04.3" [High,In progress]08:49
infinitytseliot: Comment #708:49
infinitytseliot: Easily visually verifiable that nvidia-opencl-icd-331 doesn't provide the package that the newer ones are replacing.08:49
tseliotinfinity: ok, thanks08:56
infinitytseliot: If you can fix that up ASAP, we can re-verify and get it all sorted today.09:05
tseliotinfinity: sure09:10
tseliotinfinity: is it ok if I add opencl-icd to conflicts/replaces and keep the new virtual package?09:25
infinitytseliot: Yeahp.09:26
tseliotinfinity: ok09:26
Laneyflexiondotorg: fixing the LO icon deps in your metapackage, since that's also uninstallable on arm64 and ppc64el09:30
tseliotinfinity: actually, I've just remembered that I removed that conflicts/replaces in 331 to fix LP: #1247736 . My guess is that the user who reported the problem was using an nvidia-opencl-icd-331 (331.38-0ubuntu7 ?) that is older than what is available in trusty updates (331.113-0ubuntu0.0.4). You can see the difference here: http://paste.ubuntu.com/11958577/09:32
ubot93Launchpad bug 1247736 in nvidia-graphics-drivers-304-updates (Ubuntu Trusty) "[SRU] nvidia-opencl-icd-* should not conflicts/replaces on opencl-icd" [Undecided,Confirmed] https://launchpad.net/bugs/124773609:32
infinitytseliot: Hrm?09:38
infinitytseliot: It either need to conflict/replace that virtual package, *or* needs versioned conflict/replaces on the older nvidia-opencl-icd-331{,-updates} packages.09:39
infinitytseliot: If the opencl-icd conflict is a bug, then do the latter.09:39
infinitytseliot: If you're overwriting files, you need to express that.09:40
tseliotinfinity: yes, the latter seems like the only option09:40
infinitytseliot: I assume nvidia-opencl-icd-331{,-updates} >> 340 are actually empty packages?09:41
infinitySo you can just conflict/replace against those (<< 340)09:41
tseliotinfinity: yes, those are empty transitional packages09:42
tseliotgood point09:42
tseliotI introduced that change in 331.38-0ubuntu7.109:42
infinity(<< 340) is easier to type. :P09:43
tseliotright09:43
tseliot:)09:43
infinityAnd also more obviously correct to a human reviewing, like me. :P09:44
tseliotI assumed admins were superhuman beings ;)09:44
tseliotinfinity: something like this should work, right? http://paste.ubuntu.com/11958630/09:48
infinitytseliot: That looks reasonable, yeahp.09:49
infinitytseliot: x however many packages need that fix.09:49
infinity2, 4, I dunno, the nvidia stuff seems to multiply.09:50
tseliotinfinity: right, it's not my fault though ;)09:51
tseliotinfinity: I should debuild with -v to include the diff from the previous release in -proposed too, right?10:03
infinitytseliot: Ideally, yes.10:04
tseliotinfinity: ok, good10:04
tseliotinfinity: I'm going to add unversioned conflict/replaces for 304 and 304 updates, since, apparently, I've never fixed 1247736 in trusty for these packages10:11
flexiondotorginfinity, Are you on duty?10:15
infinityflexiondotorg: I'm barely awake, so I guess it depends on what you're asking.10:16
flexiondotorginfinity, Sleep well :-)10:16
flexiondotorgLaney, So ubuntu-mate-wallpapers-common should not have been dropped. I'll check my sources and see what happened there.10:18
flexiondotorgLaney, Thanks for steering me in the right direction.10:18
Laneynp10:19
flexiondotorgLaney, I've found the issue. ubuntu-mate-common was erroneously removed in a commit.10:34
flexiondotorgLaney, I'll can raise a new debdiff bug for ubuntu-mate-artwork. Should I rev the version in the resubmission?10:34
tseliotinfinity: ok, uploaded10:35
Laneyflexiondotorg: yes, it's already uploaded so we need a new version10:37
flexiondotorgOK, understood. Sorry I messed this up. Any chance you'd be able to sponsor the new upload when I submit it?10:37
Laneysure10:38
flexiondotorgLaney, because I'd really need this in Alpha 2. I appreciate I'll need to issue a rebuild at some point.10:38
flexiondotorgLaney, Thanks very much. I'll go and make the debdiff.10:38
infinitytseliot: Were 346* the only ones with the bug?11:33
infinitytseliot: Both accepted.  Please re-verify, including testing the broken upgrade path you just fixed, so we can get these in.11:39
lordievaderstgraber: You around?12:03
stgraberlordievader: sorta12:04
lordievaderDid you see my message of yesterday?12:04
tseliotinfinity: thanks, I will re-verify that. As for the other packages, 340 has transitional packages for 331, so it should be ok. Installing 340-updates over (non -updates) 331 will probably fail though.12:54
flexiondotorgLaney, sorry for the delay. Here is the ubuntu-mate-artwork debdiff. Thanks for helping.13:24
flexiondotorghttps://bugs.launchpad.net/ubuntu/+source/ubuntu-mate-artwork/+bug/147935213:24
ubot93Launchpad bug 1479352 in ubuntu-mate-artwork (Ubuntu) " ubuntu-mate-artwork 0.4.12 new nelease [debdiff attached] " [Undecided,New]13:24
flexiondotorgdidrocks, If you're not too busy could you sponsor this upload please?14:30
flexiondotorghttps://bugs.launchpad.net/ubuntu/+source/ubuntu-mate-artwork/+bug/147935214:31
ubot93Launchpad bug 1479352 in ubuntu-mate-artwork (Ubuntu) " ubuntu-mate-artwork 0.4.12 new nelease [debdiff attached] " [Undecided,New]14:31
flexiondotorgdidrocks, I really need this so I can rebuild the Ubuntu MATE 15.10 Alpha images.14:31
flexiondotorgdidrocks, Laney is aware of this but I'm guessing he got busy.14:31
didrocksflexiondotorg: hey, I didn't download the package yet, but there is no .install? You copy them in the right dir yourself in debian/rules?14:34
flexiondotorgdidrocks, The .install for ubuntu-mate-artwork-common was present in 0.4.1114:35
didrocksah, you just forgot the declaration in debian/control14:36
flexiondotorgdidrocks, But the entry in debian/control got nuked by an erroneous commit.14:36
flexiondotorgSo, 0.4.12 just reinstates the control entry that should have been there all long.14:36
flexiondotorgdidrocks, Thanks for looking :-)14:36
didrocksflexiondotorg: yw! doing a quick build before uploading :)14:37
flexiondotorgdidrocks, Cool.14:37
Laneygo didrocks14:37
Laneyflexiondotorg: I was having lunch.14:37
flexiondotorgLaney, well I know you're busy.14:37
flexiondotorgLaney, I done some smoke test for Ubuntu MATE.14:37
flexiondotorgLaney, But I want to rebuild and make sure this artwork and meta package stuff is all good too.14:38
Laney'k14:38
flexiondotorgLaney, I assume that ubuntu-mate-meta will automatically promote from proposed when ubuntu-mate-artwork 0.4.12 is in the relese pocket?14:38
Laneylaney@nightingale> rmadison -swily,wily-proposed ubuntu-mate-meta                                                                                                                                    ~ ubuntu-mate-meta | 1.127 | wily/universe | source14:39
Laneylaney@nightingale>                                                                                                                                                                                   ~14:39
Laneyit already did14:39
flexiondotorgLaney, Ah, great.14:40
flexiondotorgLaney, Thanks.14:40
didrocksflexiondotorg: sponsored, and kudos for the trailing comma after the last depends :)14:44
flexiondotorgdidrocks, I found wrap-and-sort :-)14:44
flexiondotorgdidrocks, Thanks for helping out.14:44
didrocksoh, wrap-and-sort is adding the trailing comma? NICE! :)14:44
didrocksyw14:44
flexiondotorgwrap-and-sort -t -a14:44
Laneyhaha14:46
didrocksI swear I didn't add the -t option to it. But that would have been the case if I thought about it |o|14:46
jamespageinfinity, any chance you can release the horizon fixup (bug 1476417) to updates? it LGTM15:49
ubot93bug 1476417 in horizon (Ubuntu Trusty) "[SRU] FloatingIpManager in neutron.py missing is_supported method" [Critical,Fix committed] https://launchpad.net/bugs/147641715:49
jamespagepretty please :-)15:49
flexiondotorgLaney, ubuntu-mate-artwork 0.4.12 is now in the released pocket :-)16:24
flexiondotorgLaney, If I rebuild the iso images will the build pick up that new version? Or is there something else I need to wait for?16:25
Laneyflexiondotorg: If rmadison shows you the new version in wily then it'll get that16:25
* flexiondotorg goes to learn what rmadison is and how to use it.16:26
LaneyI gave you an example line just there ^^^16:26
* flexiondotorg knows rmadison. Whoa.16:27
LaneyIt queries a remote machine to tell you about package versions16:27
flexiondotorgLaney, Thanks. rmadison still shows 0.4.10. So I'll wait.16:27
Laneynod16:28
infinityjamespage: Done.19:25
flexiondotorgRiddell, wxl How is your 15.10 Alpha 2 testing going?19:32
lordievaderflexiondotorg: Speaking for Riddell, good :) tested lots yesterday, no crucial problems found.19:35
flexiondotorglordievader, Excellent. I've found myself in the same happy situation with Ubuntu MATE. One note worthy issue but noth critical :-)19:36
* flexiondotorg makes mental note to contact lordievader tomorrow.19:37
lordievaderI do have one problem though, since the Kubuntu team is away at Akademy I volunteerd to do the lead on this one. But as of yet I do not have the rights to make the image ready... :(19:38
infinitylordievader: You can just tell Laney your images are ready and he can mark them.19:39
lordievaderOke, I will remember that. Going to test some more tommorow. Thanks :)19:40
flexiondotorglordievader, Looks like I could also make you image as ready when you get to that point :-)20:27
lordievaderAh, cool. Then I'll bug one of you two tommorow about it ;)20:34
lordievaderflexiondotorg: ^20:34
flexiondotorglordievader, Sure, no problem :-)20:34
flexiondotorglordievader, I owe Riddell a favour or three20:35
lordievader:)20:37
jderoseinfinity: seems something is a bit off in the *lts-vivid X bits still https://bugs.launchpad.net/system76/+bug/147952421:18
ubot93Launchpad bug 1479524 in totem (Ubuntu) "Totem Crashes at launch from missing libwayland-egl" [Undecided,Confirmed]21:18
jderosethat's with todays 14.04.3 daily ISO, sha1sum 61afb95256980077b41415f2986ee75d2b3222cd21:19
infinityjderose: So something has an undeclared dep?  Lovely.21:19
infinitytjaalton: Around?21:19
infinityjderose: So, (minus version bumps, obviously), this is the delta from .2 to .3: http://paste.ubuntu.com/11962290/21:22
infinityjderose: Pretty clearly missing that one library, but I assumed this was intentional on the part of the X maintainers, so we'll have to sort out why. :P21:23
jderoseinfinity: gotcha. while i have you, do you know why there is no "libegl1-mesa-drivers-lts-vivid" package to replace "libegl1-mesa-drivers-lts-utopic"? is this deliberate or a goof?21:26
infinitySo, looks like libopenvg1-mesa and libegl1-mesa-drivers legitimately no longer exist in the new mesa.21:27
infinityBut libwayland-egl1-mesa-lts-vivid definitely exists, it's just that nothing pulls it in.21:27
infinityjderose: newer mesa Provides mesa-drivers.21:28
infinityjderose: So, I guess whatever was in that package was pulled in.21:29
jderoseinfinity: so something that truly does depend on libwayland-egl1-mesa-lts-vivid hasen't properly declared it as such... and after that's fixed, it will fix the ISO manifest automatically?21:29
jderosegotcha, thanks21:29
infinityAnd it was drivers that depended on wayland, so I guess that dependency needs to move around a bit.21:30
jderose(i don't know much about the ISO building process BTW, so sorry if that's a dumb/obvious question)21:30
jderoseah, gotcha21:30
infinityjderose: Yeah, this isn't something we should hack around in the ISO mastering process, if a package is legit missing a dep, we should fix it. :)21:30
jderoseagreed :)21:30
infinityA bit curious that this wasn't an issue with vivid, though...21:31
infinityUnless the backport broke the deps a bit.21:31
* infinity goes to look at the vivid manifest.21:31
infinityHrm, *something* pulled it in in vivid.21:33
infinityWait.  I've confused myself.21:34
infinityAhh, it's pulled in incidentally in vivid by gtk depending on it.21:38
infinityWhich clearly won't be true in trusty.21:38
infinityjderose: Can you reproduce that bug and verify that installing libwayland-egl1-mesa-lts-vivid fixes it?21:42
jderoseinfinity: sure, just a sec...21:47
jderoseinfinity: yup, `apt-get install libwayland-egl1-mesa-lts-vivid` fixes it21:50
infinityjderose: Alright.  We'll either get it fixed in mesa, or I'll pull some tricks, but we'll figure it out.21:54
jderoseinfinity: awesome, thanks! :)21:55
infinitytjaalton: When you pop your head in, please have a look at https://bugs.launchpad.net/ubuntu/+source/mesa/+bug/147952422:00
ubot93Launchpad bug 1479524 in mesa-lts-vivid (Ubuntu Trusty) "Totem Crashes at launch from missing libwayland-egl" [Undecided,Confirmed]22:00
infinityRAOF: Or you ---^22:02
infinityOh, hrm.  I used to install libegl1-mesa-drivers manually for lts-utopic.  Maybe you'll both tell me I should just install this manually now instead.22:03
* infinity grumps and admits that's probably the answer he'll be given.22:05
infinityBut we'll see.22:05
ari-tczewThere are a lot of packages related to KDE, which have been _not_ yet migrated from -proposed. e.g. kf5unitconversion, is there any blacklist for such packages?22:12
cjwatsonThat seems to fail Occam's Razor22:13
infinityAlso, I don't see kf5unitconversion in proposed...22:13
infinitySo you might need to be a bit more specific.22:13
infinitykunitconversion?22:13
cjwatsonMuch easier to actually look at http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html22:13
ari-tczewinfinity: yes, kunitconversion, sorry22:13
cjwatsonRegression in the kdelibs4support autopkgtest22:14
cjwatsonDunno if that actually has anything to do with kunitconversion, I haven't unpicked that; but it would probably be easier to make the tests pass than to invent conspiracy theories about blacklists :-)22:15
infinitycjwatson: Conspiracies are more fun.22:15
jderoseinfinity: when you say "I used to install libegl1-mesa-drivers manually for lts-utopic", do you mean as per the instructions here - https://wiki.ubuntu.com/Kernel/LTSEnablementStack22:16
cjwatsonThere is the facility to block packages, but if that were in use then you would see it logged on the above page22:16
cjwatsonhttps://wiki.ubuntu.com/ProposedMigration has overall documentation22:17
jderoseinfinity: or do you mean in terms of the ISO mastering22:17
cjwatsonOh, and there are a bunch of other problems, because kunitconversion needs ki18n and that has a couple of other autopkgtest regressions22:17
infinityjderose: I mean for the livefs mastering.22:19
infinitycjwatson: I thought you were sleeping.22:19
cjwatsonI sure am, I type by telepathy.22:20
jderoseinfinity: okay, gotcha. just to make sure i'm understanding things correctly, totem should in fact depend on libwayland-egl1-mesa[-lts-*], either directly or indirectly, right? like if you installed from the server ISO, the expectation is that `apt-get install totem` would do something sane, in theory yield a workable totem install?22:23
infinityjderose: I doubt it's totem itself that's trying to load that library.22:24
infinityjderose: Odds are it's mesa's fault.22:24
infinityjderose: Oh, you did say "or indirectly".  So, yeah, I'd imagine that mesa has an undeclared dependency there because someone was trying to be overly clever with a package split and didn't notice the consequence. :P22:25
infinityjderose: Alternate fix would be to make whatever's dl()opening that library be more conditional and less violently faily.22:26
infinityAnyhow, the band-aid solution might just be for me to pre-install it, as I did for mesa-plugins in .222:26
infinityWe'll see.22:26
infinityjderose: I'll have a healthy argument with the X folks and have it fixed -- one way or another -- by tomorrow.22:27
jderoseinfinity: okay, gotcha. for me the ISO/livefs build process is still much mystery sauce, so i'm just trying to understand things a bit better. thanks for graciously humoring my questions! :D22:27
infinityjderose: In an ideal world where all packages behave, the livefs build process should just be "apt-get install ubuntu-minimal ubuntu-standard ubuntu-desktop" && take package list && "apt-get install ubuntu-live" && take new package list, so we have install/live diff, compress and ship.22:29
infinityjderose: The world isn't always entirely ideal.  But we aim for it. :P22:29
jderoseinfinity: thanks for the pseudo-code there... that hugely, hugely helped me understand the process.22:31
jderoseand it's good to have ideals to strive for, even if you don't always reach them :)22:31
* jderose catches the bus22:51
bdmurrayCould somebody have a look at my apport SRU in the vivid queue?23:03
infinitybdmurray: What's the magic word?23:14
infinitybdmurray: Pretty sure your bug ref is bogus.23:16
bdmurrayinfinity: hmmph, thanks23:28
tewardinfinity: could you please do me a favor and take a look at the nginx sru in the trusty queue?  That's a fairly-easy-to-reproduce bug, and init failing because it can't parse pidfile is *bad*... no rush, though, i'm fine waiting for a few days longer, since I use the nginx PPAs on my trusty systems.  others, unfortunately, are still hit by this.23:30
tewardif you're busy i'll wait :023:30
infinityteward: Can you fix your changelog to remove the trailing " *\n"?23:37
infinityteward: Not that that's a problem, but it's icky and shows a lack of attention to detail. :P23:38
infinityteward: Also, the reference to "debian/control:" seems wrong, that's not the file you changed...23:39
bdmurrayinfinity: I've reuploaded apport.23:39
tewardinfinity: reject, i'll reupload23:41
tewardinfinity: i'll make those fixes :)23:41
tewardehhh, i don't need to increment the version again do I, given that it was rejected...?23:50
* teward may have made a failure if it needs another version number bump >.<23:50
bdmurrayteward: No, you don't need to the bump the version since it was never published.23:50
infinityteward: You d... What he said.23:50
tewardbdmurray: that's what i thought, but i've failed in the past :)23:52
tewardthe headache of staring at ESXi's command line, coupled by manually rebuilding two VMs is draining my brain :)23:52
tewardthat one should look better23:54

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!