/srv/irclogs.ubuntu.com/2016/04/19/#ubuntu-release.txt

slangasekmgz: so... somehow, when I raise an mp, lp tells me you do not have rights to view ~juju-qa/ubuntu/xenial/juju/xenial-2.0-beta4/ ?00:30
slangasekoh, maybe it's angry because I didn't say lp:00:32
mwhudsonslangasek: hi, so you're still poking at this juju autopkgtest thing too?00:44
superm1slangasek: after mythtv migrates from proposed would that be a safe time that a seed change to ship-live would be effective too for a re-spin?01:14
superm1i /think/ the fix for https://bugs.launchpad.net/ubuntu/+source/mythbuntu-meta/+bug/1571781 is just seeding another package that somehow didn't resolve01:14
ubot5`Launchpad bug 1571781 in mythbuntu-meta (Ubuntu) "Nvidia driver not installed" [Undecided,Fix committed]01:14
slangasekmwhudson: not actively poking at the moment; had tried to see if the tests would pass for me in a different setup, to no avail01:41
slangaseksuperm1: why waiting until it migrates before changing the seed?01:45
slangaseksuperm1: I'd generally say it's better to change the seed sooner rather than later, since it doesn't need to block on the mythtv migration and that way you're not holding up any respins01:50
mwhudsonslangasek: ok01:54
superm1slangasek: sorry wasn't clear - the seed change was already done02:09
superm1i wasn't sure how fast migration was going to happen02:09
superm1so i wanted to make sure the respin only happened after the migration was done02:10
slangasekah02:10
slangasekyeah, that should be enough time for it to propagate02:10
infinitysuperm1: When rmadison says it's in the release pocket, that'd be safe.02:14
infinitysuperm1: Don't trust LP's "published" status for this.02:15
cyphermoxflexiondotorg: yes, I'm in Canada02:49
cyphermoxoh crap, please reject that, it wasn't meant for here. ^04:36
Logancyphermox: happens to the best of us :P05:29
Loganfun fact, you can edit your dput.cf so that you must specify a host while dputting05:30
slangasekmwhudson: hmm why does the manual provider care about the test user's shell?05:40
pittislangasek, mwhudson, mgz: I suppose it won't hurt if I run http://bazaar.launchpad.net/~mwhudson/ubuntu/xenial/juju/mwhudson on the production infra?05:54
pitticyphermox: hey, still here? mind reuploading ubiquity with a non-ugly version number? (2.21.59ubuntu1~mtrudel1 → 2.21.60); or was this mis-uploaded and supposed to go into a PPA?06:02
pittiinfinity: I reviewed unapproved stuff that's not on the images; left the three ones that are to avoid stomping on your feet06:06
pittiinfinity: if you want ubiquity, I'd rather reupload with a real version number, but I suppose we want the casper fix at least06:06
pitti(and then take appstream when casper gets accepted)06:06
Loganpitti: [00:36:39]  <cyphermox>oh crap, please reject that, it wasn't meant for here. ^06:08
Logan(re ubiquity)06:08
pittiLogan: ah thanks,  rejected06:08
slangasekmwhudson: ignore the question, I see why ;)06:13
pittislangasek, mwhudson: hm, running tests from that bzr branch still fails, I'll try building the package from that bzr branch too06:14
LocutusOfBorginfinity, hi, can you please accept virtualbox again? the cherry-picked patch has been included in a stable release06:22
LocutusOfBorgor anybody else :)06:31
* LocutusOfBorg leaves, but reads replies on irclogs06:56
mwhudsonpitti: sigh07:29
mwhudsonpitti: there's no diff to speak of between the non-debian/tests code in that branch and what's in -proposed07:30
mwhudsonslangasek: :)07:30
pittimwhudson: ah ok, so building packages indeed wasn't necessary; I just followed up to the bug with the full log07:30
jtayloris there still a chance bug 1571811 gets looked at? even if its just a quick no07:30
ubot5`bug 1571811 in openblas (Ubuntu) "FFe for openblas 2.18" [Undecided,New] https://launchpad.net/bugs/157181107:30
mwhudsonslangasek: the experience of ssh-ing to a user with /bin/false for a shell is pretty confusing07:30
pittimwhudson: does your test rely on lxd already being installed and the user being in the lxd group perhaps?07:30
pittimwhudson: (we run autopkgtests in minimized VMs)07:30
mwhudsonpitti: i ran the tests in whatever adt-buildvm-ubuntu-cloud -r xenial spat out :-)07:31
pittimwhudson: ok, that's minimized, no preinstalled lxd07:31
mwhudsonpitti: but it could be something like that i suppose07:31
pittiunless you use an non-current version of autopkgtset07:31
mwhudsonlxd is listed in the test deps07:32
mwhudsonpitti: i'm up to date on xenia;07:32
pittiif you use current xenial or trusty/wily-backports, lxd should not be present in the testbed07:32
pittimwhudson: oh wait, I mis-read07:33
pittiGet https://10.0.8.1:8443/1.0/profiles: Forbidden07:33
pittithis isn't a problem of local "lxd" group membership and the unix socket, this is over the bridge07:33
pittiso maybe missing remote auth or whatnot07:33
mwhudsonuh yeah07:33
mwhudsonpitti: lxd is installed in the test bed fwiw07:33
mwhudsonpitti: http://paste.ubuntu.com/15925658/07:34
* pitti runs your branch locally in qemu07:35
pittimwhudson: hm, that looks like a non-current version of autopkgtest then07:35
mwhudsonpitti: i have 3.20.3 apparently07:36
pittimwhudson: hm,  * setup-commands/setup-testbed: Purge lxd and lxc.07:38
pittithat was in 3.2007:38
pittimwhudson: when did you build that VM?07:38
mwhudsonpitti: 6 hours ago?07:38
pittihmm, that sounds like worth a bug report07:38
pittibut anyway, at first sight it doesn't actually look like that's the problem07:38
mwhudsoni would agree with that07:39
pittian EPERM on the unix socket could be related to that07:39
mwhudsonso my successful run has 2016-04-19 02:20:15 DEBUG juju.tools.lxdclient client.go:73 connecting to LXD remote "juju-remote": "10.0.8.1"07:40
pittiooh, wait07:40
mwhudsoni don't know what 10.0.8.1 is07:40
pittimwhudson: yeah, it locally passes for me too07:40
mwhudsonpitti: ip conflict or something like that?07:41
pittimwhudson: presumably that's the network you put lxdbr0  on07:41
pittimwhudson: but this sounds like it could be a real IP in the data center07:41
mwhudsonpitti: "you"? :)07:41
pittimwhudson: which curiously is the very reason why lxd stopped configuring the bridge with hardcoded IPs by default and now requires you to set it up yourself07:41
mwhudsoni don't know a lot more about all of this than you do07:41
pittimwhudson: yes, the bridge is unconfigured by default; I suppose the test installs an /etc/default/lxd-bridge somewhere07:42
mwhudsonyeah07:42
mwhudsongrep for 10.0.8.1 :-)07:42
pittiyep, that one07:42
mwhudsonit's in setup-lxd.sh07:42
mwhudsoncan we exploit the feature in very new lxd where it guesses for you?07:43
mwhudsonor at least re-implement that07:43
pittibut in the DC there is no "real" 10.0.8.107:43
pittiso it's not that after all I guess07:44
mwhudsondammit07:44
pittimwhudson: let's move to #u-devel for further debugging to not clutter the release coordination even further07:45
mwhudsonpitti: +107:46
jibelmorning08:01
jibelit's surprisingly quiet for a release week08:02
pittibonjour jibel !08:02
pittijibel: yeah, it's suspicious08:02
jibelBonjour pitti08:02
pittithe ubiquity issue yesterday was laughable compared to what we're used to08:02
pittiwhat did we miss?08:02
jibelit means Thursday will be hell :)08:02
* pitti is scared08:02
pittijibel: OOI, how are upgrades looking?08:02
pittiI'm sure we've missed some bits at least there08:02
jibelpitti, automated tests didn't find anything, neither did davmor208:03
jibelI'll dig into launchpad08:03
jibelsomeone reported that ubuntu-gnmoe fails to start after an upgrade from Trusty08:03
davmor2jibel: I didn't think ubuntu-gnome was an option in trusty08:04
davmor2oh it's been around longer than I thought :)08:05
jibeldavmor2, bug 157193908:07
ubot5`bug 1571939 in ubuntu-release-upgrader (Ubuntu) "Ubuntu GNOME fails to boot after Trusty -> Xenial upgrade" [Undecided,New] https://launchpad.net/bugs/157193908:07
davmor2jibel: there is one last test for upgrades I want to perform, OEM install upgrade just to be sure it doesn't do anything daft08:07
jibeldavmor2, bug 1571679 I know you enjoy secure boot08:10
ubot5`bug 1571679 in ubuntu-release-upgrader (Ubuntu) "fwupdate-signed doesn't get installed on upgrade" [Undecided,New] https://launchpad.net/bugs/157167908:10
davmor2jibel: I wonder if this is part of the mokutils stuff and if these users had secureboot enabled with 3rd party drivers installed, I can test that with ease 17246808:11
jibeldavmor2, it's an upgrade not an installation08:11
davmor2jibel: yeap still kicks in08:12
davmor2jibel: and it might be disabling the modules now too08:12
davmor2jibel: that would affect wifi networking if it is broadcom and fwupdate-signed if they didn't disable secureboot08:13
pittijibel: well, at least we  still have the "OMG need to land juju 2.0" issue, so we aren't totally free of urgent things :)08:44
pittithis calms me somewhat08:44
davmor2pitti: aaaahhhh bless you think that is the only thing that still needs to land I love your optimism it's nearly infectious :D09:06
pittidavmor2: no, it's not optimism -- I'm scared that we don't (yet) know about the catastrophes that are still coming :)09:07
seb128hey there09:08
seb128we need to rename "gnome-software" to be "ubuntu-software", part of it is changing the .desktop Name=09:09
seb128we have two options09:09
seb128- rename the current one (easier) and impact the flavors as well (Ubuntu GNOME stated they would prefer to keep the upstream name)09:09
seb128- have a new.desktop and play with OnlyShowIn to have it only for Unity, but that would require to land unity with it to change the default and have user config migration09:10
seb128do you have a preference?09:10
seb128willcooke, ^ fyi09:10
seb128I think we are leaning toward the first one09:12
seb128easier and less potential to create issues09:12
jtaylorunseeded universe is still free-for-all in terms of sync or do I need approval?09:13
infinityjtaylor: See the freeze announcement.09:14
cjwatsonstgraber,xnox: germinate fixed in git and on snakefruit etc.; uploading to unstable, will sync that when I can09:14
jtaylorinfinity: so free for all?09:14
infinityjtaylor: Ish.  With a sanity clause.09:14
jtaylorok thanks09:14
xnoxcjwatson, lovely09:15
pittihey infinity, good morning09:29
pittiinfinity: I'm handling juju-core09:29
pittiinfinity: just running last tests locally09:29
cjwatsonlazr.https://bugs.launchpad.net/bugs/157189909:29
ubot5`Launchpad bug 1571899 in ubiquity (Ubuntu) "Kubuntu Daily Image Ubiquity cannot connect to X server :0" [Undecided,New]09:29
cjwatsonoops09:29
pittiinfinity: you or someone accepted casper, so I guess free for another round of reviewing/accepting bug fixes on images?09:29
cjwatsonsorry, mispaste.  can we squeeze lazr.restfulclient in?  fixes python3 support bug, but is on a couple of images09:29
zequenceinfinity: Saw you and cjwatson poke in our seeds cause of the build failure yesterday, so hopefully that is fixed. iso.qa.ubuntu.com says still building. Would like to trigger another one. Any help there, please?09:31
pittiah, someone accepted juju-core already; well, tests look good so far, so it's probably fine09:39
cjwatsonlazr.restfulclient> ta09:40
mwhudsonpitti: how long will it be until the autopkgtests run?09:41
mwhudsoni guess i'll be in bed by then09:41
pittimwhudson: package just finished building, so needs a publisher run; I'd say 45 minutes, give or take09:42
pittimwhudson: but go to bed, I'll have an eye on them09:43
pittimwhudson: if it still fails, it should be fairly shallow, and I know enough of what this is trying to do09:43
pittii. e. I should be able to handle it09:43
mwhudsonpitti: yeah, mostly just curious09:43
mwhudsonit's only 2143, it's not that late yet :)09:43
pittimwhudson: local run is at future-provider, other tests passed09:44
pitti"local" == manual run in the production env, I mean09:44
flexiondotorgseb128, Regarding the software center naming.10:16
flexiondotorgseb128, Are you proposing changing the string in the .desktop file or the filename of the .desktop file?10:16
LaneyString10:16
flexiondotorgOK.10:16
flexiondotorgWHat about upgrades from 14.04? Will you now have to Ubuntu Software centres?10:17
flexiondotorgAlso, although Ubuntu MATE doesn't ship with Ubuntu Software Centre nor GNOME Software Centre, both are installable via Ubuntu MATE Welcome.10:18
LaneyI think it's Ubuntu Software and Ubuntu Software Center (awesome)10:18
cjwatsonSounds like a great way to reduce user confusion10:19
flexiondotorgSo, please have MATE; in any OnlyShowIn setting.10:19
flexiondotorgIf you decide to make that change.10:19
davmor2seb128, Laney: out of interest why rename it at all, currently it is just called software10:19
LaneySomething about reducing confusion10:19
cjwatsonHo ho ho10:20
LaneyThey want to rename the package too, but I'm pushing back on that10:21
LaneyWe'll see10:21
pittiinfinity: ^ would be great if you could review/accept this apt SRU (trusty/wily), as bug 1560797 is still causing upgrade havoc10:39
ubot5`bug 1560797 in apt (Ubuntu Wily) "apt does not configure Pre-Depends: before depending package" [High,In progress] https://launchpad.net/bugs/156079710:39
* pitti will write SRU info now10:39
infinitypitti: Will look.10:39
xnoxpitti, can i have an autopkgtest for upstream hosted in git on sourceforge?10:43
pittiinfinity: SRU info added10:43
xnoxor does it need like remote triggers off somewhere? e.g. shall i mirror it to github or launchpad and try to do webhooks for upstream autopkgtest10:43
pittixnox: we auto-trigger systemd's autopkgtest for upstream systemd pull requests on github, so from SF should not be fundamentally different10:44
pittixnox: but yes, you need something on the SF side to actually trigger the test, if you want this to happen automatically10:44
pittigithub has really nice webhooks for doing CI, not sure if SF learned anything in that regard in the past 5 years10:45
pittixnox: also, -> #u-devel I figure?10:45
davmor2jibel: so it looks like those upgrade issues were what I thought they were looking at the issues I just hit, looks like if you don't disable secureboot through mokutil the system goes crazy,  I'll run a bunch of tests around it now.11:07
jibeldavmor2, thanks11:10
davmor2netboot is missing from the iso tracker11:19
davmor2jibel: is that something you can add?11:19
xnoxdavmor2, it will be fixed shortly, when next respin is done.11:20
davmor2xnox: thanks dude11:22
flocculantxnox: are we expecting an imminent respin? cos something appears to be up with studio's build - tracker shows them rebuilding but https://launchpad.net/~ubuntu-cdimage/+livefs/ubuntu/xenial/ubuntustudio shows built hours ago - was the same ~ 5 hours ago11:23
flocculantthough the imminent bit is me with xubuntu hat on :)11:23
maprerican you please ignore the virtualbox's autopkgtests and let virtualbox hit the release pocked?  (message proxied for locutusofborg)11:23
mapreripitti: ↑11:31
pittimapreri: yep, saw it, at it (we already have a hint, just needs a version bump)11:31
pittimwhudson, slangasek: http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html#juju-core is just landing \o/11:31
pittiall green11:32
mapreripitti: ok, ta :)11:32
infinitypitti: I'm fixing vbox to not have that bug, BTW.  Well, to not have half of that bug. :P11:41
superm1infinity: okay can you respin myth?11:56
infinitysuperm1: You can!12:01
superm1I can?12:01
infinityWell, if you have access to do so on the tracker.12:02
infinityBut lemme do it.12:02
infinityCause I should take a lock on that soon anyway.12:02
superm1Oh I didn't know I might have access. Ok thanks12:03
cyphermoxgood morning!12:33
flexiondotorgo/12:40
flexiondotorgcyphermox, My image upload died and I'm at work now and can't resume until later :-(12:40
flexiondotorgCan someone here take car of the upload please? https://bugs.launchpad.net/ubuntu/+source/ubuntu-mate-welcome/+bug/157163512:40
ubot5`Launchpad bug 1571635 in ubuntu-mate-welcome (Ubuntu) "ubuntu-mate-welcome 16.04.9 bug fix and translations [dsc attached]" [Undecided,Fix released]12:40
flexiondotorgAg, wrong link.12:41
flexiondotorgPlease can someone upload this please - https://bugs.launchpad.net/ubuntu/+source/ubuntu-mate-artwork/+bug/157212012:41
ubot5`Launchpad bug 1572120 in ubuntu-mate-artwork (Ubuntu) "ubuntu-mate-artwork 16.04.7 bug fix release [dsc attached]" [Undecided,New]12:41
cyphermoxflexiondotorg: bah, I'll get VB on some other computer and try to reproduce there12:42
cyphermoxor you know, on a laptop with spinny disks12:42
flexiondotorgcyphermox, Give me a minute...12:42
flexiondotorgcyphermox, I can't resume it :-(12:48
flexiondotorgGoing to take another 10 hrs.12:48
flexiondotorgSorry.12:48
cyphermoxflexiondotorg: don't worry12:48
cyphermoxflexiondotorg: I know what is wrong, I just haven't managed to reproduce it yet12:49
flexiondotorgVirtualBox. Install Ubuntu 15.10 with full disk. No encryption. No LVM.12:49
cyphermoxit's very clearly systemd too happy to bring up swaps, and we "just" need to find a way to disable that only when in ubiquity12:49
flexiondotorgSnapshot that. Then try install Ubuntu 16.04 daily over the top.12:49
flexiondotorgcyphermox, Agreed.12:49
flexiondotorgMy best work around is "sudo swappoff -a" then start the install and dash through the first few menus.12:50
cyphermoxright, but partman already does a swapoff just before doing the swaps12:51
cyphermoxso it's a timing thing, you might be lucky, or might not12:51
cyphermoxI can still add swapoffs in ubi-partman.py but it probably won't fix it more12:51
flexiondotorgcyphermox, It won't fix it. Because unless you do a Ubiquity speed run systemd will beat you to the punch.12:53
infinitytjaalton: Do you plan to fix xorg-lts-transitional?13:02
tjaaltoninfinity: fix how? it worked here13:13
infinitytjaalton: Check excuses.  You depend on a bunch of packages that don't exist.13:14
tjaaltonah13:14
tjaaltonoh, indeed13:15
jcastrodoes anyone mind if I add a Juju section to the updated packages section of the release notes?13:20
tjaaltoninfinity: hrm, I need to run out now, will fix it later today. guess I can drop the transitional -dbg packages at least and then let apt remove those on upgrade?13:20
rbasakjcastro: please do. SHould that be in https://wiki.ubuntu.com/XenialXerus/ReleaseNotes#Ubuntu_Server?13:20
infinitytjaalton: If apt will remove them.13:21
tjaaltonright, i'll test on a vm13:22
jcastrorbasak: well, the client/UI bits had the biggest updates so I was going to put it below php713:22
rbasakjcastro: oh, I see. I wonder if people will look just at the server section for this kind of thing or not (even though it is also a client). Maybe PHP etc should be down there too.13:24
rbasakjcastro: anyway, I'm bikeshedding, sorry. Let's get the text in and worry about where it belongs later.13:24
superm1infinity: i tried to seed libc6-i386 into mythbuntu.xenial ship-live to fix bug 151781, but it appears to have had no effect in the respin, i'm unsure how to figure out what is wrong there to prevent it from coming in13:26
ubot5`bug 151781 in kde-guidance (Ubuntu) "KDE resolution or screen changes makes Nvidia binary not to work" [Undecided,Won't fix] https://launchpad.net/bugs/15178113:26
superm1bug 1571781 that is13:26
ubot5`bug 1571781 in mythbuntu-meta (Ubuntu) "Nvidia driver not installed" [Undecided,Fix committed] https://launchpad.net/bugs/157178113:26
tewardrbasak: do we still need a release notes for http/2 (Apache)?13:29
tewardoops wrong channel my apologies13:29
rbasakSeems like the right channel to me :)13:30
pittiinfinity: may I nudge you about bug 1560797 again? I'd hope that we can expedite this SRU to intensely test this by Thursday and then release it13:30
rbasakYes we do, but I have some things I need to take care of first.13:30
ubot5`bug 1560797 in apt (Ubuntu Wily) "apt does not configure Pre-Depends: before depending package" [High,In progress] https://launchpad.net/bugs/156079713:30
infinitypitti: apt?  Looking in a bit.13:30
tewardrbasak: ack13:30
infinitypitti: ^13:38
pittiinfinity: cheers13:38
* mvo hugs pitti and infinity for this fix13:48
seb128^ I expect this one to be controversial, happy to reply to questions13:48
seb128willcooke, ^ the gnome-software rename one, just fyi13:48
superm1cyphermox: i think i figured out why libc6-i386 isn't ending up in the pool on ISOs - "Promoted libc6-i386 from ship-live to boot to satisfy grub".  is there actually a reason for grub (not grub2) to still be called out in platform.xenial's boot seed still?13:49
cyphermoxnot that I can think of, but that would make libc6:i386 *be* in pool13:50
cyphermoxdirectly available on the CDs, basically13:51
superm1cyphermox: well i'm trying to get libc6-i386 (not libc6:i386) to end up on amd64 mythbuntu ISO's13:51
superm1and the reason it's not happening is that libc6-i836 is a dependency for grub13:52
cyphermoxsorry, that's what I meant13:52
cyphermoxship-live makes it go in the pool/ directory on the iso13:52
superm1right that's the goal13:52
superm1but it's not ending up there even if explicitly seeded13:52
cyphermoxwell it's not because of that message13:52
cyphermoxlemme see13:53
superm1http://people.canonical.com/~ubuntu-archive/cd-build-logs/mythbuntu/xenial/daily-live-20160419.log is the log13:53
pittiinfinity: we'll still do a respin, right? for ^ (bug 1552539)13:55
cyphermoxI fail to read...13:55
ubot5`bug 1552539 in casper (Ubuntu) "Ubiquity Erase Disk and Install Fails to create Swap Space" [Critical,Fix committed] https://launchpad.net/bugs/155253913:55
cyphermoxsuperm1: sorry, I just can't read properly it seems13:56
cyphermoxso yeah, libc6-i386 gets promoted to the 'boot' seed because you have  * grub [amd64 i386]                      # lilo will be in supported  in that seed file13:57
superm1well that comes from platform.xenial13:57
superm1so that's why i was asking if it's actually used anywhere (to know if it could be removed)13:57
cyphermoxyep13:57
cyphermoxI don't know13:57
cjwatsonthat stuff will need to be in sync with livecd-rootfs I think13:58
cyphermoxcjwatson: or debian-cd?13:59
cyphermoxok, I see, ship-live depends on boot and live, fun.13:59
infinitypitti: Yeah.13:59
cjwatsonpossibly debian-cd/cdimage as well14:01
infinityWhy would anything else need changing to remove grub from boot?  We surely don't use grub anywhere anymore...14:02
cjwatsonone would hope14:02
cjwatsonbut it might still be added to images somewhere else14:02
infinityIf it were, I'd assume it would be because of the seed.14:02
cyphermoxlooks like it, it's in the boot seed.14:03
cjwatsonit's probably OK, just worth some grepping14:03
cjwatsonthe boot seed is historically very weird14:03
cjwatsonsubtle and quick to anger etc.14:03
cyphermoxahah14:03
infinitysuperm1: Anyhow, were that fixed, you could also revert your change (probably), since nvidia-thingee should pull in libc6-i386 without your help.14:03
cjwatsonbut I think a grep through livecd-rootfs, maybe live-build, debian-cd, and cdimage would be sufficient14:03
superm1infinity: yes14:03
infinitycjwatson: Yeah, heading there.14:03
=== maxb_ is now known as maxb
xnoxsuperm1, my dell xps 15 got offered a firmware update, it did end up in ./EFI/ubuntu/fw/fwupdate-j01CFS.cap14:33
xnoxsuperm1, however on reboot it did not boot into flash update14:34
xnoxi can go into flash update manually and find that file, but i thought this should happen automagically14:34
superm1xnox: lets talk in PM to debug what happened14:34
xnoxor not?14:34
xnoxack14:34
xnoxinfinity, Laney https://bugs.launchpad.net/ubuntu/+source/ubuntu-release-upgrader/+bug/157167914:43
ubot5`Launchpad bug 1571679 in ubuntu-release-upgrader (Ubuntu) "fwupdate-signed doesn't get installed on upgrade" [Undecided,New]14:43
Laneymerci14:44
slangasekpitti: juju-core tests all fixed? nice!15:00
pittislangasek: yep, that was an interesting journey15:00
pittiinfinity: ok to accept lxc with a tests-only fix to make the tests green again? (will still rebuild the binaries, of course)15:09
pittiinfinity: or maybe the other way around, can you poke me, Laney, stgraber etc. to stop reviewing stuff from the queue once you want to lock down?15:09
stgraberthat lxc is to make apw happy basically :)15:10
* apw likes to be happy15:11
pittiwell, and to stop new kernels from breaking lxc too :)15:11
stgraberwe know lxc itself is good because we saw its tests pass before, but a change to the adt image broke things recently so that should make everything green again15:11
infinitypitti: When I'm locking the world down, I'll put in a britney block.15:11
infinitypitti: So if pepole keep reviewing, it'll all get stuck, and we can sort it out post-release.15:11
pittior rather, currently lxc depends on which of the clouds it runs on15:11
pittiinfinity: ah, fair enough; they'll end up as SRUs then15:11
zequenceWho can I poke about the tracker being stuck? We can't do rebuilds for Ubuntu Studio since it got stuck yesterday.15:17
stgraberI'm about to board a plane so can't really help, but infinity should be able to have rebuild-requests cancel anything that's pending which will fix the web UI15:22
cyphermoxinfinity: stgraber: should we have a cpc product in the iso tracker?15:33
slangasekcyphermox: no15:41
cyphermoxok15:43
slangasekcyphermox: this is not a "cpc" product at all, and the fact that it's showing up that way currently on cdimage is a bug15:44
cyphermoxslangasek: I thought it should be in ubuntu-server indeed; but what I really meant is, should I be seeing a product entry for the raspi2 preinstalled images under server in iso.qa.u.c, or anywhere else?15:45
slangasekah ;)15:46
cyphermoxsince I slapped it on my device and booted it, I might as well tick the tested box ;)15:46
slangasekthen yes, probably, but not until the image publishing is sorted properly ;)15:46
cyphermoxyeah15:46
cyphermoxinfinity fixed up my initial merge, and then there was some extra magic to make it so ubuntu-cpc -> ubuntu-server15:46
slangasekthat part doesn't seem to be landed yet however?  at least there's no http://cdimage.ubuntu.com/ubuntu-server/daily-preinstalled15:47
infinitycyphermox: There will be a tracker project in a bit when I fix the other thing.15:47
cyphermoxslangasek: no, it's currently under /ubuntu-cpc/daily-preinstalled15:48
cyphermoxinfinity: do you want me to finish the work of fixing that up? preferrably with passing tests? ;)15:48
infinitycyphermox: If you want to figure out how to move it from -cpc to -server and submit another MP, go for it.  If this one fails the testsuite, I'll fly to Montreal and stab you a little.15:49
cyphermoxjust a little?15:50
infinitycyphermox: Just a tad. :)15:50
Laneygentle caress with a pointy thing15:50
* cyphermox is tempted to make the ubuntu-cdimage merge with the fix, plus an added test to test the fix, but leaving the test failing :)16:31
slangasekinfinity: where do we stand at the moment regarding candidate image mastering?  per discussions elsechan, it's been observed that snapd didn't get into the platform seed; doing this for 16.04 was discussed with flavor leads but the execution fell through.  If we seed it right now into desktop-common are you going to have a sad?16:31
davmor2jibel, cyphermox: hey guys so I seem to hit an issue with upgrades.  The issue is as follows. Install 14.04, install binary drivers, upgrade, during upgrade you get the mokutils password setup, However if you don't see the menu the driver seems to go into some mad loop, you can't access the gui if it is a gfx driver or use wifi if it is that driver, but your tty fills with errors so you have no way to trigger16:32
davmor2mokutils to fix it for gfx issue16:32
Laneyslangasek: He's out, but the metas are all being rebuilt anyway for fwupdate-signed16:34
cyphermoxinfinity: https://code.launchpad.net/~cyphermox/ubuntu-cdimage/cpc-in-server/+merge/29229716:34
cyphermoxthis does pass tests here.16:34
slangasekLaney: "are being" meaning currently or at some point in the near future?16:35
cyphermoxdavmor2: I'm not sure I follow what you mean. did you file a bug with screenshots or something?16:35
Laneyslangasek: Built on laptop but not uploaded yet16:35
cyphermox"if you don't see the menu" ?16:35
slangasekLaney: ok; am twiddling seeds now, do you happen to know the full list of metapackages that need uploading and would you like to help divide and conquer by chance?16:37
davmor2cyphermox: no, no you miss understand, if you don't realise that you need to hit a key or you go off and make a drink while the system reboots so you miss it or don't modify it then it breaks16:37
Laneyslangasek: Adam's got a nice alias in his shell history that I'm sure he could re-run quite easily16:38
slangasekLaney, infinity: snapd added to the seed16:39
slangasekso it will get picked up in meta rerolls16:39
cyphermoxdavmor2: you mean, do the upgrade, reboot, miss the MokManager blue prompts, it times out and you get no drivers?16:39
davmor2cyphermox: bingo16:40
cyphermoxright.16:40
Odd_Blokeinfinity: There was murmuring about a kernel with fixes needing to land; which kernel was that?16:43
infinityOdd_Bloke: -2116:44
infinityOdd_Bloke: In proposed.  We're working on promoting it.16:44
infinityOdd_Bloke: It'll be in the release pocket tonight.16:44
Odd_BlokeOK, cool.16:44
infinityslangasek: Right, I'm fixing some other seed bits, and will re-roll the world shortly.16:46
davmor2cyphermox: the odd thing is that this only happens on upgrade. IE 14.04 install nvidia-binary, ignore the mokutil efi menu, and TTY7 is black with a white arrow and that is as much as loads16:46
davmor2cyphermox: because the tty 1-6 is flooded with errors you can't login16:47
cyphermoxok16:47
davmor2cyphermox: also because the mokutil only shows the one time you need to be able to login to trigger it again.  I managed to do it via rescue mode but it is not fun or pretty16:49
bdmurrayinfinity: I've the Wily SRU for bug 1560797.  Will having the new apt in -updates reach enough people though?16:50
ubot5`bug 1560797 in apt (Ubuntu Wily) "apt does not configure Pre-Depends: before depending package" [High,Fix committed] https://launchpad.net/bugs/156079716:50
seb128slangasek, Laney, if you do a seed change, can you gnome-software->ubuntu-software for desktop?16:54
Laneynot before the package exists16:54
slangasekseb128: let's wait until that's in the archive for that, as it hasn't landed yet16:54
seb128k16:54
zequenceinfinity: Any chance you could fix the problem we are having at the ISO tracker? We can't do rebuilds. It's been stuck att (rebuilding) for 24h, since the build failure yesterday.17:03
seb128slangasek, Laney, infinity, ^ new take, without renaming the .desktop but installing one to "/usr/share/ubuntu/..." with the same name, that path is at the start of XDG_DATA_DIRS in Ubuntu sessions so going to be preferred in Unity17:08
seb128oh also creating the .desktop by copying the upstream one and caling sed in debian/rules to change the Name/Icon17:08
seb128seems to work fine from my local testing17:08
seb128willcooke, ^ just for info17:09
Laneyseb128: I need to add another commit17:10
seb128Laney, queue is public, feel free to grab and merge another change or stack another upload on top17:11
Laneyseb128: it'll be a little bit easier if you push that to bzr17:11
Laneyif you have that17:11
seb128no I don't17:11
seb128attente gave me a ppa the other day and I forgot about the vcs17:11
seb128need to reconciliate things17:11
seb128I can try to have a look but I've a tennis match in 45 min and didn't get dinner yet :-/17:12
Laneyok17:12
davmor2cyphermox: oh oh oh pitti just fixed https://bugs.launchpad.net/bugs/1552539 can we get that in asap?17:12
ubot5`Launchpad bug 1552539 in casper (Ubuntu) "Ubiquity Erase Disk and Install Fails to create Swap Space" [Critical,Fix released]17:12
cyphermoxI know17:13
cyphermoxnot sure it really will fix it, we'll see17:13
slangasekLaney: shall I hold off on review until you fix up, or shall I review/maybe-accept this one and then review again your next commit?17:13
LaneyI'll upload it in a minute, so you might as well wait17:14
seb128Laney, I'm fixing the vcs, give me 30s17:15
Laneyseb128: already done locally, no worry17:15
seb128oh ok17:15
seb128thanks17:15
seb128going to get some food then :p17:15
seb128Laney, slangasek, I would appreciate if you could handle the seed change when the new package gets in, need to have dinner and go in ~30min17:16
seb128otherwise I can have a look later tonight when I'm back17:16
slangasekseb128: yes, I can have a look17:16
seb128thanks17:16
pittidavmor2, cyphermox: yeah, this is mostly just my best bet so far, so testing appreciated; it can be simulated reasonably well on the current image with breaks=casper-bottom and doing that rm there17:33
superm1infinity: did you have a conclusion on grub (re boot seed) from looking at livecd-rootfs, debian-cd, cdimage et'al?17:37
infinitysuperm1: I'll get there.17:39
superm1infinity: okay.   i just wanted to say if it's gotta stay or it's hard to confirm if it's gotta stay, can another way to fix this be to seed "grub" in mythbuntu.xenial's ship-live seed to pull libc6-i386 into the pool?17:40
infinitysuperm1: Won't help.17:40
Laneyslangasek: check yo'self17:45
infinityslangasek: And/or wreck yo'self17:45
slangasekLaney: I am here17:46
slangasekcheck complete17:46
LaneyHAHA17:46
slangasekadt-run --apt-source slangasek17:46
LaneyDONT CHECK ANYTHING17:46
LaneyINFACT STOP LOOKING17:46
slangasekahhh autotools wut17:46
infinity-0ubuntu1-0ubuntu1 is brilliant.17:47
slangasek"new upstream snapshot"17:47
Odd_Bloke-0ubuntu1, Laney 2016: An effective commentary on the true meaningless of version strings when faced with the complexity of life17:49
Laneylalala17:53
cyphermoxis that the version number length award for this cycle?17:59
rcjrolls off the tongue18:00
cyphermoxI will not let gir1.2-networkmanager-1.0 0.9.7.995+git201301311547.17123fc-0ubuntu1 down so easily.18:01
LaneyI got a warning from lintian with the duplicate 0ubuntu118:03
Laneybut I was like YEAH WHATEVER18:03
LaneyWHO CARES OLD MAN18:03
* ogra_ waits for the rollback ... with the +really-$version 18:03
Odd_Bloke+really-0ubuntu1-0ubuntu118:04
cyphermoxthere wouldn't be reallys for non-upstream revisions18:04
ogra_3.20.1+git20160419.2.b8d2e40.ubuntu-xenial+really+git20160419.1.d7dc006.ubuntu-xenial-0ubuntu118:04
infinityslangasek: Please to re-review gnome-software so we can get that in and twiddle der seeden.18:05
ogra_"twiddle der seeden." ... sounds like you have a dutch moment today :)18:05
slangasekLaney, seb128: not sure this debian/rules sed of debian/ubuntu-software/usr/share/ubuntu/applications/org.gnome.Software.desktop in place will dtrt when we *do* have translations?18:06
infinityAccording to Laney, "Dunno, won't fix it, **** off."18:06
slangasekLaney: also, /usr/share/ubuntu/applications does not exist on my system at all currently, so what does this do?18:06
LaneyIt's prepended dynamically18:06
Laneyby $something18:06
infinityslangasek: Evidently, there's some hierarchy where the session name is checked first.18:06
Laneyaccording to $someone18:07
slangasekinfinity: that sounds an awful lot like "please reject"18:07
infinitySo someone says.18:07
infinityI feel like this needs some testing once built. :P18:07
infinityBut we have the power to do that here.18:07
* infinity stares at the man to his left.18:07
slangasekthis non-DRY sed is also quite upsetting to me18:10
slangasekanyway, accepted18:10
LaneyI *think* that translations will be okay since dh_translations adds X-Ubuntu-Gettext-Domain18:13
Laneyand XDG_DATA_DIRS is set up by an Xsession.d script AFAICS18:15
Laneyit's at least right here18:15
Laneyhttp://people.canonical.com/~laney/weird-things/lifeisgood.png18:24
Laneyquality18:24
cyphermoxthat's a lot of software18:25
Odd_BlokeTwice the software, still free.18:25
zequenceWe would really like a rebuild for the Ubuntu Studio ISO, but can't do it ourselves, since the tracker is stuck at rebuilding since yesterday. Could someone help us out here, please?18:25
infinityslangasek: New kernel appears to function the way we think it should WRT mok thingees now.18:26
infinityslangasek: The UI for all of that is *vile*, though.18:26
tewardinfinity: hate to add something to your list for FinalFreeze breaking, but, nginx 1.9.15, Bug #1572223, has some fixes to HTTP/2 handling to address Chrome issues, and two other bugfixes in HTTP/2.  It changes a runtime dependency for nginx-extras (Universe) which doesn't affect nginx-core, nginx-common (Main).  Happy to drop the nginx-extras changes if you NACK that, but the HTTP/2 changes should land.18:27
ubot5`bug 1572223 in nginx (Ubuntu) "[FFe] Update nginx to 1.9.15" [Wishlist,New] https://launchpad.net/bugs/157222318:27
teward(that came out today)18:27
infinityteward: I thought we had http/2 disabled?  Or am I stuck in the past?18:28
infinityteward: But will look at it tonight.18:28
tewardinfinity: ACK'd by the Security Team for 1.9.1418:28
cyphermoxkernel mok thingees?18:28
infinityteward: Ahh, kay.18:28
davmor2cyphermox: that's the technical term for it don't mok ;)18:29
tewardinfinity: HTTP/2 in Apache is still disabled, but nginx's implementation was OK'd for enablement18:29
tewardwhich is one of the things the server team has wanted to land for Xenial18:30
tewardso, keeping the protocol handling is the main 'break the freeze' reason for nginx.18:30
infinityteward: Check and check.  Will look in a bit.18:30
tewardno rush :)18:30
* teward goes and pokes at his builders which are choking on armhf builds18:31
Kamilionsweet.18:35
Kamilionit'll be fun to roll out letsencrypt and http2 on a fresh nginx, I have some 12.04 machines that are sorely due for an upgrade series.18:35
tewardKamilion: indeed, though you still need Wily+ for *good* HTTP/2 - PPA for Wily, Xenial-native (1.9.14 or newer) for HTTP/2 to truly work - PPAs're going to lose HTTP/2 soon for pre-OpenSSL1.0.2 versions...18:40
Kamiliongood to know -- I'm on their stable PPA normally18:41
slangasekinfinity: it was implied up-thread that you would be doing mass rebuilds of meta packages; is anything blocking that? do you need an extra body?18:42
infinityslangasek: I've got them all sorted except ubuntu, which is waiting on these bits to land.  It's sorted, though.18:44
infinityslangasek: We're going to wait for the kernel and ubuntu-software to settle and then update metas, respin images, and go get drunk.18:44
slangasekinfinity: sorted how? I don't see them uploaded yet, and I think they ought to be18:45
infinityslangasek: Was going to upload them all at once, but meh.18:45
slangasekinfinity: you're looking at at least another half hour delay for ubuntu (gnome-software build, new accept, proposed-migration); surely it's better at this late hour to not block the rest of the flavors on this18:46
infinityslangasek: Sure, I'll upload all !Unity flavour metas.  Sec.18:47
apw"is that your cursor" time18:47
tewardheh18:48
infinityslangasek: Uploaded.18:52
slangasekinfinity: thanks, I shall reviewinate18:54
phillwHi good people, is there an estimate for the global respin with the new kernel? (The testers are nagging and straining at the leash to get their hands on it!!! )18:56
tewardphillw: tell the testers they have to learn patience18:58
teward:P18:58
apwphillw, the kernel is copying out as it is currently awol18:58
slangasekinfinity: ubuntustudio-meta-0.154/debootstrap-version lulwat18:59
phillwteward:  we are to run with linux-headers-4.4.0-21 ?18:59
infinityslangasek: Dropped it to the xenial version so ./update didn't have a fit. :P18:59
infinityslangasek: (previous upload was from a sid machine, clearly)19:00
* slangasek nods19:00
slangasekinfinity: do you know why the ubuntu-mate refresh seems to have downgraded snapd to a recommends?19:02
slangasekI guess ubuntu-mate had seeded it themselves already?19:02
infinityslangasek: Yeah, they had it seeded in a seed that can't do recommends.19:03
slangasekand gnome-software built/accepted19:03
KamilionI just learned that plymouth-theme-ubuntu-text is the default theme. I'd been assuming it had been broken and *falling back* to the text theme this whole time (after experiencing some of the older fancy animated themes in the past)19:06
KamilionI humbly retract my accusation that plymouth has been 'broken' for my machines19:06
slangasekinfinity: and all the other metanana accepted19:06
davmor2slangasek: wasn't that a Bananarama song?19:07
cyphermoxKamilion: it's only the default if you can't otherwise show the prettier graphics19:08
cyphermoxKamilion: otherwise you'd normally get plymouth-theme-ubuntu-logo19:08
tjaaltoninfinity: hope xorg-lts-transitional is fixed now.. I didn't drop packages, instead made them depend real ones19:17
infinitytjaalton: Works for me.19:17
Odd_Blokeinfinity: I see a migrated kernel; shall we pull the trigger on a (potential) cloud-images RC, or is it worth holding off for anything else you know of?19:29
infinityOdd_Bloke: Lemme look if there's anything else you're waiting on.19:31
Odd_Blokeinfinity: Danke.19:31
infinityOdd_Bloke: You're waiting on at least lxc and snapd still.19:34
Odd_Blokeinfinity: Ack.19:35
Odd_Blokeinfinity: Is there an easy way for me to see whatever you just looked at?19:35
pittiinfinity: new linux is off the release now?19:36
infinitypitti: s/the/to/?19:37
infinitypitti: If so, yes.19:37
pittiinfinity: oh, it landed19:37
pittino, I actually meant "will not go in", I didn't see that it got unblocked19:37
rbasakinfinity: FYI, ^^ was the second last reverse dep for src:mysql-5.6. Remaining is pinba, which I'll look at with Skuggen tomorrow.19:48
rbasak(and myodbc which is pending removal)19:49
rbasakI'm a little worried about pinba, but we'll see tomorrow.19:49
SkuggenYeah, it uses the source itself to build, and there are significant changes19:51
infinityrbasak: \o/19:51
infinityrbasak: Right down to the wire on this one. :P19:51
rbasakinfinity: yeah, too close for my liking :-/19:52
phillwInstalling side by side the alternate installer reports fail at "Select and Install Software" Are there any logs that I could possibly get to flesh out the bug report (I'm using KVM, so do have guestfish to pull files out).20:03
slangasekwhy do publisher runs seem to get slower at the end of the cycle right as there are fewer things it needs to publish?20:18
pittiOdd_Bloke, infinity: lxc and snapd are currently landing, FTR20:18
infinityslangasek: Because we become less patient.20:18
pittibecause more people stare at it?20:18
slangasekinfinity: I promise you I am impatient the rest of the cycle also20:19
slangasekpitti: why were the snapd test regressions bypassed/bypassable?20:19
slangasekI guess because they now include a test that looks for a package that doesn't exist on these archs?20:19
pittislangasek: this isn't a regression in 2.0.2, but a race condition/bug in the store; it happened before and mvo investigated it with Laney already20:19
infinityslangasek: I've got the ubuntukylin-meta review, I synced them with the ubuntu seeds today, so checking to see if the outcome was sane.20:20
pittislangasek: a fix is being deployed to the store, but as this existed before and this blocks image builds we expedited it20:20
pittislangasek: see #u-devel discussion 50 mins ago20:20
slangasekok20:21
slangasek+  * Added tuxmath to ubuntu-edu-primary-recommends [s390x]20:22
slangasekimportant20:22
infinitySUPER IMPORTANT.20:22
slangasekbut at least we got mono fixed for all archs, so we can have tomboy on our edubuntu arm64 deploys20:22
infinityHellz yeah.20:23
apwparty time20:23
infinity\o/20:23
infinityo/20:23
infinity\o20:23
slangasekrbasak: myodbc removal isn't in some way blocking on me, is it?  You asked about RoM from Debian, but I haven't decided yet what to do there and this shouldn't block its removal from xenial20:26
slangasek(if myodbc later sneaks back into Ubuntu via sync from Debian, that by definition means the FTBFS has been fixed, so yay)20:27
slangasekinfinity: I hereby declare that the plural of 'meta' is 'metaxa'20:28
apwtjaalton, your xorg-lts-transitional seems to be upsetting britney .... lots of uninstallables ...20:30
infinityslangasek: Isn't meta already the plural, and each individual one is metum?20:32
slangasekinfinity: are you trying to stand in the way of my metaxa?20:33
cyphermoxwho'd date?20:33
cyphermox*dare20:33
knomemeti20:34
knomemetæ could work as well.20:35
cyphermoxslangasek: you'll be able to get plenty of metas very soon though.20:35
rbasakslangasek: as long as an archive admin says it's blocking on Debian's removal or something, no it's not blocking.20:43
rbasakErr, doesn't say, that is.20:43
mvopitti: thanks again for the 2.0.2 hint, fwiw, autopkgtest on amd64 worked now and autopkgtest on the i386 and ppc64el are running20:48
pittimvo: yeah, just saw the same, thanks20:48
mvopitti: I expect them to work as well after the store update20:48
mvopitti: maybe one more retrigger, I need to check if all snaps are already in20:48
* mvo does that now20:48
pittihit it hard! :-)20:49
mvoha!20:49
* mvo does so :)20:50
slangasekinfinity: hakuna metata, the metaxa are published in xenial (minus edubuntu+ubuntukylin+ubuntu)20:57
slangasekand I'm updating ubuntu-meta now20:57
tjaaltonapw: still?20:58
* pitti waves good night20:59
tjaaltonbah21:01
tjaaltoninfinity: soo, can you purge xorg-lts-transitional from proposed so that the i386 xspice packages get cleaned21:05
slangasekor, I won't upload ubuntu-meta because somebody else uploaded it, somehow before ubuntu-software was published, and bypassed the unapproved queue <eye>21:09
slangasekok ;)21:09
slangasekinfinity: did you guys already skedaddle?  not sure what plans you had for respins today...21:39
slangasek("already" - it is quite late there, of course)21:40
flexiondotorgLaney, http://people.canonical.com/~laney/weird-things/lifeisgood.png21:58
flexiondotorgI'm so confused.21:58
infinityslangasek: I planned to respin when I got back from drinks, which I've done (the getting back, not the respinning)22:00
cyphermoxinfinity: console-setup upload I promised incoming, up to you to take or not.22:01
cyphermoxru shows the right glyphs, vi is *better* but not quite fixed yet; but at least more readable.22:02
slangasekinfinity: right, could've saved you the trouble... :)22:02
infinitycyphermox: console-setup requires a d-i respin, so it'll wait until tomorrow to see how high the carefactor is.22:03
cyphermoxyeah, no problem22:03
cyphermoxin the meantime I'll test one other non-ascii language, and see if I can fix vi for reals.22:04
infinitycyphermox: Can you give me a matching ubiquity in the queue with a package refresh, and we'll sort it out tomorrow?22:04
cyphermoxyes22:04
infinityDanke.22:05
cyphermoxoh wait22:05
cyphermoxno, I can't get ubiquity to refresh for console-setup unless it's in proposed.22:05
cyphermoxwell..22:05
cyphermoxyeah, I suppose I can trick it to be happy22:05
infinitycyphermox: I can get it in -proposed, that's fine by me.22:06
* infinity refiews.22:06
infinityOr reviews.22:06
infinityI haven't had alcohol, I promise.22:06
cyphermoxpfff22:06
infinity(I had all the alcohol)22:06
cyphermoxdon't drink and respin isn't a rule is it?22:06
infinity+[ "$CODESET" != guess ] || CODESET=''22:06
* flexiondotorg hides the wine22:06
* infinity wraps his head around that.22:07
cyphermoxwait, that's wrong22:07
slangasekis it wrong? looks ok to me22:07
infinity+[ "$CODESET" != guess ] || CODESET=''22:07
infinity+if [ -z "$CODESET" ]; then22:07
slangasekif confusingly written22:07
cyphermoxwhat if you set CODESET to what you want alredy?22:08
cyphermoxoh22:08
infinitySeems like it would be better written as "if [ "$CODESET" != guess ] || [ -z "$CODESET" ]"?22:08
cyphermoxI fail at logic.22:08
infinityErr.22:08
infinityWait.22:08
cyphermoxnot my code, in any case22:08
infinityThe changelog claims guess or unset is what we want.22:08
infinityThat doesn't say that, though..22:08
cyphermoxso that line will set CODESET to '' if it's == guess22:09
infinityRight.  It's just weird logic for a drunk brain.22:09
cyphermoxand then if -z, it will do the magic.22:09
cyphermoxyeah22:09
cyphermoxI'm not drunk though, no excuse22:09
infinityHence "if [ "$CODESET" = "guess" ] || [ -z "$CODESET" ]"; then" would read better, I'd think.22:09
cyphermoxI think it really was just copied straight from setupcon22:10
infinityWhich, indeed, the FONTSIZE test does...22:10
* cyphermox looks again22:10
infinityAhh, cargo culting is a valid excuse. :)22:10
cyphermoxyeah, copied22:10
infinityI'm all for duplicate code matching exactly.22:10
infinityKay.22:10
infinityThe logic works, it's just my brain that doesn't.22:11
infinityAcceptiferated.22:11
cyphermoxwell, the whole thing works, before, you get white squares. after, you get backwards Ns ;)22:11
infinityDeliver me a ubiquity while I sleep.22:11
cyphermoxzug zug22:12
infinityslangasek: unblocked your qemu.  Can you respin server when that lands?  I'll kick off !server now.22:13
infinitytjaalton: xorg-lts-transitional NBS cleaned up, you should be good if :9 fixes all your woes.22:15
infinitypitti: mvo seems to have left us, but can you two make sure his snapd hasn't actually regressed and nick highlight me with a yay or nay?  It's already migrated, but I want to know we've not migrated crap. ;)22:17
slangasekinfinity: yes. what's the proper way for me to respin at this point?  just a build from commandline?22:17
infinityslangasek: Ticken ze boxes on der ISO tracker und clicken ze rebuild.22:17
slangasekinfinity: ok22:17
infinityder boxen?22:17
infinityWhatever.22:17
infinityMy faux German is suffering.22:17
slangasekinfinity: accusative plural ;)22:17
infinityI'll accuse your plurals.22:18
* infinity waits for the current publisher run to be done.22:23
infinityLa dee da.22:23
infinityslangasek: Oh, also, I haven't fixed up the tracker part of the cpc bits, so when you're about to respin server, you can also do "for-project ubuntu-cpc cron.daily-preinstalled" in parallel to test the merge you pulled from cyphermox.  If it fails to be useful, I can fix in the morning.22:26
infinityAaaand, !server respin is now in progress.22:28
infinityOh sweet, we have new NBS issues 2 days before release too.22:32
infinityslangasek: Can you also hunt down someone cloudish (I'm guessing stokachu) and figure out WTF is up with the openstack-* stuff on NBS?22:33
infinity(kernel NBS is cleaning now)22:34
infinityslangasek: Thanks in advance, etc.  You can be my manager again in two days.22:34
phillwinfinity: is that going to be the server release with 4.4.0-20 kernel? Or should we expect another one?22:35
slangasekinfinity: did glibc 2.23 regress setenv() to crash on a NULL value where it didn't previously? https://errors.ubuntu.com/oops/6483e4a6-02f2-11e6-8b73-fa163e54c21f22:39
slangasekinfinity: ack to all the highlights in scrollback22:39
infinityslangasek: Re: setenv() > maybe, though that seems unlikely.  Remind me in /msg in a few minutes when I walk away and stop obsessively checking IRC, and I'll poke that in the morning with a stick.22:42
stokachuinfinity: whats wrong?22:43
stokachuhow can i turn that frown upside down22:43
slangasekstokachu: your latest openstack package, which has made it into xenial \o/,  drops the openstack-{landscape,multi,single} binary packages; these packages still have reverse-dependencies in the form of cloud-install-{landscape,multi,single}22:43
slangasekas shown on http://people.canonical.com/~ubuntu-archive/nbs.html22:43
slangaseker, check that22:44
slangasekthose are also built from the same source package22:44
slangasekso the revdep problem is further up the stack, at...22:44
slangasekat... I don't see it22:44
slangasekso this may be a false-positive on the NBS report and maybe I just need to delete me some binaries?22:44
slangasekok here we go22:45
stokachugo for it22:45
slangasekthere is a dead-end cloud-installer source package in xenial22:45
slangasekwhich is not uploadable because its binaries are superseded by openstack22:45
infinityRight, that likely all needs to go then.22:46
slangasekbut this means nbs is unhappy22:46
infinityBut I thought the transitional stuff was meant to live until post-16.04?22:46
infinityI guess not?22:46
infinityCause current openstack drops it all.22:46
slangasekit does, but I know nothing of the transitioning22:46
stokachuso now that everything relies on juju2 and conjure, openstack and bigdata only need to rely on conjure22:47
infinity1.0.0 is where it all went away.22:47
slangasekstokachu: but, would there be users who had cloud-install-* installed on trusty who will get a sad on upgrade because these packages dead-end rather than upgrading them to the openstack binary package?22:47
slangasekat least, I am sure that the cloud-installer source is obsolete and needs culling; doing that bit22:48
infinitystokachu: Is there an upgrade path from 14.04, or should there be?  You had an upgrade path from trusty all the way to wily, then dropped it right before the next LTS.22:48
phillwHi good people, I see a rebuild of lubuntu alternate, is there a server rebuild with new kernel due?22:48
infinityslangasek: Ignoring that /msg highlight, assuming it's the reminder I asked for. :P22:49
stokachuthere is not a viable upgrade path from openstack 0.99 + juju122:49
infinityKay.22:49
infinityI mean, not "kay" in the sense of "okay", but "kay" in the sense of "oh, well I guess that's how it is, then".22:49
infinityPerhaps better summarized as "Thanks Obama".22:50
slangasekphillw: search scrollback for "server"22:50
stokachuopenstack 1.0.0 is a new release in a sense of non compatible backward changes22:50
stokachuinfinity: dont you mean The Hillary?22:51
slangasekstokachu: just to confirm, this means that a user on trusty who had cloud-installer-* installed should not expect to have the openstack package installed on upgrade?22:51
stokachuslangasek: correct22:51
slangasekok22:52
slangasekthen the rest will sort itself out imminently22:52
stokachuslangasek: i think balloons wanted to try and do another upload of juju22:53
balloonsslangasek, yes, this has the dpkg-divert stuff we talked about yesterday. It's the juju-1.25 branch22:53
infinitystokachu: The upload can happen, whether it gets into the final images will rely on other respin triggers, it's not a valid enough reason on its own.22:53
infinityballoons: ^22:53
stokachuinfinity: ok thats what i wanted to confirm22:53
phillwslangasek: I'm on pidgin... limited search tools. I was alerted that there was a respin due with a new kernel. So, if you would be so kind as to answer, it would be much appreciated as I've got a couple of test cases to run and want to check against server for https://bugs.launchpad.net/ubuntu/+source/debian-installer/+bug/157230622:54
ubot5`Launchpad bug 1572306 in debian-installer (Ubuntu) "Side by side fails at "Select and Install Software" step" [Undecided,New]22:54
balloonsinfinity, ack22:54
infinityBut do upload, the world is blocked in proposed-migration right now, so we can sort it as we go.22:54
stokachuballoons: yea ill be around22:54
stokachuslangasek: i also have a package name change :(22:54
stokachuwas handed down on friday22:54
stokachuconjure becomes conjure-up22:54
stokachuwhich means bigdata and openstack would rely on that22:55
jderosetjaalton: are you open to rebuilding mesa against llvm 3.6 for the sake of installable Xenial ISOs on certain Nvidia hardware? https://bugs.launchpad.net/oem-priority/+bug/156415622:55
ubot5`Launchpad bug 1564156 in System76 "xenial: invalid opcode when using llvmpipe" [Critical,Triaged]22:55
stokachuballoons: ill be around just ping me when juju 1.25 is good22:56
jderosetjaalton: and the post release, we can figure out the underlying problem, try to get mesa back on llvm 3.8 as an SRU?22:56
slangasekstokachu: the sooner the better, though you're not up against the wall yet since none of this is seeded22:56
stokachuslangasek: i see conjure is out of proposed so it will need a breaks/replaces << 0.0.8?22:56
balloonsstokachu, I'm happy with it, so I think let's build22:57
stokachuslangasek: https://github.com/Ubuntu-Solutions-Engineering/conjure-up/blob/master/debian/control22:57
slangasekstokachu: if it's a straight rename, conjure-up Conflicts/Replaces/Provides: conjure is preferred22:57
stokachuok22:57
jderoseinfinity: what are your thoughts release-management-wise on https://bugs.launchpad.net/oem-priority/+bug/1564156 ?22:58
ubot5`Launchpad bug 1564156 in System76 "xenial: invalid opcode when using llvmpipe" [Critical,Triaged]22:58
infinityjderose: A day and a half before release, my thoughts amount to "ugh".22:58
infinityjderose: But I'm removing myself from my keyboard, so I'll let more awake people examine it.22:59
slangasekjderose: I think that's my cue22:59
jderoseinfinity: i likewise have "ugh" thought about it :P unfortunately, i was at a trade show last week22:59
infinityjderose: I'm not against a respin event for mesa, but it would need some rather broad testing to make me comfy.22:59
jderoseslangasek: much thanks :)22:59
slangaseksurely this shouldn't be a last-minute regression, though? i.e. if this was widespread, shouldn't it have been picked up sooner in the cycle?23:00
* infinity unlaptops and walks off.23:00
jderoseinfinity: agreed. for the record, system76 is more than willing to test this on as much hardware as possible as quickly as possible :)23:00
slangasekjderose: most likely scenario at this point: we have to release note it and fix it for .123:00
slangasekdowngrading to llvm-3.6 for the build isn't really an option23:01
jderoseslangasek: why is that? as far as i can tell, llvm-3.6 is in main. what is the specific issue there?23:02
slangasekbut, let me see if there's anything obvious in the build history23:02
slangasekerrm, you're right it is in main, and wow is *that* a bug to have two versions of llvm in main23:02
slangasekjderose: the switch to llvm-3.8 was for bug #1535500; dropping this means having to downgrade the supported level of OpenGL, 2months+ after the change was made in the devel cycle23:03
ubot5`bug 1535500 in mesa (Ubuntu) "Enable OpenGL 4.1 with radeonsi driver on xenial" [Wishlist,Fix released] https://launchpad.net/bugs/153550023:03
slangasekso it's not a simple "just rebuild with old compiler" change23:04
jderoseslangasek: okay, gotcha23:04
slangasekjderose: would you be able to check whether this problem exists with older llvm-3.8 builds of mesa? (11.1.2-1ubuntu1 and later)23:04
slangasekthis might let us identify a regression in llvm-3.8 (if we're lucky)23:05
jderoseslangasek: sure, i can definitely check that. any advise on the best way to do this? (I'm assuming those older binaries aren't around anymore in the archive)23:06
slangasekjderose: you can browse links on https://launchpad.net/ubuntu/+source/mesa/+publishinghistory and pull binaries 1-by-1 from launchpad23:07
slangasek(sorry, no aptable index)23:07
jderoseslangasek: gotcha, can do... checking 11.1.2-1ubuntu1 now...23:09
slangasekjderose: the last 3 builds of mesa were all built with the same version of llvm-3.8, and the 3rd from last update was a minor change for glibc 2.23 compatibility; we should be able to pin this down with some certainty to an llvm vs. mesa induced regression23:09
slangasek(I have my suspicions but won't interfere with your testing)23:09
dokoadded a note about the toolchain updates to the release notes23:26
dokothat changelog for juju-core-1 doesn't look very complete23:35
mgzdoko: oh yeah, I was going to stick the pre-fork changelog on the bottom23:36
dokoI'll leave to slangasek23:37
dokoI'll leave that to slangasek, even23:38
mgzdoko: wait, I *did* stick the pre-fork changelog on the bottom23:38
mgzdoko: so, you just mean the 1.25.5 entry?23:38
dokomgz, maybe you forgot to build with the -v option?23:38
mgzdoko: I didn't actually build the uploaded package, I don't have rights23:39
mgzballoons: ^23:39
slangasekok, NBS cleared, stokachu is off the hook23:40
stokachu\o/23:54
* balloons looks around23:56
balloonsso stokachu, who is holding the short straw?23:57

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