| xnox | pitti: slangasek: don't we in fact have enough of things ported to satisfy the requirements implied by the versioned dependency? e.g. such that all that's needed is lowering the versioned dependency to the appropriate ubuntuX?! | 00:05 |
|---|---|---|
| stgraber | infinity, cjwatson, slangasek: so if either one of you has a few minutes, I've got a very simple reproducer for the 12.04.5 alternate issue. "apt-get install xserver-xorg-lts-trusty ubuntu-desktop^" | 00:50 |
| stgraber | hmm, actually, failing too but not entirely identical, I suspect doing those two in two steps will give us the same output (trying that now in a container). | 00:53 |
| rcj | stgraber, just an fyi, I'll be doing the release for cloud images since utlemming is at a sprint. | 03:00 |
| stgraber | rcj: ok, I'll ping you when I start working on the announcement. I'm not expecting a release before at least mid-afternoon US/Canada eastern time tomorrow. | 03:04 |
| rcj | stgraber, sounds good. and if you forget and ping ben instead don't worry, his nick triggers on my client | 03:05 |
| stgraber | haha, nice :) | 03:06 |
| === jseutter_ is now known as jseutter | ||
| pitti | slangase`: the reverted dependency bump was harmless in the end; Debian bumped it to ensure to get an NFS fix with systemd, but that's not a big concern right now in utopic | 04:36 |
| pitti | xnox: I don't remember us backporting the NFS fix in sysvinit; but it's not urgent, we'll just land the merge | 04:37 |
| pitti | xnox: NFS under systemd under utopic isn't an urgent matter right now | 04:37 |
| cjwatson | pitti,slangase`: What wasn't harmless, as it turned out, was introducing the systemd-sysv binary package :) | 05:37 |
| pitti | cjwatson: indeed, thanks for the upload | 05:37 |
| cjwatson | That meant that libpam-systemd's systemd-sysv | systemd-shim dep tried to resolve systemd-sysv first, failed, and gave up in a confused heap | 05:37 |
| pitti | right; that worked before because we didn't build systemd-sysv | 05:37 |
| cjwatson | I'm pretty sure my upload wasn't the right answer, but it seems to have unblocked build-deps at least for now | 05:37 |
| cjwatson | As Adam said, having systemd-shim preferred somewhere else might be a better plan | 05:38 |
| pitti | cjwatson: I think the only "better" answer is to make systemd-sysv installable, but I don't think we want this to be the preferred dep until we actually want to switch over | 05:38 |
| pitti | and that might still be a year or so out | 05:38 |
| pitti | so keeping this dep swap seems correct to me | 05:39 |
| cjwatson | hm, ok | 05:39 |
| pitti | cjwatson: Steve/James want systemd-sysv for some images of their's, so they asked for building it now | 05:39 |
| pitti | (tracked at bug 1351306) | 05:39 |
| ubot93 | bug 1351306 in sysvinit "Cannot uninstall upstart and install systemd-sysv" [Undecided,In progress] https://launchpad.net/bugs/1351306 | 05:39 |
| cjwatson | Yeah, familiar with that image | 05:41 |
| cjwatson | Today is Debian import freeze; I'll shut down auto-sync later today | 05:55 |
| === RAOF_ is now known as RAOF | ||
| === yofel_ is now known as yofel | ||
| === slangase` is now known as slangasek | ||
| Mirv | could someone remove qtconnectivity-opensource-src 5.3.1-1 from utopic-proposed so that I could upload a 5.3.0 update instead? | 07:27 |
| Mirv | doko asked me about the various Qt 5.3.1 packages in utopic-proposed and I responded they could be all removed so to make the eventual 5.3.x upgrade via landing PPA instead | 07:28 |
| Mirv | qtscript actually slipped its 5.3.1 to release pocket too, but that didn't hurt. there is a newer Debian version though in -proposed. | 07:29 |
| infinity | Mirv: And why don't we want the newer versions? | 07:31 |
| infinity | Mirv: It's in sync with Debian, that's generally considered a good thing. | 07:32 |
| Mirv | infinity: the Debian's 5.3.1 build-deps on qtbase 5.3.1, and at this point of RTM it seemed risky to update even to newer minor version. some of the modules also can't be upgraded if qtbase stays at 5.3.0, since there are run-time version checks in Qt. | 07:33 |
| infinity | Mirv: Weren't we planning to fork the archive for these very sorts of reasons? | 07:34 |
| infinity | Mirv: That said, new minor versions are probably not a terrible idea at all, just need some testing. | 07:34 |
| Mirv | infinity: yes, it should happen still this week, but I'm not sure about how quickly whan can start to put stuff normally to ubuntu after that. | 07:34 |
| Mirv | infinity: minor versions of Qt have broken some things subtly before, so for RTM that's a no-go | 07:35 |
| infinity | Mirv: changes in any package break things subtly, we can't stop development on everything that's in the image/sdk. | 07:35 |
| Mirv | but regarding -proposed, the testing will need testing via landing PPA anyhow, so it wouldn't hurt not to have those 5.3.1 packages removed from -proposed and land everything (synced or not) from the landing PPA | 07:36 |
| infinity | Mirv: So, yes, I agree that it might want evaluation, and that the forking probably needs to happen soon, but if you're going to single out Qt, I can give you dozens of other packages changed in the last week in Ubuntu to worry about. | 07:36 |
| Mirv | with 5.3.0 I however removed the certain packages from landing PPA before publishing, so that the packages in -proposed would be reused, so that's also possible | 07:36 |
| infinity | Mirv: (Not to freak you out, just saying, this is the process and the model) | 07:36 |
| Mirv | infinity: Qt is slightly problematic because everything is based on it, and if I would be to start upgrading to 5.3.1, it could critically slow developing other things down, since everything depending on eg. qtbase-abi-5-3-1 needs to be locked for the Qt transition. and that's not including the subtle but possibly serious bugs that could be introduced. | 07:37 |
| Mirv | so timeline wise, it doesn't fit, but it should be ok as soon as we have the archive forked | 07:38 |
| infinity | Mirv: So, when the archive is forked, do your 5.3.0 uploads in the forked archive? :) | 07:38 |
| infinity | Mirv: Given that it's all FTBFS in proposed, it's not hurting anyone. | 07:39 |
| infinity | But given that it's all FTBFS, I guess I can remove it too. *shrug* | 07:39 |
| Mirv | infinity: actually this upload is not for Ubuntu, but more like for Kubuntu. Ubuntu doesn't even use it, but the package if basically Bluetooth package without working Bluetooth support at the moment. | 07:39 |
| infinity | It'll autosync the next time it's updated in Debian, though. | 07:39 |
| infinity | Because we're in sync. | 07:39 |
| cjwatson | RTM's only supposed to get stuff that's been already tested in Ubuntu, generally | 07:39 |
| infinity | Which is good. | 07:39 |
| infinity | Not bad. | 07:39 |
| Mirv | infinity: oh, there's autosync after sync period for packages that are already in sync? | 07:39 |
| * cjwatson fails to parse that question | 07:40 | |
| Mirv | anyway, I just figured out KDE people might want a working Qt Bluetooth module, even though we don't use it | 07:40 |
| infinity | DebianImportFreeze is, actually, today. :P | 07:40 |
| cjwatson | Right | 07:40 |
| infinity | So, I suppose if we remove this and Colin turns off the cron job, only a manual sync will bring it back. | 07:41 |
| cjwatson | I'll flip cron this evening | 07:41 |
| infinity | But you can argue with the Kubuntu folks, since I bet they want it. | 07:41 |
| Mirv | wow, when it has been moved to be so late :) I thought it was in June. ok, that explains it. | 07:41 |
| infinity | And rightly so. | 07:41 |
| cjwatson | We've been pushing it later as the economics work out much better this way. | 07:41 |
| Mirv | yes, I agree it's much nicer this way, thanks for that | 07:42 |
| mlankhorst | cjwatson: did you find out more about bug 1351262 ? | 07:42 |
| ubot93 | bug 1351262 in xorg-lts-transitional "precise alternate installations fail with unmet deps due to the conflict ' xserver-xorg-lts-trusty : Conflicts: libgl1-mesa-dri (>= 0~)'" [Critical,Confirmed] https://launchpad.net/bugs/1351262 | 07:42 |
| cjwatson | mlankhorst: No, sorry, have had no time | 07:42 |
| mlankhorst | ok np | 07:42 |
| cjwatson | I wasn't expecting to be involved in 12.04.5 ... | 07:42 |
| mlankhorst | well I don't know how to create iso's | 07:42 |
| cjwatson | You don't need to | 07:43 |
| cjwatson | 01:50 <stgraber> infinity, cjwatson, slangasek: so if either one of you has a few minutes, I've got a very simple reproducer for the 12.04.5 alternate issue. "apt-get install xserver-xorg-lts-trusty ubuntu-desktop^" | 07:43 |
| cjwatson | 01:53 <stgraber> hmm, actually, failing too but not entirely identical, I suspect doing those two in two steps will give us the same output (trying that now in a container). | 07:43 |
| zequence | stgraber: Yes, we will participate in 12.04.05 | 07:43 |
| mlankhorst | cjwatson: ah ok :-) | 07:43 |
| cjwatson | I'm pretty certain trying to create ISO images would be a dddistraction for this | 07:44 |
| mlankhorst | yeah | 07:45 |
| mlankhorst | but how do I get libgl1-mesa-dri/glx removed from the ubuntu-desktop task? | 07:45 |
| cjwatson | You can't | 07:45 |
| cjwatson | We can't change tasks after release | 07:46 |
| cjwatson | This is why we use metapackages for point releases | 07:46 |
| cjwatson | (So yeah, actually stgraber's test isn't hugely meaningful, then) | 07:46 |
| cjwatson | mlankhorst: ok, what image is this affecting, and how can I reproduce the problem? | 07:47 |
| mlankhorst | but isn't that what the alternate cd does? select the ubuntu-desktop task and add the lts-trusty stuff | 07:48 |
| mlankhorst | cjwatson: it affects the alternate cd, I've tested with 20140805.1 because later had the no modules found bug | 07:48 |
| cjwatson | oh, well, we do an independent germinate run for the alternate CD, we're not reliant on archive tasks | 07:48 |
| cjwatson | mlankhorst: How can I reproduce this problem? | 07:50 |
| mlankhorst | cjwatson: simply install the image in a vm, it will fail with that error | 07:50 |
| === doko_ is now known as doko | ||
| cjwatson | if the kernel modules are out of sync then I think I need to fix that first | 07:51 |
| mlankhorst | https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1353086 | 07:52 |
| ubot93 | Launchpad bug 1353086 in ubiquity "no kernel modules found" [Undecided,Confirmed] | 07:52 |
| cjwatson | gar why is that on ubiquity | 07:53 |
| * cjwatson moves | 07:53 | |
| jibel | mlankhorst, really, I think kernel module is fixed in 20140806 | 07:55 |
| mlankhorst | ok | 07:55 |
| mlankhorst | I' ll try | 07:55 |
| jibel | cjwatson, before you start anything on "no kernel module found" let me confirm if it is fixed or not | 07:55 |
| cjwatson | Certainly looks like it should be fine | 07:56 |
| cjwatson | uname -a matches udebs | 07:56 |
| mlankhorst | ok | 07:56 |
| jibel | cjwatson, mlankhorst I confirm that on alternate 20140806 kernel and modules match, and the remaining issue is the installation of the task ubuntu-desktop | 08:02 |
| mlankhorst | jibel: yeah I can reproduce the error message by simply doing apt-get install --install-recommends xserver-xorg-lts-trusty libgl1-mesa-{dri,glx}{,-lts-trusty} libglapi-mesa{,-lts-trusty} | 08:04 |
| mlankhorst | in a pbuilder chroot | 08:04 |
| mlankhorst | so the real problem is having libgl1-mesa-dri, libgl1-mesa-glx and libglapi-mesa on that iso | 08:06 |
| Riddelll | stgraber: what's new in the images today? | 08:06 |
| jibel | Riddelll, last respin was a fix to dmidecode to support smbios > 2.7 and not crash ubiquity | 08:11 |
| Riddelll | thanks | 08:12 |
| === Riddelll is now known as Riddell | ||
| jibel | psivaa, precise lamp is still unstable, you didn't push your fix? | 08:25 |
| psivaa | jibel: ohh, sorry you must have missed my message about it: | 08:27 |
| psivaa | Aug 01 11:06:50 <psivaa>jibel: https://code.launchpad.net/~ubuntu-server-iso-testing-dev/+junk/server-tests-precise is being a junk branch, i've created another junk with these changes | 08:27 |
| psivaa | Aug 01 11:06:51 <psivaa> http://bazaar.launchpad.net/~psivaa/+junk/precise-lamp-test-fixes/revision/271 | 08:27 |
| jibel | psivaa, I know you have a fix in your junk but it'd be more useful in production :) | 08:28 |
| psivaa | jibel: the production tests are themselves are using a junk branch ( which i agree is questionable) and i can not push to that branch | 08:29 |
| psivaa | jibel: i can create another project for this but that'd require all the precise jobs to be regenerated | 08:30 |
| jibel | psivaa, now you can | 08:30 |
| psivaa | jibel: great. thanks. will push it | 08:30 |
| cjwatson | mlankhorst: right, I suspect I'll need to tweak cdimage or possibly the seeds for this, but just trying to carve out enough sprint time to test it | 08:31 |
| mlankhorst | ok | 08:31 |
| mlankhorst | thanks | 08:31 |
| jibel | psivaa, are server tests not using utah? | 08:32 |
| psivaa | jibel: precise ones dont use uta | 08:32 |
| psivaa | *utah | 08:32 |
| jibel | psivaa, ah, I didn't realize it was still using the old ubuntu-server-iso-testing. And who was maintaining it for the last 2 years? | 08:33 |
| pitti | cjwatson: ah, systemd-sysv is known to be uninstallable, hence britney doesn't let it propagate; is it possible to override that, or do we need to wait on the sysvinit merge? | 08:33 |
| jibel | (I guess the answer is no one :)) | 08:33 |
| psivaa | plars: i dint think it needed much maintenance for precise tests, except couchdb failing a couple of times or so.. i with the help of retoaded do fix them | 08:34 |
| xnox | pitti: it's best to fix it, since increasing total uninstallable count can lead to britney later trading one uninstallable for another and causing havoc... | 08:34 |
| jibel | psivaa, anyway, thanks for the fix. let me know when I can retrigger LAMP tests | 08:35 |
| cjwatson | pitti: Not sure | 08:36 |
| cjwatson | It's possible, but as xnox says possibly problematic | 08:36 |
| cjwatson | Hm, I wonder what forces xserver-xorg-lts-trusty to be pulled in in the first place | 08:36 |
| pitti | cjwatson, xnox: ok; I suppose we could do another sysvinit upload with just that change, and I'll re-re-re-re-redo the sysvinit merge | 08:36 |
| psivaa | jibel: np. in the landing meeting. will do it in ~20 mins | 08:37 |
| cjwatson | mlankhorst: I think you could fix this by changing the preferred libgl1 alternative in xserver-xorg-core-lts-trusty | 08:37 |
| cjwatson | It's currently libgl1-mesa-glx | 08:37 |
| xnox | pitti: shouldn't it just be s/sysvinit-core// ? cause sysvinit-core provides halt/init/poweroff/reboot/runlevel/shutdown/telinit all of which systemd-sysv is suppose to provide and in fact declares replaces on it.... | 08:38 |
| xnox | wait. read it wrong. | 08:38 |
| pitti | xnox: it's because our current initscripts still hard-depends on upstart and mountall | 08:38 |
| mlankhorst | cjwatson: ok I'll look, last I checked the packaging had the correct one at least, maybe it needs a similar hack? | 08:38 |
| mlankhorst | oh DERP | 08:38 |
| xnox | pitti: ah, i see. | 08:38 |
| pitti | xnox: which is fixed in my pending merge; I'll prepare an upload with just that | 08:38 |
| pitti | to get that out of the way | 08:39 |
| mlankhorst | probably happens at shlibdeps time | 08:39 |
| mlankhorst | libglx adds it, lets see if it happened in saucy too | 08:42 |
| mlankhorst | odd, it should have been | 08:42 |
| pitti | xnox: ok, done in https://launchpad.net/ubuntu/+source/sysvinit/2.88dsf-41ubuntu17 | 08:42 |
| xnox | cjwatson: stgraber: wubi for 12.04.5 for what it's worth http://people.canonical.com/~xnox/wubi/precise/wubi-r281-precise-signed.exe | 08:44 |
| cjwatson | xnox: hm, I guess I need to install that and any images that have it need to be rebuilt ... | 08:45 |
| cjwatson | I lose track | 08:45 |
| xnox | it was signed and done yesterday, but i didn't get any emails from RT =/ =( | 08:46 |
| xnox | so it could have been on yesterday's respin, oh well. | 08:46 |
| cjwatson | copied, http://people.canonical.com/~ubuntu-archive/wubi/precise/stable symlink repointed | 08:48 |
| psivaa | jibel: i've triggered the lamp jobs. and sorry i have no idea why i pinged Paul in the earlier message :/. must be muscle memory | 09:01 |
| psivaa | jibel: and the dashboard is happy now :) | 09:04 |
| mlankhorst | can someone drop xorg-server-lts-trusty_1.15.1-0ubuntu2.1~precise1 ? it will conflict with the fixed xorg-server | 09:19 |
| mlankhorst | it's in precise's NEW queue | 09:21 |
| === pete-woods1 is now known as pete-woods | ||
| mlankhorst | cjwatson: ^ enjoy | 09:41 |
| === bluesabre_ is now known as bluesabre | ||
| xnox | cjwatson: ubiquity in precise doesn't look like it was rebuild since the last https enablements: apt-setup 1:0.55ubuntu4.1 base-installer 1.122ubuntu7.4 choose-mirror 2.39ubuntu4.1 debian-installer-utils 1.88ubuntu2.2 | 09:53 |
| xnox | do we need those in ubiquity? | 09:53 |
| cjwatson | I'm not sure, check the changelogs? is it just https? | 09:55 |
| cjwatson | I'm not hugely bothered about https there if it otherwise works | 09:55 |
| jibel | psivaa, thank you | 10:06 |
| psivaa | jibel: yw | 10:06 |
| wgrant | cjwatson: Do we care about the special case in edit-acl's query command which uses all_distro_archives if no archive is given? The other commands all just default to the primary archive. | 10:28 |
| cjwatson | wgrant: It's sort of useful to be able to see them all, but maybe it should be a separate option instead | 10:39 |
| cjwatson | (Because otherwise I forget about the existence of partner or something) | 10:40 |
| === Pici` is now known as Pici | ||
| === ogra_` is now known as ogra_ | ||
| doko | Riddell, cjwatson: does one of you look at the kde migration? | 11:10 |
| Riddell | doko: I've looked at it and my eyes started bleeding, I don't know where else to look | 11:11 |
| Riddell | doko: I removed some nepomuk packages because that's gone in this version of KDE SC | 11:11 |
| Riddell | but it didn't help | 11:11 |
| mlankhorst | jibel: can you accept xorg-server-lts-trusty? it should fix your iso, if lucky | 11:16 |
| jibel | I cannot do that, I am just QA not archive admin :) | 11:17 |
| mlankhorst | ok | 11:17 |
| cjwatson | mlankhorst: that's a biggish diff against current precise - do you think it can be verified in time? | 11:41 |
| mlankhorst | er? | 11:41 |
| mlankhorst | it's 4 lines | 11:41 |
| cjwatson | I was expecting a diff for just this change | 11:41 |
| cjwatson | http://paste.ubuntu.com/7979052/ | 11:42 |
| mlankhorst | 2:1.15.1-0ubuntu2~precise2 | 11:42 |
| cjwatson | Er, hm, where's that I wonder | 11:42 |
| cjwatson | Oh, unapproved not new | 11:42 |
| mlankhorst | weird, how did that end in unapproved :P | 11:42 |
| cjwatson | I think that LP bug was fixed | 11:42 |
| mlankhorst | ah | 11:42 |
| cjwatson | mlankhorst: William confirms that's fixed now | 11:44 |
| mlankhorst | just delete the one in NEW, I'll see if I can get in the 2.1 later | 11:44 |
| cjwatson | I did | 11:44 |
| cjwatson | Bah, sru comment didn't happen because wrong source, will fix | 11:46 |
| cjwatson | ah comment but not bug source | 11:47 |
| cjwatson | status. whatever | 11:47 |
| jibel | jamespag`, if you want to smoketest iscsi on 12.04.5 server, it's the only test missing with maas | 11:49 |
| === pete-woods is now known as pete-woods|lunch | ||
| === pete-woods|lunch is now known as pete-woods | ||
| * stgraber reads backlog | 13:29 | |
| mlankhorst | so with xorg-server-lts-trusty in -proposed, time to respin alternate? | 13:32 |
| mlankhorst | # apt-cache show xserver-xorg-core-lts-trusty |grep glx | 13:33 |
| mlankhorst | Depends: xserver-common-lts-trusty (>= 2:1.15.1-0ubuntu2~precise2), keyboard-configuration, udev (>= 149), libc6 (>= 2.15), libdrm2 (>= 2.3.1), libgcrypt11 (>= 1.4.5), libgl1-mesa-glx-lts-trusty | libgl1, libpciaccess0 (>= 0.12.902), libpixman-1-0 (>= 0.30.0), libudev0 (>= 161), libxau6, libxdmcp6, libxfont1 (>= 1:1.4.2) | 13:33 |
| mlankhorst | that seems to work at least | 13:33 |
| stgraber | mlankhorst: no, we need to move it to -updates first | 13:36 |
| mlankhorst | ah | 13:36 |
| mlankhorst | ok let me see if it runs locally | 13:37 |
| stgraber | mlankhorst: can you just install that one from -proposed in a chroot or something to confirm it installs fine? | 13:37 |
| stgraber | mlankhorst: if it does, I'll consider that verification-done and will copy it to -updates | 13:37 |
| mlankhorst | it installs, but I'm still on precise so I can check if it runs too | 13:37 |
| stgraber | xnox: so wrt wubi I'm a bit lost, did we still let people install using wubi in 12.04? if so, I suspect we indeed need a respin of everything that includes it... | 13:38 |
| * stgraber doesn't like the idea of a mass respin this late, it'll be very hard to still release today if we do | 13:38 | |
| xnox | stgraber: i believe in 12.04 it is still supported. I do know that without update the old wubi will break, as it will keep on looking for 12.04.4 | 13:44 |
| mlankhorst | stgraber: go ahead, verification-done | 13:44 |
| xnox | stgraber: i thought cjwatson was pondering on respins, not sure if he did that or not. | 13:44 |
| mlankhorst | still boots -core on my system | 13:44 |
| stgraber | xnox: damn, ok... | 13:45 |
| stgraber | xnox: I'm not seeing any respins, so I guess I'll release mlankhorst's upload to -updates and then do a mass respin | 13:45 |
| stgraber | zequence, Riddell, jibel: bad news, we're heading to another respin in about an hour. Affected will be ubuntu alternate and anything which contains wubi (fix for 12.04.5) or xserver-xorg-core-lts-trusty (to be in sync with the source image) | 13:47 |
| Riddell | good thing I haven't started testing today yet :) | 13:48 |
| stgraber | so at this point, I'm expecting a very very late 12.04.5 release, still kinda hoping to make it today but I guess tomorrow isn't entirely out of the question at this point :( | 13:49 |
| jibel | stgraber, for desktop its just the cd not the livefs that is affected or are you rebuilding the livefs too? | 13:49 |
| stgraber | jibel: I suspect we have xserver-xorg-core-lts-trusty in the desktop livefs | 13:51 |
| jibel | :( | 13:51 |
| jibel | right | 13:51 |
| jibel | the day will be long | 13:51 |
| stgraber | realistically, if you already tested the current one, just re-test wubi and test a single install, if that works, you're good. It's a LTS release and I know we haven't released any other thing in between, so if the image isn't corrupted, past results should stand. | 13:52 |
| jibel | stgraber, realistically I won't be able to do much more than this. | 13:53 |
| jibel | server, netboot and core looks good at least :) | 13:54 |
| stgraber | yay, some good news! | 13:55 |
| stgraber | can you mark those as ready on the tracker? | 13:55 |
| * stgraber figures out the list of needed respins | 14:04 | |
| stgraber | ok, so we have some good news there, though not for jibel | 14:09 |
| stgraber | Edubuntu DVD => full respin | 14:10 |
| stgraber | Kubuntu desktop => iso respin (same livefs) | 14:10 |
| stgraber | Ubuntu alternate => full respin | 14:10 |
| stgraber | Ubuntu desktop => full respin | 14:10 |
| stgraber | Ubuntu DVD => full respin | 14:10 |
| stgraber | Ubuntu Studio DVD => iso respin (same livefs) | 14:10 |
| stgraber | Now I just need to figure out how we do an iso respin only in the new world order | 14:11 |
| stgraber | oh, I guess just not passing --live should do the trick | 14:12 |
| stgraber | cjwatson: are you around to confirm? | 14:12 |
| Riddell | stgraber: kubuntu alternate no respin? | 14:27 |
| stgraber | nope, not needed as it doesn't contain wubi and doesn't have the HWE X stack | 14:28 |
| * Riddell gets on with testing | 14:28 | |
| stgraber | and the ones I listed above as iso respin only, means that you should just give a very limited spot check to make sure the livefs didn't get corrupted, but any existing testing you've done should still apply | 14:30 |
| cjwatson | stgraber: confirmed | 14:46 |
| cjwatson | stgraber: for alternates, suggest running: germinate -s ubuntu.precise -d precise,precise-updates,precise-proposed -a amd64 --no-rdepends | 14:46 |
| cjwatson | and check that libgl1-mesa-{dri,glx} aren't in desktop-common or desktop, only the -lts-trusty variants | 14:47 |
| stgraber | cjwatson: only seeing the lts-trusty ones, so looks like we're good | 14:52 |
| cjwatson | Yay | 14:56 |
| stgraber | ok, so rebuilding the wubi-only images now since those are trivial, still waiting for xserver-xorg-lts-trusty to publish | 14:59 |
| * stgraber grabs the new kubuntu image to confirm wubi matches xnox's build | 15:19 | |
| * stgraber spins up a new ubuntu alternate | 15:23 | |
| stgraber | confirmed that wubi on the image and wubi on people.u.c have the same md5, so things look good | 15:27 |
| stgraber | Riddell: you should be all set for 12.04.5 at last | 15:27 |
| Riddell | stgraber: lovely | 15:29 |
| === psivaa is now known as psivaa-afk-bbl | ||
| stgraber | jibel: there you go, hopefully those will actually work ^ | 15:34 |
| stgraber | zequence: those studio images should be your 12.04.5 images unless you notice a major issue with them. As I said above, only the iso was respun, not the livefs, so a simple spot check is enough if you already tested the previous one. | 15:35 |
| jibel | hopefully | 15:35 |
| stgraber | current status is that I'm doing an alternate test install to confirm the X issue has been resolved, once that's confirmed, I'll respin the rest of the affected images | 15:44 |
| jibel | stgraber, don't waste your time on this I'm on it | 15:45 |
| stgraber | jibel: how far are you with your install? | 15:45 |
| stgraber | here it just went past the point where it'd usually fail in the past! | 15:51 |
| jibel | stgraber, here too | 15:51 |
| stgraber | starting the respins of edubuntu dvd, ubuntu desktop and ubuntu dvd | 15:51 |
| stgraber | all building now | 15:52 |
| stgraber | and also building a new wubi live image (forgot that one in my earlier list) | 15:53 |
| stgraber | zequence, Riddell: What will the URL be for your respective announcements? | 16:09 |
| Riddell | http://www.kubuntu.org/news/kubuntu-12.04.5 | 16:11 |
| stgraber | draft of the announcement: http://pad.ubuntu.com/12-04-5-announcement | 16:14 |
| Riddell | groovy | 16:15 |
| Riddell | stgraber: when do you want to release it? | 16:15 |
| stgraber | hopefully still today, or at least today for North America. My hope is for within the next 6 hours | 16:16 |
| stgraber | but that entirely depends on testing | 16:16 |
| * stgraber does the usual wiki mangling for the point release | 16:17 | |
| Riddell | groovy, I've marked kubuntu as ready | 16:17 |
| stgraber | yay! | 16:18 |
| rcj | stgraber, no respin planned for cloud images, just fyi but you probably figured that. :) | 16:18 |
| Riddell | but it's only me who's tested them so could do with some others, hoping for some helpers later on | 16:18 |
| stgraber | rcj: yeah, I figured you weren't affected by crazy windows or X issues :) | 16:18 |
| stgraber | Riddell: same thing for Edubuntu, I'm the only tester for this point release, though I'm not too concerned as it's not massively different from 12.04.4 | 16:19 |
| Riddell | stgraber: sgclark is going to be the release contact for Kubuntu this evening, do give her a ping when it goes out so she knows to publish the news story | 16:27 |
| sgclark | hello! | 16:27 |
| stgraber | Riddell, sgclark: ok, will do! | 16:29 |
| stgraber | wiki should be good now except for the list of changes from 12.04.4 which I'm still compiling | 16:34 |
| stgraber | cjwatson: if you're still around, DIST=precise cron.source is failing somehow... | 16:35 |
| stgraber | actually, let me try to call it the right way first, maybe that'll do the trick :) | 16:41 |
| stgraber | hmm, no, this is the right way (looking at bash history from what infinity did for 14.04.1) | 16:42 |
| stgraber | ok, let's try an ugly workaround for now | 16:47 |
| === charles_ is now known as charles | ||
| stgraber | who's around from Ubuntu Studio? | 17:07 |
| zequence | stgraber: I am | 17:07 |
| zequence | stgraber: Our release link http://ubuntustudio.org/2014/08/ubuntu-studio-12-04-5-point-release | 17:08 |
| stgraber | zequence: thanks | 17:08 |
| stgraber | ok, so we just need to wait for Edubuntu to finish building + me to do a quick re-test on both and then hopefully jibel will have somehow tested everything else by then ;) | 17:08 |
| jibel | I asked someone to test wubi, and I'm reviewing desktop images | 17:09 |
| stgraber | cool, thanks | 17:12 |
| highvoltage | cool | 17:36 |
| highvoltage | stgraber: I also just finished my workshop but will be driving home in a few mins. everything ok? | 17:36 |
| stgraber | highvoltage: yeah, the 12.04.5 candidate images for Edubuntu just finished building, downloading them now | 17:39 |
| highvoltage | it's weird it feels like last week that 12.04.4 was released. time flies too fast. | 17:41 |
| highvoltage | ah I'm thinking of the 14.04 point release | 17:41 |
| stgraber | yeah, we had 14.04.1, alpha-2 of 14.10 and 12.04.5 really close by | 17:42 |
| jibel | and B1 in 3 weeks | 17:54 |
| jibel | stgraber, alternate is okay but I didn't test LTSP | 17:55 |
| jibel | stgraber, I'm taking a break now and be back later to review desktop and DVD | 17:56 |
| jibel | and hopefully update results of Wubi | 17:56 |
| mlankhorst | is the alternate cd built? | 17:57 |
| jibel | mlankhorst, yes and tested | 17:57 |
| mlankhorst | works? | 17:58 |
| jibel | mlankhorst, after installation I've a desktop, so the test is positive :) | 17:58 |
| mlankhorst | good | 17:58 |
| jibel | mlankhorst, thanks for your help. | 17:58 |
| mlankhorst | np | 17:58 |
| jibel | stgraber, I won't have results for Mac. Installation fails on the Mac used by the team | 17:59 |
| stgraber | jibel: ok | 18:03 |
| stgraber | jibel: we'll see if someone else can give us a positive result for it, otherwise we'll just skip +mac I guess | 18:03 |
| === chuck_ is now known as zul_ | ||
| stgraber | xnox: around? | 19:48 |
| stgraber | right, so wubi is wrong and we'll need another iso-only respin for that | 19:59 |
| stgraber | xnox is currently busy (just called him) but should be able to come online and get a new build in the next 30min or so | 19:59 |
| stgraber | I'll try to build it myself here though as it doesn't seem to be very difficult to do | 20:00 |
| stgraber | Riddell, sgclark, zequence, jibel: sorry but this will mean another mass rebuild hopefully in the next hour. Note that this is only going to be the ISO and not the livefs, so a simple boot test to confirm your iso isn't corrupted will be sufficient. | 20:02 |
| stgraber | affected products are: edubuntu dvd, kubuntu desktop, ubuntu desktop, ubuntu dvd and ubuntu studio dvd | 20:03 |
| jibel | stgraber, and a test of wubi of course | 20:08 |
| stgraber | xnox: pycrypto download url is broken and you need to have grub-pc installed apparently... | 20:15 |
| stgraber | and it now failed for yet another reason... retrying the whole thing again | 20:16 |
| stgraber | yay, got a wubi.exe | 20:26 |
| stgraber | and that does say 12.04.5! | 20:26 |
| xnox | stgraber: hey | 20:27 |
| xnox | stgraber: online now | 20:27 |
| xnox | stgraber: oh =( yeah pycrypto i was getting it from a new location recently. | 20:27 |
| xnox | =) | 20:27 |
| xnox | stgraber: and i guess i owe doing testing past respin.... | 20:27 |
| xnox | yeap, my 281 locall signed binary says .4 & 280 =( | 20:27 |
| xnox | i did what ev did the other time. upload wrong revision wubi and go off motor cycling. | 20:28 |
| xnox | sorry. | 20:28 |
| stgraber | xnox: also, pylauncher was pretty unhappy here, I had to port it from os.system to subprocess.call for it to work | 20:29 |
| stgraber | anyway, pushed both fixes back to the branch in case anyone ever needs to do that again | 20:30 |
| stgraber | and now very much hoping someone from IS has a minute to sign that thing | 20:30 |
| xnox | stgraber: top contributor to wubi! Congrats, so I'm guessing you will be building it from now on =) 2 years of that was enough for me ;-) | 20:31 |
| xnox | ev handed over to me, when he build it with wrong version..... history seems to repeat itself | 20:31 |
| stgraber | jibel, zequence, Riddell, sgclark: ok, so what do you prefer, another testing blitz in an hour or so and release today or just go get some rest and we pick this up tomorrow and release a day late? | 20:32 |
| jibel | stgraber, I don't mind retrying in 1h | 20:32 |
| sgclark | does not matter to me | 20:32 |
| jibel | and it's just 8:30AM on Baker Island we have plenty of time to retry and release on the 7th | 20:32 |
| stgraber | ok, good, so I'll ping all of you as soon as I get hold of someone at Canonical IS to sign that new wubi | 20:33 |
| stgraber | jibel: in case you can get someone to pre smoke-test it for you: http://people.canonical.com/~stgraber/wubi-12.04.5.exe | 20:33 |
| sgclark | I don't think many will be around for testing on our part today, late already for most | 20:33 |
| stgraber | jibel: it's not signed so Windows may complain about it | 20:33 |
| stgraber | sgclark: so since we won't be rebuilding the livefs, you'll just need someone to boot the two affected images and make sure they're not corrupt. | 20:34 |
| jibel | stgraber, I'll wait the iso just in case someone has the funny idea to put the previous version or sign it with the wrong key and who knows what else can happen now | 20:34 |
| sgclark | stgraber: where are these images? sorry. thought I was just doing a post :) | 20:35 |
| stgraber | sgclark: queuebot will announce them here when they're built and you'll be able to download them from http://iso.qa.ubuntu.com/qatracker/milestones/320/builds but I should have enough time to give you a hand with testing those whenever we're actually ready to build them | 20:37 |
| sgclark | ok thanks | 20:37 |
| stgraber | going away for a few minutes (food), hoping to get a response from IS by the time I get back | 20:37 |
| stgraber | xnox: at some point between ev and you, I've been doing wubi builds, but that's a part of my life I prefer to pretend never happened :) | 20:38 |
| stgraber | we've got a signed wubi: https://chinstrap.canonical.com/~rfinnie/wubi-12.04.5-signed.exe | 20:53 |
| stgraber | moving it into place and starting the rebuilds | 20:53 |
| stgraber | images are rebuilding now, ETA is ~15min | 20:58 |
| stgraber | jibel, zequence, Riddell, sgclark: ^ | 20:58 |
| mlankhorst | will it still be ready in time for today? :P | 20:59 |
| stgraber | jibel: new md5sum is 9a83deb1fa55e4b4811730880ce71788 | 20:59 |
| stgraber | mlankhorst: for me, sure, for Europeans, probably not :) | 20:59 |
| jibel | 1h to release on the 7th in France | 21:01 |
| mlankhorst | yeah | 21:03 |
| * xnox goes for the dvds | 21:18 | |
| stgraber | and of course as usual edubuntu will publish last :) | 21:22 |
| sgclark | ok so which are the two affected? | 21:22 |
| stgraber | sgclark: the two I rebuilt, Kubuntu Desktop amd64 and Kubuntu Desktop i386 | 21:22 |
| stgraber | sgclark: http://cdimage.ubuntu.com/kubuntu/precise/daily-live/20140807.1/ | 21:23 |
| sgclark | ty | 21:23 |
| stgraber | sgclark: that's precise-desktop-amd64.iso and precise-desktop-i386.iso | 21:23 |
| xnox | hm. | 21:23 |
| * xnox is trying to run wubi.... | 21:24 | |
| stgraber | xnox: does it at least say 12.04.5 now? | 21:24 |
| xnox | it fails to unpack for me, but i have newer wubi here installed again. Let me reclone pristine windows 7. | 21:24 |
| stgraber | hmm, the .exe did start fine here | 21:26 |
| * stgraber grabs the smallest iso he can find to test | 21:26 | |
| * xnox is slowly cloning a fresh 30GB win7 VM | 21:27 | |
| jibel | I asked nuclearbob to verify wubi on a real windows host | 21:31 |
| jibel | yay alternate ltsp passed | 21:36 |
| stgraber | signature is valid here and it appears to start, though that's a Windows server that's running out of disk space so it's not going very far :) | 21:37 |
| stgraber | jibel: that's impressive, it usually always fails for a reason or another! :) | 21:37 |
| stgraber | well, there was an obvious fail with LTSP but I fixed it two days ago (usual upload to switch it to the new kernels) | 21:38 |
| jibel | stgraber, usually it fails because it picks the wrong nic | 21:38 |
| jibel | lets do another round of desktop and dvds | 21:38 |
| * stgraber grabs Edubuntu and re-tests those two one more time | 21:39 | |
| stgraber | well, first have to wait for them to sync to cdimage I guess... | 21:40 |
| jibel | stgraber, do we agree to say if an image boots and installation is successful then all the test passed on previous image are still valid? | 21:41 |
| stgraber | jibel: yes | 21:41 |
| stgraber | livefs is identical so if it's not corrupt, we're good | 21:41 |
| jibel | good | 21:41 |
| xnox | dvd/i386 is all good for wubi | 21:43 |
| xnox | (in a fresh windows7 vm) | 21:43 |
| stgraber | yay! | 21:44 |
| xnox | ubuntu amd64 desktop & dvd boot in UEFI modes fine. | 21:49 |
| jibel | and in bios mode too :) | 21:50 |
| * xnox likes trusty hwe stack. Got widescreen in qemu =) | 21:51 | |
| jibel | ubuntu desktop is ok, untested screen reader and vmware easyinstall | 21:53 |
| stgraber | finally getting there! :) | 21:55 |
| jibel | Mac images are untested. Only test we did, didn't pass the partitioning. It's a no-go for QA | 21:56 |
| stgraber | ok, I won't be releasing them then | 21:56 |
| * xnox has no macs =( | 21:56 | |
| stgraber | if someone complains about it and does testing for us, we can always release them later | 21:56 |
| * xnox grabs studio images | 21:56 | |
| stgraber | xnox: thanks! I haven't heard from the studio guys in a while so I was actually planning on doing the spot check for them once I'm done with Edubuntu. | 21:57 |
| * xnox ponders about how pointless it is to GPG sing MD5SUMS file | 21:58 | |
| xnox | well SHA256SUMS is also signed, so it's alright. | 21:58 |
| stgraber | yeah, I wouldn't really recommend anyone use MD5SUMS to check authenticity, using it to check integrity is fine though (but then you don't care about the signed one) | 21:59 |
| xnox | i like studio theme, but i'm getting OCD about all the minor ux bugs that i fixed in ubiquity since, in precise. | 22:03 |
| stgraber | :) | 22:03 |
| xnox | live session is fine on amd64, finishing up install and then will do i386. | 22:03 |
| stgraber | edubuntu amd64 live and ltsp are good, doing a standard install to confirm the media is clean | 22:04 |
| jibel | ubuntu dvd amd64 is ok | 22:05 |
| sgclark | kubuntu amd64 good, now testing i386 | 22:07 |
| stgraber | I'm doing pre-publishing of everything that's on the tracker except for ubuntu mac images | 22:08 |
| xnox | i can draw a doodle in gimp post ubuntustudio installation -> perfect. | 22:09 |
| xnox | stgraber: we are not respining livefs for openssl security fix just pushed out....?! | 22:10 |
| xnox | where just is 2h ago. | 22:10 |
| xnox | (for updates, 3h for security pocket) | 22:11 |
| stgraber | nope | 22:12 |
| xnox | good =) | 22:12 |
| stgraber | if we had to respin every time there was an openssl security fix... :) | 22:12 |
| xnox | ... we'd never be able to release </troll> =) | 22:13 |
| stgraber | pre-publishing done | 22:14 |
| stgraber | xnox: ok, so what did you test so I update the tracker with those? | 22:15 |
| xnox | stgraber: yes. | 22:15 |
| xnox | studio i386 in progress, will finish soon. | 22:15 |
| xnox | i've updated tracker with green test results for things i've tested. | 22:15 |
| stgraber | oh, I see, marking that one ready then | 22:15 |
| xnox | ubuntu dvds should be ready | 22:15 |
| xnox | as well | 22:15 |
| stgraber | I'll let jibel flip the magic switch for those since he's around :) | 22:16 |
| jibel | done | 22:16 |
| stgraber | cool, so just waiting for studio i386, kubuntu i386 and both edubuntus (which take a bloody long time to install...) | 22:17 |
| stgraber | I'll soon start final publishing of everything to let time for cdimage mirrors and bittorrent to catch up, by the time that's done, I should be done with edubuntu | 22:18 |
| jibel | 13 contributors have provided 290 results and covered 217 testcases | 22:19 |
| xnox | will we have to test torrents as well.... or they are now "can be assumed to be operational" | 22:19 |
| jibel | and 1 contributor is a bot | 22:19 |
| stgraber | xnox: I usually load a bunch in my client here and wait to see when they start working. The server is freakishly slow so it needs a good hour for the initial hashing. | 22:19 |
| stgraber | so with that said, my current ETA for release is midnight UTC (1 hour and 40min from now). | 22:20 |
| stgraber | Riddell, sgclark, zequence: ^ | 22:20 |
| sgclark | ok. Riddell is long asleep by now I suspect but I will be here, almost done with i386 | 22:21 |
| xnox | ubuntu studio dvd i386 is ready, all good. | 22:23 |
| sgclark | kubuntu i386 good | 22:27 |
| stgraber | ok, beginning publishing now | 22:28 |
| stgraber | 1.5h should be plenty for bittorrent to catch-up | 22:28 |
| jibel | testing report sent | 22:29 |
| stgraber | jibel: perfect, thanks for staying up so late! | 22:29 |
| jibel | I've had my share of testing for today | 22:29 |
| jibel | stgraber, if you don't need me for anything else I'm heading to bed | 22:30 |
| jibel | good night everyone ! | 22:30 |
| stgraber | jibel: nope, all good, good night! | 22:30 |
| xnox | yeah, me too. good night. sorry for the mistake =( | 22:31 |
| stgraber | xnox: good night! Thanks for helping with the testing :) | 22:32 |
| stgraber | rcj: FYI, release in 1h15 (midnight UTC) | 22:45 |
| rcj | stgraber, sounds good, thanks. | 23:06 |
| stgraber | and at last, we've got test results for everything. | 23:15 |
| stgraber | I just pushed out the images to the mirrors, so that should be syncing about now and bittorrent should have enough time to catch up in the next 45min | 23:15 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!