[00:06] skaet: do you have a few minutes to spare? [01:28] slangasek: Sorry, i've been afk. It's a well tested upload, that isn't *entirely* beta critical.. as in, does not warrant a respin. It is an opportunistic include, which cannot impact other testcases. [01:29] NCommander: ^^ [01:30] Daviey: it sounds to me like NCommander's biggest concern was that even taking it opportunistically would be a problem for him on ARM maas validation for beta-2 [01:31] slangasek: considering the current version in the archive is known not to work, i can't see why it would upset NCommander [01:31] note, that rbasak as purely been working on arm support for that upload for the last 2 weeks. [01:32] So this one, has a higher chance of some success.. compared to known suckage [01:32] ah, well, then I don't know [02:20] Daviey: the lack of documentation upsets me. Nowhere at a quick glance can I see that MAAS as it was in the archive was totally broken. [02:21] * NCommander was at dinner [02:21] (well, I knew it didn't work on ARM, but I didn't know it was totally foobar'ed [02:21] ) [04:17] all builds finished, lubuntu ac100+armhf image now accessible from tracker. [04:17] * skaet --> zzz === doko_ is now known as doko [07:45] NCommander: The fact that you had no idea it was broken, troubles me more TBH. === henrix_ is now known as henrix === henrix is now known as henrix_ === mmrazik is now known as mmrazik|early_lu [09:04] there somebody who could help or point to somebody could help with problems during ubiquity? [09:05] xubuntu is affeted by both bug #924909 and bug #101087 and we think this has something to do with the ubiquity config [09:05] Launchpad bug 924909 in xfwm4 "Windows have grey traces in Ubiquity" [Medium,Confirmed] https://launchpad.net/bugs/924909 [09:05] Launchpad bug 101087 in silva "request for get_ordered_ids()" [Wishlist,Won't fix] https://launchpad.net/bugs/101087 [09:09] knome: you probably meant something else for the second one? [09:09] yes... [09:09] bug #1010487 [09:10] Launchpad bug 1010487 in ubiquity "Xubuntu - black windows" [High,Confirmed] https://launchpad.net/bugs/1010487 [09:10] is there some settings how ubiquity/xfce is run or similar we could look at? [09:12] knome: #924909 is funny. There is an upstart job which launches ubiquity in the greeter mode. But it looks like the background is not repainted, maybe something needs to be done there to ask thunar to repaint itself? [09:12] bin/ubiquity-dm in the ubiquity source might be worth looking at, assuming this is in the "Install Xubuntu" mode and not "Try Xubuntu without installing" [09:13] You can generally tell whether something's a ubiquity-dm bug by seeing if behaviour in those two modes differs [09:13] (significantly) [09:13] knome: cjwatson reply is better =) [09:13] We might quite plausibly not be launching some bit of your desktop infrastructure [09:13] the bugs both confirm that [09:13] I didn't look :-) [09:13] cjwatson, yeah, the grey bits is only in "install" mode [09:14] and probably the other bug as well.. [09:14] though somebody reported to have it on "live", but not when running "install" from live [09:14] cjwatson, bin/ubiquity-dm where? :) [09:15] As I said, in the ubiquity source [09:15] bzr branch lp:ubiquity [09:15] oh, right. just woken up.. sorry! :) [09:16] have it on "live", but not when running "install" from live> that sounds confused as ubiquity doesn't run in "live" mode unless you run install ... I suspect they meant "install" mode for the formere [09:16] *former [09:16] yeah. [09:16] probably [09:16] knome: the second one might be a dupe of bug 744283 but I am not entirely sure. There are changes to two files in that branch (gtk_ui.py & ubiquity.ui you could replace them in place and see if helps, replace on the tty1 and do $ sudo restart lightdm to retest the "greeter" mode) [09:16] Launchpad bug 744283 in ubiquity "Steps "Preparing to install" and "Erase disk" are unreadable with high-contrast theme enabled" [Medium,In progress] https://launchpad.net/bugs/744283 [09:17] :o [09:17] we do need consistent terms for "greeter", "try ubuntu" & "install ubuntu". Cause live/alternate are the traditional counter parts =) [09:18] desktop/alternate are the traditional counterparts [09:18] live -> desktop and install -> alternate were simultaneous changes [09:19] I sometimes call them "live session" and "only-ubiquity" modes [09:19] ok. [09:19] bit jargonish which is why I tend to use the boot menu names when talking to non-ubiquity-developers [09:20] And of course only-ubiquity and maybe-ubiquity are not quite identical. Terrible naming there [09:20] maybe-ubiquity is the greeter mode (i.e. what you get if you let an Ubuntu desktop image boot without touching it) [09:20] cjwatson: which are not visible by default & no logo either. See the video from the 924909. Why is there no isolinux ubuntu logo shown?! only unless you click Esc... [09:21] gray logic ftw =) [09:21] heh [09:22] That video shows the "press key for options" rebus icon [09:22] Which is as expected [09:22] If you don't press a key there, you get maybe-ubiquity mode [09:22] Which I think is what you're calling "greeter" [09:23] yeah. [09:23] It's sort of an uneasy compromise [09:23] I think it would be nice to show ubuntu logo & the rebus. But I am sure there are reasons for it =) e.g. wrong resolution, low color depth, etc.... [09:23] Design didn't want to show the boot menu by default; I wasn't happy with showing text before we know the user's language [09:24] And yes, the Ubuntu logo is kind of a no-go there because the aspect ratio is very likely to be wrong [09:24] And syslinux is a bit too primitive to be able to do much about that [09:24] At least not while preserving my sanity [09:26] well remaining sanity bits, need preserving =))))))) totally agree with you there! [09:26] anyway, thanks for the help [09:26] we'll look at this and hopefully fix it soon! [09:30] * iulian is not getting any mails from Launchpad when he comments in bugs where ubuntu-release is subscribed. [09:31] Any idea why? [09:32] iulian: what's your bug mail preferences? you can choose not to receive spam about your own updates =) [09:32] xnox: I've no idea what my bug mail preferences are. :) [09:32] Where do I check that? [09:32] It used to work before. [09:32] iulian: it's somewhere on your personal page http://pad.lv/~iulian ? === mmrazik|early_lu is now known as mmrazik [09:33] iulian: also #launchpad support on #launchpad? =)))))))))))) ;-) [09:34] iulian: Follow xnox's advice and go to #launchpad for Launchpad Mail help:) [09:34] you don't get your own comments back [09:34] you can =) [09:34] I do =) [09:34] Laney: We used to get them. What happened? [09:35] eww [09:35] 1046206 dupe of 1010487 ? I am suspecting a thunar change [09:35] I see a somewhat relevant sounding setting on http://launchpad.net/~/+edit [09:35] I cannot remember each bug I commented on and it makes reviewing things a bit harder. :( [09:35] OK, let's see. [09:36] I've got the "Send me bug notifications..." ticked already. [09:37] * iulian sighs. [09:37] * iulian goes to #launchpad. [11:15] jibel: bug 1046175 debdiff attached, fixed in lp:ubiquity . Tested locally created 4 primary, delete 4th primary, create 3 logical.... the partition type is now shown in the create dialog. [11:15] Launchpad bug 1046175 in ubiquity "[regression] Manual partitioner only creates primary partitions" [High,Confirmed] https://launchpad.net/bugs/1046175 [11:16] release team please see bug 1046175 and check if you want to: release note it, respin opportunity or trigger. [11:17] the manual partitioner currently creates only primary partitions. so you are limited to create maximum 4 partitions when starting from scratch. If starting from existing system you can only create (4 - # of existing primary partitions) [11:22] I think that's at least worth uploading so that we can take the decision about whether to respin for it (potentially) in your absence [11:22] Worth noting that many systems have three primary partitions from the factory so it can be very difficult to install at all without using logical partitions [11:23] ok. [11:23] in the mean time I will continue on implementing manual lvm controls. [11:23] xnox: Have you tested the behaviour when editing an existing partition as well? [11:30] cjwatson: yes. [11:30] although I do not understand why partman decides to check for bad blocks while doing edit =/ remove+add are instant, yet edit are not. [11:31] I wouldn't expect bad blocks checks until you commit [11:33] edit decided that "previous changes need to be committed", although there were no previous changes (just restarted ubiquity). [11:33] change 3 partitions / repartitioned. [11:33] and the primary/logical options are not shown on edits. [11:41] * ogra_ wonders whats the reason that http://iso.qa.ubuntu.com/qatracker/milestones/232/builds/22564/testcases/1167/results tells him the last update was on may 9th [11:41] :P [11:41] (read: cant we use sane date strings ?) [11:42] +1 [11:43] ogra_: drupal has no clue about browser locales =) [11:43] heh [11:43] * xnox +1 for ISO dates 2012-09-05 [11:44] xnox, well, i dont need a date in my personal locale necessarily ... a more international way would already suffice [11:44] and yeah, ISO would be good [11:44] ogra_: you made a slight assumption there that american is not internation way ;-) [11:45] no, i made a hard claim ;) [11:48] Yeah, no real excuse for not using ISO dates in an international project === henrix_ is now known as henrix [12:11] Yummy. [12:12] I'd have thought -dpkg needed re-merging or syncing [12:12] (the interface changed) [12:14] Right, should we reject this one and sync it instead? [12:14] think so [12:15] test build the debian version and see [12:16] Already started building it. [12:24] good morning [12:25] Laney: Status: successful. I'm rejecting the one in the queue right now and sync it. [12:25] Morning skaet. [12:25] good afternoon iulian :) [12:26] Laney, cjwatson - pad looks quiet. Any concerns emerging from latest set of images? [12:27] Riddell, has anyone been able to duplicate https://bugs.launchpad.net/ubuntu/+source/casper/+bug/1046244? [12:27] Launchpad bug 1046244 in casper "plasma-desktop crashes with SIGFAULT on boot" [Undecided,New] [12:28] skaet: bug 1046175 [12:28] Launchpad bug 1046175 in ubiquity "[regression] Manual partitioner only creates primary partitions" [High,Confirmed] https://launchpad.net/bugs/1046175 [12:28] xnox: ^- still awaiting an upload for that, I think ... [12:28] ok. [12:34] skaet: no I couldn't recreate that plasma crash but I'm asking for more testers [12:35] Riddell, thanks. [12:47] iulian: please take a look at silently and ghc-syb-utils (my uploads, so I can't do them myself) [12:52] Laney: Done. [12:52] I reckon that ghc.html page will look nicer. :) [12:56] cheers [12:58] queuebot is on holiday again. [12:58] Daviey: go! [12:59] Is he bringing a new robot from the clouds? [13:00] Daviey, can you provide more background on the MAAS upload on the pad, and a bit more background? [13:01] in general - ie. what are the implications if we don't respin, and just make it available as SRU? [13:02] SRU? [13:02] * skaet needs more coffee [13:03] first update after we put out beta 1. Had concept of 0-day SRU in my head, where 0-day was for Beta 1, but term SRU is inappropriate. Development Update? DU [13:03] * xnox 's machine locked up. On reboot I have 15 apport windows..... *sigh* I am submitting them all! [13:05] just "update" is fine [13:07] fair 'nuf ;) [13:11] bug 1046228 [13:11] Launchpad bug 1046228 in ubiquity "Quantal Desktop AMD64 installation failed with: ubiquity.install_misc.InstallStepError: Plugin console_setup failed with code 1" [Undecided,Confirmed] https://launchpad.net/bugs/1046228 [13:13] cjwatson, xnox - is one of you looking at this as well? [13:13] * skaet --> vet, back in 2 hours. [13:13] not yet. needs follow up as nobody else is reproducing this yet. [13:13] maybe the reported did something non-standard to get that. [13:14] but then again. I am not that good with console-setup [13:16] xnox, i reported the bug, but did not do anything non standard, [13:17] xnox, let me try a coupld more times to see if i could nail the steps to reproduce [13:17] I'll have a look once I get some I/O back from this GRUB build [13:17] But the move to 3D unity only means I'm kind of hamstrung in terms of doing much in the way of installer debugging in kvm [13:18] psivaa: If you're running it again, please add the "debug-ubiquity" boot parameter [13:18] That will produce more helpful logs [13:18] psivaa: You also didn't say which language, location, and keyboard layout/variant you selected [13:19] Actually, never mind, those bits are in the log [13:20] cjwatson, i could add the debug-ubiquity but i need to look up how to do that [13:20] f6 at boot menu [13:20] cjwatson, thats good enough thanks v much [13:20] skaet: i'm not sure it requires more content.. it's an opportunity include. [13:20] probably best to add maybe-ubiquity debug-ubiquity to make it as similar as possible to the previous scenario === jbicha is now known as Guest14986 === Guest14986 is now known as jbicha_ [13:37] xnox, skaet , I filed bug 1046323, which I think should be marked in red in the release notes if not fixed. [13:37] Launchpad bug 1046323 in ubiquity "Ubuntu installs on first drive without asking when 'Encrypt disk' is selected on a dual drive system " [Undecided,New] https://launchpad.net/bugs/1046323 [13:37] jibel: are both disks same size? [13:37] jibel: or one bigger than the other. [13:38] xnox, yes, same disks [13:38] jibel: then does it matter which one to install onto? since we can't install on to both. [13:39] jibel: if they are different size, maybe it does matter. [13:39] xnox, what if there is an existing system on sda and want to install to sdb ? the content of sda is lost [13:39] jibel: if existing OS are detected, it's a different code path and you should always see selection. [13:40] xnox, if the user doesn't select to encrypt the drive, there is an additional page to select the target drive [13:40] jibel: plus you would be choosing side-by-side install, which is differe from wipe disk and install. [13:40] jibel: no, there isn't. It became optional, if there are no choices for the users to make. [13:40] jibel: as per design spec. [13:41] I recently commented on the armhf bug that testcases need adjustment. [13:41] xnox, I have the choice, there are 2 disk drives on the system [13:42] both are empty, both are the same size, and can be enumerated in random order. so you don't know which is sda and which is sdb => you have no choice =) [13:42] I am not sure if selection should be presented. [13:43] xnox, hm install precise on disk1, precise on disk 2, then quantal with entire drive + luks [13:43] don't tell me it selects to wipe a random disk [13:43] jibel: is that what you did? [13:43] xnox, it was not precise but pretty much [13:44] jibel: cause the entire drive becomes => "delete $existingos and do clean install" [13:44] jibel: if you have existing system and you want to preserve it, you should select "install side-by-side" [13:44] then you should be able to choose which drive to install onto. [13:45] xnox, ok, let me try again [13:45] and by the way you can currently only use "lvm & crypt" options on "wipe & install" [13:45] it should not offer those on dual boot options [13:45] they will be gray. [13:46] the reason it's ambigious where to do the dual boot at: disk level, crypt level or lvm level. [13:48] jibel: please take screenshots / read carefully (prefferably in english, If you want to file bugs about it. Maybe there is something lost in translation?!) [13:48] or both locales.... [13:49] xnox, I'm reinstalling lucid on sda, precise on sdb and will tell you what happens with no lvm, lvm only, lvm+luks. [13:50] jibel: using automatic installer? [13:50] jibel: hmmmm =/ i expect something ugly will happen. [13:51] cjwatson: skaet ^^^ upload for bug 1046175 [13:51] Launchpad bug 1046175 in ubiquity "[regression] Manual partitioner only creates primary partitions" [High,Confirmed] https://launchpad.net/bugs/1046175 [13:51] xnox, I expect it too, and why I think it should be documented [13:52] I could have windows on my first drive, and want to try new shiny features of Quantal on an additional drive. [13:53] jibel: it's different code paths if it is existing $windows + $ubuntu vs ($windows + $other_linux or 3+ OS) [13:55] jibel: if you have lucid/sda and precise/sdb. You should see "You have multiple OS installed, what do you want to do?". The option that will offer crypt or lvm, should have a Red Warning label saying everything will be destroyed. [13:55] not sure about exact text of that option. [13:56] psivaa: any more luck reproducing your console-setup crash? [13:57] xnox, not yet, but trying [13:58] xnox, ok, I'll try with windows too [14:00] plars: So, skaet's away for a bit, but she asked me to talk to you about bug 1046175 [14:00] Launchpad bug 1046175 in ubiquity "[regression] Manual partitioner only creates primary partitions" [High,Confirmed] https://launchpad.net/bugs/1046175 [14:01] psivaa: i wonder if something went very wrong in that install.... see my last comment on your bug. That's from ubiquity syslog [14:01] plars: This is a pretty nasty bug that I reckon renders the installer unusable for a number of classes of users, and is certainly deeply confusing [14:01] cjwatson: I'm looking at that one right now and talking to jibel about it [14:01] plars: I can't personally speak to its testing, but xnox has tested his own fix [14:01] plars: What do you think about the chances of a respin? [14:01] 'cos I know QA love that [14:01] lol [14:02] cjwatson: heh, well... if we kick off a respin now, we are looking at late night for most of the team but me by the time it's ready [14:02] Yeah, and we need to accept the new ubiquity and get it built and published [14:02] It's not ideal [14:02] cjwatson: yeah, I know [14:03] xnox, yes i saw that, but this is a vm install so not sure what exactly might have gone wrong there [14:03] * xnox doesn't consider bug 1046175 respin worthy. maybe opportunity candidate. I'm not in release team or anything like that. [14:03] Launchpad bug 1046175 in ubiquity "[regression] Manual partitioner only creates primary partitions" [High,Confirmed] https://launchpad.net/bugs/1046175 [14:03] Perhaps I should get it built and published anyway? At the very least, we aren't going to build images without this fix [14:03] Yeah, I think I'll do that [14:04] psivaa: anything can happen, I had virt-manager loosing track of it's storage pools and producing very interesting virtualisations of hardware failures ;-) [14:04] I was actually about to talk to jibel to see if we have any standard criteria for respinning at milestones that differ from release. Certainly for release we would need to respin, but for beta, in my thinking, is there a big difference between release noting something like this and telling people that if this is a big deal for you, you should pick up tomorrow's daily build instead for testing? [14:04] having it in the archive as an opportunity target sounds good [14:04] indeed [14:04] plars: Depends how many support mails we want to answer :-) [14:04] The day after a milestone, daily builds are often broken, of course [14:04] But we'll see ... [14:10] xnox, hmm ok, could not reproduce it yet, the vm installation on that particular system is very slow though [14:11] psivaa: most of my testing is in kvm, and I didn't see that one yet. I blame solar winds =) [14:12] oh, fun, Edubuntu only has xserver-xorg-video-dummy installed, no other video driver. That certainly explains why I'm not getting any X session... [14:12] stgraber: yeah I noticed that yesterday too and wondered what was going on [14:12] (I assumed it was a global ubuntu problem, is it not?) [14:13] I don't think so or we'd have a lot more people panicing ;) [14:13] xserver-xorg-video-all seems to be missing, trying to figure out why [14:14] xnox, hmm possibly, let me try if that wind is persistent :) but owing to its lack of frequency, i dont think its very important for today [14:15] highvoltage: hmm, might be because we're shipping arkose which ships xpra which depends on xfvb which depends on xserver-xorg-video-dummy [14:15] highvoltage: xserver-xorg-video-dummy meets the "xorg-driver-video" criteria part of xserver-xorg so xserver-xorg-video-all never gets installed [14:16] ah, weird that we didn't run into it before. [14:17] I guess the dependencies/recommends got updated with the new X stack that got uploaded. [14:17] something must have changed recently, either xvfb/xpra using xserver-xorg-video-dummy or some change on xserver-xorg-video-dummy so it now provides xorg-driver-video... [14:19] trying to figure out what the right fix is... I can see why someone would wnat to install xserver-xorg with only xserver-xorg-video-dummy as I'm actually doing that for some containers/auto-upgrade-testing... I'll just add xserver-xorg-video-all to edubuntu-desktop for now [14:28] Laney, skaet, cjwatson: I'm going to push a new edubuntu-meta to fix that missing X driver issue we're having (seed change pushed already), then will respin with whatever other fixes will have landed by then [14:30] are we expecing a new ubiquity? [14:30] you might want to wait for that if so [14:30] stgraber: Let's wait until ubiquity has published [14:30] Laney: It's building or publishing or something [14:30] stgraber: But yes, makes sense to upload edubuntu-meta in the meantime [14:31] cjwatson: ok, will add the ubiquity fix as rebuild trigger for Edubuntu then [14:31] ah, it's being release noted for everything else? [14:32] I don't think we've decided [14:32] just going by the pad [14:36] ^ can someone please review edubuntu-meta? (diff should be tomboy replacing gnote and xserver-xorg-video-all added) [14:36] doing [14:37] the gnote -> tomboy part is bug 1046345 which I mentioned in the seed commit but forgot in the upload [14:37] Launchpad bug 1046345 in edubuntu-meta "[ffe] Change from gnote to tomboy in Edubuntu" [Medium,Confirmed] https://launchpad.net/bugs/1046345 [14:37] (yay tomboy) [14:38] done [14:38] Laney: thanks [14:43] could someone spare a bit of time to look at the powerpc builds for lubuntu? Both are 'fail', two different bugs. [14:43] bug http://launchpad.net/bugs/1041625 [14:43] Launchpad bug 1041625 in openchrome "X not starting after install [openchrome]" [Critical,Confirmed] [14:43] bug http://launchpad.net/bugs/1040544 [14:43] Launchpad bug 1040544 in ubiquity "Installer dialog does not come up" [Undecided,New] [14:51] ahh, there has been a comment on the X bug :) [15:01] grumble, ppc buildd has been hit with lots of PPA builds today [15:09] tumbleweed: sorry... I know ppc can be a pain in the neck. [15:15] plars, cjwatson, Laney - looks like we'll be getting in a security fix as well. [15:16] for what? [15:16] phillw: not your fault. I just don't want to accept too many unseeded things until the backlog has cleared [15:22] skaet: which packages are affected? we have the fix already? [15:23] Is it a security fix that *has* to be on the live session rather than being upgraded latere? [15:23] *later [15:23] isn't it kernel?! [15:23] which package is it? [15:23] Well, I don't know, I haven't been told [15:24] we've been uploading userpace things to quantal-proposed. they don't need to be on the CD unless someone explicitly asks for it [15:25] stgraber: can you add http://cloud-images.ubuntu.com/quantal/20120905/ to the tracker for B1? [15:25] utlemming: sure [15:25] stgraber: thank you kindly :) [15:26] Can't find: us-west-2-amd64-hvm [15:26] utlemming: is that something new? ^ [15:27] stgraber: yes, that's new. It came online the day of A3 [15:27] jdstrand, openjdk6 and openjdk7 [15:27] those are probably coming from doko [15:27] yes, jamespage has been working with him on testing. [15:27] skaet: Why do those need to be updated on the images? 12.04 is vulnerable too, surely [15:27] utlemming: ok, I'll need a bit more time for that one then [15:27] skaet, yes, as mentioned earlier [15:28] but I don't see why they need to be on the image per se-- but the server team might want it... [15:28] If it's vital that it be in 12.10 beta 1 for some reason, I'm not sure why logically we aren't scrambling to do a snap 12.04 point release [15:29] cjwatson, I believe that's in progress as well. [15:29] cjwatson, 12.04 has the openjdk security updates [15:29] doko: In the 12.04.1 images? [15:29] skaet: A snap point release? [15:29] That's kind of a big deal, surely we need to talk about that [15:30] We didn't even do that for the OpenSSL vulnerability of death [15:30] cjwatson, agreed. No discussion of a snap point release for 12.04 yet. [15:30] sorry, didn't get this with the point release [15:30] There's a sync (octave) sitting in the queue. It takes more than 1 hour to build on amd64/i386 and roughly 4 hours on arm* and ppc. If you need the buildds to be free, then let it stay in the queue. [15:30] I understand that the Java plugin has serious vulnerabilities, but web stacks have vulnerabilities all the time - we generally just let people upgrade afterwards and move on [15:31] The java browser plugin wasn't shipped in 12.04.1 anyways and isn't on any media for quantal [15:32] utlemming: there you go ^ [15:33] Yeah, indeed - so at least the Java vulnerabilities I've heard about don't seem urgent for images [15:33] stgraber: most appreciated :) [15:33] Well, icedtea-6-plugin is on the Kubuntu DVD actually [15:33] Ah, but we've stopped building that by the looks of things [15:34] server and edubuntu? [15:34] AIUI the vulnerabilities are only critical for things that ship icedtea*-plugin [15:34] i.e. Java on the security boundary rather than YA language interpreter [15:34] can someone from ubuntu-archive please merge: https://code.launchpad.net/~stgraber/ubuntu-archive-tools/ami-add-us-west-2-amd64-hvm/+merge/122903 ? [15:42] cjwatson, hmm, no, the ones from last week are in the jdk [15:43] doko: Please expand [15:43] doko: Are they attacks by malicious applications on the language runtime/VM? That's what I understood [15:43] http://lwn.net/Articles/514274/ [15:43] doko: In that case they only matter if the application comes from a different security context [15:44] We don't care about the possibility that a local application might exploit a local VM to do something malicious with its own user privileges [15:44] Only things on security boundaries matter, such as the web plugin [15:44] well, ok [15:52] Laney, cjwatson, skaet: edubuntu-artwork and ubiquity are published on all architectures, can we respin Edubuntu now? [15:53] stgraber, ok for respinning Edubuntu [15:53] Laney, will you trigger it, or do you want me to? [15:57] plars: so I gather there's some debate about whether to respin desktop, and we should talk about it here [15:57] we let in the ubiquity fix kind of as an opportunity target, or at least because there was no reason not to [15:58] I understand QA time is tight but I think it's very bad to release with a broken manual partitioner [15:58] cjwatson: I had just been reading the backscroll, it sounded like most here were in agreement that it should be updated, but no respin? [15:58] prepared to be overruled, but we shouldn't all assume we agree without clear discussion :) [15:58] oh, I thought you meant about the java bug [15:58] skaet, Laney: as neither of you seems to have done it yet, I'm triggering the rebuild [15:58] stgraber, ack. [15:58] I don't think the Java bug needs a respin, no, but the ubiquity issue remains [15:59] cjwatson: which ubiquity bug specifically? do you mean https://launchpad.net/bugs/1046175 [15:59] Launchpad bug 1046175 in ubiquity "[regression] Manual partitioner only creates primary partitions" [High,Fix released] [15:59] Yes [15:59] or https://launchpad.net/bugs/1046167 [15:59] Launchpad bug 1046167 in ubiquity "Cannot use manual partitioner if an LVM partition already exists (dup-of: 1042647)" [Undecided,New] [15:59] Launchpad bug 1042647 in ubiquity "[FFe] [UIFe] Manual Partitioning LVM" [High,Confirmed] [15:59] No, 1046175 [15:59] The only image for which that's relevant that also contains OpenJDK is the Edubuntu DVD, BTW [16:00] skaet, Laney: doh... I forgot to set ARCHES... so looks like we'll get an armhf build for Edubuntu (not published though), sorry for the extra time it'll take to build that one for nothing... [16:00] I've been trying to reproduce 1046175 and so far, not able to [16:00] xnox: ^- [16:00] maybe because I was testing it on mac? [16:00] Yes [16:00] Macs use GPT which does not have a primary/logical distinction [16:00] Testing that bug on Macs is a waste of time :) [16:01] ah, ok I had wondered if there was a diff there.. (I just got this mac from jibel, never used one before) [16:01] I wish people would stop doing silly things like buying Macs. [16:01] * xnox if only everyone used GPT like I also do..... [16:01] Oh don't get me wrong it's a much better partition table format [16:01] But [16:01] cjwatson: there's another one to be aware of too... [16:02] https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1046323 [16:02] Launchpad bug 1046323 in ubiquity "Ubuntu installs on first drive without asking when 'Encrypt disk' is selected on a dual drive system " [High,Confirmed] [16:02] jibel and xnox were going round on that; I didn't keep up [16:02] this is IMHO expected behaviour. But I still need to review jibel's screenshots. [16:03] I'm not sure I agree, but I'm not sure now is the time for that debate either :-) [16:03] I only offer lvm/crypt on clean installs, not dual boot configs / reinstall / upgrades etc. [16:03] xnox: my understanding was that if you have two disks, and you are trying to install on the second drive with encrypted disk, you're first drive is going to get wiped instead [16:03] It seems to lack defence in depth against data loss [16:03] i need to review jibel's pdf before doing any fixes. [16:04] plars: ubiquity partitioning does not currently offer automatic option "install on the second drive with encryption". [16:04] ok, so it's back to just bug 1046175 then [16:04] Launchpad bug 1046175 in ubiquity "[regression] Manual partitioner only creates primary partitions" [High,Fix released] https://launchpad.net/bugs/1046175 [16:04] that one is uploaded. [16:05] didn't there used to be an option to get the latest ubiquity before starting the install? [16:05] or is that gone now [16:05] if there was, it was way before my time. [16:06] There is, but it relies on a remote trigger to activate [16:06] xnox: ubiquity/auto_update.py [16:06] ok. [16:08] cjwatson: I think this would surely be critical if it were final release, but for beta1, would we have normally done a respin in the past with <24 hours for something like this? [16:08] * plars just started on the team this week, so I don't yet have a good feel for the history of things like this [16:08] certainly have in the past [16:09] depends how far in the past you mean I suppose, I only have records in wetware memory [16:09] cjwatson: my concern is that there may not be enough time to rerun all the tests [16:09] well, I don't know, as I say I'm prepared to be overruled if you think there isn't enough time [16:10] although I think it is a process failing if there isn't enough time, because we've certainly done shorter-notice respins than this in the past, and we've cut down on our image sets so that QA would be quicker [16:10] skaet: what time do we actually release tomorrow? [16:10] cjwatson: what do you mean by "cut down our image sets"? [16:10] I mean if we stop shipping alternates and QA takes twice as long, that seems kind of odd [16:11] cjwatson: could be ok, as I said, just started, don't have a good sense of time on this yet [16:11] plars, its variable, we certainly have time for a round of testing from europe in the morning, as well as the rest of the afternoon here. [16:12] stgraber: OK. I was never going to do it because you said you would. ;-) [16:12] skaet: ok, sounds good then... it's just desktop respin right? [16:12] I don't want to steamroller this through by force of personality or whatnot; if other people don't think it's desperately important I'll step back [16:12] xnox said he didn't feel it was respin-worthy [16:12] Anyone else? [16:12] cjwatson, I'd rather it get fixed and out there as well, and since there is still time to retest, that's my preference [16:13] plars, yes, just desktop respin. [16:13] for Ubuntu. [16:13] +1 then [16:13] well it affects all images that use gtk-ubiquity so: ubuntu desktop, lubuntu desktop, xubuntu desktop, ubuntu studio desktop, edubuntu (is already respinning) [16:13] Riddell, gilir, knome - do you want respins to pick this up for yours? [16:13] * skaet snaps with xnox [16:14] does not affect Kubuntu [16:14] Yeah, doesn't appear to affect Kubuntu [16:14] * xnox snap [16:14] xnox, cjwatson, ack. Riddell - relax ;) [16:14] * Riddell relaxes [16:14] skaet: the lubuntu testers will dive back in if warned. [16:15]