[00:00] Kubuntu tests are complete. [00:06] Thanks ScottK [00:07] ogra_ how are those netinst tests looking? [00:07] nearly done with armhf [00:09] excellent, thanks. Please mark them on the tracker as they complete. :) [00:09] indeed [00:09] ;) [00:47] hmm... iso tracker points to stale wubi image for those tests. [00:54] skaet: Is it ok to add Juju Devs to the draft release notes this cycle? [00:54] bkerensa - yes please! [00:54] thanks [00:54] Thank you! [00:54] :) [00:56] * skaet -->zzz now, tomorrow comes early. ;) [01:21] skaet, omap armel netinst failed [02:41] ogra_: Was it https://bugs.launchpad.net/ubuntu/+source/debian-installer/+bug/1040393 ? [02:41] Launchpad bug 1040393 in debian-installer "omap netboot partition too small for flash-kernel backup procedure" [Undecided,New] [02:41] ogra_: I kinda never got around to testing a fix for that. Really should do. [02:44] * infinity sets up raring in Launchpad. [02:46] * ScottK moved the release schedule already. [02:52] * ScottK made ~raring-backports. [02:57] * infinity decides to sleep. [03:18] Quantal relevant ubuntu-archive bugs processed. [03:19] is it too late for a process removals run? [03:20] Yes. [03:20] Anything we do now ought to be manually reviewed/requested. [03:20] (I just went through all those). [03:20] should I go through and file them manually? [03:21] Let me see what I get when I run it. [03:22] Running it now. [03:50] goodnight everyone. [03:54] im looking at doing the 64bit iso tests for ubuntustudio from http://iso.qa.ubuntu.com/qatracker/milestones/240/builds/26218/testcases ...did i miss it? is it too late? [03:54] holstein: it's not too late [03:55] holstein: you may want to check with #ubuntu-qa, I believe that's where most of the ISO testing coordination happens. [03:55] highvoltage: 'night [03:55] slangasek: thanks! [03:58] micahg: Unless I ran it wrong, it didn't come up with anything. [04:01] ScottK: could be someone already ran it [07:08] morning [07:09] Bah, just noticed the server image is rather larger than it needs to be since packages are duplicated inside and outside the squashfs :-( [07:09] Too late to fix now ... [07:09] I swear I tested that originally [07:29] stgraber, I tried edubuntu live. The session menu is empty on the client, is it expected ? === yofel_ is now known as yofel [07:37] infinity: hey: https://bugs.launchpad.net/ubuntu/+source/gnome-session/+bug/1036752 FYI, got 3 confirmations :) [07:37] Launchpad bug 1036752 in compiz "[quantal] [regression] Gnome Classic has no compiz plugins loaded" [High,Fix committed] [07:43] * cjwatson tries to reproduce the server powerpc bug on iso.qa [07:50] jibel: it's a bug but one we'll have to release note... I'm not sure what's the source of the bug as I didn't get it when running from a physical thin client on an installed (non-live) LTSP [07:51] jibel: I'll release note bug 1028972 for Edubuntu [07:51] Launchpad bug 1028972 in indicator-session "Empty session menu in ltsp client" [High,New] https://launchpad.net/bugs/1028972 [07:52] hggdh: bug 1067970 is a bug in http://people.canonical.com/~cerdea/preseed.cfg - please could you drop "d-i base-installer/kernel/override-image string linux-image-server" from that? It's effectively a no-op for x86 now and breaks (a) other architectures and (b) amd64 systems requiring secure boot. [07:52] Launchpad bug 1067970 in debian-installer "ubuntu server ppc quantal server fails on linux-image-server" [Undecided,New] https://launchpad.net/bugs/1067970 [07:53] stgraber, I also had a crash of nautilus on the client on amd64 and bluetooth-applet.orig on i386 [07:55] skaet: ^- FYI, I've investigated the failure on Ubuntu Server powerpc and it's a test bug not a software bug [07:55] infinity, didrocks: the compiz bug affects Edubuntu as gnome-classic/gnome-fallback is a supported desktop for us (not default but available as an installer option), so having this in -updates for release is important for us. Thanks [07:56] stgraber: well, that's what I try to have since yesterday :) [07:57] didrocks: I know, just putting some extra pressure on it as it's more than the initial "only affects people who manually installed gnome-session-fallback" ;) [07:57] stgraber: it's bad enough for people upgrading ;-) [07:57] we agreed yesterday to move it if we have 2 confirmations [07:57] so I think we'll get it today [08:16] slangasek: ubuntu-qa is the Quatar LoCo team channel which has been now claimed inline with Ubuntu irc channel namespace policy (i kid you not). It's ubuntu-testing for Q.A. these days ;-) [08:17] cjwatson: Are we pre-published? [08:19] Should be, I did that all yesterday [08:19] I can run it again to make sure if you like, though [08:19] Maybe not amd64+mac [08:19] I'd probably better rerun [08:20] stgraber / didrocks: compiz released to -updates. [08:20] infinity: excellent! thanks :) [08:22] infinity: thanks [08:25] infinity: ok, definitely prepublished now [08:31] and .manifest re-pruned === tkamppeter_ is now known as tkamppeter [08:32] * cjwatson starts archiving old images [08:34] beta-1 archived [08:42] how are things looking this morning? === henrix_ is now known as henrix [08:57] peterm-ubuntu: pretty reasonabe [08:57] +l, even [08:57] everything fine apart from my keyboard ;-) [08:58] cjwatson so any sense of the day? to you have an ETA for any stages? [09:40] peterm-ubuntu: not anything terribly exact as yet; will take at least a few more hours to get all our ducks in a row [09:41] jibel,plars: so, Ubuntu desktop is looking good on iso.qa; any chance of testing signoff on that? [09:41] cjwatson ok… was wonedering if you pushed to mirrors and then X hours later we were ready [09:41] peterm-ubuntu: http://wiki.ubuntu.com/ReleaseProcess, though the numbers are a bit fungible [09:42] peterm-ubuntu: we do have to push images to mirrors some hours beforehand, yes [09:42] peterm-ubuntu: what I'm trying to figure out right now are which ones are signed off or can soon be signed off for publication [09:44] cjwatson great… thanks [09:44] Anyone know where we are with server image signoff? Daviey isn't around [09:45] I want to get those two sets signed off ASAP so that I can push them to releases [09:46] Until then I think I'm pretty much blocked [09:53] If they've been tested, I'll sign 'em off, it's not like Daviey needs to bless them. [09:53] server testing is still a bit partial [09:53] incomplete mandatory tests [09:54] On, indeed. [09:55] Can we juju deploy some more testers? [09:59] ogra_: any news on bug 1067952? [09:59] Launchpad bug 1067952 in debian-installer "quantal omap armel netinst installation fails" [Undecided,New] https://launchpad.net/bugs/1067952 [10:00] cjwatson, sorry, was up until 4, just got up, will start anoter install in a minute [10:00] * cjwatson nods [10:00] *another [10:00] Unfortunately, the fact that the board hung (and it may be unrelated) makes it tough, but I was guessing the failure was actually bug #1040393 [10:00] Launchpad bug 1040393 in debian-installer "omap netboot partition too small for flash-kernel backup procedure" [Undecided,New] https://launchpad.net/bugs/1040393 [10:01] I'd like to start some publishing reasonably soon, but I don't want to have to keep track of lots of fine-grained what-I've-published-and-what-I-haven't [10:01] Which comes and goes depending on the exact kernel/initrd size, as we're borderline. [10:01] is there a workaround? [10:01] cjwatson: Other than server, what are we waiting on? [10:01] anything that isn't listed as "ready" [10:02] cjwatson: The only workaround would be manually mangling flash-kernel to not do the backup bit. [10:02] so Lubuntu, Ubuntu desktop, Ubuntu Studio, Wubi, Xubuntu [10:02] Right now I mostly care about desktop/server/Wubi since those have longer mirroring requirements [10:02] The obvious fix is bumping the netboot vfat filesystem size, but I didn't get around to it. I'm going to SRU that into precise at some point, though. [10:03] knome: ping [10:05] stgraber, dong [10:15] knome: what's the state of xubuntu testing? it's one of the products we're awaiting sign-off for [10:16] stgraber, ack from me [10:16] just got one upgrade test filed [10:16] skaet: just to confirm, all the ones tested are okay to go. Including ppc-alternate which has been release noted for the 1 failure [10:17] thanks phillw, I'll mark them ready then. [10:17] * smartboyhw hopes to see scott-work asap for the release for Studio:D [10:17] Looks like the testing is fine, no failures:D [10:17] thanks smartboyhw - good to know. :D [10:17] knome: ok, marking them as good to go. Can you update https://wiki.ubuntu.com/QuantalQuetzal/ReleaseManifest? [10:17] sure [10:18] done [10:18] skaet, we are ready anyway, just we need scott-work to say the word:D [10:19] thanks stgraber [10:19] and skaet! [10:19] smartboyhw: if he isn't around by a few hours from now, I expect we'll just take that as sufficient signoff [10:19] cjwatson, please do:D [10:20] Wait for few hours and then... [10:20] * skaet nods [10:21] ok, slangasek secured jasoncwarner_ 's ok on the desktop as long as it tested out ok. [10:21] jibel, any concerns on ubuntu desktop from your perspective? [10:22] cjwatson, skaet looks good to me but I'll let testing signoff to plars as he is the qa testing lead. [10:23] stgraber: did you ever test whether it was possible to install using the netboot mini.iso on SB? [10:23] jibel, fair 'nuf. We still seem to not have completed the server testing - is there a plan in place there? [10:24] cjwatson: no, does mini.iso come with grub-efi? I thought it only contained isolinux [10:24] skaet: URL for Edubuntu will be http://www.edubuntu.org/news/12.10-release [10:24] thanks stgraber. :) [10:24] stgraber: it does [10:24] knome, what's the one for us to use for Xubuntu? [10:24] stgraber: thinking it might be a good fallback for flavours we didn't have time to enable [10:24] cjwatson: oh, nice. I'll try that after lunch then. [10:25] skaet, let me chek [10:25] +c [10:25] smartboyhw - do you know the one we should be using for ubuntu studio in the announce email? [10:26] skaet, http://xubuntu.org/news/12-10-release/ [10:26] skaet, ubuntu-studio-users@lists.ubuntu.com [10:26] skaet, maas, gema is discussing about it with jamespage . he will update the tracker. iscsi root on kvm, I don't know the status. [10:26] jibel, I'm not even sure where that test came from? [10:26] jamespage, "Install Server iSCSI Root (virtualized KVM)", do you know if it has been tested ? [10:26] skaet, here [10:26] https://lists.ubuntu.com/mailman/listinfo/Ubuntu-Studio-users [10:27] jibel, the test case is vague - I think its probably covered by the other two iscsi root tests [10:27] which I have run :-) [10:27] and if you want to also address the developers, then https://lists.ubuntu.com/mailman/listinfo/Ubuntu-Studio-devel will be a good place [10:27] slangasek: would https://en.wikipedia.org/wiki/Unified_Extensible_Firmware_Interface#Secure_boot perhaps be a better reference than the current link target in the release notes (which goes to an agreement page for downloading the full spec)? [10:27] unless you're trying to make a point about semi-closed specs :) [10:27] thanks knome, smartboyhw :) [10:28] skaet, the problem: We don't have people doing the amd64 upgrade test for Ubuntu Studio and I'm worried that we are not going to be able to do it before delaying release........ [10:29] skaet, real difficult, we don't have much testers, and I can't do it due to revision......:( [10:31] smartboyhw - lets discuss in #ubuntustudio channel, and see if we can figure out some options then. [10:31] skaet, ok [10:38] * stgraber -> lunch [10:39] * ogra_ is already in pkgsel on omap armel netinst, infinity you might be right with the SD partition size issue [10:40] ogra_: Seems likely. Not something we'll fix for release, but I'm happy to SRU it once I've actually tested bumping the partition size. [10:41] ogra_: Seems like the right thing to do anyway, since it'll make netboot installs look more like ISO-based ones. [10:42] yup [10:44] ogra_: But yeah, if you can confirm it's my bug, just dupe it and remind me to fix it next week. It's broken in 12.04.1 too, cause our initrd grew just enough to break it. [10:44] ouch, ok [10:45] (And it's only a matter of time before it breaks for omap4, it's just pure luck that the omap4 kernels are slightly smaller) [10:46] smartboyhw: You still need help with studio/amd64/upgrade? We can spin one up here. [10:46] infinity, if you can yes. I will definitely thank you for that:D [10:48] infinity, can you add a note about netboot armel not working to the known issues on https://wiki.ubuntu.com/QuantalQuetzal/ReleaseNotes/CommonInfrastructure? [10:48] ^ or ogra_ ;) [10:49] cjwatson, wubi testing looks pretty much complete - ok to mark it ready? [10:53] oh yeah, I'm contact for that :) [10:53] marked ready [10:53] thanks. can you sign off the manifest too? :) [10:55] skaet: somebody appears to have done so ... [10:56] looks like slangasek did [10:56] For all the community-supported flavours we are the latest to sign off.....really badf [10:56] *bad [10:56] skaet: It's armhf+omap that's broken, actually. But I'm not sure it's worth release noting that a best-effort community netboot image is currently broken. [10:56] infinity, +1 [11:01] infinity, iso tracker has bug 1067952 listed next to armel+omap. No results listed next to armel+omap4. Nothing listed next to armhf+omap. What is the bug for armhf+omap? [11:01] Launchpad bug 1067952 in debian-installer "quantal omap armel netinst installation fails" [Undecided,New] https://launchpad.net/bugs/1067952 [11:04] skaet: If ogra verifies the bug is the same one we had in precise, it's all omap, but not omap4. [11:04] skaet: armel versus armhf is irrelevant in this case. [11:05] erm [11:05] no errors [11:05] o_O [11:05] it rebooted fine and is just starting up [11:06] :D [11:06] * ogra_ waits for a prompt [11:06] ah., there [11:06] ogra_: Oh. Then maybe your crash was bogons. Even better. [11:06] ogra@beagle:~$ dpkg --print-architecture [11:06] armel [11:06] ogra@beagle:~$ [11:06] yeah, i'll mark it as passed [11:07] and blame cosmic rays [11:07] LOL [11:07] ogra_: Shiny. Invalidate your bug too, then. :) [11:07] * ogra_ finds it funny that the least supported image seems to have caused the most work in arm world [11:07] ogra_: And while I'll fix the d-i size thing, it's clearly not hitting Q right now, I'm okay with that. [11:07] yeah [11:08] (Given that it hit 12.04.1 but not 12.04.0, it's obviously just barely borderline) [11:08] infinity, so are you testing the upgrade for Studio? [11:09] smartboyhw: apw's doing an install/upgrade right now. [11:10] infinity, ah thanks, I am doing it also at the same time see who's faster [11:11] infinity: bug 1068071, FYI [11:11] Launchpad bug 1068071 in launchpad "Need facility to redirect Ubuntu uploads to non-release pocket" [Undecided,New] https://launchpad.net/bugs/1068071 [11:16] So, I'd like to get at least a little bit ahead on publishing, and publish the following products which are entirely marked as ready: Edubuntu, Kubuntu (including Active), Lubuntu, Ubuntu Core, and Xubuntu [11:17] I'd rather leave Wubi until we're ready to publish Ubuntu Desktop and Ubuntu Server as well, in order to avoid craziness on releases.u.c [11:17] ok [11:17] Any objections? [11:17] cjwatson: Soundeth good. [11:17] none from me [11:17] sounds fine. [11:17] all good [11:18] * smartboyhw finds it fine [11:18] Hmm, or I could fix script bugs first - publish-image-set doesn't seem to know about Kubuntu Active [11:18] cjwatson, whoa! Really? [11:18] * smartboyhw is super surprised:D [11:18] slightly surprising but I wouldn't rate it "whoa!" :-) [11:19] It's a pretty bodgy script full of regexes [11:19] cjwatson, personal comments:P [11:20] cjwatson: Can't we just have people upload to raring-proposed instead? [11:20] cjwatson: Err, what? It knew about it during milestones... [11:20] Oh [11:20] Actually I'm just unable to read [11:21] wgrant: Realistically, that won't happen reliably [11:21] wgrant: I'd rather have the rewrite. [11:21] How does one upload to raring, if raring is raring-proposed? [11:21] One doesn't [11:21] One copies into it [11:21] wgrant: Besides, it has precedent from Debian, they always rewrite uploads to stable releases. [11:21] (That's why I suggest doing this in archiveuploader) [11:22] Once we have britney in operation, I want raring to be locked and not directly uploadable to, basically [11:22] Sure, and that's a good way to encourage people to upload to raring-proposed instead :) [11:22] Otherwise we'll be playing whack-a-mole for months [11:23] wgrant: I'd much rather fix it technically than socially. Plus, it's just prettier. [11:23] OK, we *could* just lock it; it produces ugly changelogs though, and is a bit of a waste of most developers' time [11:23] Heh [11:23] It's hardly prettier [11:23] But perhaps [11:23] When do we expect raring to open? [11:23] We can add the DB column tomorrow if we need [11:23] wgrant: Soonish. [11:23] I mean for general uploads [11:24] Toolchain should be ready in a few days, but getting britney up and running will take a bit longer [11:24] wgrant: Less soonish. [11:24] Right [11:24] Mid next week maybe; adding CI to the distro is a fairly reasonable justification for it taking slightly longer than usual [11:24] Heh, indeed [11:24] Adam and I plan to hack on this once 12.10 is out [11:24] wgrant: I still need to get glibc sorted, doko's got compilers mostly in hand, I want to do some packaging toolchain bits, plus we want to get britney ready (or at least try). [11:24] Right [11:25] If distroseries.default_pocket isn't totally unreasonable, I expect there'd be a certain amount of fallout in the LP test suite that would take a while to fix [11:25] Although we wouldn't have to set default_pocket nonzero except in some archiveuploader tests, really ... [11:25] There'd be no fallout at all [11:25] Because the default will still be release [11:26] doctests. but yeah [11:26] Meh, the doctests won't need touching, hopefully [11:26] Just a few new unit tests [11:26] Every time I add a DB column it breaks some BS doctest somewhere I didn't expect. :-) [11:26] Heh [11:26] (Well, model for the column, anyway) [11:28] Right, anyway, starting to publish the images above now [11:28] Anyway, should be a pretty painless change. [11:28] We are DBAless next week, but I have DB review privs now so we can push it through quickly if need be. [11:29] That's what I was hoping. I realise it's more effort than, say, not changing LP at all, but I do think it'll help [11:29] Hm, maybe forbid direct uploads to release if default_pocket is nonzero? Or is that too hackish? [11:30] Hmm? [11:30] Maybe I'm trying to generalise this too much and it should just be a column meaning "send it all to -proposed and forbid release" [11:30] If you're doing that, then you might as well just have a forbid_release col [11:30] Right, and don't bother generalising to default_pocket [11:30] Just hardcode -proposed [11:30] That sounds reasonably sensible [11:31] Given that we already have lots of hardcoded pocket assumptions [11:31] I just want not to break other distributions that don't have britney set up, even if derived distros in LP are largely theoretical [11:31] If someone wants to use a non-Ubuntu pocket definition we're going to need code changes anyway [11:31] "largely theoretical" ← putting it nicely [11:31] that's me [11:32] So yeah, I'd just add a flag which redirects in archiveupload. You may need to specially consider the suite name in the upload path, which everyone tends to forget about. [11:32] Tests should be simple :) [11:32] specially consider> how do you mean? [11:33] cjwatson: I forget where that override is applied [11:33] And if you want to entirely forbid release uploads, you'll need to make sure you check after that override [11:33] Oh, you mean the business where you can upload to an (S)FTP path including the suite name? [11:34] Yeah, I always forget about that, fair point [11:34] Right [11:34] Debian ended up having to have a testing-proposed-updates back door, but I don't think we'll need that since we have PPAs [11:34] An upload path of /ubuntu/raring overrides the .changes Distribution [11:34] Right [11:35] And we won't be applying a 10-day delay or anything like that [11:35] So it should get backed up a lot less [11:35] Hm, I wonder if I should also fix the way the copier gets credited for uploads [11:36] Otherwise ubuntu-archive-robot is about to have a great deal of karma [11:36] lalala [11:36] Hey dholbach :D [11:36] hi smartboyhw [11:36] Impersonation is something we haven't really worked out [11:36] I was wondering about assuming that auto_approve=True means they don't care about credit [11:36] Since that approximately means "I have an archive admin hat on" [11:36] Maybe [11:37] Even so, the way From is handled for copies is pretty bizarre [11:37] I had thought of reusing the sponsored field [11:37] But then it's sponsorship all the way down [11:37] An unending mess [11:38] Half the time it seems to be From: , ; half the time it's just From: [11:38] Copy announcements have always been bizarre [11:38] I think it depends on which of the three ways they happen [11:38] Presumably due to public/private e-mail addresses or whatever [11:38] Nah [11:38] Soyuz doesn't care about private email addresses [11:38] Pretty sure these are all due to sru-release ... [11:38] Well, doesn't care about email address privacy [11:38] I had somebody thank me a day or two ago for fixing all his bugs [11:39] Turns out it's just because I release a lot of SRUs [11:39] Seems a tad unfair [11:39] Heh [11:39] So, I'll maybe try to have a look into that since it's about to get really annoying [11:42] wgrant: by three ways, do you mean accept copy from queue, direct copy, delayed copy? [11:42] cjwatson: Something like that [11:42] unembargo-package was also slightly strange, but it's dead now [11:42] I guess the first *might* be involved, though shouldn't be for most of mine [11:43] Accept from +queue uses rSP somewhat generously to avoid issues with private addresses [11:43] We don't filter them out, we just ignore the security [11:44] I'll probably remove delayed copies after the first round of autosyncs [11:45] Since they'll prove that +copy-packages won't lag too much if it's made async for everyone? [11:45] cjwatson: my laptop won't boot the mini.iso at all. Selecting it in the boot list will just bring me straight back to the list, so it doesn't look like it's a secureboot failure, more like efi failing to find something that boots [11:45] wgrant: right [11:45] wgrant: the fix for which I just finished QAing [11:45] So I saw [11:46] stgraber: hm. oh well. we'll work that out for raring I guess - just means we don't get to release-note it [11:46] Which is unfortunate since now we'll get accused of aiding and abetting some kind of conspiracy against flavours [11:46] oh well [11:47] I thought I'd tested it in OVMF, so I suppose it *could* just be your laptop [11:47] cjwatson: yep. That machine is also able to do IPv4/IPv6 secureboot PXE, I'll definitely need to try that stuff for raring, not sure if our current netboot.tar.gz includes the needed bits [11:47] stgraber: doesn't have the tftp module [11:47] I was too much of a coward to shove that in for 12.10 [11:47] in principle it should work after that though I expect there'll be a ton of details [11:48] yeah. I'll add testing that stuff to my raring work item list so I make sure I get a couple of workitems on some blueprint for it [11:49] skaet, xubuntu amd64 upgrade passed. [11:49] yay! thanks knome [11:49] first batch of released images syncing out [11:51] ^ sorry, local mistake (gra, stupid script), rejecting the dups [11:52] cjwatson: I guess we'll have to manually work around bug #1040572 [11:52] Launchpad bug 1040572 in launchpad "initialize_distroseries doesn't handle pocket permissions" [Low,Triaged] https://launchpad.net/bugs/1040572 [11:52] "We need to fix this before opening Ubuntu 13.04." [11:53] wgrant: whoopsie, go me [11:53] wgrant: Err. I could try to fix that today I guess ... [11:54] using http://people.canonical.com/~stgraber/permissions/teams and looking for "pocket" should make it easy to extract the pocket permissions to re-apply [12:03] jamespage: heya, can you, or someone else on your team take a look at the iscsi tests on -server and the maas tests (at least the missing juju one)? [12:04] apw, infinity how is the upgrade testing? (Sorry for bothering again:P0 [12:06] plars, struggling with the maas+juju one ATM as people with the relevant access and knowledge in test labs are not awake [12:06] plars, as far as I am concerned the iscsi testing is done - the remaining test cases are dupes of the auth/unauth test cases. [12:07] jamespage: ok, if it's been tested from your end, mind marking it done for sake of completeness then? [12:08] plars, boxes ticked with appropriate comments for iscsi [12:09] wgrant: In progress - do you think we might get a deployment later, or is a cowboy more likely if I can get it done? [12:09] jamespage: thanks! who's the best person to ping about the maas+juju when they come online? or are they all traveling? [12:09] skaet, how much time left to do the upgrade testcase? I need around approx 30 minutes [12:09] plars, I'm dealing with it [12:09] lots travelling [12:09] cjwatson: We're in a good position to deploy whenever you need [12:09] jamespage: yeah, so I hear. Thanks a ton! [12:09] Raring Ringtall...Good name for R... [12:09] When you need a review give me a yell [12:10] Hopefully within the hour [12:10] smartboyhw, you have it. [12:10] cjwatson: We can cowboy if ops prefer, but with pepo in ndt now it seems silly [12:11] skaet, actually I can't even determine the time but anyway... [12:13] cjwatson: (you really don't have to fix this yourself if you're too busy, although I guess at this stage you're probably not) [12:17] wgrant: mostly waiting for testing right now so I might as well [12:17] but thanks [12:17] Besides, I meant to do it months ago [12:17] Yeah [12:18] The hardest bit of all these changes is tests... [12:19] That might be true for people for whom SQL is their first language ;-) [12:19] cjwatson, who are they?:P [12:19] wgrant [12:19] I need to port all that stuff to use the new bulk insert helpers [12:19] Makes it slightly less hideous [12:19] Slightly [12:20] I'm afraid I'll have to make it more hideous at least temporarily [12:20] Well [12:20] Aren't you just adding a single new statement elsewhere? [12:21] Since packageset and pocket are mutually exclusive, and that's inside the packageset loop [12:21] Oh, I forgot about that exclusivity [12:21] OK, that makes it marginally less unpleasant [12:21] :P [12:22] Right, it makes it pretty much trivial [12:22] Probably two new statements though, since there's the distroseries IS NULL case (preserve) and the distroseries = parent case (set to child) [12:22] "one_target" CHECK (null_count(ARRAY[packageset, component, sourcepackagename, pocket]) = 3) [12:23] Oh right, pocket perms can exist for the whole archive [12:23] Forgot that [12:23] But why do you need to consider the distroseries IS NULL case at all? [12:23] Just filter to distroseries = parent [12:23] duplicate them exception with distroseries = self [12:23] done [12:23] Oh, yeah, NULL doesn't need copying [12:23] Duh [12:24] Yeah, we're not creating a new archive [12:27] infinity, https://wiki.ubuntu.com/QuantalQuetzal/ReleaseManifest <- can you add the date on those community images you're signing off on as ready to ship. [12:28] skaet: he is afk =) [12:29] thanks xnox [12:30] Those guys in #ubuntu-release-party are getting impatient now [12:31] skaet, when can I start to upgrade to Raring?:P [12:32] They're a separate channel so that we can ignore their impatience :-) [12:33] :) [12:33] thanks for the reminder smartboyhw - have joined now. [12:33] skaet, lol [12:34] skaet, cjwatson: things on desktop are good to go from our point of view [12:37] thanks plars, marking them now then. [12:39]