[00:00] cjwatson: any idea what went wrong with my ubiquity upload? [00:00] http://starsky.19inch.net/~jr/tmp/ubiquity_2.13.14.dsc [00:00] "Rejected: [00:01] ubiquity 2.13.14 in raring (source has no binaries to be copied)" [00:03] Riddell: those ubiquity files need chmod +r =) Forbidden at the moment. [00:03] Riddell: do you want me to try an upload? [00:05] xnox: permissions changed [00:05] xnox: yes that would be good [00:05] xnox: Eh, hang on [00:06] Let me investigate Riddell's failure first [00:07] That is a truly strange error [00:07] ok. [00:10] The question is why proposed-migration thought it was suitable to copy [00:10] Because it wasn't built yet [00:11] I had a force on it [00:11] could that have caused it to go early? [00:12] Yes, you broke it, please don't do that [00:13] Please in general don't force things unless there is no alternative [00:13] "I'm in a hurry" isn't a good enough reason [00:13] * cjwatson tries to work out if it is possible to clean up this mess [00:14] how about "I want it in the beta"? [00:14] Not good enough [00:14] The only excuse for a force is that proposed-migration won't accept it due to some problem you've checked with other members of the release team first [00:14] They really need peer review [00:15] As it is, you *delayed* having this in the beta [00:15] Riddell: well "i want it in the beta" sure, but you didn't get it in beta, and instead powerpc & armhf have not build. [00:15] If you want to bypass the general freeze migration, use "unblock", never "force" [00:16] xnox: the latter *may* be fixable [00:16] interesting =) how? [00:17] I'll tell you if it works :) [00:17] ah, sorry [00:18] * xnox is pondering if it's trigger a rebuild, or sneakily copy-binaries back into the archive. [00:23] No, unfortunately that didn't work. I changed Riddell's hint to be unblock rather than force, then undeleted the packages from -proposed in the hope that I could then trigger a rebuild of armhf/powerpc and proposed-migration would copy it to release again when those were done. [00:24] However, that isn't enough to resurrect a SUPERSEDED build record. [00:25] It might be possible to recover this with DB surgery, but I'm not willing to attempt that when it isn't strictly necessary. I suggest reuploading. :-( [00:30] stgraber: libreoffice is currently blocked in -proposed; we ought to take the update, since it fixes coinstallability of language support [00:30] slangasek: agreed [00:30] stgraber: can you handle the unblock? (since you already have a hint file handy :) [00:30] slangasek: yep, will do [00:30] ta! [00:33] cjwatson: I'll upload a new version then [00:35] Thanks [00:36] You'll need to bump the unblock version in your hint of course [01:00] cjwatson: I was under the mis-understanding that AMD64+Mac was bug 1153992, with that working for alternate (lubuntu) there is still no build for the AMD64+Mac. I'm sorry if I 'nag', but do you have any update on it? [01:00] Launchpad bug 1153992 in Ubuntu CD Images "Raring server and precise d-i installations fail at the clock configuring step" [Critical,Fix released] https://launchpad.net/bugs/1153992 [01:17] phillw: It's on the Raring Daily milestone. As to why it's not on the Raring Beta 1 milestone on iso.qa, that would be something for you to discuss with stgraber. [01:17] phillw: The cdimage side of things appears to be working fine. [01:18] phillw: that'd be because the product lead didn't ask for +mac to be on the manifest [01:18] cjwatson: thanks, I'll catch up with stgraber and ask him to mirror it over (I'm guessing it is due to this 'freeze' thingie :P ) [01:18] phillw: 1153992 was indeed entirely unrelated to this matter. [01:18] the manifest was generated from what was released in 12.10 and apparently you didn't have +mac back then [01:18] stgraber: he did not? [01:19] or it wasn't addded to the manifest for 12.10 anyway [01:19] stgraber: yes, due to no lack/absence of +mac testers. [01:19] stgraber: https://help.ubuntu.com/community/Lubuntu/GetLubuntu [01:19] We did, as far as I can see? [01:20] xnox: we have a very dedicated +Mac tester, it was he who alerted me to the missing ISO. [01:20] phillw: https://wiki.ubuntu.com/QuantalQuetzal/ReleaseManifest disagrees [01:20] phillw: anyway, if you get gilir to ask for +mac, I'm happy to add it [01:21] stgraber: without wanting to start a row, http://cdimages.ubuntu.com/lubuntu/releases/quantal/release/ says we have. [01:22] That's a different matter. [01:23] Builds != releases [01:23] stgraber: I'm actually in the rare position of being able to ask for it :) I knew what he did to me would be handy one day...https://launchpad.net/~lubuntu-product-managers [01:23] Wait, that is a release, wtf [01:24] well, I suppose this is just one more way the old manifest scheme didn't work [01:24] cjwatson: you've lost me there. We had builds, we have had releases, what is the need to ask for it to be resurrected 3/4 the way through 13.04? [01:24] I'm on my phone, accuracy is optional [01:25] we should kill off the old manifest scheme ... oh wait, we did :) [01:25] ;) [01:26] asides of that, can we please have the +Mac on the beta 1 testing area? Pretty please :D [01:26] I suspect that the tracker and the manifest got out of sync near the tail end of the 12.10 release and nobody noticed. [01:27] cjwatson: as I recall, we were still testing PPC stuff when you were moving them to 'final release'. It was some what of a crazy day :) [01:27] Or possibly I got confused while doing the publishing. [01:28] shrug [01:28] Not likely to be debuggable from this distance in time [01:28] cjwatson: heck, they all got out there and bug reports were what was in the release notes.... [01:28] No release is perfect [01:29] if it were, you'd all be out of a job :) [01:29] oh no, I'm sure we'd still have enough work even if the releasing process was perfect [01:30] stgraber: I was not referring to the release process, I was referring to what is released :P [01:31] But, I will leave you good people in peace. thank you, as always, for tolerating me on here :) [01:36] or maybe not... Have I used up my allotted time to ask about things? [01:37] I'll email it, as it something I cannot see an answer to as a bug. [01:53] stgraber: / cjwatson can you remove the notice "Respins upcoming for bug 1153992; will affect at least most server/alternate images." from http://iso.qa.ubuntu.com/ Thanks :) [01:53] Launchpad bug 1153992 in Ubuntu CD Images "Raring server and precise d-i installations fail at the clock configuring step" [Critical,Fix released] https://launchpad.net/bugs/1153992 [02:53] I'm waving through the grub binary in Unapproved so I can then have this migrate to raring === james_ is now known as Guest90685 [03:58] respinning Edubuntu for new grub2, that will hopefully fix amd64 [04:20] cjwatson: edubuntu-dvd-amd64 on kapok.buildd finished at 2013-03-13 04:16:05 (success) [04:35] You know what would be awesome? If people could say “verifed on precise” or “verified on quantal” rather than just ‘verified’ on bugs SRUd to both Precise and Quantal. [04:45] Hm. We don't _really_ support Oneiric now, do we. [04:45] * RAOF eyes off the SRUs awaiting verification for 200+ days [04:46] technically for about another month [04:57] !!! There are 3 packages in the oneiric-unapproved queue‽ [04:57] RAOF: I am only a bot, please don't think I'm intelligent :) [04:59] infinity: Were you the one who didn't want be rejecting gcc-4.6 from the precise-unapproved queue because you wanted it there as a reminder to fix it? [05:00] RAOF: Yeah, but go ahead and reject it, and doko and I can discuss it later. [05:04] Oh, god. Webservices. [05:14] Oh dear lord, unity-2d's mega sru. === mmrazik is now known as mmrazik|afk === tkamppeter_ is now known as tkamppeter === mmrazik|afk is now known as mmrazik [07:34] stgraber: woo [08:33] infinity: hey. grrr, it seems that the killing all process fix didn't work. Do you mind killing https://launchpad.net/~ubuntu-unity/+archive/daily-build/+build/4365076 please? [08:53] Hi, I am Jack from UbuntuKylin team. We are wondering how could we customize the first show when installing? we created a branch lp:syslinux-themes-ubuntukylin, but don't know how to use it... [08:55] JackYu, hey, what do you call "first show"? the installer slideshow? [08:56] or the screen at the very beginning of the boot? [08:57] seb128: the screen at the very beginning of the boot. [08:57] is the first screen of install process [08:59] seb128: how could we customize it? thanks. [09:00] JackYu, I'm not sure, but cjwatson can probably help you (not sure if he's around yet) [09:01] seb128: thanks. I will ping him:) [09:01] JackYu, you can considered him pinged [09:01] he reads this channel and his nickname got mentioned [09:01] I'm sure he will read it when he's around ;-) [09:02] oh, i see. so I will wait for his response. [09:07] JackYu: I would rather not have a separate package - all other flavours share a single implementation [09:07] JackYu: do you just need to replace the graphic? [09:08] Hi watson, we also want to replace the dialog for F1 [09:09] cjwatson: yes, we need replace the graphic and the F1 [09:11] cjwatson: so we submit the graphic and the content to you? is it ok? [09:13] I'm not sure you can replace F1 straightforwardly [09:14] It would depend what you wanted to do with it [09:14] maclin: FYI Watson is my family name - my personal name is Colin [09:14] or just "cjwatson" is fine [09:15] I'm sorry [09:15] For the graphic, yes, send that to me [09:15] cjwatson: we only replace the 'Ubuntu' to 'UbuntuKylin' in F1 [09:16] I'll have to check how feasible that is right now. I don't think we replace it in other flavours [09:16] In some places it's talking about the Ubuntu project as a whole [09:17] sure, thanks. if not easy, just keep the F1. [09:18] I will send the graphic to your email asap. [09:18] If we change anything there, it should be some kind of general change for all flavours, rather than a specific hack for UbuntuKylin [09:18] IMO [09:19] Are you still seeing that problem with ubuntukylin-default-settings not being installed on the target system? That was pretty surprising ... [09:20] maclin: That's OK, I know ordering conventions differ [09:20] But better to say early :) [09:21] yes, it works ok when we build locally [09:22] we tried ubuntukylin-default-settings locally, it works well. [09:22] I mean the cdimage.u.c builds, not something local [09:23] en, we are also wondering why it not work on cdimage... [09:24] OK, so I'll try to reproduce that this morning [09:25] Thanks. do you have any logs on the iso build process? [09:28] BTW, we are working to update some packages. I will file bugs to these exceptions:) [09:31] http://people.canonical.com/~ubuntu-archive/livefs-build-logs/raring/ubuntukylin/ [09:31] http://people.canonical.com/~ubuntu-archive/cd-build-logs/ubuntukylin/raring/ [09:31] former for the squashfs part, latter for the main ISO [09:32] well, thanks [09:32] hmm, weird black bar on the boot menu [09:33] JackYu: do you want to have the same behaviour as Ubuntu where you have to press a key if you want the boot menu, but otherwise it boots straight into a graphical "try" or "install" screen? [09:36] maclin: regarding bug 1153917, it's always a good idea to report separate problems as separate bugs; version updates need to be tracked separately [09:36] Launchpad bug 1153917 in UbuntuKylin "UbuntuKylin Default Settings do not worked as expected" [Undecided,Confirmed] https://launchpad.net/bugs/1153917 [09:38] cjwatson: We prefer the same behavior as Ubuntu... [09:40] OK, I can arrange that [09:40] Drat, why doesn't this laptop have KVM support turned on [09:45] I find the following log: [09:45] The following packages will be REMOVED: [09:45] libreoffice-help-zh-cn ubuntukylin-default-settings [09:46] Are you sure that's current? [09:46] I thought I'd fixed that [09:46] Ah, yes, it is [09:47] In that case this'll be fixed by the new libreoffice [09:47] So I think a plain rebuild will fix this - I'll sort out your boot menu and then do that [09:47] cjwatson: I think it maybe about 'apt-get autoremove' libreoffice-l10n-zh-cn [09:49] No [09:49] It was a Conflicts in libreoffice which we analysed and fixed the other day [09:49] But it took a while to build everywhere [09:49] Because libreoffice [09:50] I'm building new images now [09:50] it's great:) [09:50] OK. Another question: this is found in the log of 20130311.2 while the image we test is 20130311.3. Is that OK? [09:53] Yes, the livefs build IDs and the CD build IDs are not necessarily in sync [09:53] Because it's possible to build a new image without building a new livefs [09:56] Good, we will expect new image to test, thanks cjwatson! [10:08] what rationale did stgraber use for his beta-1 block? there are packages seeded in flavours participating in the beta that he hasn't blocked, and people are requesting FFes for [10:09] tumbleweed: whatever the product owners asked to block. [10:09] no more / no less. [10:10] I think stgraber said it was incomplete ... [10:10] right, and based on what ScottK had before [10:10] what am I getting at, is should we be deferring FFes for a couple of days, or adding to the block list [10:10] but there are first-time flavours this time around, and hence the block possibly needs to be expanded. [10:11] also, some of these things are insane. did people forget about FF again? [10:11] (first-time with blocks that is) [10:12] Laney: do you think the verification in bug 1041432 is sufficient for both precise and quantal? you just applied the general v-done tag [10:12] Launchpad bug 1041432 in gst-plugins-good0.10 (Ubuntu Quantal) "Any application which uses v4l2src element can be froze when recording video (e.x. cheese)" [High,Fix committed] https://launchpad.net/bugs/1041432 [10:12] reviewing "New Unity stack" are we? :-) [10:12] Laney: well, it came past :P [10:13] cjwatson: I couldn't rememeber what the format was (v-d-precise?) and figured that the promoter would just do precise and then flip it back [10:13] Don't rely on that :) [10:13] tweaked the tags now [10:13] merci [10:14] And yes, v-done-precise [10:26] Package: gnome-shell [10:26] Task: ubuntu-gnome-desktop [10:26] oh good, that's in place [10:27] tumbleweed: I think I'm going to approve stuff and block it [10:28] Laney: thanks. I've got intermittent access to my e-mail atm [10:28] e.g. shotwell [10:28] yeah, I'm ok with the yorba stuff [10:28] nautilus sounds sane too [10:29] will leave unity to you :-) [10:29] pffff [10:32] seb128: ^ so you don't need to hold off uploading that stuff as I've blocked it until after the beta [10:32] Laney, thanks [10:33] or I would have... diverged branches [10:34] * Laney just found some worcestershire sauce which expired in 2008 [10:34] probably fine, right? [10:34] for sure. [10:43] stgraber: could you spin up new kubuntu images? [10:47] Riddell: What are the changes, so I can mention them on the tracker? [10:48] (I expect stgraber isn't up yet) [10:49] cjwatson: ubiquity fixes, kubuntu-settings and kde-workspace [10:51] Riddell: Kubuntu Active too, or just desktop? [10:52] cjwatson: yes active too please [10:53] OK, marked for rebuild [10:54] And rebuilding now [11:01] have anyone come across this bug testing today's images yet? https://bugs.launchpad.net/ubuntu/+bug/1154539 [11:01] Launchpad bug 1154539 in ubuntu "[raring-beta1] Live system does not boot to LightDM when a disk is attached with swap" [Undecided,New] [11:15] cjwatson: I have filed two separate bugs for Chinese-calendar and indicator-china-weather to Upgrade to latest version , i.e., bug #1154542 and bug # 1154523. Would you please help us to check them? [11:15] Launchpad bug 1154542 in chinese-calendar (Ubuntu) "Upgrade to chinese-calendar 0.7.6 in Raring" [Undecided,New] https://launchpad.net/bugs/1154542 [11:15] JackYu: I'm afraid I have no experience with either package - you probably want somebody a bit more desktopy? [11:16] Or maybe whoever uploaded them last [11:17] I mean, the bugs themselves are fine [11:17] cjwatson: it said that 'Once the Feature Freeze Exception has been ACK'd by a member of the Release Team, the status will be changed to TRIAGED. ' [11:17] Do you have a conventional tag for bugs that UbuntuKylin is interested in? [11:17] Ah, yes [11:17] You might want to put '[FFE]' or something at the front of the bug subject to indicate that that's what you're looking for [11:17] so we need somebody to TRIAGED them:) [11:17] I'll amend these ones now [11:18] Ah, thanks:) [11:23] cjwatson: the other one is bug # 1154523 :) === mmrazik is now known as mmrazik|lunch [11:28] bug #1154523 [11:28] Launchpad bug 1154523 in indicator-china-weather (Ubuntu) "Upgrade to indicator-china-weather 1.0.4 in Raring" [Undecided,New] https://launchpad.net/bugs/1154523 [11:29] Yeah, I know, I was just interrupted by the computer recycling man turning up [11:29] :) [11:31] study[cjwatson] -= 4 * monitor [11:32] JackYu: Don't worry about it for this time, but for future reference, excluding the .bzr directory from your diffs would be very helpful [11:32] If it's in bzr anyway, you could just use 'bzr diff' with appropriate arguments, rather than creating a -new directory and diffing the old-fashioned way [11:32] But, failing that, --exclude=.bzr [11:39] Triaged now but I have some comments on 1154523 that really ought to be addressed [11:43] JackYu: http://doc.bazaar.canonical.com/latest/en/mini-tutorial/ is good intro into what bzr is, what it can do, and how does it help. === mmrazik|lunch is now known as mmrazik [12:01] cjwatson: nice. I got it. will follow your suggestion next time:) [12:51] cjwaton: we have downloaded the newest image and found that 1) the default-setting package still dose not take effect, e.g., the wallpaper is still the same as ubuntu; and 2) the screen at the very beginning of the boot when installing doesn't show as we expected. [12:52] tumbleweed: The goal of the block (and it's coverage is not as complete as it should be) is to block transitions of packages on the participating product ISOs. I gave stgraber an initial hint for Kubuntu + base. He's added other packages for other flavors, AIUI. No, you don't need to defer FFes. In theory stuff should just sit in proposed until the block is listed with no harm. [12:54] cjwatson: but this time, we did not find any error in the log. [12:56] JackYu: I noticed the boot screen problem myself and have just committed a fix [12:56] (gfxboot has fairly limited PCX parsing; I ran 'mogrify -colors 256' over it to fix it) [12:57] thanks for that. [12:57] ScottK: ok, I didn't have a hints file at the time, but I created one now (assuming that doesn't still need a bzr up on teh other side). so I'll block things I approve, as needed [12:57] JackYu: As for the wallpaper, please can you describe the exact sequence of steps you took all the way from the boot screen? There are a few possibilities and I need to make sure I'm checking the right one. [12:58] It seems odd to me that the calendar pops up in the middle of the desktop when you start. Is that intentional? [12:58] tumbleweed: You do have one. When I bzr pull the hints directory, I got your file. [12:58] tumbleweed: bzr up is automatic; you're fine [12:58] cjwatson: ta [12:58] yes:). But we have changed in the latest version of calendar. [12:59] ScottK: there's the config side to tell britney about it (in the early, days, I believe that was manual) [12:59] Oh. [12:59] Didn't know. [13:00] JackYu: I seem to get the proper wallpaper in my simple test (i.e. boot live CD in "Try UbuntuKylin without installing" mode) [13:01] cjwatson: the wallpaper is in the default-settings package, and default-setting will install ubuntukylin-theme. but they did not show... [13:01] JackYu: But it does show for me, so if you want me to investigate it then I need a precise sequence of steps to reproduce the problem. [13:01] really? [13:01] We're obviously not doing the same thing. I've told you what I did; now you tell me what you did [13:02] we run the iso using virtual box and try ubuntukylin without installing, but the wallpaper doesn't show [13:03] Can I have a screenshot? [13:03] a moment [13:03] Also, i386 or amd64? [13:04] i386 [13:04] Likewise === Pici` is now known as Pici [13:15] JackYu: So, after I move the calendar out of the way, it looks like this for me in virtualbox: http://people.canonical.com/~cjwatson/tmp/ubuntukylin.png [13:15] That's your wallpaper, isn't it? [13:15] cjwatson: email you [13:16] Yes! [13:16] I'll just respin to fix the boot screen [13:16] you are using which image? [13:17] JackYu: http://cdimage.ubuntu.com/ubuntukylin/daily-live/20130313/raring-desktop-i386.iso [13:17] ok, we will try this one again. [13:17] The link on iso.qa looks correct too [13:18] sure. thanks. [13:19] (I typically use rsync though) [13:23] * stgraber waves [13:23] yo [13:24] JackYu: there we go, that fixes the boot screen [13:24] (20130313.1) [13:24] no changes to the live filesystem [13:25] ok, we will try the latest one:) [13:38] JackYu: So this is why I wanted you to give me exact steps - you didn't tell me that you'd let the boot menu proceed without pressing a key, and then selected "Try UbuntuKylin" from the graphical menu; as opposed to pressing a key at the boot menu and selecting "Try UbuntuKylin without installing" [13:41] cjwatson: without pressing a key, it goes to the selection screen. [13:42] Indeed. But you didn't tell me that that was what you were doing :) [13:43] And that was what I needed to know. I worked it out from the screenshot .. [13:43] Do you see what I mean now? [13:44] * ogra_ hands cjwatson hos spare crystal ball [13:44] *his [13:45] oh, i see. sorry for your time:0 [13:45] Well, no, I'm not trying to be sarcastic, just trying to make bug reporting more effective :) [13:48] thank you. I will do it better next time:) [13:54] JackYu: OK, ubiquity bug, fix committed for the next upload (2.13.16) === highvolt1ge is now known as highvoltage [13:54] should only have affected the wallpaper during installation / live session, not the rest of it [13:56] cjwatson: yes. [14:02] can I get someone to review bug 1154601 fairly quickly? [14:02] Launchpad bug 1154601 in edubuntu-live (Ubuntu) "[FFe] Disable the edubuntu-server bits on the Edubuntu media" [Undecided,New] https://launchpad.net/bugs/1154601 [14:03] edubuntu as it currently is isn't in a state we'd feel like releasing for beta1 so I'm preparing a bunch of changes that I hope will address that. If I make it any worse, we'll simply skip beta1, if I somehow get it all right, we'll have something worth shipping. [14:03] Sorry for the last minute notice but I only just found a few minutes to look at it now after highvoltage did an initial test of the images generated yesterday... [14:04] stgraber: acked in the bug [14:04] cjwatson: thanks [14:04] Damn, too slow, ScottK beat me [14:04] ;-) [14:05] Seemed like an easy decision to make. [14:05] Anyone on u-release have an opinion on this new Unity stack FFe? [14:06] It seems to me exactly the kind of thing we said we'd say no to when we moved feature freeze later in the cycle. [14:07] I'm really annoyed that they haven't learned from the past cycles... and I don't think they can get away with the "vUDS and release discussions" argument because they're essentially telling us they'll only be ready in a month... [14:09] ScottK: I think we should push back hard, yes. We said we would [14:09] stgraber: I think for Edubuntu, it might make sense to skip beta1 for tomorrow, and then aim to have a beta quality image early next week [14:10] (probably not even worth while to make a big deal about it either) [14:11] highvoltage: well, I'll still do the cdimage, artwork and installer fixes now and respin. We should have another image early this afternoon and can decide then. I know that if I don't do those changes now they'll get moved to the bottom of my TODO and we'll be in the exact same state for beta2 [14:11] stgraber: imho the swap partition problem is a blocker [14:11] https://bugs.launchpad.net/ubuntu/+bug/1154539 [14:11] Launchpad bug 1154539 in ubuntu "[raring-beta1] Live system does not boot to LightDM when a disk is attached with swap" [Undecided,New] [14:14] highvoltage: I'm trying to reproduce that one now. I find it hard to believe we would have missed this if it was as simple as "any swap partition" [14:15] stgraber: yeah, hopefully that's not the case [14:28] highvoltage: appears to be pretty consistently reproducable... I'll have to test with Ubuntu next [14:28] highvoltage: the reason why I never noticed is that I usually use virtio devices and casper doesn't know about those, so simply skips them [14:33] highvoltage: Ubuntu isn't affected... odd... [14:35] highvoltage: I'm going to ignore that problem and fix everything else first, then respin and hope the problem will just vanish ;) [14:35] highvoltage: the only thing I can think of that may cause casper to do weird things on Edubuntu is the multiple squashfs and I just fixed that by removing the server one from the cdimage scripts [14:50] stgraber: ok. could you confirm it on edubuntu? [14:53] highvoltage: yeah, I definitely have it on edubuntu [14:58] infinity: hey, around? :) [15:02] highvoltage: I'm reverting edubuntu-live to the 12.10 version, that should fix all our current issues [15:03] highvoltage: that should fix the lightdm config bug in the process === Ursinha-afk is now known as Ursinha [15:18] highvoltage: revert of edubuntu-live uploaded. I'll trigger a new build once it lands. [15:19] highvoltage: unless we broke anything else, this should bring us back to where we were with 12.10 === Ursinha-afk is now known as Ursinha [16:36] highvoltage: edubuntu rebuilding now [16:37] * Laney stabs affects-metoo for process bugs [17:05] cjwatson: I noticed that http://cdimage.ubuntu.com/ubuntu-gnome/daily-live/current/ is live, thank you! [17:06] Oh, yes, I set that up earlier today - didn't notice you coming online [17:06] I don't think we plan to support powerpc at this time, I'm not sure if we have people to test the amd64+mac images or not [17:06] jbicha: Do you want the boot menu to behave like Ubuntu's (the "press a key" screen)? [17:07] and then into ubiquity with a try/install choice if you don't press anything? [17:07] Yeah, I was going to check on architectures [17:08] jbicha: So http://paste.ubuntu.com/5611239/ ? [17:08] joshuahoover: hey, please do not update the FFe bug without asking [17:08] joshuahoover: adding lenses we are not waiting to install by default this release [17:08] cjwatson: yes to those archs [17:09] didrocks: my understanding was we want those in by default [17:09] joshuahoover: it's not the list communicated by PS to me [17:09] joshuahoover: did you check with them? are they ready? is the packaging fixed as we did for the others? [17:10] you know it's OK to have multiple FFE bugs if there's some disagreement about whether things are connected ... [17:10] didrocks: i will double check, but i was talking to thostr_ earlier and he wanted me to double check the list for client scopes that are said to be ready to go [17:11] joshuahoover: ok, tell him to send me an email, I spent a day updating 40 components, so if you want those in, please update with the same fixes I had to do on the others [17:11] joshuahoover: I'm removing them meanwhile, feel free to update if the double checking is positive :) [17:11] didrocks: will do, sorry for the confusion :) [17:11] joshuahoover: no worry, as the bug is polemic already, I prefer we avoid noise ;-) [17:12] heh [17:21] jbicha: OK, I've applied that and removed the amd64+mac and powerpc images [18:11] jokerdino: Am now. [18:15] infinity: what do you want me to do to get openssl accepted into quantal|precise-proposed? write autopkgtests to abuse webservers over https? anything else? [18:16] xnox: Talk to Rob Herring to see what testing he's done? [18:17] xnox: I'm not sure how autopkgtests factor into it, unless you happen to have an autopkgtest environment on ARM (I thought it all relied on KVM or something). [18:18] infinity: does this mean I get to go and see motocorss races with him? [18:18] * xnox ponders if I got wrong Rob Herring...... [18:21] xnox: :P === highvolt1ge is now known as highvoltage === mmrazik is now known as mmrazik|eod [19:39] hallyn: hey, just so you know, we found a very very odd bug on Edubuntu :) [19:40] hallyn: mountall will fail at boot time if cgroup-lite is installed and you have a swap in /etc/fstab [19:40] hallyn: preventing the media from booting completely [19:40] hallyn: removing the swap from /etc/fstab OR removing cgroup-lite fixes the issue [19:40] * slangasek squints [19:40] is this related to the new change to add cgroups to /lib/init/fstab? [19:41] a SWAP in fstab? [19:42] that seems random [19:42] slangasek: not impossible. We may be getting some kind of race between mountall and cgroup-lite to mount /sys/fs/cgroup. [19:42] yes, since cgroup-lite is 'start on mounted MOUNTPOINT=/sys' [19:42] hallyn: actually, looking at it, we probably want to change the start condition for cgroup-lite. that maybe the problem [19:42] And mounting (or not) your swap partition is altering the timing of the race? :) [19:42] there's certainly a race there [19:43] hallyn: we'd want MOUNTPOINT=/sys/fs/cgroup [19:43] right, cgroup-lite's upstart job should be changed, and cgroup-lite should add a versioned depends on mountall [19:43] slangasek: good, we seem to agree then ;) Changing that in my test VM to see if that fixes it [19:44] stgraber: good point, now that that is always a mountpoint, we can do that, cool. [19:44] stgraber: if it fixes it for you do you mind pushign the fix? [19:44] (or we can add a versioned Breaks: on old cgroup-lite, if people think cgroup-lite Depends: mountall is wrong) [19:44] infinity: it's a very very reliable race though. I reproduced it 10-15 times here and highvoltage reproduced it using a different VM solution on a different machine, apparently just as reliably :) [19:44] stgraber: Cute. [19:44] hallyn: if that works, I'll push the fix immediately, then respin Edubuntu so we have a working image for beta1 [19:44] slangasek: The Breaks is definitely more correct. [19:45] infinity: is it? The new version of cgroup-lite will have a no-op upstart job in the absence of the new mountall [19:45] still it might be worth figuring out what is the actual cause of the hang... cause that feels like a workaround for a bad bug [19:45] so it might be that both are equally correct [19:45] slangasek: Oh, if the job ends up actually depending on mountall, then the dep is fine. [19:46] infinity: well, cgroup-lite being "start on mounted $anything" implies a dependency on mountall [19:46] eventually cgroup-lite should make its way to Debian and be used with something else than upstart, but for now, versioned Depends works for me [19:46] slangasek: I was thinking of this more from the Debian multiple swappable init systems perspective, where forcing mountall on people is wrong, but if our cgroup-lite is actually useless without mountall, the dep is correct. [19:46] slangasek: Moreover, it's all academic, since mountall is Essential in Ubuntu anyway. [19:46] stgraber: nah, eventually cgroup-lite should make its way to Debian and Debian should be using upstart for everything [19:47] infinity: no it isn't :) [19:47] upstart is not Essential, and neither is mountall [19:47] slangasek: Transitively, isn't it? Oh, sorry, required, not Essential. [19:47] ah, required, yes [19:48] ok, MOUNTPOINT=/sys/fs/cgroup works fine. I'll push that along with a versioned Depends on mountall [19:48] Hrm, have we actually removed anything that depends on upstart out of the Essential set? I was sure it used to be. [19:48] That's shiny news for people who want to fiddle with init swapping at some point. [19:48] highvoltage: ETA ~2h for working Edubuntu images [19:48] (Though I'd rather do that in Debian anyway) [19:51] infinity: i think i added upstart-job dependency to shadow, which i should remove probably. [19:52] xnox: Not to login, which is the only Essential part of shadow. [19:52] xnox: you added it manually? [19:53] anyway, the fix for that is to make upstart-job go away [19:53] slangasek: we've been through this.... dh_installinit did that for me. [19:53] slangasek: No, he added it by giving passwd an upstart job. :P [19:53] by 1) fixing remaining init scripts to work with insserv, 2) un-neutering insserv, 3) merging sysvinit [19:53] remember the locked shadow dbs bug..... [19:53] xnox: right, thought so, but maybe my memory was playing tricks on me :) [19:53] xnox: yeah, I remember the job in question - was just wondering about "added upstart-job dependency" and making sure you didn't do it manually [19:54]