=== nhandler_ is now known as nhandler === orly_owl_ is now known as orly_owl [00:17] StevenK: Based on your changelog entry, I'll ack that. [00:19] ScottK: Cool. I saw the bug fix, compared the versions, and thought it was easy enough and serious enough to fix [00:19] Drat, I generated the .changes file wrong [00:20] StevenK: Reject it quick then because I already asked for it to be accepted (or don't worry about it) [00:20] Who can I poke for another motu-release ack? [00:20] It's a very simple change that fixes two big bugs in Wine :) [00:20] ScottK: It isn't a real problem [00:21] Yes. [00:21] YokoZar: What is it? [00:21] ScottK: the wine bug you acked earlier ;) [00:22] Oh. That wasn't in the upload you did already? [00:22] Urgh. [00:22] Yeah it was [00:22] Sorry [00:22] The upload just needs to be built [00:22] Err approved [00:22] It needs accepted and I've asked. Maybe StevenK will do it ... [00:23] https://bugs.launchpad.net/bugs/223989 [00:23] Ubuntu bug 223989 in wine "wine: incorrect "Browse C:\ Drive" launcher" [Medium,Fix committed] [00:23] Will I? :-) [00:24] StevenK, Perhaps. It's a mystery. Do you have a coin? Are both sides heads? [00:27] If anyone has any idea WHY this upload was necessary it would be helpful, by the way. If I put a Gnome launcher in the Applications menu that says "xdg-open $home/foo" it is now being turned into "xdg-open /home/user/$HOME/foo" and then failing [00:27] I'm not sure if that's a bug or a deliberate change [00:27] Running xdg-open $HOME/foo from a terminal still works as expected [00:38] YokoZar: Without experimenting, I wouldn't expect environment-variable substitutions to be carried out in launchers; thus "xdg-open $home/foo" would be expanded as "$(CWD)/$home/foo", since it's relative. === macd__ is now known as macd [00:44] RAOF: ahh, ok. It does do ~ though. [00:44] Really? That's also somewhat of a shell-ism. It surprises me less than env substitution, though. [00:45] It's kinda annoying :) [00:52] StevenK: Pretty please on wine? [00:58] StevenK: Never mind. [00:58] YokoZar: It's accepted. [01:02] persia: You're correct. Universe goes ahead of Language Packs. [01:02] ScottK, Even today? I thought it might get tuned the other way in the rush for images. [01:03] persia: Yes. Just tested by the strongswan build just accepted already building on i386 [01:05] That seems a risky plan, but maybe it's less risky to throttle uploads than to fiddle with priorities at this point. it's what, about 8 hours to hard freeze? [01:05] 5 ish by my math, but it's not a hard stop necessarily. [01:06] It probably is for stuff on images [01:06] Hm. Can I upload 4 packages in 5 hours? [01:06] Is there anything on images pending? [01:06] (Then I hit tenth spot on UTU) [01:07] StevenK, Grab some RC bugs : they aren't on images generally. === ssweeny is now known as chairy [01:14] * StevenK wonders if UTU credits syncs correctly === chairy is now known as ssweeny [01:15] StevenK: I think it does, although not all of mine that someone tagged to ~scottk instead of ~kitterman. [01:15] Heh [01:15] It looks like we can fix 2 rcs with two syncs [01:16] The aren't long builds are they? [01:17] StevenK, UTU credits syncs properly iff the archive admin gets the Changed-By: line correct in .changes. [01:17] ScottK: I've not downloaded them yet to check. [01:18] Right. There's four. [01:19] OK. [01:20] ScottK: I'll churn over them nowish [01:20] OK. [01:20] persia: Did you look at the virtualbox question I asked earlier? [01:20] Do we care? [01:22] ScottK, To me, virtualbox is an annoyance, and I don't know any good way to fix it. From what I hear, it doesn't DKMS well, and everything else leaves it semi-broken. [01:22] Oh geez. [01:22] The orig tarball for gcl is 8MB, and the diff.gz is 14MB [01:22] I'd leave that for last. [01:23] Unless you really want to mess with it, ask NCommander. He loves gcl. [01:23] Yeah, it takes ~ 30 minutes to build [01:23] StevenK, try decompressing that diff.gz [01:23] NCommander: Dun wanna [01:23] StevenK, it bloated up to 70MB, and the complete source to binutils, and parts of GCC [01:23] and no patch system [01:23] Its ALL inlined [01:24] * NCommander hands StevenK acid for his eyes [01:24] Oh, the burning feels like cleaning [01:24] Oh. Native / non-native. If wgrant isn't right, the solution is to fake-bump the upstream version (2.0.4-dfsg1-0ubuntu1). Needs testing against a PPA. [01:24] StevenK, there is a reason I won't touch gcl [01:25] * StevenK kicks off test builds for three of them [01:25] persia: I'm fairly certain I don't care at the moment. Should I? [01:25] ScottK, No good reason to care, that I see. [01:26] THanks. [01:26] One built [01:26] blueyed should be dinged, but that's about it. Not worth SRU to fix, and it'll get SRU'd at the next kernel bump anyway, and can be fixed then. [01:27] Yep. [01:30] * StevenK files sync requests [01:31] s/ts$/t/ # currently [01:31] ScottK: Bug 289703 [01:31] Launchpad bug 289703 in gmediaserver "Please sync gmediaserver 0.13.0-3.1 (universe) from Debian unstable (main)." [Wishlist,Confirmed] https://launchpad.net/bugs/289703 [01:33] StevenK: Ack'ed. Go for it. [01:34] ScottK: I'll bundle them, or try to [01:43] ScottK: And bug 289706 [01:43] Launchpad bug 289706 in rxvt-unicode "Please sync rxvt-unicode 9.05-4 (universe) from Debian unstable (main)." [Wishlist,Confirmed] https://launchpad.net/bugs/289706 [01:46] StevenK: Ack'ed. [01:53] ScottK: And bug 289709 [01:53] Launchpad bug 289709 in luatex "Please sync luatex 0.28.0-2 (universe) from Debian unstable (main)." [Wishlist,Confirmed] https://launchpad.net/bugs/289709 [01:54] StevenK: Done. [01:55] ScottK: Are you irritated by me yet? :-) [01:55] Nope. [01:56] Besides, I'm reviewing one I'm going to need someone to accept in ~20 or 30 minutes. [01:59] YokoZar: Any idea why Soyuz things wine is supposed to be built on hppa? [01:59] That isn't going to work, is it? [02:00] AFAIK Windows doesn't exist on hppa so I have some doubts you can find an hppa .exe :) [02:00] (and then wine has just no reason to exist) [02:01] * StevenK checks P-a-s [02:01] * ScottK was going to point young YokoZar in that direction (IIRC I've mentioned it before). [02:01] Yeah I've seen it before [02:01] That's neatly odd. P-a-s looks right [02:02] Maybe YokoZar wasn't nice enough to lamont. [02:02] Maybe I should just change the control file? [02:02] If it's in p-a-s, that won't affect things. p-a-s takes precedence. [02:02] Wine will (eventually) be built on other arches though, for running winelib apps [02:03] It's easy enough to get p-a-s changed when the time comes. [02:03] Yeah but if it "looks right" then there's something else weird [02:07] ScottK: Bah, my pattern broke; bug 289714 [02:07] Launchpad bug 289714 in sdcv "Please sync sdcv 0.4.2-8 (universe) from Debian unstable (main)." [Wishlist,Confirmed] https://launchpad.net/bugs/289714 [02:07] Too slow. [02:07] I was waiting for it to build and install! [02:09] StevenK: Ack'ed. [02:10] * StevenK processes all four of them [02:14] Wahoo. All the non-dead buildd's running. [02:18] * stgraber wonders why "Arch: all" are always built on i386, they could be built on any of the other archs as well ... (and that'd avoid that 355 builds pending thing) [02:19] stgraber: Generally it's not a major issue. We have an extra i386 buildd to compensate, but normally we don't get a full language pack dump right before release either. [02:20] In Debian with binary uploads, such packages are never built on the buildd's and are built on whatever arch the maintainer uses. [02:22] stgraber, It's selected as being most likely to match the majority of users. There are a few packages that broke because of the selection, but an unfortunately large number of arch:all packages make assumptions about the architecture of the build host, so it's just safer to have it be i386. [02:23] Woot. I haz tenth spot [02:27] geser, nhandler Quick : StevenK is winning. Grab some RCbugs fast before the archive closes! [02:28] Hmm. Actually, given the timing, I suspect there's little chance of that. Advantages of being in signficantly positive timezones. [02:28] persia: That is ok. I'm not going to get 40 packages uploaded. I'll beat him in Jaunty [02:29] nhandler, You're in -7 right? [02:29] Haha [02:29] persia: -5 right now [02:29] nhandler: Did you see what I managed in Gutsy? [02:29] No StevenK. Let me check [02:30] 413, very nice StevenK [02:30] * persia vaguely wonders why it only goes back to Dapper [02:30] persia: Did they have -changes mailing lists pre-dapper? [02:30] * wgrant notes that whoever keeps putting 'Done' comments should probably demote the bugs as well. [02:30] Yes. Perhaps the archives are going away. The Feisty uploads list is empty now. [02:32] On a side note, how long does it usually take for the new repositories to open after a release? [02:32] * wgrant only has complete archives from Edgy onwards. [02:32] They generally open during/after UDS, but UDS is really late this time. [02:33] I really hope they open before UDS. I don't know if I could stand waiting that long [02:33] I suspect they will. [02:33] Given that Intrepid isn't an LTS, and UDS is late. [02:33] nhandler: A week to ten days after it is added to Launchpad [02:34] Thanks StevenK. I'll be waiting. And congrats on becoming #10 on UTU [02:34] StevenK: Would you please accept kvirc when it appears. [02:35] ScottK: Certainly [02:35] StevenK: Thanks. [02:35] ScottK, can I get an ACK from you? [02:35] I never remember them opening *after* UDS. Once during, but usually the week before or so (although I also hope this will be ~ mid-november) [02:35] ScottK, https://bugs.edge.launchpad.net/ubuntu/+source/linux-rt/+bug/289683 [02:35] Ubuntu bug 289683 in linux-rt "Please rebase linux-rt on 2.6.27-7.14 for intrepid" [Undecided,New] [02:36] * TheMuso looks. [02:36] * persia downloads the debdiff in hopes of a positive answer [02:38] persia: Cory already ack'ed it and I think we can (barely) fit it in build wise, so I'd so go ahead. [02:38] I think after that and kvirc though we need to wait a while and see how fast i386 drains. [02:39] * wgrant will be glad when the buildds are pooled. [02:39] Damn, that's a fair way behind. [02:39] NCommander, And I don't need to bump lrm-rt, right? [02:39] Yeah, well dump 400+ language packs in ~8 hours before final freeze and that's what you get. [02:40] persia, I think you only need to bump lrm-rt if you break the ABI [02:40] * NCommander checks the manual [02:40] That's correct. [02:40] lrm only depends on the ABI. [02:40] Arch all on i386 makes sense, but IMO langpacks should be done globally to speed them up. :) [02:40] langpacks are also normally done in PPAs. [02:40] But that won't matter soon, as the buildds will be same, I hear. [02:41] PPA will just be lower priority? [02:41] I think so. [02:41] There's a bug open to work out the priorities. [02:41] Does this also mean more arches for PPAs? [02:41] No, since not all arches can be virtualised. [02:41] persia: PPA is a Launchpad project. Ubuntu is not. Take a guess how it works out. [02:41] so we'll have parts of Ubuntu building in Xen ? [02:42] The plan seems to be to unify lpia/i386/amd64 buildds into one multi-arch primary/PPA pool. [02:42] ScottK, I'm not worried about that. Ubuntu has a fair bit of weight when it comes to certain classes of uploads. [02:42] Odd thing about my UTU rankings. Feisty was the first one I was involved in development in - 34, gutsy - 15, hardy - 15, intrepid (so far) - 15. [02:43] wgrant: That makes a lot of sense. [02:43] UTU? [02:43] TheMuso: Definitely. [02:43] Ubuntu Top Uploaders [02:43] ScottK, You were all excited in Feisty. [02:43] wgrant, link? [02:43] http://thc.emanuele-gentili.com/utu/utu_intrepid.php [02:43] It's just odd coming out the exact same numerical rank for 3 consecutive releases. [02:43] Particularly since I quit for part of one of those. [02:44] wooo, 38 [02:44] I'm four packages behind persia [02:44] ScottK, You weren't very good at quitting. You uploaded stuff even when you were most inactive. [02:44] Yeah. [02:44] NCommander, It's *really* not hard to get ahead of me. [02:44] I'm one of the lowest-upload-count people of those who watch RCbugs or mdt. [02:44] (or NBS for that matter). [02:45] 57 uploads isn't anything to stick your nose up at [02:45] stgraber: kvirc should have appeared for you now. [02:45] RCbugs is looking fairly good this release. Thanks to those that had time to look at it! [02:45] Someone needs to dethrone Riddell for jaunty [02:45] * persia gives three cheers to sebner for chasing RCbugs ever since DIF [02:45] ajmitch's improvements to it make it a lot easier to focus on what needs focusing on. [02:45] Blame kde-i18n packages. :p [02:46] TheMuso: No. Blame Launchpad. [02:46] Well, and nobody else sponsoring Kubuntu stuff. [02:46] Oh, wrong topic. [02:46] I thought we were on buildd backlog again. [02:46] ScottK: hmm, kvirc ?? You probably got the wrong s* :) [02:46] ScottK: Those were actually my improvements. [02:46] Ah. [02:47] ScottK: I was basically saying Jonathan is as high as he is because of the tens of i18n uploads. [02:47] wgrant: Thank you for those. [02:47] TheMuso: Yep. I finally got that. [02:47] stgraber: ? Yes, kvirc. [02:47] ScottK: Why are we blaming Launchpad for buildd lag? [02:48] I wonder if hppa will catch up if somebody doesn't bounce primeor. [02:48] Because of the difficulty extracting langpacks. [02:48] Ah, true. [02:48] wgrant: Because the translation process is a byzantine mess, templates didn't get approved until late, and then Rosetta was finicky and didn't like anything about KDE4 for some important languages. [02:48] ScottK: hmm, I don't use kvirc (and I don't use KDE). Are your sure you're talking to the right person ? [02:48] And then LP hung at least 5 times during exports. [02:48] Which stopped other builds. [02:49] wgrant: I think Rosetta design is more broken than soyuz. [02:49] Rosetta has to deal with a lot of stuff. [02:49] stgraber: Ah. Sorry. Bad tab completion. That was meant for StevenK. [02:49] An awwwful lot of stuff. [02:49] But yes, it needs fixing. [02:49] ScottK: Sorry. kvirc accepted [02:49] StevenK: Thanks. [02:50] * StevenK got distracted [02:50] No problem. [02:50] StevenK: persia should have a new kernel for you to accept here in a minute. [02:50] Are we hardfreezing once the langpacks are through? [02:51] wgrant: Someone suggested 0600 UTC to slangasek and he suggested that was about the right time. [02:51] Hmmm. I wonder if they'll finish by then. [02:52] There's just not much more time on the buildds, unless someone has something that doesn't need i386. [02:53] persia: The kernel goes on a derivative CD and the lang packs go on almost no CDs, so I'd say go ahead, but I don't plan on approving anything else. [02:53] StevenK: There it is now. [02:53] hppa is going to be a few hours too late. [02:53] linux-rt [02:54] wgrant: We could make it weeks late if we retried more stuff. [02:54] ScottK, That kernel doesn't go on any CDs. [02:54] ScottK: Done [02:54] Er, langpacks in NEW? [02:54] ScottK, It has some bugs (like no SMP support), so it's not default. [02:54] wgrant, more languages must have passed the 40% barrier. [02:54] -syr, -pap, -arn, and -bra [02:54] I guess. They're all three-letter ones, too. [02:55] No -tlh [02:55] given where the langpacks are, we may need to stop accepting other packages sooner so that the backlog clears (or else someone should handhold the queues to make sure all the langpacks are done first that need to be) [02:55] StevenK: those are supposed to not be generated; I'll nuke them from NEW [02:55] slangasek: I think we're pretty much done. [02:55] slangasek: Ah, kay [02:55] slangasek: I'd already reached that conclusion. I'm not accepting anything else until the backlog drains some. [02:56] slangasek: I can process the other things in binary NEW if you wish [02:56] slangasek: re: handholding, that's going to be a pain in theneck [02:56] StevenK: sure, feel free [02:56] Hobbsee: the alternative to handholding is to just not accept anything else that's not critical until the queue drains [02:57] and evaluate whether we can afford to squeeze anything in after that [02:57] A HPPA buildd bit the dust :-/ [02:57] slangasek: that might be more sane. afaik, there is now *no* way to rescore builds, short of going to LP itself. [02:57] At least palmer grabbed linux-rt [02:57] Just for the avoidance of doubt : linux-rt is *less* important than the langpacks, as it's not part of the install for any users. [02:57] (It's the fastest i386 buildd) [02:57] StevenK: Lucky [02:57] Well, if it's already building. [02:58] It is. [02:58] virtual buildds should at least be able to take load off the non-virtual ones :-/ [02:58] Isn't it only the one flavour? [02:58] StevenK: being grabbed by the fastest buildd doesn't help when the backlog is > 3 buildds wide and hours deep :) [02:58] Or somebody could assign the 20 recently moved PPA buildds to primary i386 for the next couple of hours. [02:58] But I suppose it's not quite that easy. [02:58] * ScottK enjoys the empty "Bugs related to MOTU Release Team" page. [02:59] slangasek: Hehe :-) [02:59] Er, yeah. Where did the 2 billion PPA buildds go? [02:59] I believe they got moved. [03:00] ScottK, Nice job to get that clean. [03:00] But you can imagine that infinity is busy setting them up non-virtually if you want. :P [03:01] Given the time in Calgary, I bet he isn't. :-P [03:02] how many hours until final freeze? [03:02] Ten minutes shorter now [03:02] NCommander, 3, but the buildds have to drain. [03:02] so if the buildds are still building, we don't get screwed? [03:02] Not really in this case. [03:03] NCommander: That isn't an invite to get them to keep building [03:03] hppa is screwed regardless. [03:03] If the buildds are still building, the release manager will likely bark at MOTU Release. If they drain with a publisher cycle to spare, we can push a few more RCbugs. [03:03] ScottK, they are a builder down, what can we do [03:03] hppa has been screwed all cycle [03:03] On i386 we'd be short some language packs. [03:03] NCommander: Even if everying pending built it would still be screwed. [03:03] hppa has been screwed since it was revived in Feisty. [03:03] Being short language packs is bad, because they go on the DVD images. [03:03] It had thousands of pending builds when we released Feisty, IIRC. [03:04] It's substantially worse now than on Hardy. === not_rly is now known as orly_owl [03:04] wgrant, hppa only had a couple hundred outstanding builds for gutsy. Best release since Breezy, I'd say. [03:04] Hardy had a nice freeze to fix things up. [03:04] If my math was correct, it should land about right. [03:04] Why was it revivied in feisty? [03:04] If not, then slangasek can yell at me. [03:04] NCommander, Someone wanted it enough to put buildds in the DC. [03:04] NCommander: Because it was gone for a release or two. [03:05] NCommander: Because lamont loves hppa [03:05] NCommander: No different to how we love PowerPC, and want to see the port survive. [03:05] The buildds had been there forever, but the port was missing for Edgy at least. [03:05] It was missing because I can't remember what threads thing was totally broken on hppa at the time. [03:06] NPTL [03:06] Now it's only sort of broken. [03:06] IIRC [03:06] Yeah. [03:06] That's the one. [03:06] Yeah, it's NPTL [03:06] * TheMuso has memories of NPTL, relating to audio and realtime/Jack. [03:06] From before his Ubuntu days. [03:06] Launchpad is going to fall over in an hour or so, isn't it? [03:06] Please don't. [03:07] * Hobbsee clubs wgrant [03:07] wgrant: dont' give it ideas. [03:07] wgrant, fall over? [03:07] Please don't be letting out those negative waves. [03:07] * NCommander should grab a coffee and watch LP chug along [03:07] NCommander: That's what usually happens when you've been clubbed. [03:07] Hm. two minutes per langpack [03:08] That's not so bad. Initial estimates were 4-6 minutse. [03:08] StevenK: measured over how long an interval? [03:08] One whole langpack [03:08] lol [03:08] So don't listen to me :-) [03:08] so fairly inexact [03:08] * NCommander just hopes that no langpacks FTBFS [03:08] although I note it is now taking 32 seconds to load https://bugs.edge.launchpad.net/ubuntu, so i'm wondering if it's under high load again or something [03:09] NCommander: It's a bunch of cp and debhelper [03:09] StevenK, when there is a will, there is a way [03:09] Hobbsee: ah, it's not just my crappy internet ? [03:09] stgraber: i don't think so [03:09] Hobbsee: hum; if you notice that this persists, please yell and I'll try to escalate before we crash & burn [03:09] stgraber: i'm getting that whether i proxy from the US, or whether I go direct from au [03:09] slangasek: which problem? [03:09] the "bugs iz slow" [03:10] slangasek: ahhh. Will do, although i usually reserve that for the "slavescanner and friends have collapsed" [03:15] lpia is back to idle, so that's a good sign [03:16] yay [03:16] * NCommander is off to get some coffee and food [03:16] fairly meaningless, the backlog has all been i386 for langpacks [03:16] slangasek: you can't convert some ppa i386 machines to other ones, can you? [03:16] slangasek: It means kvirc ought to finish soon at least, freeing up another buildd for langpacks [03:16] Hobbsee: I can't convert anything [03:16] kvirc is very nearly done [03:17] slangasek: hmm. It occurs to me that I probably can. I'd just hope not to break the world. [03:17] I wouldn't have gone for that one except it was an exploitable security bug. [03:17] * Hobbsee glances at all the multiple pages of switches [03:17] Builder/+admin is required. [03:18] I'm going to watch the queues for a spell to try to estimate a completion time for the queue; if my estimate worries me I would consider poking someone [03:18] wgrant: i've got that. [03:18] But I think somebody would explode you. [03:18] slangasek: Find the lowest priority build. [03:18] wgrant: that's what I was suspecting. Just thinking it's an option. [03:18] * ScottK reviews his disaster preparedness plans. [03:18] wgrant: hmm? [03:18] slangasek: Builds have an estimated start time. [03:18] So if you find the one at the end of the queue, you should be able to work something out. [03:19] Queued: 4 hours ago [03:19] Estimated build start: in 6 hours [03:19] Woot. Below 300 [03:19] o_o; [03:19] but I don't know what that estimated build time is based on, whereas I /can/ estimate the rate of progress :) [03:19] slangasek: probably multiplying by the page number or something [03:19] ;) [03:19] It takes into account historical build time, but that of course dies because of palmer. [03:19] Which is markedly faster than the others. [03:20] * persia gets more annoyed at being mailed edge URLs. Just because the person leaving a comment is a beta tester doesn't mean I don't want access to the buttons. [03:20] persia: You could always log into edge. [03:20] persia, what buttons? === paul__ is now known as Elbrus [03:21] I thought edge did auto-redirect for non-beta users? [03:21] persia: If it's bugmail, it's LP doing that and there is an open bug on it. [03:21] RAOF: It doesn't. [03:21] Not for a long time. [03:21] beta did, edge doesn't. [03:21] That sucks more than I thought, then. [03:21] wgrant, Actually, I can't. I'm not on the whiltelist for that (by choice). [03:21] Yeah. That's it. [03:21] persia: It's not whitelisted any more. [03:21] ScottK, Yeah. I've been active on that bug. [03:21] persia: beta.lp.net was, but edge hasn't been for 18 months. [03:22] Ah. Stiil, edge tends to have lots of broken stuff : I like to find out about it all at once : reduces my sense of pain. [03:22] Personally I find discovering bad stuff on edge more depressing. [03:23] If it's in production, they say "sorry, too late it's in production". [03:23] I can understand that. [03:23] i'm currently finding that edge behaves like production now anyway, so reconsidering my subscription to edge. [03:23] If it's on edge, they say "No, we don't care how much it hurts, we aren't changing it" [03:23] * StevenK hasn't found edge so painful [03:23] And that's more depressing. [03:23] I find it's good to use edge so I can complain about stupidity and get it fixed. [03:25] kvirc finally finished! [03:25] wgrant: You mean like re-implmementing the reload button in a hyperlink? [03:26] ScottK: It's not quite that, but in most cases you're right. [03:26] I don't think anybody will work out its intended purpose if they need to use it. [03:26] I actually ran across a case yesterday where that would have been useful. [03:27] But that bug should be fixed by redirecting. [03:27] wgrant: Yes, but I generally copy/paste off that spot several times a day. [03:27] wgrant, Yes. [03:27] 7 langpacks done in a 10-minute interval [03:27] ScottK: I generally use the address bar, as it avoids mousing. [03:27] slangasek: Note that we have another buildd now. [03:27] So they fix another bug in a totally obscure and unusable way and .... [03:27] And we might have another one it not too long. [03:27] slangasek, So ~350 minutes left? [03:27] My typing is atrocious today. [03:28] assuming the build time is fairly constant across languages, that's 7 hours to completion [03:28] I think we need more buildds. [03:28] With ~400 langpacks on the DVD, I'm guessing rescheduling stuff won't help much. [03:28] Well, we need a more sensible way to handle langpacks. [03:28] At least English is building now. [03:28] * persia *really* dislikes langpacks changelogs [03:29] wgrant: which is the another buildd? I still only see 3 building langpacks [03:29] We can't exactly implement a more sensible way right now. [03:29] wgrant, No. Certainly not today. [03:29] slangasek: rothera only started building langpacks a couple of minutes before you gave you 10-minute average. [03:29] oh, there's still non-langpack stuff out there, mumble [03:29] (linux-rt building now on palmer) [03:30] Right. [03:30] That's the last of it. [03:30] Unfortunately, linux-rt takes 45-60 minutes depending on the buildd, which will be a bit more. [03:30] persia: well, no one can cancel it, so it's going to have to build. [03:31] Yep. it's my fault for waiting too long for the rumoured 2.6.27-7-15. Sorry. [03:31] +cancel is such a useful page. [03:31] wgrant: you mean the (12) Not implemented yet? [03:31] I think it'll land just about right. [03:31] wgrant: but it's nice and shiny... [03:32] Hobbsee: That one. [03:32] and i'ts been improved from what it was. It just never actually got the backend work done on it [03:32] Oh, with the clock change in the UK, I suspect we're probably OK, it's just always tight. [03:38] ScottK, Sorry. cut&paste To: line. [03:39] ? [03:41] * persia is actively reading mail for the first time in a while [03:43] I'm subscribed to both of the mail lists you sent it to, so no problem. [03:44] back [03:45] linux-rt is finishing up. [03:45] OK. That was 4 in 10 minutes with two buildd's on the problem. Maybe I was optimistic. [03:47]