[00:01] somerville32: I have no idea why the livefs buildd isn't picking up the new xubuntu-desktop seed, sorry. this will have to wait for one of the more experienced archivists to wake up and explain it to me [00:02] slangasek, try rebuilding -meta? [00:02] rebuild it why? [00:02] the one in unstable already has the correct dep [00:02] sorry, s/unstable/hardy/ [00:03] * Fujitsu was wondering why our metapackages would be in Debian. [00:03] because that's what happens when the channel only gets a quarter of my brain [00:05] Fujitsu: np, changes look great! thanks, I'm pushing the merge now [00:06] keescook: Great! Thanks. [00:07] * ogra starts his rsyncs to have the fresh isos tomorrow and goes to bed ... [00:07] slangasek, Is it possible 0-4 is being pulled in some other way? [00:07] somerville32: no, I've just checked that [00:07] slangasek, "make clean"? :P [00:08] somerville32: it was only a direct dependency because of the seed. I understand there's an extra cycle involved with updating the livefs's view of the meta packages, but I don't see why that wouldn't have been satisfied already - and anyway, if it isn't I don't know how to fix it [00:09] keescook: Aha, I see the merge seems to e done. [00:09] +b [00:10] yup, and I saw the LP "merge request" and marked it 'done' too. not sure what that really does. :P [00:10] `Merged 1 minute ago as revision 842', my branch now says. I presume that's what it did. [00:10] ah, no, I did that bit manually [00:11] That was the 34th merge ever, so I doubt anybody actually knows what it it's meant to do... [00:12] s/it // [00:12] hahah [00:19] stgraber: I don't seem to know how to enable the "upgrade" tests [01:42] siretart: libxine1-main in dapper recommends a bunch of gnome stuff but doesn't depend on any of it [02:10] dear kde. please use smaller packages. kthxbye [02:14] Haha [03:25] hi Hobbsee [03:27] hey aj [03:27] hey ajmitch [03:27] :) [03:28] Hobbsee, [03:28] Hobbsee, livefs failed again [03:29] Hobbsee, slangasek tried to fix it but doesn't know what is wrong [03:32] darn, who broke kubuntu too? [03:33] somerville32: lifefs archives look out of date. germinate's taken the change correctly, and it's mirrored correctly on the germinate output from this morning... [03:34] hello ajmitch Hobbsee somerville32 :) [03:34] hey joejaxx! [03:35] lamont: ping? [03:35] somerville32: poke elmo or lamont. this is a canonical problem. [03:35] Hobbsee: What appears to be broken? [03:36] Hobbsee: Also, hi, I'm bac. [03:36] back. too. [03:36] infinity: heya! and on a sane tz too! [03:36] Unable to type, but back. [03:36] infinity: which machine builds the livefs'es? [03:36] infinity: Hey! LTNS [03:36] Hobbsee: terranova for i386... Which arch is having issues? [03:36] infinity: both amd64 and i386 [03:37] infinity: xubuntu-meta is showing the correct thing, as is the germinate output on p.u.c, yet terranova and the other appear to be using the old package lists [03:37] Heya joejaxx [03:37] King is the amd64 livefs builder [03:38] * infinity looks. [03:38] infinity: it should be seeing xubuntu-desktop 2.53, but i'm betting it's only seeing 2.52, as it hasn't updated in a while. [03:38] thanks [03:38] Hobbsee: Oh, only xubuntu is having issues? [03:38] infinity: it appears so. [03:38] infinity: i'd say someone else has broken kubuntu, for another reason [03:39] infinity: but only xubuntu had teh -meta uploaded so late [03:39] afaik [03:39] Hobbsee: xubuntu has some hardcoded packages in livecd.sh. Did someone update the package when they updated the seeds? [03:39] E: Couldn't find package libgoffice-gtk-0-4 [03:39] ^--- That? [03:39] infinity: what would be the package of livecd.sh? [03:39] That's the hardcoded package. [03:39] infinity, we removed it from our seeds :/ [03:40] infinity, and slangasek said -meta was updated [03:40] *sigh*. and in keeping with doing non-standard things, are they going to switch all their packages to yada yet? [03:40] somerville32: The seeds had a comment in them to update livefs when that line changed. Someone should have done so. [03:40] Hobbsee: livecd-rootfs. [03:40] The livefs can't just use the seeds? [03:41] oh, nice. that didn't show up on my grep. === Shely is now known as iLeft [03:41] StevenK: It was a sticky and complex "apt resolves this incorrectly unless you kick it" thing. [03:41] infinity: right, can you please fix it? yes, it's that. [03:41] infinity: Ugh [03:41] infinity: i thought that's why they explicitly seeded it? [03:41] and so that was enough? [03:41] somerville32: Was that removed entirely from the seeds, or updated to a new SOVER? [03:42] infinity: the latter [03:42] updated to 0-5 [03:42] Hobbsee: They explicitely seed it to not seed the alternative, that won't help apt. [03:42] infinity: because apt will go for the alternative first? [03:42] Hobbsee: Right. [03:43] somerville32: you now anything about the libgoffice-gtk package? [03:43] Hobbsee: Do you have main upload rights these days? [03:43] Hobbsee: I'm not around my key yet (still in a moving mess), so I can't upload. [03:43] infinity, Hobbsee is a release manager :P [03:43] somerville32: Yes, I know she is, but those two are orthogonal things. [03:43] infinity: She's also a core-dev [03:43] somerville32: She was an RM long ago, and certainly wasn't core-dev at the time. [03:43] StevenK: Thanks. [03:43] infinity: Moving? Ugh. [03:43] Hobbsee: A cookie if you fix that for me. [03:43] infinity, I was unaware. [03:44] StevenK: Melbourne -> Calgary. [03:44] EWW [03:44] infinity: My condolences. [03:44] infinity, Is Melbourne in Canada? [03:44] somerville32: Aye. [03:44] Melbourne is in *Australia* [03:44] infinity, I'm in Fredericton [03:44] imbrandon: i was core-dev at the time i was playing RM. [03:44] er, infinity [03:44] infinity: and yes, i do [03:45] somerville32: Err, Melbourne's in Australia. Somehow, I read that as "is Calgary in..." [03:45] somerville32: i'm not, slangasek is. [03:45] (Technically, I'm an RM too, don't tell anyone) [03:45] infinity, Welcome to Canada :) [03:45] heh [03:45] * Hobbsee is still part of -release [03:45] infinity: Hope you didn't forget anything. :-P [03:45] Hobbsee: The fix is just s/0-4/0-5/ on livecd.sh, if you feel the urge. [03:45] * somerville32 urges Hobbsee [03:46] StevenK: Forgot everything. Taking a month off to move is painful on the memory. [03:46] somerville32: I was told today by an goffice upstream that libgoffice-gtk should not exist [03:46] infinity: When I say forgot anything, I meant like personal effects and stuff you left here [03:46] somerville32: do you know anything of the history of it? was it a part of the grand gtk/gnome split? [03:47] somerville32: do this prac for me hten :) [03:47] StevenK: Left a bunch of stuff behind intentionally. [03:47] StevenK: Furniture that we couldn't get rid of, mostly. [03:47] Ah [03:47] StevenK: Sucks, but oh well. :/ [03:47] * StevenK nods. [03:47] Hobbsee, Is "prac" some Aussie slang for homework? [03:47] no, practical [03:47] LaserJock, I don't know anything of the sort. [03:47] Hobbsee: If you want me to fix livefs-rootfs, I can. [03:47] LaserJock: It exists to strip libgoffice of GNOME deps, yes. [03:47] LaserJock, I'm just looking to get our desktop cd to build ;] [03:47] StevenK: i've got it here [03:48] Hobbsee: Kay [03:48] LaserJock, (would assume what infinity said) [03:48] infinity: I was told by upstream that that is not a great thing to do [03:48] LaserJock: Yes, but that's why it's free software. [03:48] HAh [03:49] as the gtk-only stuff was only there for Win32 [03:49] Oh, *EW* [03:49] so there could be lots of breakage [03:49] LaserJock: (goffice "works better" with GNOME deps, but Xubuntu is all about minimalism) [03:49] GTK on Win32 is pure crack [03:49] infinity: I realize, but I wonder if the Xubuntu guys knew what they were doing on this package [03:49] infinity: FYI, i had core dev. i just didn't have drescher access. [03:49] (for doing RM) [03:49] I'm concerned that it doesn't do what they thought it did [03:50] Hobbsee: Ahh, my mistake. Memory's fuzzy in my old age. [03:50] Hobbsee: You may now point, laugh, and call me an old fart. [03:50] nah, you're just old and decrepit [03:50] (I'm still having a hard time coping with having turned 30 2 months ago...) [03:51] LaserJock, It was done before my time. Ask mr_pouit [03:51] somerville32: I guess if nobody complains it's not a big issue, but I'm told that goffice will not print without the gnome stuff [03:51] for instance [03:51] LaserJock: goffice appears to function in Xubuntu. *shrug*... How well it functions is a matter to discuss with someone who uses fewer terminals and more GUIs than I do. [03:52] yeah, I guess we can just let the bugs decide ;-) [03:52] somerville32: what does Xubuntu use for office suite? [03:53] infinity: somerville32 uploaded. [03:53] Abiword and Gnumeric, as apt-cache depends xubuntu-desktop shows. [03:53] Hobbsee: Poor guy [03:54] infinity: thanks for pointing out my (and slangasek's) inability to read. [03:54] LaserJock, Abiword and Gnumerics [03:54] somerville32: hmm, then there could be issues [03:54] Hobbsee: Perhaps if the seed comment was in ASCII blink? [03:54] haha [03:55] somerville32: I might ask upstream a little more about it and shoot Xubuntu an email if it sounds like something to look into [03:55] LaserJock, thanks [03:55] Hobbsee: ? [03:56] imbrandon: sorry, wrong nick completion. i wanted infinity. [03:56] ahh okies [03:56] ponies ? [03:57] * LaserJock uses his laser to atomize every pony in sight [03:57] You missed one [03:57] Ponies! [03:58] * infinity ponders putting "< Hobbsee> ... i wanted infinity." in a fortune file somewhere. [03:58] It's about as close as I'll ever get to a compliment from you, I think. :P [03:59] * Hobbsee slaps infinity [03:59] infinity: i occasionally give compliments :P [03:59] Lies. Subterfuge. [03:59] Even more occasionally, they aren't back-handed compliments [03:59] * StevenK hides [04:00] Okay, life without PGP or SSH keys is actually rather crap. [04:00] infinity: where is livecd.sh and why does it hard-code this instead of pulling it via the seed? [04:00] slangasek: A) livecd-rootfs, B) apt hates me. [04:03] slangasek: If you just do a straight task-based *-desktop installation with apt, it will pull in the bloaty libgoffice instead, so we hardcode the lighter one. [04:03] slangasek: A bit of a pain, but oh well. [04:03] I hear you. [04:03] When the RCMP took my desktop (voluntarily - I did nothing wrong), I sulked [04:04] slangasek: I could add a tiny biy of logic to parse some apt-cache output to pick the right SOVER or something, but it changes once a year or so, so I didn't much care. [04:04] infinity: oh, so even though it's embedded in the seed, apt pulling in the task manages to get the wrong lib somehow? [04:04] slangasek: Yeah [04:04] somerville32: er, are you saying that the RCMP is in possession of your PGP key? [04:05] slangasek, Umm... [04:05] infinity: right, nasty then [04:05] slangasek, Possibly one of them... [04:05] And you upload to the archive using that key? [04:05] he cant upload [04:05] Ah [04:06] Umm... [04:06] * somerville32 just got a really really bad feeling in his stomach [04:06] We should change that. I'd like to see what the mounties would upload. [04:06] Haha [04:06] I had more than just pgp keys on there :/ [04:06] somerville32: If someone else gets your private key, *revoke it ASAP* [04:06] How do I revoke it without the private key? [04:07] You have a seperate revocation certificate backed up. Don't you? [04:07] No. [04:07] RCMP? [04:08] Royal Canadian Mounted Police [04:08] ion_, Canada's FBI + more [04:08] Way cool, the storm has connected as TheMuso. [04:08] I had ssh keys for certain servers owned by a certain company we all know. Should I get those changed ASAP or is it not that big of a deal? [04:09] Uh, yeah? [04:09] Why are you even asking this? :P [04:09] I thought that would have been fairly obvious. [04:09] StevenK: lol [04:09] Why did they take your computer? [04:09] Maybe somerville32 is a terrist. [04:09] lol [04:10] No no, I didn't do anything wrong [04:10] I just didn't think I had to worry about them having that, lol [04:10] * Hobbsee suspects that certain powers that be are going to rethink about giving somerville32 access [04:10] So why then? :-) I’m just curious. [04:11] I'm actually horribly embarrassed at the moment. I shouldn't have said anything. [04:11] somerville32: you'd get crucified even worse if people found out, after it was known to have been comprimised. [04:12] Well, I guess I know better now. I wasn't uploading packages or anything back then so I didn't think it mattered. [04:12] (or maybe I was, I can't really remember) [04:12] If it's in the web of trust, it matters. [04:13] But it never occurred to me how precious they might be [04:13] hum. come to think of it, why don't i have a revovication certificate for my new key? [04:13] Hobbsee: Because you didn't create one? [04:13] Hobbsee: Because you're a slacker? [04:13] * somerville32 goes to take care of this mess. [04:13] somerville32: it's not a very useful means of identifying you as a person if other people are in possession of it, is it? :) [04:13] Hobbsee: Send me your private key and your passphrase, and I'll generate a revcert for you. [04:13] StevenK: i did for my first, i forgot for my second :P [04:13] infinity: :P [04:13] Bwahaha [04:14] Hobbsee: I'm a touch-typist, I promise I won't actually read your passphrase as I type it. *nod* [04:14] He reads sbuild code, he lies! [04:14] infinity: Could you do that for me as well? I’ll post the info you requested to a pastebin. Thank you. [04:14] Bwahaha [04:15] ion_: *giggle* [04:15] Is it safe to put a backup of my private key, ssh-keys, and revok certs on my thumb drive? [04:15] I carry it with me always [04:15] somerville32: Encrypt the whole mess (valume level, or each file), and sure. [04:15] s/valume/volume/ [04:15] infinity, Can just encrypt those files instead? :P [04:16] * StevenK has an 10Mb encrypted FS on his USB key [04:16] slangasek: sup? [04:16] * RAOF thinks that is a fine idea, and looks at doing that himself. [04:16] enc-file: data [04:16] lamont: I give up, what? [04:17] somerville32: Are any of the 3 SSH keys you have listed on LP not in your possession? [04:17] slangasek: was wondering if you were still having the livefs issues, or if that's all happy [04:17] lamont: infinity tells me he's magicked it [04:17] kewl [04:17] lamont: I shamed some people into learning to read, then Hobbsee was kind enough to upload a fix, since I'm keyless this evening. [04:18] heh [04:18] * somerville32 sighs deeply. [04:18] * StevenK ponders handing infinity s/key/clue/ and then thinks better of it. [04:18] * StevenK ducks [04:18] StevenK: Behave. [04:19] StevenK: I may have been gone awhile, but I still vaguely recall how to make your life miserable. [04:19] infinity: I did think better of it. :-) [04:19] infinity: hey now, I was never looking at the seed at all, so I had no opportunity to overlook these comments. :) [04:19] StevenK: Thinking better outloud isn't quite the same thing. :) [04:20] infinity: Oh? Like you know how to do that better than I do. :-P [04:20] slangasek: Consider your slack not only cut, but in the mail. [04:22] I might go catch an early nap, so I can fight the jet-lag and deliver on this "normal core hours tomorrow" thing. [04:22] slangasek: Are you working vaguely 9-5ish most days? [04:25] slangasek: Well, assuming the answer to the above is "yes, which is why I'm not responding right now", I'll see you tomorrow morning. :) [04:25] heh [04:25] infinity: "more or less", yes [04:25] It's good to be back in a sane timezone. [04:26] But AEST is sane.... [04:26] * infinity hugs MST7MDT tightly, and refuses to let go. [04:26] infinity: if you're in the mood for taking the piss out of somerville32 more, perhaps you'll answer this. what should you do, apart from removing the keys from LP, if you don't have a revovication certificate, nor a private key for the key in question? [04:27] Hobbsee: If you have no access to the key or a revok cert, you can still get all the people who've signed your key to revoke THEIR sigs on your key. [04:27] infinity: it's a great TZ, isn't it? [04:27] infinity, no one signed it [04:27] Hobbsee: But, really, what you have is a free-floating key out there, and not much you can do about it. [04:27] Yay for the web of trust, if no one signed your key [04:27] somerville32: That's something positive, at least. [04:27] infinity: well, that's what i was thinking. presumably the only sane solution then is to attempt to drop all links from that key to you. [04:28] Hobbsee: get a farm of machines busy factoring it, and tehn you can revoke it. :0) [04:28] infinity, Are you a sys-admin? [04:28] somerville32: Depends on the day. [04:28] somerville32: We'll go with "yes", though. [04:28] lamont: it's not my key. it's somerville32's :P [04:28] infinity, Do you have access to humboldt? [04:28] Hobbsee: so? [04:28] somerville32: Oh, no I don't. [04:29] humbolt is in california. [04:29] infinity is in canada. [04:29] somerville32: Those who do are mostly in London. And asleep. [04:29] lamont: I suspect he meant root, not console. :) [04:31] Who ever Fujitsu was talking to but messaged me by mistake, my computer wasn't confiscated. [04:32] So the RCMP just took it for something to do? [04:32] They were borrowing it. [04:32] Haha [04:32] No, let them have it. Someone else did something wrong and they had evidence on that computer. It was not compromised in anyway. [04:32] what did the taser-loving RCMP do now? [04:33] somerville32: getting taken by the police == get compromised [04:33] getting, rather [04:33] ugh, my english sucks [04:33] Ok, it was compromised in that sense but I understand the ramifications at the time because I wasn't really using my keys at the time. [04:33] *did not understand [04:34] I'll call them tomorrow, get the key, and revoke it. [04:36] It isn't like I have upload rights anyhow. [04:42] somerville32: i'd be more worried about teh fact that the machines you have ssh access to may be compromised too. === mpt_ is now known as mpt === lc is now known as lcstudio === stu2 is now known as stub [07:21] Good morning [07:21] Morning pitti [07:22] pitti: do you have an ETA on the next lang pack update for Gutsy? [07:24] LaserJock: erm, one is currently in progress, I just don't get a lot of testing feedback on the -proposed packages [07:24] hey StevenK [07:25] pitti: how about for Feisty? [07:25] LaserJock: dito, in -proposed [07:25] and are the still done roughly monthly? [07:25] *they [07:28] there was a big lag, but we are planning to go back to that rhythm, yes [07:28] ok great [07:29] I'm getting gcompris cleaned up and just wondered when I could tell upstream to expect the translations [07:41] good morning [07:41] Good time of day. [07:42] hey ion_ [07:42] Howdy [08:44] From the lack of a new daily for Ubuntu Studio, I am assuming the cdimage cron has been turned off. [08:50] i have to say, this is pretty neat [08:50] http://www.bononia.it/~zack/blog//posts/2007/11/rss_powered_pts.html [08:52] pwnguin: Nice! [08:54] Greetings everybody! [08:57] * Hobbsee waves [08:57] Hey Hobbsee. [08:57] Evening Hobbsee [08:57] * Hobbsee curses cars that break down [09:00] hey Hobbsee [09:00] heya Fujitsu, TheMuso and pitti! [09:01] ogra: ping [09:22] Fujitsu: you think LP already uses something like that for sync requests? [09:22] automated sync, i mean [09:26] pwnguin: No. The autosyncer just looks at the differences between the sources files. It polls. === \sh_away is now known as \sh [09:29] hey seb128 [09:30] hello dholbach [09:36] hi [09:58] cjwatson, there was a mention of you talking about demoting xubuntu packages to universe in #xubuntu-devel. Was there much to discuss on this (that it should be moved to the ubuntu-devel ML), or is it pretty much going to be happening. [09:59] superm1: I'd like to do it, but would also like to have something resembling consensus [09:59] cjwatson, did you forsee any negative consequences by doing so off hand? [10:04] superm1: no technical consequences; the obvious consequences have the nature of marketing [10:04] superm1: IMO they could be effectively spun either way :) [10:04] yeah that's what I had thought [10:05] cjwatson, I'll bring it up on ubuntu-devel then soon and a consensus can get reached there [10:06] is there a Xubuntu developers list? [10:06] yeah i was going to CC both lists [10:10] Riddell: yes, I think we need to loosen the dependencies a bit [10:12] siretart: hey === iLeft is now known as iWebPy [10:13] siretart: do you still work on emacs22 in ubuntu? what are the reason again why the package is different from the debian one? [10:14] seb128: I'd hazard a guess at us considering GFDL free. [10:14] * Fujitsu recalls that to be correct. [10:14] Mithrandir: well, we could have the non-free package to main and Recommends it [10:14] yes, it was the gfdl documentation, and we follow the packaging of romain francoise, emacs upstream and former maintainer of emacs22 in debian [10:15] tjaalton, bryce: Debian removed xserver-xorg-video-i810; should we follow that or keep it? [10:15] Mithrandir: which would make the documentation installed [10:15] and it would drop the delta to a mere dependency change [10:16] siretart: any reason that debian doesn't use this packaging if it's better? [10:16] pitti: we probably should finally do that, now that intel upstream is again improving the driver [10:16] I was just discussing this with pitti again because of bug #172389 [10:16] Launchpad bug 172389 in emacs22 "Please remove emacs22-common-non-dfsg from the archive" [Undecided,Incomplete] https://launchpad.net/bugs/172389 [10:16] all installations default to intel anyway [10:16] tjaalton: btw, I just had that discussion with kwwii: how do we ensure that xorg.conf points to intel on upgrades where i810 was used before? [10:16] seb128: sorry, I'm pretty busy at work, and I need to look again at the package for details [10:16] tjaalton: this should be done anyway, I guess [10:17] pitti, tjaalton: I concur [10:17] bryce: wow, didn't expect you to be awake at this hour :) [10:17] pitti: the debian -intel ships i810_drv.so symlink [10:17] tjaalton: ah, great, so it'll just work [10:17] heh [10:17] heh, it's only 2am ;-) [10:18] siretart: no hurry, that was just curiosity because the situation seems to be suboptimal and we were not sure what would not work with using the debian packaging and adding a recommends on the documentation to make it installed on Ubuntu [10:18] bryce, tjaalton: kicked then, thanks [10:19] btw, those who suffered intel lockups; there's a proposed patch for the server, and it'll probably get in upstream 1.4.1 branch soon [10:19] Hobbsee, Fujitsu: ^^ [10:20] seb128: I'm aware of the situation, and we are working on unifying the debian and ubuntu emacs packages (but granted, I need to take some more time to investigate it more closely) [10:20] siretart: ok, looks good, thanks [10:20] pitti: ^ [10:20] yup, saw it [10:20] no idea what to do with teh removal request now [10:21] tjaalton: Lockups? I've never experienced those, as far as I know. The crashes are gone, however. [10:21] but IMHO it's better to just let the package be where it is and add teh conflicts/replaces: [10:21] right [10:21] (C/R: needs to happen anyway) [10:21] tjaalton: \o/ [10:22] Fujitsu: well, it's an EXA issue, and I think you are using XAA now [10:23] that still crashes, though [10:23] tjaalton: Ah, true. With EXA it's all very slow, but that was about it. [10:23] Fujitsu: ok, that's another thing ;) [10:25] tjaalton: Has anybody worked out why that happens, or if a fix is on the way? [10:25] * Hobbsee wonders if it's worth trying EXA again [10:25] Fujitsu: if you mean that scrolling is slow, it's known [10:26] tjaalton: And rendering of new windows. [10:27] ugh, libtool from Debian experimental breaks with dash [10:27] * cjwatson scratches head [10:27] Nice. [10:27] I can see *why* [10:28] it's testing shell features with the shell that configure's using; configure re-execs itself with bash, but when libtool runs it has different logic and decides that the existing shell is fine [10:33] TheMuso: orca was removed in Debian due to being obsolete and unmaintained; is that true for Ubuntu, too? [10:34] pitti: orca or gnome-orca? [10:34] orca | 0.2.3 | hppa [10:34] orca | 0.2.3ubuntu1 | source, amd64, i386, ia64, powerpc, sparc [10:34] I think gnome-orca is the thing that TheMuso cares about [10:34] pitti: If the source package is orca, then its nothing to do with a11y. [10:35] ah, ok [10:35] Description: Router Monitoring and Graphing System [10:35] whoops, sorry :) [10:35] gnome-orca is the package for GNOME, named gnome-orca for that reason. [10:35] pitti: np [10:38] siretart: ffmpeg was removed in Debian "[auto-cruft] obsolete source package"; that's not true for us, I guess? [10:39] ah, that: [10:39] ffmpeg-free # ffmpeg -> ffmpeg-free rename in Debian (needs merge) [10:39] (on sync blacklist) [10:40] Riddell: dolphin is now built by kdebase source in Debian, and the source got removed; will you follow that? [10:41] (I don't remove it for now, since it looks maintained in ubuntu) [10:41] * Fujitsu wonders why KDE people seem intent on merging everything into one source package. [10:42] superm1: based on the discussion between somerville32 and cjwatson earlier, it appears that canonical doesn't provide commercial support for xubuntu-desktop anwyay? [10:43] Hobbsee, well it's listed on the xubuntu.org help page [10:43] that they do. perhaps they don't have any clients for it however [10:43] superm1: who wrote the xubuntu.org help page? [10:44] Hobbsee, that I couldn't tell you. I joined xubuntu-devel after it was already there. [10:44] superm1: i'm suspecting the answer is "not someone from canonical"? [10:45] Hobbsee, well I don't know very much of the history behind Xubuntu's conception, so its possible. [10:46] Hobbsee: Xubuntu started as a community thing, if I recall. It still is, in some ways [10:47] It is very much still a community thing, isn't it? [10:47] StevenK: that's what i thought, too. [10:48] Fujitsu: yeah, there are no canonical-assigned devs to it [10:48] Hobbsee: That's what I thought. [10:48] well then you have to really wonder how that official canonical support thing ended up on the bottom of that page [10:49] superm1: Somebody uneducated noting that the packages were in main, I would expect. [10:49] Fujitsu: ++ [10:49] superm1: this is why i question the credibility of that page [10:49] All the more reason to demote it.. [10:49] (and perhaps look at the rest of the site) [10:49] hello superm1 and Hobbsee [10:49] I replied to the xubuntu demoted to universe mail [10:49] hi luisbg_ [10:49] hey luisbg_ [10:49] I agree [10:50] superm1: I don't understand your releases.u.c. mirror space concern -- Xubuntu images have never been there, have they? [10:50] * Hobbsee accepts the mail [10:50] fudge, i sent mine from teh wrong email. [10:51] Hobbsee: Is it still queued? [10:51] * Fujitsu can't see it. [10:51] pitti, this was raised to me by somerville32 as a concern voiced earlier. i was just going off what i had heard. [10:51] it appears that they aren't though. you are correct. [10:52] Fujitsu: no. i just noticed it when going thru the queue :) [10:52] Fujitsu: this is the advantage of sending mail to a list that you're an admin of :) [10:53] to me the biggest concern is that any development of main packages require a lot more bureaucracy.. unless you are a core dev [10:53] Heh. [10:53] and that can make xubuntu development slower [10:53] having been the one fixing xubuntu stuff over the past couple of days, this is very true [10:54] superm1: right, I just replied to the mail, too [10:54] * Hobbsee wonders if the livefs, etc, logs will appear for the other canonical-built universe-based distros on p.u.c/~ubuntu-archive anytime soon [10:55] how canonical supported is xubuntu anyway? [10:55] based on yesterday's logs, ti's not [10:55] Moving xubuntu to universe means that more people can help out with it, and more MOTUs can more easily sponsor uploads for it. [10:55] the security guy has done a grand total of one xubuntu-based secuirty upload. [10:56] IMO its a good idea. i'll reply to the mail when I see it, if its on -devel. [10:56] hey TheMuso [10:56] c [10:56] hey luisbg_ [10:56] =) [10:56] TheMuso, yeah it's on -devel === cprov-out is now known as cprov === bluekuja_ is now known as bluekuja [10:58] this reminds me... when is ubuntu studio going to appear in the "other ubuntu editions" list? [10:59] siretart: ping [11:01] siretart: why does libxine1, which appears to be the main library, depend on gnome-based stuff? [11:02] siretart: (thru libxine-plugins). this is obviously somewhat suboptimal for kde-land. [11:02] Hobbsee: to the best of my knowledge, indeed, Canonical doesn't offer commercial support for Xubuntu [11:03] Hobbsee: he said he needs to losen the dependencies [11:03] cjwatson: right. ah, pitti's said that in writing as well [11:04] Riddell: oh, that was the discussion earlier? got it. [11:05] So... main is officially supported, but not? [11:09] Fujitsu: only for security updates [11:09] Fujitsu: commercial support covers products, not components [11:09] it's a bit fuzzy, admittedly [11:10] does anybody has some tool to reassign launchpad bugs all bugs on a product to an another one? like when a source package is renamed [11:10] p-lp-bugs? [11:11] I'd be interested to see how canonical support contracts measure up with RHEL support [11:11] pitti: it has a builtin command for that? that's likely a small piece of code but I was just wondering if somebody had it handy before starting reinvented the wheel [11:11] Fujitsu: this is exactly why I would like to move Xubuntu to universe, in order that the description of what is supported becomes reasonable and comprehensible again [11:11] realist, there was a review in a magazine some months ago ... [11:11] ogra: Linux Format [11:11] ah, right [11:12] seb128: no, you need to write a script; I don't know an existing tool which does that [11:12] realist: said review indicated that we were tied with RH support [11:12] pitti: ok, thanks [11:12] cjwatson: RIght, it makes sense to have things make sense. [11:18] pitti: `When you install software from the main component you are assured that the software will come with security updates and technical support.' [11:19] If Canonical doesn't provide support for all of main, then how is the technical support different from universe? [11:19] Fujitsu: hm, that's not entirely true then; security yes, but not commercial supprot [11:19] doesn anyone know why upstart-logd has no replaces line for bootlogd ? i thought it should replace it ... [11:19] (that's on ubuntu.com/community/ubuntustory/components) [11:20] pitti: Last night you said you'd be going through NEW today.. Is that still the plan? (Not trying to rush you, I just want to know what I can expect) [11:20] Fujitsu: well, it does not mention commercial support, so I think that's ok [11:21] soren: yes, it is; still doing removals [11:21] * soren hugs pitti [11:21] pitti: Then what support is there to distinguish it from universe? [11:21] pitti: You rock! [11:21] asac: Wouah did you see thunderbird came before evolution in the full circle survey? [11:21] Fujitsu: security's, and distro team's [11:21] Fujitsu: where they say they support it, vs where they say they don't supprot it. [11:22] Can anyone tell me which issue of Linux Format had the Canonical support review? [11:23] lool: haha [11:23] Hobbsee: Um? [11:23] realist: would have thought the correct answers would be any of: "JFGI", "offtopic", or other, similar answers. [11:24] Fujitsu: universe says it's unsupported. [11:24] last i checked, anyway [11:24] Hobbsee: JFGI? [11:24] Hobbsee: Right, but I don't see technical support guaranteed for all of main, unless security support counts as it. [11:24] realist: long from is http://justfuckinggoogleit.com/ [11:25] realist: of course, those answers would be in politer form, but... [11:25] Hobbsee: considering it was just mentioned, I figured they'd know which issue [11:26] Fujitsu: true, but it says that it does (like you say) [11:40] realist: August 2007, I believe === dholbach_ is now known as dholbach [11:44] cjwatson: cheers :-) === ogra1 is now known as ogra [11:52] c [11:52] ugh wrong tab [11:59] TheMuso: C? use Python! :) [11:59] heh [12:07] Is Python as fast as C? [12:08] realist: it's in the order of 30 times slower than C programs (I measured this with a prime factorization function, which might not be representative) [12:08] realist: but it's also 30 times faster to actually get done something in Python than in C (development-wise) [12:09] so the use cases are different [12:09] google for 'python c speed comparison' or some such [12:10] realist: for many tasks, execution speed does not matter at all, but robustness, correctness, and fast/easy development does matter; that's where scripting languages and Python in particular excel [12:11] Well, I prefer Python for my day to day work (sysadmin) [12:12] * persia vaguely complains that a 486 tablet is no as useful as it once was [12:12] pitti, can you trigger package rebuilds? [12:12] xhaker: yes [12:13] https://edge.launchpad.net/ubuntu/+source/eclipse/3.2.2-4ubuntu1/+build/457225 [12:13] pitti, i386 failed and right next lpia succeeded. :) [12:14] if you could try it again on i386 i'd appreciate it [12:14] * pitti cranks [12:14] done [12:15] * xhaker is joyful \o/ [12:25] When is iso testing expected to start? [12:25] Or is that not happening for alpha 1? [12:25] TheMuso: some 12 hours ago :) [12:25] see topic [12:25] Oh. I would have thought there would be more activity in here. :) [12:26] haha right. One tends to miss such changes if one idles in the channel, and the backlog is large enough to miss topic changes. [12:27] TheMuso: don't worry :) if you can help out with testing, that would be great [12:27] pitti: Once I get the latest image synced, I'll at least be able to do an alternate install test. [12:30] hi folks, how am I supposed to write a debian watch file for a software which is realeased on launchpad (let's say, plone)? [12:30] pitti: But I am also thinking about Ubuntu Studio images. We would like to make an alpha 1 release, even if it means telling people which daily image to test with. However, yesterday's daily image dir on cdimage has no isos, and no image has been built today, understandably [12:30] pitti: Would it at all be possible to get an Ubuntu Studio image set built? [12:33] kobold: 19:37 < RainCT> https://launchpad.net/freevial/+download/ https://launchpad.net/freevial/(?:.*)/(?:.*)/\+download/freevial-(.*)\.tar\.gz [12:33] pochu: thanks a lot [12:33] Adjust it for your project :-) [12:34] kobold: thank RainCT and morph_, I've grepped my logs :) [12:34] xhaker: A while ago you asked me why I was passing on the pilot-link merge (thanks for taking care of it). The reason is I only touched it so we could get rid of sylpheed-claws. I don't have any actual interest in the package. [12:35] kobold: a bug report in launchpad about the odd url (which includes the series afaik) might be interesting though... [12:35] xhaker: Also, please remove the bug numbers from your new debian/changelog entry when you merge as Launchpad happily closes the bug again and sends another round of bugmail. [12:36] ScottK, hmm.. remove.. or add? I see.. Close LP ##### syntax automatically closes those bugs right? [12:37] xhaker: LP: #xxxxxx [12:37] xhaker: Don't repeat the closure of already closed bugs. You can check the source.changes file to see what would be closed, and make sure they are new closures. [12:38] xhaker: What pitti said, but if it's present on a subsequent upload LP will "close" it again and spam everyone subscribed to the bug (that's an LP bug). [12:38] oh, for merges? [12:38] pitti: Yes. [12:38] xhaker: yes, if you describe the remaining Ubuntu changes, yuo should use "LP #xxxx" (no colon) [12:38] TheMuso: building you an image now [12:38] so that you keep the reference, but don't trigger the autoclose mechanism [12:39] cjwatson: Thanks a bunch. [12:39] should be ready in 15 minutes or so, assuming it actually buidls [12:39] builds [12:39] cjwatson: Going on the last two days images, and merging our seeds earlier today, it should. [12:39] thanks for clarification folks. Sorry for the bugmail :) [12:51] cjwatson: Thanks again, they've built it seems. === dendro-away is now known as dendrobates [12:56] MOTU Q&A session in 4 minutes in #ubuntu-classroom === dholbach_ is now known as dholbach === asac_ is now known as asac [13:27] * pitti hugs pedro_ for all the OO.o verifications [13:28] * pedro_ hugs pitti back [13:28] still one to do [13:29] pedro_ rocks :) [13:30] * pedro_ hugs ogra [13:30] :) [13:31] TheMuso: excellent === cjwatson_ is now known as cjwatson [13:33] pitti: what does cf. mean? [13:34] Riddell: isn't that the English way to say "in the name of.."? [13:34] oh, noes, I mixed that up, sorry [13:34] pitti: usually "pp" [13:34] I thought it was "compare to" [13:34] right, sorry [13:34] although I don't know what that actually means :) [13:35] per procura? [13:35] Per Procurationem [13:35] (By agency of) [13:36] I'm impressed (so long as you didn't cheat and look it up on wikipaedia :) [13:36] * persia cheated, but not with wikipedia [13:37] cf. is compare, yeah [13:38] (Lat. confer) [13:47] * soren looks at the NEW queue and hugs pitti [13:47] soren: working hard on it :) [13:48] Erk.. -ENOCOFFEE [13:52] thegodfather: why did vmware-server recently get dumped into gutsy partner repo? [13:53] (more specifically, why does the vmware-server there appear to depend on libssl0.9.7) [13:54] Hobbsee: Why is that strange? [13:55] soren: because it doesn't actually install :) [13:55] but it belongs ito the partner repo [13:55] Hobbsee: Well, that's now what you're asking :) [13:55] Hobbsee: s/now/not [13:56] soren: there was a reason it got removed from feisty-commercial. i'm wondering why it was then re-added, if the installability wasn't fixed. [13:56] soren: so, the question about why it's now there, although obtuse, si still legit :) [13:56] Hobbsee: feisty? gutsy, surely. [13:56] persia: i thought it got removed from feisty-commercial? [13:56] or is the feisty version fine? [13:57] feisty was fine last I checked, but that was a while ago. Looking now. [13:57] * Hobbsee can never remember what's in which release, as she tends to use the dev ones [13:57] Hobbsee: Er.. [13:57] Hobbsee: It's still in feisty-commercial? [13:58] I'm running the version from feisty-commercial, actually. [13:58] soren: right, OK, so i'm on crack, and it's gutsy that's the problem [13:58] (why oh why aren't they all in the rmadison scripts?) [13:58] even the backports lists are - why not partner? [13:58] Partner is not really a part of Ubuntu. [13:59] ...I seem to have missed a point somewhere. What's the issue with the vmware-server in the gutsy partner repo? [13:59] Hobbsee: feisty is fine. [14:00] soren: It depends on libssl0.9.7, which isn't shipped in gutsy [14:00] persia: Oh. [14:00] persia: That's.. Um... Interesting :) [14:01] soren: Should be fixed by a recompile, but most of us can't :) === pkl is now known as pkl_ === cprov is now known as cprov-lunch [14:18] <_MMA_> slangasek: Are you around? [14:18] most unlikely, he went to bed very late [14:18] <_MMA_> Ahh... Ok. [14:21] superm1: lol @ nuvexport copyright: "... Foundation; either version 2 of the License." [14:25] pitti: did I answer your question about dolphin? [14:25] pitti: it's only built from kdebase in debian experimental, not unstable, but I'm not following that (we call that version dolphin-kde4) [14:26] right, ok [14:44] * Hobbsee starts doing bits of NBS [14:44] pitti, hi [14:44] Riddell: erm, why did kaffeine-xine get NBS'd? [14:45] Hobbsee: to keep closer to debian it's include in kaffeine binary [14:45] oh right [14:50] bah, i can't fix that anyway then === \sh is now known as \sh_away [15:02] hi tkamppeter === cprov-lunch is now known as cprov [15:09] where to findthe proposed repos ?? [15:11] phaidros: System -> Administration -> Software sources [15:11] phaidros: Updates tab [15:13] pitti: nup, I just needed it on a server in sources.list and wasn;t able to guess the right url :) [15:13] and in de mirrors it seems not to be there [15:14] phaidros: ah (you need to mention your level of expertise :) ) [15:14] so I just use main mirror now, and its there :) [15:14] thanx anyway [15:14] phaidros: de.archive. has them, though [15:14] phaidros: just add another line with gutsy-updates -> gutsy-proposed [15:14] really? update was yuckin' [15:14] phaidros: I use de., too, and it works fine [15:14] I check that later, have to fix urgent mailserver-foo o.O [15:14] thx pitti [15:15] pitti, you use .de on your dev machine ? [15:16] i'd be to impatient to wait for the mirror run every time i upload stuff [15:16] ogra: sure [15:16] deb http://de.archive.ubuntu.com/ubuntu hardy main restricted universe multiverse [15:16] deb http://archive.ubuntu.com/ubuntu hardy main restricted [15:16] i know these, i use them on stable machines ... [15:16] ogra: i. e. I use de. for the bulk, and archive.u.c. for the latest crack in main [15:17] and use pinning ? [15:17] No pinning is required. [15:17] ogra: no, apt prefers the topmost mirror which has the latest version [15:17] ah, i didnt know [15:17] I don't need COTD for universe, a day of delay is fine for me [15:17] yeah [15:19] Had the mirrors been implemented better, there would be almost no delay. [15:20] A F2F network between mirrors, where the software watches directories to be synced and syncs them on the minute something changes. [15:23] doko, i'm grabbing the xaos merge if you dont mind [15:24] ogra: please do [15:25] how to find a package which contains a file named something with "db" and "recover" ? [15:30] phaidros: grep in Contents.gz? [15:30] http://archive.ubuntu.com/ubuntu/dists/gutsy/Contents-i386.gz etc. [15:31] apt-file may help with that [15:34] soren: hm, would you mind reuploading open-vm-tools without the 2.3 MB autom4te.cache/? [15:34] pitti: It's from the orig.tar.gz. [15:34] pitti: Right? [15:34] pitti, pfft, that makes uploading to fast : [15:35] soren: right [15:35] soren: ah, you didn't create the tarball? [15:35] pitti: You want me to repack the tarball for that? [15:35] No, it's an actual upstream tarball. [15:35] soren: only if you actually packaged it yourself [15:35] Ok, good. [15:35] ah, ok; nevermind them (D'oh, though) [15:35] * soren was kind of baffled for a few seconds there [15:35] s/them/then/ [15:35] :) [15:36] pitti: Oh, I see why you thought so. The version number doesn't really suggest a real release, but that's how they do it,apparantly. [15:37] hm, the copyright doesn't point out the files which are licensed under BSFD [15:37] BSD [15:37] argh argh typing [15:37] . o O { 7 hours of archive admin in a row is detrimental for brain and typing sanity) === heno_ is now known as heno [15:41] pitti: yow [15:45] * pitti jumps for joy \o/ [15:45] source NEW is empty [15:46] and binary NEW is down to the two 2.6.24 kernel things we stall [15:47] do yo want nor eNEW crack ? [15:47] pitti: well done! [15:47] *more [15:47] ogra: yes he does [15:47] heh [15:47] * pitti jumps to ogra's throat [15:48] pitti: you can't kill him. [15:48] pitti, i was plannig to get libflashsupport in :) thats a really tricky one .... but lucky you i'm not done yet :) [15:48] * ogra coughs up pitti [15:53] ogra: are you using git.0pointer.de/repos/libflashsupport.git? [15:55] tkamppeter_: did you get some feedback about your dnssd backend? I wonder whether I should commit it to Debian already [15:56] crimsun, no i wasnt doing any code stuff yet, i'm aware i should use lennards version though ... but for now i was looking into possibilities to build it with tls support and was checking the license since these seem theto be the two reasons it didnt go into debian yet [16:14] pitti: please give-back camlimages. Thanks. [16:15] geser: done === ant30_ is now known as ant30 [16:47] lamont: I think I find your ia64 upgrade issue, if you could give it a go? wget http://people.ubuntu.com/~mvo/update-manager/update-manager_0.81.1.tar.gz , unpack it and run ./gutsy (if its a server with --mode=server) [16:48] lamont: and let me know please if it still explodes :) [16:48] ah, iz server. [16:48] so ./gutsy --mode=server [16:50] mvo: I need to head officewards, will test it later today and let you know [16:51] lamont: sure, thanks! [16:51] pitti: please give-back: cduce e16menuedit ecasound2.2 dieharder [16:53] pitti: have I mentioned how much I appreciate that you can give things back? [16:53] lamont: not recently :) [16:53] geser: done, btw [16:54] thanks [17:02] lamont: urgs, wrong url. I write the instructions how to test it up in the bugreport [17:02] pitti: please give-back: vr zipper.app textedit.app terminal.app [17:02] pitti: is the bug with + in the version in your script fixed? [17:03] mvo: ok. [17:07] geser: yes, it is [17:08] geser: done [17:08] thanks again [17:08] * pitti should write an IRC plugin for this :) [17:08] * ogra sees cjwatson's fuse megre and thinks we should really package our own like he thinks every release if he sees that big delta [17:10] lamont: instructions udpated :) [17:10] mvo, mind if i do the tftp merge ? [17:11] ogra: go ahead, I think I did the tftp-hpa already, but it may be updated already [17:11] hmm, mom still shows it [17:11] ogra: I have a tftp-hpa_0.43-1.1ubuntu1_source.upload file here [17:12] Subject: Accepted tftp-hpa 0.43-1.1ubuntu1 (source) [17:12] its 0.48 :) [17:12] aha! [17:12] yeah, wrong cycle [17:12] * mvo blushes [17:12] ogra: go ahead :) [17:12] ogra: it's a big list, but it's not really as bad as it looks IMO, nor is it a difficult diff to maintain [17:13] ogra: hm, there is still something fishy, hold on a sec [17:13] cjwatson, well, its nearly all in the debian dir :) [17:13] ogra: sure, but definitely not worth maintaining our own [17:13] s/all/everything/ [17:14] using Debian's gives us the advantage of seeing merges as they arrive, which I think is non-trivial [17:14] ogra: I have it here, I just haven't uploaded it. very strange. I upload it now [17:14] cjwatson, ah, right ... [17:14] mvo, ok [17:14] ogra: I even submited the diff to debian [17:14] * mvo scratches his head [17:15] pitti: if it didn't require that you locally tell it, we'd have to revoke your access... [17:15] * pitti points out the ":)" [17:16] mvo, you need more sleep [17:20] bryce: upgrading to hardy has broken VT switching for me; was I right to attach this to bug 131751, or should I file a new bug? [17:20] Launchpad bug 131751 in xorg-server "Unable to switch Virtual Terminal with C-A-F[1-6] on Intel-based new laptop" [Undecided,New] https://launchpad.net/bugs/131751 [17:21] cjwatson, i saw the same in ltsp yesterday ... [17:21] (havent filed it yet though) [17:22] it only hapens if X was up once ... if i boot directly to console all is fine [17:22] yes, it's fine from gdm's X server too [17:35] gdm's still the same, or I'd suspect that it was starting up the server differently ... [17:36] i wonder on what crack was Hobbsee [17:36] i can install vmware-server just fine in gutsy [17:36] pitti: please give-back: rocklight [17:37] done [17:39] thegodfather: Really? Was it a fresh gutsy install, or an upgrade from feisty? [17:39] persia: i did both.. [17:39] pitti: and 6:15 is very early even when I didn't go to bed very late. :) [17:39] slangasek: hehe -- good morning [17:39] thegodfather: Hmm.. Right. I'll download gutsy vmware-server again, but I could reproduce earlier. [17:40] persia: can somebody atleast show me the error? [17:40] thegodfather: dependency missing: libssl0.9.7 [17:40] * thegodfather scratches his head [17:41] blah. archive.canonical.com is slow (from here) [17:42] i don't understand [17:42] i installed this thing today [17:43] one sec.. [17:44] ok, does someone know why xserver-xorg-video-all is currently uninstallable then? [17:44] pitti? [17:44] pitti: and good morning :) [17:44] uh? eww, again? it wasn't last time I checked [17:45] slangasek: at least it doesn't seem to affect the current CDs [17:45] * pitti checks [17:45] pitti: it affects ubuntustudio, which is why the question has arisen [17:46] and it's the current 7ubuntu2 that's listed as uninstallable, as of an hour ago [17:46] hang on, trying in a small hardy chroot [17:46] thegodfather: I'm having trouble reproducing it again. When I try in a chroot, it tells me it can't present the license, and when I try in my workstation, it says I have something else I need to purge (and I don't find it available for purging). My apologies. [17:47] xserver-xorg-video-all: Depends: xserver-xorg-video-i810 but it is not installable [17:47] E: Broken packages [17:47] slangasek: E: Package xserver-xorg-video-i810 has no installation candidate [17:47] persia: i was able to reproduce it in a chroot [17:47] eww, that's gone from the archive, and we don't want it any more either [17:47] thegodfather: Excellent. [17:47] ? [17:47] pitti: i guess we did drop libssl0.9.7 from gutsy.,. right? [17:47] pitti, ergh [17:47] I'm fairly sure we do still want i810 [17:47] thegodfather: If I remember the transition properly, it's just a rebuild. [17:47] pitti: er, and you took it away before the alpha images were blessed? :) [17:47] -intel still has regressions on some older hardware [17:47] mjg59: it was removed from Debian, and tjaalton and bryce ack'ed it [17:48] persia: we did rebuild vmware-server with the small exception that we don't have the source code [17:48] persia: all those Depends are hardencoded in the packaging [17:48] pitti, we need it (se #c) [17:48] *see [17:48] pitti, thats why kyle backpoted it last cycle [17:48] thegodfather: Yep. The rebuild has to happen for the blobs, and then the packaging can be adjusted. [17:48] ok, I copied it back [17:49] not sure it works with the current X we have in hardy right away though [17:49] ok, thanks [17:49] <855 probably wants to still be i810 [17:49] tjaalton/bryce said that X defaults to intel nowadays, hmm [17:49] That probably wants fixing, then :) [17:50] pitti: err, hardy had 2:1.7.4-0ubuntu6, not ubuntu5 [17:50] just copying back ubuntu5 probably won't be great [17:50] xserver-xorg-video-i810 (2:1.7.4-0ubuntu6) hardy; urgency=low [17:50] * Rebuild against the new xserver-xorg-dev. [17:50] * debian/control: provides xserver-xorg-video-2. [17:50] I think the easiest plan is to reupload ubuntu6 as ubuntu7 [17:50] It's my understanding that -intel hasn't really been tested on i845 (especially with LVDS), and I'm not sure of the i830 situation [17:51] ok, I'll do that then [17:51] pitti: I can't reproduce bug 172807 with the current image. Are you sure your test was current? [17:51] Launchpad bug 172807 in cdebconf-keystep "automatic keyboard layout detection loops and messes up screen" [Undecided,Confirmed] https://launchpad.net/bugs/172807 [17:52] at least not on i386 [17:52] cjwatson: hm, I tested on amd64 [17:52] but that shouldn't make a big difference [17:52] cjwatson: but if I had the old image, network-manager would have been broken, too [17:52] so I'm fairly sure it was current [17:52] (and I didn't get the busybox failure either) [17:52] I suppose 64-bit awfulness is *possible* [17:53] it's usually best to check /var/lib/dpkg/status in the running installer to confirm the version [17:53] -i810 reuploaded [17:53] slangasek, mjg59: ^ [17:53] pitti: thank yous [17:53] sorry for the misunderstanding [18:26] https://wiki.ubuntu.com/HardyHeron/Alpha1 is still light on both features and caveats; is there anything else folks would like documented there? (If you can dodge my wiki lock for ongoing structural edits) [18:28] slangasek, do you really think its worth doing that for alpha 1 ? i mean beyond putting there "be careful this is an alpha release, things are supposed to break dunrign development !" .... [18:28] slangasek: yeah, if you think it's important enough I'll add a note about keyboard selection not working post-install due to the lack of an xorg.conf. [18:28] evand: probably reasonable :) [18:29] ogra: we've had /something/ in the past, and in particular documenting caveats prominently is useful to avoid duplicate bug reports [18:29] ogra: I'm not asking for it to be a masterpiece of documentation, but it's going up on the ubuntu.com website as soon as it's done :) [18:29] i dont think we had anything before gutsy ... thats relatively new [18:30] oh, on the website ? wow [18:30] ogra: c.f. http://www.ubuntu.com/testing/herd1 [18:30] that's not gutsy :) [18:30] no, indeed [18:30] wow, i didnt even know that [18:31] Burgundavia's secret cabal :) [18:34] slangasek: broken VT switching might be worth mentioning, if it affects more than just Intel [18:35] cjwatson: reference please? [18:35] mjg59, I know of bug 114331 and bug 135093; are there any other -intel with pre-855 bugs you know of? we'd like to alias i810 to intel like Debian, because there's still a lot of people using i810 and running into issues [18:35] Launchpad bug 114331 in xserver-xorg-video-intel "xserver starts but display is corrupt (i830M,845G)" [High,Incomplete] https://launchpad.net/bugs/114331 [18:35] Launchpad bug 135093 in xorg-server "xserver-xorg-video-intel does not work with 845G" [Unknown,Confirmed] https://launchpad.net/bugs/135093 [18:37] bryce: Before 855, it's possible to attach basically arbitrary LVDS encoders to the chipset - the BIOS knows how to talk to them, but -intel doesn't use the BIOS so needs to have code for each of them individually. As far as I know, the entire set isn't currently supported. [18:37] slangasek: bug 131751 [18:37] Launchpad bug 131751 in xorg-server "Unable to switch Virtual Terminal with C-A-F[1-6] on Intel-based new laptop" [Undecided,New] https://launchpad.net/bugs/131751 [18:37] old bug, but both ogra and I only started seeing it with hardy [18:38] ok, thanks [18:41] mjg59: ok thanks, I'll follow up with intel to see if they'll be adding the bios replacement calls for those old chipsets [18:54] pitti: please give-back: fmultivar foreign ftrading foptions [18:56] tjaalton: any objections to me taking care of the syslinux merge? I need to make a change to the gfxboot patch anyway. === jpatrick_ is now known as jpatrick === pwnguin_ is now known as pwnguin [19:12] geser: done [19:14] pitti: tell me if you get bored of my give-back requests and I'll pester an other build-admin :) [19:14] geser: no, don't worry; with my script it takes me more or less just a cut&paste [19:16] good, then please give-back cairo-ocaml [19:17] done [19:17] thanks === marcheu_ is now known as marcheu === nxvl_work_ is now known as nxvl_work [19:51] Riddell, slangasek, superm1: ah, -video-810 is in, so it's installable again [19:52] pitti: please give-back gnustep-dl2 [19:52] done [19:53] pitti: yep, retrying ubuntustudio now [20:06] when I look at https://edge.launchpad.net/ubuntu/hardy/+source/wmshutdown do I see it right that it got build twice? [20:06] see also https://edge.launchpad.net/ubuntu/hardy/amd64/wmshutdown [20:07] and I remember asking for a give-back for it (as it FTBFS) [20:09] hrm, an edit to the alpha release notes that I can't credit because I can't find his real name [20:10] geser: heh, yeah; interesting bug, probalby only revealed itself becuase this has only been built once *ever* all the way from warty ;) === cprov is now known as cprov-out [20:18] pitti: you mean a bug in soyuz? [20:22] geser: probably [20:25] is it worth reporting? as this would mean, that I shouldn't upload a -1build1 version (wmshutdown is listed in the NBS file for libglib1.2) to not break the current state [20:50] * lamont is reminded that he wishes kde used smaller packages (so that they build quicker), or didn't upload so often. [20:54] Vernadsky is acting strange... It's taken at least 6 minutes now to unpack python2.5. :/ [21:02] soren: it's probably just the log thingy which is trailing because the queue builder ran, or something like that. [21:03] Mithrandir: Ok... At any rate, building that package should take < 1 minute. [21:03] Mithrandir: Oh, now it's doing something else! [21:04] Heh.. sbuild output: Build needed 00:00:07, 1660k disk space [21:04] ..but launchpad says: Finished: 6 minutes ago (took 30 minutes) [21:05] I’d guess there’s a cron job that runs every 30 minutes, and the granularity for the latter value is that. [21:06] soren: LP sometimes confuses itself, not only its users. [21:07] hi all [21:07] I've got some problem with my simple c++ program [21:07] using ubuntu gutsy [21:08] string str; [21:08] cin >> str; [21:08] cout << str << endl; [21:09] svschwartz << "Please read the topic"; [21:10] ok [21:10] thnx [21:10] hi folks [21:13] Hi === cjwatson_ is now known as cjwatson [21:17] Mithrandir: You're full of answers tonight. :) Can you answer why a package with this control file: http://pastebin.ca/803054 has sparc, powerpc, ia64, and hppa builds on https://edge.launchpad.net/ubuntu/+builds?build_text=open-vm-tools&build_state=pending ? [21:18] soren: because PaS, not the package controls what it gets build records for? Talk to lamont or infinity about getting overrides added to PaS [21:18] PaS? [21:18] packages-arch-specific [21:19] it's a file inherited from pre-wanna-build times in Debian [21:19] http://cvs.debian.org/srcdep/Packages-arch-specific?cvsroot=dak [21:19] argh, I do wish gdb could attach to Xorg usefully [21:19] slangasek: and still used, for some reason. [21:20] Mithrandir: what would you like instead, OOI? [21:20] slangasek: believe what the package claims? [21:20] Mithrandir: some maintainers are idiots? :) [21:20] slangasek: then I think we should twap them until they fix their bugs. [21:21] It would be nicer if the source.changes listed the architectures it would make sense to buld the package for. [21:21] soren: the .dsc does, iirc [21:21] Oh.. Why not just depend on that then? [21:21] Er.. No, it doesn't. [21:22] (being able to add new ports easily? ... just guessing) [21:22] Mithrandir: P-a-s is nice because porters know better than arbitrary maintainers what their ports need/can use, so it cuts down on round-trips in the general case [21:22] Mithrandir: anyway, I do thwap maintainers who have buggy arch declarations, when I see them [21:22] soren: does too? [21:23] (principally, because in Debian being in P-a-s doesn't prevent a rogue uploader...) [21:23] Mithrandir: Mine does'nt? [21:23] soren: your lists "any", I suspect? [21:23] Oh. Er.. Yes. [21:23] I was mentally scanning for i386 amd64, but clearly didn't find it. [21:23] Hm... Why would it say any? Because there's an arch: all package in there? [21:24] no, that would be "all" [21:24] slangasek: It's a mixed package. [21:24] "any" means "this package is allowed to be built (or attempted) for any arch" [21:24] slangasek: There are two [i386 amd64] packages and one [all] [21:24] soren: if I only have Architecture: i386 amd64 for all my binary packages in debian/control, I get Architecture: i386 amd64 in ../*.dsc [21:25] Why does Arch: any land in the .dsc? That seems odd. [21:25] soren: is this in the archive? what package? [21:25] slangasek: all + "i386 amd64" gives you Architecture: any in the .dsc [21:25] heh, ok [21:25] open-vm-tools. Just got out of source NEW. [21:25] just tested [21:26] Mithrandir: sounds buggy to me :) [21:26] slangasek: http://pastebin.ca/803054 <--- the control file. [21:26] slangasek: arguably, yes. Though what should it say, then? [21:26] Arch: i386 amd64 all, maybe. [21:27] Mithrandir: that's what I would expect [21:27] but there may be Deep Reasons why that's not done [21:28] I thought the deep reason was that nobody understood the correct semantics of Architecture when all is involved [21:28] (including all the stuff that parses it) [21:28] k :) [21:28] cjwatson: that might well be the reason, in which case I'm in favour of deciding on a set of semantics, implementing that in dpkg and fixing what breaks. [21:28] that probably involves Work though [21:30] if you want to know the build-arch from a .dsc, it includes changing to list it for all binary packages from one source package :/ [21:31] so any for any set of binary packages not resolving to a single architecture seems indeed fine [21:31] Mithrandir: IIRC aj did try to get it sorted out once [21:31] but I may not RC [21:31] sistpoty: sorry? It should just be the union of Architectures listed in debian/control, right? [21:31] cjwatson: you're aware you've fallen off of #ubuntu-release? You're missing out on my publish-release-induced speculations :) [21:31] oops [21:32] * cjwatson falls back on [21:32] Mithrandir: why? (or rather what would you gain from this?) [21:33] sistpoty: you could build on the right set of architectures without PaS [21:33] hm... [21:34] since you could then go "if ! grep -q $(dpkg --print-architecture) foo.dsc; then # skip this source ; else # build this ; fi [21:35] ogra: this display problem seems to be specific to edubuntu; bug 173130 [21:35] Launchpad bug 173130 in ubuntu "edubuntu hardy 64bit live cd issues" [Medium,Incomplete] https://launchpad.net/bugs/173130 [21:35] Mithrandir: indeed, should work :) [21:42] heno, well, we had some issues with the console in 32bit ... probably thats realated and just exposes differently [21:44] ogra: ok.ħave you had reports of others running 64 bit edubuntu live daily? (that's all we need for alpha 1 really) [21:44] not really, you and davmore were the only testers i'm awareof ... [21:45] but testing alpha1 isnt that serious for edubuntu since we'll change the whole CD architecture anyway soon [21:45] -server is going away ... [21:45] the desktop will move over to addon [21:45] and ltsp goes into ubuntu [21:46] Mithrandir: do you know who could know why LP decided that a package needs to build a second time? [21:46] geser: You mean it successfully built, then got rebuilt again in the next release, and failed to upload? [21:47] heno, its nice to know it oughly works and to know the drawbacks but dont put extra effort into it ... the world will change :) [21:47] *roughly [21:48] Fujitsu: see https://edge.launchpad.net/ubuntu/hardy/+source/wmshutdown and https://edge.launchpad.net/ubuntu/hardy/amd64/wmshutdown [21:48] Fujitsu: but it's similar to what you describe [21:48] https://edge.launchpad.net/ubuntu/+source/wmshutdown/0.2-1 [21:48] It is what I describe. [21:49] Oh, you're right, the upload succeeded. That *is* strange. [21:49] But it seems to be the same underlying problem. Soyuz had a bug a while back that broke its detection of builds that needed to be retried in the new release. === heno_ is now known as heno [21:56] Mithrandir: it's used because developers are idiots^W not aware of porting issues. [21:57] pas... sorry. old discussion [21:58] I imagine everyone grokked that context ;) [22:01] I wonder... what happens when the sparc buildds try to build the package that [i386 amd64 all] only? [22:02] Shouldn't P-a-s stop them? [22:02] hehe [22:02] soren: if PaS says i386 amd64, then it never tries [22:03] if it doesn't exclude sparc, then it dies at dpkg-gencontrol(?) time [22:03] or dpkg-builddeb. I forget which [22:03] lamont: Right now it doesn't say anything. It's a brand new package and there are bulid records for sparc, hppa, ia64, etc. [22:03] what package? [22:03] open-vm-tools [22:03] and the expectation is that there won't be support for those other architectures anytime in the forseeable future? [22:04] Never ever. [22:04] so just i386 amd64? [22:05] It's only supposed to build on i386 and amd64, yes. [22:05] %open-vm-tools: i386 amd64 # [ANAIS] [22:05] ..and I (naïvely) thought that was all that would be tried now that I specified the architectures in the control file. [22:06] nah. it tries, fails, and then the porters decide to either port it, or have elmo/lamont/infinity put it in PaS :-) [22:06] PaS is all about porters [22:08] and committed [22:08] as for when LP will see the new version? no clue [22:09] I thought there was some bit that would cause sbuild to recognize a package was ANAIS based on debian/control, but I don't remember now [22:09] #15 * * * * cd /srv/launchpad.net/builddmaster/srcdep/ && cvs -q up [22:09] slangasek: I know it will bail out if all packages are arch: all and it isn't. [22:09] that cron job is commented out on drescher because it's been moved to cesium, but I imagine it's the same frequency there [22:10] slangasek: mebby. not in ours, I didn't think [22:11] lamont: I do remember having seen FTBFS logs where the package failed out with an arch error /before/ building the package [22:11] yeah... [22:11] Fujitsu: if all packages are arch: all, it should never be given to the buildd at all? [22:12] slangasek: Perhaps in Debian. [22:12] nope [22:12] slangasek: how'd you get to the binary then? [22:12] in Debian, if the source only generates arch: all binaries, it never makes it to wanna-build [22:12] We don't do binary uploads here. arch: all is built on the i386 buildds. [22:13] soundconverter_0.9.7-1.dsc: i386 not in arch list: all -- skipping [22:13] (which causes hassle at least for one package: the bios of quemu for ppc) [22:13] That's from sbuild. [22:13] sistpoty: openhackware? [22:13] slangasek: which makes life interesting for the arch-all-producing source packages that only build on, say ppc. [22:13] Fujitsu: hm? [22:14] what sistpoty said [22:14] IIRC openhackware only builds on PPC, but is arch: all, which causes problems. [22:14] Fujitsu: but LP doesn't give the package to /all/ the buildds for building a single arch: all set, does it? [22:14] slangasek: No, just i386. [22:14] Fujitsu: ah... so there's more than one [22:14] slangasek: uh... we call that a "feature" [22:14] Fujitsu: right, so still not the case I was describing [22:14] sistpoty: Is there another qemu PPC BIOS? [22:15] hrm... [22:15] actually, not sure what LP does. [22:15] Fujitsu: I thought there was a dedicated qemu-something package... but I'm not too up to date on this *g* [22:15] our WB impl gave it to everyone, and then removed it everywhere once i386 built and uploaded it [22:16] LP gives arch: all sources only to i386. [22:16] which is much better. [22:17] Except that it introduces a bottleneck, even with the extra buildd. [22:17] Isn't that because traditionally the i386 buildd is the fastest? [22:18] it's because if you're going to assume an arch that can build arch: all for you, i386 is the one to pick [22:19] the solution is obvious. Find an all machine to use as the buildd [22:19] slangasek: Isn't it a bug if it doesn't build everywhere? [22:19] jdong: unfortunately my all machine suffers from a missing libjustdoit ;) [22:19] jdong: I want one. :-) [22:19] ion_: me too :) [22:20] Fujitsu: powerpc bioses are not the only things that will require an arch-specific toolchain to build arch: all data [22:22] Fujitsu: and while in some cases this will simply be a bug in the toolchain on the other archs, those bugs tend to take longer to get fixed on archs other than i386 (even amd64, because the free software community has 20 years of experience writing code that don't run on 64-bit) [22:23] hehe [22:23] slangasek: True. [22:25] who is receiving email sent to archive@ubuntu.com ? [22:29] there was a very brief time during warty (oxford, to be precise), where ppc was building arch all. about 1 week, iirc. [22:29] i386 was, at that time, undergoing a bit of an archive event. [22:29] lamont: sweeet [22:30] lifeless: the really bad part about that was that elmo and I had the trash-one-arch-and-rebuild down to a science. [22:30] I wonder... did libsgutils1-dev go into universe or something? [22:30] Haha. [22:30] https://edge.launchpad.net/ubuntu/+source/libgpod/0.6.0-3/+build/458615 [22:31] Fujitsu: we only did it for i386 once. [22:31] or was it twice? [22:31] twice [22:31] lamont: It got promoted in hardy, by the look of things. [22:31] lamont: How did you manage it? [22:31] Fujitsu: (1) make sure you haven't released yet. [22:32] (2) do all the right stuff to rebootstrap yourself using what you have to build a brand new set of stuff. [22:32] kinda like the -autotest builds we still do, only keeping the result, instead of throwing it away. [22:32] Oh, I know that, but how did you get into a situation where you needed to? [22:32] first one was a little compiler option-forcing "oops." [22:33] the second one was compliments of Via. [22:33] Hah. [22:33] well, gcc and Via [22:33] change gcc to not emit the instr that doesn't work on those via chipsets, and then rebuild the archive. [22:33] Via chips didn't like your binaries? [22:33] An. [22:33] Fun. [22:34] s/An/Ah/ [22:34] or maybe it was "fix the gcc bug that happened to work everywhere but those Via chips." the details are hazy. [22:35] and then 4.10 shipped, and we aren't allowed to have archive events any more because of rule (1) above. [22:35] * lamont -> kids [22:48] any archive admins up? I'd like to petition the removal of x264 from multiverse/hardy [22:48] I think it was a mistake, x264 is distributable in Multiverse [22:48] jdong: Petition *against*, you mean? [22:48] Fujitsu: meh minor wording details ;-) [22:48] it's been 26hrs since my last nap [22:58] quick question that i figured a developer would know- do packages in universe have to meet the same licensing requirements as those in main, or can universe packages have restrictive licenses? http://www.ubuntu.com/community/ubuntustory/components leaves me a little unclear [22:59] seekay: Multiverse packages can have somewhat more restrictive licenses, but those in universe follow identical guidelines to main. [22:59] ok that's what i figured multiverse was for- just wanted to make sure [23:00] this text in particular made we wonder... "In universe you can find almost every piece of open source software, and software available under a variety of less open licences, all built automatically from a variety of public sources." [23:00] Where did you find that? [23:00] http://www.ubuntu.com/community/ubuntustory/components [23:01] perhaps that was written during a time when universe allowed less open licenses [23:02] seekay: There has been no such time. [23:02] then perhaps the text was written incorrectly the first time :) [23:03] that's more likely, yes :) [23:03] There is some other misleading text on that page, mainly about the benefits of main. [23:06] thanks for the info Fujitsu, just wanted to make sure universe packages had the same license requirements as main, but it sounds like someone needs to give that page a second look === doko_ is now known as doko [23:40] * emgent heya === Skiessl is now known as Skiessi