[00:02] ScottK, How soon should I have it finished? ie. Are you looking for someone to do it right now? [00:03] Sooner the better. [00:03] * ScottK just noticed the security advisory for Debian being released. [00:05] link? [00:05] somerville32: debian-security-announce@lists.debian.org [00:05] Er, http://www.debian.org/security/ [00:09] somerville32: http://www.debian.org/security/2008/dsa-1464 [00:09] Kmos, 404 [00:10] somerville32: not yet there.. [00:10] only 1463 [00:11] somerville32: http://lists.debian.org/debian-security-announce/debian-security-announce-2008/msg00022.html [00:11] Yea, I already found that, thanks [00:11] :) [00:18] is urgency always low? [00:18] Like, whats the point of it? :P [00:18] <\sh> ScottK, claws-mail 3.2.0 works like a charm... [00:23] <\sh> somerville32, debian needs it...security updates e.g. have a higher priority and the buildds and archives are working differently...+ [00:24] <\sh> somerville32, #debian-security@oftc (or white when he's here) can elaborate on it... [00:26] Ok, test building for Gutsy [00:30] Should I file a bug for each release or just one big bug? [00:39] keescook, ping :P [00:52] ScottK, ^^ [00:57] \sh, ping [00:58] <\sh> emgent, hi...saw your updates :) checking later today...:) [00:58] cool ;) [00:58] \sh, thanks^2 [00:58] <\sh> I'm sitting here with a buddy and trying to flash a freecom fsg-3 appliace with linux :) [00:58] heehhe ok ;) [01:02] How does one compile a package with debug symbols? With "CFLAGS = -g -O0" ? === santiago-php is now known as santiago-ve [01:06] DEB_BUILD_OPTIONS=nostrip,debug [01:07] crimsun: thanks [01:14] somerville32: One big bug [01:14] bug #183389 [01:14] Launchpad bug 183389 in syslog-ng "[SECURITY] CVE-2007-6437 prone to denial of service attack" [Undecided,New] https://launchpad.net/bugs/183389 [01:15] I had to reconfigure my gutsy pbuilder so I'm just doing the test build now [01:15] I have a debdiff for Gutsy [01:15] somerville32: Tasks for Gutsy/Feisty approved. Does this not apply to the Dapper/Edgy versions? [01:16] The CVE does not say it affects 1.x [01:16] But I intend to look [01:16] OK. [01:24] ScottK, Feisty doesn't appear approved [01:25] Is now. [01:33] ScottK, Should I have the bug # in changelog? Is Launchpad smart enough to close the correct task? [01:34] somerville32: Yes it is and you should. [01:35] ScottK, Should I update the maintainer field if there were no previous ubuntu changes? [01:35] somerville32: In Feisty/Gutsy yes. For earlier releases no. [01:41] Argh. advi still hasn't lost its libungif dependency. [01:42] Lets see if it was rebuilt too soon... [01:42] ScottK, According to Debian's security announcement, 1.x isn't affected [01:42] "The old stable distribution (sarge) is not affected." [01:42] ScottK, However, the version in dapper and edgy is newer than Sarge [01:43] Heya gang [01:44] Hiya bddebian [01:46] Hi somerville32 [01:48] ScottK, Dapper is not vulnerable it appears [01:49] somerville32: OK. Nominate it for Edgy and I'll approve then. [01:49] ScottK, already did [01:50] * ScottK hits refresh (again) [01:50] somerville32: Approved [01:52] For a no-changes rebuild you just add (or bump) the build number, right? [01:53] Aye [01:54] Cool. I'll just fix advi then. It seems the last rebuild was too soon to actually do the libungif -> libgif transition. A rebuild in my sbuilder fixes it. [01:58] \sh: you're kidding - he's doing that again is he? [01:58] after he promised about a month ago, that he wouldn't? [01:59] Hobbsee: what? [01:59] Hobbsee: And we all know what that's worth. [01:59] LaserJock: Kmos messing with MOTU's sync request bugs after having been told specifically to leave them alone. [01:59] that's what I thought [01:59] he did 3 of mine last night [02:00] ScottK: it says that daniel is wrong, and that kmos is incapable of contributing. [02:01] Agreed. [02:01] ScottK: then again, it's already been proven that he remembers things for under an hour. [02:01] or can't reapply concepts, using actual thought. either way. [02:01] what surprises me is how he's actually reasonably competent with picking up stuff that didn't build. [02:01] although I wouldn't say it's that annoying to have him set wishlist I does create uneeded bugmail [02:01] as in, launchpad bugs. [02:02] LaserJock: i've had him revert my changes to bugs. that's not annoying - that's incorrect [02:02] What was his rationale for disengaging from the agreement? [02:03] somerville32: didn't follow the listed process, apaparetnly - (confirmed vs triaged) [02:03] oh goody, he's even got a written warning in here. and still he chooses to ignore it. [02:03] LaserJock: BTW, someone else pointed me to the Golden Pony. Thanks! :-) Too funny. [02:03] bddebian: ;-) [02:04] But what was his rationale for disengaging from the agreement he made? Why did he decide to do something that he promised not to? [02:04] somerville32: I don't think he would agree that he's done anything wrong [02:04] somerville32: i'm not sure. he never says. he still appears to remember, just not apply the concepts that he's been taught. [02:05] LaserJock: This is just the way it goes with him. [02:05] <\sh> Hobbsee, when you mean that he confirmed and wishlisted all my sync bugs, so yes...and I never ever confirmed, wishlisted etc. my sync req in the past [02:05] Has anyone said to him: Here, do you see here where you said in writing that you wouldn't do this. See over here? This is where you just did it. Why did you do that? [02:05] somerville32: i suspect he's really eager, and wants to do good, without actually caring if his work is right or not. He's also said that he doesn't feel he should get everything right, as he's not trying to go for MOTU [02:05] somerville32: good luck. try when he comes on [02:05] The key thing is he's been told not to do stuff without checking during this evaluation period and he's routinely flouted that instruction. [02:06] somerville32: Again, and again, and again. [02:06] \sh: er, you need to confirm bugs - -archive ignores unconfirmed bugs, due to people's incompetence with requestsync. [02:06] What does he say? "Oops, I forgot?" [02:06] somerville32: he doesn't answer. [02:06] usually [02:06] somerville32: as soon as he hits a hard question, he doesn't answer. [02:06] somerville32: That or, "I thought because X (usually irrelevant fact) it was different. [02:06] oh yes, i had my own ammo to add to that page, too [02:06] Then when you corner him, it's as Hobbsee says [02:07] yes, I was about to point out that confirmed+wishlist is the right state (AIUI) if you want -archive to act on it... [02:07] <\sh> Hobbsee, ok...I dojn't use requestsync tool...and that I have to confirm those bugs, i didn't even know, because the other stuff was synced automatically by archive admin..but nevertheless...I think i have to confirm them but not others who didn't test the sync at all [02:07] <\sh> brb [02:08] So kmos went around and changed sync bugs from confirmed to triaged? [02:08] slangasek: wishlist. bah humbug [02:08] somerville32: triaged back to confirmed. [02:08] Hobbsee: well - ok, I don't look at the severity either :-) [02:08] he just set mine as wishlist [02:08] * bddebian files some "wishlist" bugs [02:08] (but in principle, sync requests are wishlist bugs) [02:09] slangasek: oh of course. [02:09] slangasek: The point for Kmos is he's been specifically told on multiple occasions to leave sync bugs alone and he fails to list. [02:09] So, he didn't actually do anything wrong technically besides violate his agreement? [02:09] list/listen [02:09] somerville32: The wishlist stuff is not wrong, but un-needed. [02:09] He did ask for a sync that FTBFS and he never tested. [02:10] ScottK: I understand; I haven't even looked at the details of the current "parole" so I won't comment on that further [02:10] slangasek: No problem. [02:10] * bddebian dances around the room [02:10] Has anyone checked to see if kmos had asked? [02:10] * ScottK hides his eye [02:10] eye/eyes [02:11] * Hobbsee adds ammo [02:11] somerville32: He's several dozen times past his last chance from me. He defended his untested FTBFS sync today by saying he'd discussed it with the Debian maintainer who told him it would be fine. [02:11] somerville32: there's plenty of stuff he's done outright wrong. see https://wiki.ubuntu.com/MOTU/Council/KmosReport [02:12] * bddebian checks to see if he has a page [02:12] BTW, I haven't been looking for it myself (trying to avoid it actually), but he's unavoidable. [02:12] evening [02:12] ScottK: this comes from someone who believes that all ubuntu changes can automatically be dropped, due to it having a new debian maintainer. [02:12] bddebian: Debian has your page ;-) [02:12] ScottK: Haha, not shit eh? [02:12] Hobbsee: Of course, I forget. [02:13] s/not/no/ [02:13] <\sh> back [02:13] in any case, there's not much point discussing kmos as it's just a general downer [02:13] wow. [02:13] somerville32: asked what? [02:13] LaserJock: yes. now, where's that report? [02:13] Hobbsee, for approval regarding the wishlist thing. [02:13] somerville32: no, of course not. [02:14] <\sh> slangasek, well, it's ok when we need to confirm the sync reqs, but please if I testbuild and make functional test with the stuff, noone else should work on the files bugs [02:14] that would require actually obeying his agreement. [02:14] slangasek: You may care to know that bugging developers and archive admins to do stuff (like he was doing to you yesterday) is also on his forbidden list. [02:15] <\sh> script fixed...no need that others have to confirm ;) [02:16] Hobbsee: 3 -'s is a bit much === DelayLama is now known as DreamThief [02:16] ScottK: hum, he's not been doing a very good job of adhering to that then :/ [02:16] slangasek: No. === kitterma is now known as ScottK2 [02:16] LaserJock: was meaning one from each of us. [02:16] LaserJock: seeing as he's done it to all of us [02:17] yes, but they are minor infractions [02:17] LaserJock: He's the death of a thousand cuts. [02:17] sure, and that should show on the page [02:17] lemme guess Kmos right? [02:17] but it doesn't help to get overly upset about minor things [02:18] they should be noted but it's nothing to cry over [02:18] zul: correct, everyone's favourite topic [02:18] zul: yeah. who else is that incompetent? [02:18] ajmitch: heh [02:18] oh yay, there's actually a correct bug in here [02:18] Hobbsee: I am, that's why I left [02:18] ajmitch, you left because of kmos? [02:18] ajmitch: smart move [02:19] somerville32: no, just general atmosphere around here [02:19] :-( [02:19] That stinks [02:19] I agree [02:19] Probably because of me :-) [02:20] I don't find it _that_ bad but I have a different perspective [02:20] Yes, #d-devel is sooo much nicer :-) [02:20] But I wonder how I can help improve the atmosphere personally :) [02:20] ajmitch: it should be better when people who don't follow what they agree to actually leave. [02:20] #debian-devel is mostly ari and helix spouting crap [02:20] play WoW [02:20] at least, that's what i'm hoping [02:21] StevenK: Yep [02:21] * ScottK2 finds there is a direct correlation between Kmos presence/activity and the level of negativity on the channel. [02:21] * somerville32 tosses lavender about the room. [02:21] * ScottK2 sneezes. [02:21] * bddebian sneezes [02:21] lol [02:21] hah, jinx [02:21] Hobbsee: right, which is why when I felt like I'd be lynched for the slightest mistakes, I decided that I didn't really have much more to contribute [02:21] bddebian: You are old and slow. [02:21] I know :( [02:21] ajmitch: you have to make a fair few before you even start getting lynched. [02:21] ajmitch: most people don't actually check all of a person's changes - only if they notice them to be wrong. [02:22] ScottK2, btw, I've finished the security stuff :) [02:22] I make shitloads of mistakes and I haven't gotten lynched yet. At least not that I know of.. [02:23] ScottK2: I can't honestly say I've seen that correlation, but I admit I'm not as active as I used to be [02:23] In fact I made the mistake of trying to help Debian... [02:23] * bddebian hides [02:23] :D [02:23] it's actually kinda more unsettling to see people go after somebody else, even if it's warranted [02:23] somerville32: Great. Ping keescook Bug #183389 [02:23] Launchpad bug 183389 in syslog-ng "[SECURITY] CVE-2007-6437 prone to denial of service attack" [High,Confirmed] https://launchpad.net/bugs/183389 [02:23] uhm [02:23] we should be able to deal with things in a better way, IMHO [02:24] LaserJock, I agree. [02:24] I think we should all be happy :) [02:24] LaserJock: yeah. kickboxing. [02:24] I'm sure that whoever gets onto the MC in the next few weeks will make everything wonderful [02:24] In general, I agree, but Kmos has done too much damage. [02:24] * \sh 's happy... [02:24] \sh: :-) [02:24] \sh, lol [02:25] * bddebian is frustrated [02:25] <\sh> I just managed to break a freecom fsg-3 NAS [02:25] bddebian: woot! REVU masta! [02:26] * Hobbsee notes that no other business, school, workplace, or community would probably do this. [02:27] <\sh> ajmitch, well, I don't think it has a lot to do with the MC...it's more: well, if I confirm a bug, I actually tested it and agree that it's needed to be fixed...but if I didn't test what's in the bugreport I shouldn't put my hands on it [02:28] absolutely; if a bug isn't confirmed and you think it should be but haven't verified yourself, the most you ought to do is check with the submitter about whether it was overlooked... [02:29] <\sh> anyways, it's not nice to talk about a person when this person is not there, so we should end here and follow up this discussion during the day when kmos is awake... [02:30] \sh: But then Hobbsee isn't awake to participate :P [02:31] <\sh> somerville32, right, damn timezones :( [02:31] \sh: [02:31] * somerville32 proposes we all get together for brunch sometime to discuss the issue. Say, my house? [02:31] \sh: not really so much point discussing with him here anyway. he suddenly always happens to have lunch on, or have to go out, when anything complicated comes up. [02:31] it's very, very convenient. [02:32] JFTR, I've not said anything about him here that I haven't already said to him directly. [02:32] This is publicly logged channel anyhow. I wonder if he reads what we say about him when he is sleeping. [02:32] * somerville32 ponders. [02:33] <\sh> ok ladies and gentlemen it's 3:32 UTC+1 here and I need to sleep a bit before my wife will wake up :) [02:33] <\sh> so cu in a couple of hours... [02:33] * somerville32 waves. [02:33] Gnight \sh [02:33] <\sh> ScottK, claws-mail works and sync req filed (confirmed it actually;)) [02:34] ScottK, Do you have any suggestions on how we can improve the atmosphere for that we don't lose more contributors like we did with ajmitch? [02:34] s/for/so [02:34] somerville32: he doesn't. [02:34] somerville32: therefore he can claim ignorance. [02:34] somerville32: Yeah. Our leadership shouldn't bury their heads in the sand while someone causes hate and discontent in the hope that it will just go away. [02:35] somerville32: which means he doesn't have to respond to mail, as he repeatedly claims he hasn't seen it / is too busy / etc [02:35] somerville32: when suggested that he does less, and spends more time to get higher quality on the stuff that he is doing, he doesn't really respond, or gives a non-answer. [02:36] <\sh> somerville32, /ignore and filtering mail is helping sometimes...and if this doesn't help, a tie-up could also do some favours, when he is doing the merges/syncs/fixes all by himself ;) [02:36] lol [02:37] <\sh> but this will a bad day for universe [02:37] <\sh> anyways..now good night :) === \sh is now known as \sh_away [02:37] Maybe we could have some community events or something to help build team skills and a tight sense of friendship :) [02:37] During Gutsy there were times when 3 or 4 MOTUs were working full time to nullify the bad stuff he was throwing at the sponsorship queue. [02:37] <_MMA_> ScottK: I also feel from watching hear, reading ML posts and talking to a good many people in Boston that there is no real system setup to manage situations like this. [02:38] <_MMA_> *here [02:38] _MMA_: Agreed. This is the first time it's come up. [02:38] well, there's really nothing we *can* do in this case [02:38] it's a bit messy [02:38] ugh, we're STILL talking about Kmos ? [02:38] and frankly I just think we should move on, let the MC deal with it and get something done [02:38] '/me goes to do something else for a while [02:38] \sh_away: you can't actually ignore and filter mail when he's actively breaking the archive. [02:39] * ScottK is actually working on stuff while doing this, FYI. [02:39] somerville32: i suspect the problem would go away if the team felt that those who decided not to obey rules actually got something odne about them. [02:40] <_MMA_> LaserJock: I think the "leadership" ScottK mentions is the MC and nothings being done. [02:40] _MMA_: umm ... a lot is being done [02:40] but as far as having something to really take action with we're basically powerless [02:40] _MMA_: Been done. [02:40] There's at least hope. now. [02:41] <_MMA_> Ok. "Our leadership shouldn't bury their heads in the sand" Doesnt really elude to something being done. :P [02:42] <_MMA_> But sure, I trust ya. :) [02:42] and I don't think that's a very accurate assessment either [02:42] LaserJock: Nothing got done for months. [02:43] ScottK: and that proves "leadership burying their heads in the sand"? I think not [02:43] LaserJock: All I got were requests to give him time long after it was clear it was hopeless. [02:43] LaserJock: Nothing would be being done now if I hadn't quit listening to those requests. [02:44] but you didn't email ubuntu-motu or motu-council for a long time [02:44] it was basically dealt with by informal means [02:44] which does not say anything for leadership burying their heads [02:44] Well there were no formal means and via informal means I was asked not to do anything. [02:44] I was told it would all be fine. [02:45] formal means would be emailing -motu [02:45] at the very least [02:45] LaserJock: no, it was a "we know we need to do something about this, we just don't know what to do yet, or don't want to" [02:45] and being unhappy with the results of 1 or 2 people emailed in private is a fair cry from leadership burying their heads [02:46] LaserJock: What person in leadership here did anything about him except when forced? [02:46] Hobbsee: but that is not ignoring the issue. It's perfectly legitimate for the MC to say we don't know what to do yet. [02:46] ScottK: good question, I have *no* idea [02:46] I have no idea who you talked to [02:46] Ok, time to stop talking about kmos :) [02:47] LaserJock: until they keep saying that, and don't try to find a solution. [02:47] somerville32: Let's talk about you now ;-) [02:47] because you don't like their solution [02:47] Ok! [02:47] How do you guys think I'm doing MOTU-wise? :) [02:47] LaserJock: it was decided that he needed to go at UDS. it was another 2 full months later when it was also decided that he needed to go, but no progress was made as to how. i call that them burying their heads in the sand. [02:47] LaserJock: They had no solution until I filed a complaint and it couldn't be avoided any longer. [02:48] * ScottK isn't the MOTU warden and shouldn't have been forced to deal with it. [02:48] no [02:48] LaserJock: at least now we have a rather damning report, but because it's got some +'s on it, it'll probably be shown as "he does some good stuff. lets keep him, and keep monitoring" [02:48] but the point of MC is to not play police either [02:48] LaserJock: why not? [02:49] because that is not their charter [02:49] it is a mediation charter [02:49] If someone is being disruptive, who's responsible? [02:49] that means people have to bring complaints [02:49] if it's the only people that can kick people out for disobeying rules, then they automatically *have* to take up those reponsibilities. [02:49] ScottK: apparently, the MC is only there to mediate, and if that doesn't work, no one. [02:49] clearly the disruptive person is responsible [02:50] Hobbsee: the CC is above the MC [02:50] LaserJock: That's clear now. It's also clear that no matter what one has done, the MC will give additional chances, so there's no point in being slow to complain. [02:50] ScottK: that's ridiculous [02:50] establishing rule off of a *single* case is retarded [02:50] LaserJock: which shoves it back to the MC, yes. [02:50] LaserJock: It's the obvious consequence of the way this has been handled [02:50] LaserJock: we tried that one. [02:50] LaserJock: actually, it goes back to jono, who sends it back to daniel, who sends it to the MC. [02:51] ok, so we should work on things, I have no doubt [02:51] LaserJock: When this was brought up, I tried to say it was a bad situation to make rules based on and the MC decided different. [02:51] but the MC is not daniel [02:51] and the CC is not jono [02:52] and if you don't bring complaints to the team I don't think you can claim that they are ignoring the issue [02:52] I wasn't aware that there was a problem until shortly before ScottKs email [02:52] LaserJock: There's no way they couldn't have known. [02:53] and I don't think I'm *that* out of touch that there was a lot of things going on without my knowing [02:54] LaserJock: do you do the sponsorship queue? [02:54] we always have people who make mistakes, and we should be tolerant in as much as we can while still retaining standards [02:54] Hobbsee: yes, some, not as much anymore [02:56] LaserJock: it's just where people deliberately go and ignore the standards where the problem comes. [02:56] there is no doubt that something needs to be done, and that things could have been addressed better [02:56] but we've never had this problem [02:57] and I think it's better for us to have a positive attitude and look at what we can be doing to make Universe and Ubuntu a better place [02:58] Gah, I gotta stop doing this Debian shit. I just tried to build a revu package in my sid pbuilder.. :-( [02:58] LaserJock, +1 [02:59] * Hobbsee will have a more positive action once seeing that the MC is serious about getting rid of people who deliberately do not follow the proceedures required of them, of which they've agreed to. [03:00] I really don't understand your statement [03:00] 1) the MC has never shown itself to *not* take issues seriously [03:00] 2) we can't kick people out of something they aren't in [03:01] Kmos has no official standing so we can't really do anything but say we'd rather he not contribute [03:01] LaserJock: We can ask Launchpad to remove his ability to do Ubuntu related things. [03:01] Can we take kmos discussion to another channel, maybe #ubuntu-kmos_is_the_end_of_the_world or something? :) [03:01] we could do that, although I'm not sure the MC can do that [03:02] ScottK: they say "let the CC handle that. sounds like a person issue" [03:02] LaserJock: locking of LP account on ubuntu stuff would be enough [03:02] i think [03:03] Hobbsee: Exactly. I've been told (by Jono) that MC has authority for that. [03:03] somerville32: lol, have you registered it? I wanna be an op with supercowpowers! [03:03] :D [03:04] Hobbsee: and that is basically what the MC has said they are willing to do [03:04] LaserJock: This is why I've said I'm hopeful action will finally be taken. [03:04] LaserJock: they are now, yes. [03:04] so it would seem to me that the MC has acted fairly well with the issue [03:05] the issue was raised by ScottK [03:05] there was discussion [03:05] a reasonable plan set forth [03:05] and a nice wiki page to track things [03:05] seems like quite a bit of action to me [03:05] * ScottK thought the entire trial period was a waste of time and effort. [03:06] LaserJock: I count it action once he's stopped from making trouble. The rest is just paperwork. [03:06] that seems odd to me [03:07] it should be all about transparency [03:07] LaserJock: just the amount of time that it took to actually get the action done. [03:07] about having reasons and evidence for people [03:07] Hobbsee: and that wasn't the fault of the MC [03:07] no? [03:07] no [03:07] I think there's been plenty of transparency about all the trouble he's caused. I think it was sufficient, without further warning, for him to be booted. [03:07] show me where the MC was asked about the issue and they did nothing? [03:07] the MC had to decide what it wanted to do about the issue - how is it not then their fault for taking so long to come up with a plan? [03:07] When does the current probation period end? [03:08] somerville32: "mid jan" [03:08] somerville32: apparently they're doing a report at the moment. [03:08] We're mid January now [03:08] So the pain should be over soon [03:08] somerville32: however, it was worded as "give it 2 weeks, and we'll see", rather than "give ti 2 weeks, and we'll make a decision" [03:08] Hobbsee: so long? it took 20 days to hear evidence and get the plan in place, that's not that long [03:10] Gawd, I so wanna do urbanterror but I can't deal with a 300Mb data package :( [03:11] LaserJock: the objection is surely between the "started creating trouble, and hearing complaints" and "concocted a plan" [03:11] LaserJock: I was told it would be taken care of after UDS. Nothing happened, so I filed a complaint to force action. I realize that was all in private, so in some respects doesn't count, but I don't count the time from my request, I count it from UDS. [03:12] ScottK: right, but I do think it matters [03:12] Which gets back to my earlier comment I made about having learned the lesson that trying to work things out without formal complaints. [03:12] even UDS it was still 2 months before they collected evidence, afaik [03:12] well, this is an unusual case [03:12] which is how he got so far in the first place, yes === bigon is now known as bigon` [03:14] ScottK: well, I was told at UDS Mt. View that we'd have a new TB in 2 weeks and we still don't [03:14] we have a lot of issues [03:14] LaserJock: that's a process failure too, surely. [03:14] of course it is [03:14] but I think MC-bashing is not helpful [03:15] especially when there was nothing sent to the list [03:15] I agree it's frustrating [03:15] * somerville32 wonders how this discussion is helpful at all. It is just rehashing the numerous bitch sessions we've already had. [03:15] Yeah, let's bitch about something else :-) [03:15] somerville32: hmm, I always find it helpful :-) [03:16] LaserJock, I agree it is nice to get it out but it is taxing on other people who don't care. [03:16] yes, that's true [03:16] and if I were a new contributor and I just walked into the channel I would bolt [03:16] I'm personally disappointed too with the number of process failures but I'm wondering what we can do to actually accomplish something. Unfortunately, the powers that be don't read these backlogs to see whats up. [03:17] * pochu doesn't care about the conversation whatever you talk about :P [03:17] somerville32: what powers that be? *we* are the powers that be [03:17] * bddebian has no power [03:17] bddebian: of course you do, more than most and don't forget it [03:18] ;p [03:18] heh [03:18] LaserJock: we have no power in view of the CC. we only have power to whinge when it breaks. [03:18] look, the level of power we have within our OS is fairly unprecedented for a project this size [03:18] tru dat [03:19] we are a volunteer army of people who care about Ubuntu [03:19] Speak for yourself ;-P [03:19] we give software to thousands and thousands of people [03:19] I just care about my karma ;-P [03:19] and we can determine our future through our actions, our votes, and our words [03:19] bddebian: lol, I want karma for uploads :) [03:20] LaserJock, +1 [03:20] I probably have no karma anymore :( [03:20] bddebian: you have negative karma! [03:20] Heh, probably [03:20] Karma is still messed up [03:20] Create a spec or two and you get about a thousand karma [03:21] Do a hards day work on bugs and you'll get maybe a hundred or two [03:21] somerville32: yes, Marks says that will make people work on underutilized parts of LP [03:21] *Mark [03:21] Plus, most of my karma from last year is gone [03:21] So I have 2k from non-ubuntu stuff now since I did a binge on my Sapidlib proejct [03:22] Although overall, I've contributed way more to Ubuntu [03:22] Karma has 365 days of life, and every day it loses 1/365 of its value. [03:22] err, 1/365 of the total value [03:22] lol@guesses [03:22] * somerville32 tickles lifeless [03:24] Well, I'm hungry. How about we order in? [03:24] * pochu is off to bed, g'night [03:29] so whats the naming policy for packages when you skip over an existing debian release...for example, gutsy release at 0.1-1ubuntu1, debian release at 0.2-1 and latest upstream is 0.3, what would my hardy version # be if I want to package 0.3?? [03:30] 0.3-0ubuntu1 [03:30] aye [03:30] thx [03:33] what time will the motu school kick off? [03:33] jdong: has upstream complained about you being able to distribute 0.3 of supertux? [03:34] oh, they've changed what htey want again. nvm. [03:38] Hobbsee: no, but their wiki "suggests" vendors to ship the 0.3.x stuff in a separate package, which we already do. [03:39] and I really wish they'd stop changing distribution sites so that watchfiles would actually serve a purpose :D [03:39] Heya jdong [03:39] jdong: hehe :) [03:39] jdong: last time i got in trouble for packaging the new version, because they'd retroactively put up a "please don't put this into distros" sign. [03:39] Hobbsee: haha [03:39] i kid you not :) [03:40] Hobbsee: well we've got supertux-stable now, and I just modified the two so that they can be installed side-by-side today, so I don't see any way upstream can yell at us :) [03:40] (famous last words) [03:40] :) [03:40] yeah well [03:40] * ScottK remembers. [03:40] Hobbsee: mostly because my little sister wanted to play new supertux today ;-) [03:40] haha :) [03:41] Hobbsee: she knows how to say debuild now... and she's 9. [03:41] scary. [03:41] taught her how to use prevu yet? [03:41] hahaha no :) [03:41] I should though [03:45] jdong, wow. supertux's data is huge! [03:45] DarkMageZ: the thing is 5x bigger since 0.3.x [03:45] DarkMageZ: seems to mostly be ogg soundtracks though [03:46] yeah. i'm grabbing 0.3.1d from hardy [03:46] yay, more music ? [03:46] yeah, the tracks are slightly less repetitive at a first glance. === _czessi is now known as Czessi [04:35] Any MOTUs here willing to take a look at http://revu.tauware.de/details.py?package=photoml? This package has been through a number of revisions - technical issues were taken care of a while back, and the recent upload addresses some copyright statement problems. === bigon` is now known as bigon === boomer` is now known as boomer [05:11] * ScottK is reviewing qtsmbstatus [05:14] that sounds... vertical. [05:16] What does horizontal sound like? :-) [05:18] slangasek: Looks like it might be a new way to tickle even more bugs out of samba. [05:19] By exposing QT bugs [05:19] * StevenK runs [05:19] heh [05:19] "tickle bugs out of samba"? [05:19] Sure. It's qt4. [05:19] It's late and I'm tired. [05:20] Try to get samba do to something in a different way so different 'interesting' things happen. [05:20] bddebian: horizontal sounds like a kernel [05:20] ScottK: I think we'll call those qtsmbstatus bugs :) [05:20] ah :) [05:21] Just like all the samba won't work with my Windows $whatever bugs are samba bugs. [05:22] yes, those are samba bugs [05:22] Fair enough then. [05:24] Uploaded. [05:26] * ScottK will review qextserialport now. [05:30] That didn't take long to find a problem with. [05:31] :] [05:31] bddebian: You may want to take your advocacy off that one. I think the watch file really ought to be fixed. [05:37] ScottK: Which one? [05:38] * bddebian thinks the watch file push is BS anyway :) [05:42] bddebian: qextserialport [05:43] Maybe, but for sourceforge, if you are going to have one, you really want it to be in the right format. [05:43] bddebian: I've done updates of a number of perl modules recently and I found them really handy. [05:45] OK, removed [05:47] bddebian: But you're batting .500. I advocated one of the two I looked at. [05:49] Hah, my average is picking up ;-) [05:49] Ugh, this headache is killing me, I'm going to bed. Gnight folks [05:49] Night bddebian [05:50] * somerville32 decides to upgrade qtorrent package. [05:53] http://thegraveyard.org/qtorrent.php shows the current version as 0.9.5 but version 2.9.1 is in the archive and uscan says the latest is 2.9.3 [05:54] The website must be out of date [06:00] hi warp10 [06:00] Hi somerville32 :) [06:11] ScottK, since you're a member of the -server team, can you ACK bug #182445 please? [06:11] Launchpad bug 182445 in ebox-network "Sponsor ebox-network_0.9.3-0ubuntu4" [Wishlist,Confirmed] https://launchpad.net/bugs/182445 [06:11] Oh wait [06:11] Soren already replied :) [06:11] somerville32: I'd suggest you talk to StevenK about that one. [06:12] ScottK, soren (who is the package maintainer) already said it was okay. Should I still seek StevenK's approval for such trivial changes? [06:12] No. [06:13] I was thinking of a different package. [06:13] * ScottK is going to be soon anyway, so no time to review. [06:13] be/bed [06:13] okay [06:14] Luca already reviewed it, so you could get away with just uploading it ;] [06:15] pull trigger first, think about boom later :) [06:16] indeed [06:20] makes things go faster, but you need a grab bag of rationales :) [06:20] I keep mine in tomboy notes ;-) [06:21] * jdong pulls into hat and finds "It's an early alpha stage" [06:24] :) [06:28] somerville32: Luca should have uploaded it then. Anything I'm going to upload, I'll review. [06:30] * ScottK sends mail and crawls over in a corner to hide. [06:30] Good night all. [06:30] groovy === Allan_ is now known as Hit3k_ === Hit3k_ is now known as Hit3k [06:44] ScottK: yay! [07:06] Heh, that's a fun kwin4 bug. Disable vsync, and it won't display anything but the mouse pointer :). [07:06] Bwaha [07:07] Furthermore... that's now set. This might be the end of my kde4 exploration. [07:07] is feature, not bug :) [07:07] Hobbsee: Graphical effects so advanced, mere human eyes are unable to percieve them? [07:07] Right! [07:08] Look at those windows not exist in 3 dimensional space! [07:09] I was wondering why it seemed to be running so jerkilly, thought it was probably a symptom of nvidia's wonderful "the reported refresh rate should not bear any resemblence to the actual refresh rate" theorem. [07:10] * RAOF is sure there was some terribly good reason for the kde team to write their own opengl compositing manager. [07:11] RAOF: See "[18:08] < Hobbsee> is feature, not bug :)" for the Nvidia refresh rate. [07:11] Yeah, that's what *they* say. [07:12] heh [07:12] Including the time and IRC nick? Neat. [07:14] How do I set per-package build variables in CDBS? [07:14] [18:08] < Hobbsee> is definitely my favourite feature of all time [07:14] I'm building two packages that differ only in the value of one automagic CDBS variable [07:14] * two binary packages, from the same source one [07:16] This sounds like a recipie for frustration. [07:16] Indeed [07:17] Let's see if I can get kwin to give me an all-white screen under Xgl... [07:26] If there is a package in Debian and not in Ubuntu, do I just follow the normal sync procedure? [07:27] somerville32: except that you file the bug against ubuntu w/o a source package, yes [07:35] Yay! kwin survives Xgl. [07:39] g'morning! [07:41] Morning wallyweek [07:43] Hello. is there any problem with revu? [07:43] I can't login to revu, it seems the server has gone :( [07:43] anyone having the same problem? [07:44] yep, me too.. [07:46] oh, again? [07:47] in this very moment, I can't even access the main page :( [07:48] :( [07:49] * Hobbsee didn't take it down this time [07:50] * wallyweek feels victims of some kind of course and starts thinking sdlmame will mess hardy too :( [07:50] jejeje [07:51] * somerville32 has trouble parsing wallweek's last. [07:52] * wallyweek also thinks he's still fast asleep and types everything wrong :) [07:53] * wallyweek feels a victim of some kind of evil curse and starts thinking sdlmame will miss hardy too [07:53] Nah, you got an entire month [07:54] what's a month when you didn't do it in a whole year? :( [07:54] jeje [07:55] wallyweek, I'll help yo [07:55] how much time do we have? i thinked it was about two weeks [07:55] *you [07:56] somerville32: featurfreeze is 14th february IIRC [07:56] good morning [07:56] featurefreeze [07:56] morning dholbach [07:56] hiya dholbach [07:56] ahh.. ok.. [07:57] heya somerville32, wallyweek === bigon is now known as bigon` === bigon` is now known as bigon [08:11] pochu: Are you on amd64, by any chance? [08:12] good morning dholbach [08:13] hey geser [08:13] hey soren :) [08:13] dholbach: Good morning, Daniel :) [08:13] how's it going? [08:13] Rockin' as always :) === bigon is now known as bigon` [08:14] morning dh [08:14] bring it on! :) [08:14] morning highvoltage :) [08:15] oops, lag :) [08:18] Should I strip rpath from binaries depending on KDE4 libraries? === bigon` is now known as bigon [08:28] any news on revu? it seems still down :( [08:55] * siretart goes cycling sparky... [08:55] I think it is time to cycle _me_ :] [08:55] * somerville32 heads off to bed [08:56] * StevenK gets the cattle prod [08:59] sparky is booting [09:04] grr, hangs at "* Activating swapfile swap..." [09:04] siretart, Do you have another server to host revu? [09:05] oh wait, it just took ages. revu is back :) [09:05] :) [09:06] somerville32: well, the plan is to use spooky for that. however gutsy fails to boot from dm-raid from that machine :( [09:07] How much of a load is revu on a server? [09:09] siretart: well done! thank you! :) [09:12] somerville32: not much at all. an old sparc ultra 10 is able to handle the load without any problems [09:16] siretart: can't comment my own package: connection timeout :( [09:17] revu seems to be down again? [09:18] sommerville32: yeah, for me at least [09:20] * wallyweek is glad he pasted his comment on notepad before clicking on submit :D [09:20] no, its up again. I fiddled with the serial console and accidentally sent a 'break' over the serial line [09:21] somerville32: if you are considering helping out with (hosting) revu, please join #ubuntuwire [09:21] siretart: correct. comment addressed [09:27] If i use cdbs i have to use python in build-deb [09:27] ? [09:28] josesanch, using cdbs or debhelper wouldn't effect that [09:28] somerville32: you said you could help me... could you have a look at http://revu.tauware.de/details.py?package=sdlmame ? [09:28] somerville32: but a program that uses setup.py and is written in python has to build-dep in python. it isn't? [09:29] josesanch, yes [09:30] someville32: many thanks. [09:31] wallyweek, I will but I need to get some sleep for now :] [09:32] ok, thanks and g'night! :) === bigon is now known as bigon` === bigon` is now known as bigon [10:14] dholbach: happy birthday [10:14] geser: thanks a lot! :-) [10:25] c [10:25] argh [10:25] Hi TheMuso [10:25] Hey geser. [10:26] TheMuso> I was advised to ask you about Debian sponsorship [10:26] LucidFox: Debian sponsorship? [10:26] LucidFox: I am not a Debian developer. [10:26] * DaveMorris plugs his package for revu - http://revu.tauware.de/details.py?package=opensg It's an OSS distributed scenegraph API for doing some cool 3D graphics across multiple machines, I used it for a 46million pixel display. It's cool so please revu it. [10:29] ScottK: re: milter. If you can not depend, I like it just for less-stuff-in-main, never mind meaning no library split. [10:32] persia> I've reuploaded smplayer-themes, addressed your and norsetto's comments [10:36] LucidFox: Great. I'm not likely to review again until REVU day, but I suspect you'll at least get an advocation then if not an upload (of course, if someone has a particular interest, and reviews in advance, this increases the chances). [10:37] * persia is glad to see the "didn't get enough review" list down to three packages, from twelve ~24 hours ago. [10:38] * DaveMorris was surprised to see so many reviewed. TBH [10:38] dholbach: Are the current incumbents also up for reelection? [10:39] persia: I'd prefer if they re-applied [10:39] DaveMorris: We're getting close to Feature Freeze, so the pressure is on. In just a little bit, the MOTUs will entirely ignore REVU for almost three months. [10:39] to make it explicit they stand for election [10:39] dholbach: That sounds like a very good idea indeed. Let's hope there are more than five candidates. [10:39] * pochu waves [10:40] So during Feature freeze the packages just stack up [10:40] dholbach: so everyone's getting revoted now? [10:40] neat [10:41] I assume your working on other packages during the freeze then [10:41] soren: nope, x86 [10:41] Hobbsee: no, I just said to persia that I'd prefer if people re-applied for the job to make it explicit they stand for election [10:42] pochu: Oh, well. I thought it might an amd64-only thing. [10:43] DaveMorris: The idea is that we have (roughly) four phases of development in each cycle. First, we haul in everything that got missed for about six weeks, then we try to juggle it to be sane for about six weeks (that's where we are now), then there is about six weeks of bugfixing to make it releasable, followed by about six weeks including final release testing and polish, release, break, and UDS. [10:43] dholbach: i'm attempting to parse that into something i actually undersatnd, and i'm failing. [10:43] Hobbsee: there's no re-vote - people who wish to stand for election should reply to the mail I just sent [10:44] dholbach: as FF and UVF got merged into FF, does the motu-uvf work like before (exceptions for every new upstream version incl. minor versions updates) or on FF exceptions? [10:44] dholbach: Reply to the mail? Now I'm confused. I thought they were supposed to put their names on the wiki (which was a change from sending mail to MC members previously). [10:44] * persia votes for motu-uvf handling both [10:44] Hobbsee: I'm not per-se expecting that people run again for the job [10:45] dholbach: oh...i thought you were talking about MC. my mail was out of date. [10:45] oh ok [10:45] I was talking about UVF - persia: you too? :) [10:45] * persia was also confused about which team was which [10:45] how many new packages are normally accepted through REVU? [10:46] geser: both - it's afaik what they did last time [10:46] Hobbsee, persia: is there anything else unclear now? /me hopes not :) [10:46] dholbach: Not at all. Sorry for the noise. [10:46] dholbach: no, that makes more sense now [10:46] ok great [10:47] * dholbach refuses to fully wake up today... so it might just be me :) [10:48] dholbach: denying that you got a year older doesn't help [10:48] * persia suspects senility [10:49] * dholbach had the pleasure of getting breakfast served in bed today - after that much convenience at the start of the day do you really expect me to be up to scratch already? :-) [10:51] Does anyone care about mldonkey in Edgy? There's an SRU candidate about to exipire: if you want it, now is the time to test. [10:54] * Hobbsee curses blinken [11:11] TheMuso> Addressed issues with qdvdauthor [11:12] LucidFox: ok thank [11:12] thanks [11:20] Hi [11:21] hello pople [11:22] Does anyone know about zope3 and python2.5? Is this impossible, or just undone? [11:25] I have some questions about packaging translations. I understood that there should be no .mo files in the upstream release tarball, so I changed my setup.py to build the .mo files before installing. Now dpkg-source complains about 'binary file contents changed' regarding the .mo files when creating an Ubuntu package..? [11:26] rulus: Specifically there should ideally be no .mo files in the upstream tarball, but it is acceptable to delete them in the clean: rule, and leave them there. As dpkg-buildpackage (often called as debuild) ignores file deletions, this doesn't do anything at packaging time, so you shouldn't get that. [11:26] rulus: I suspect when you call setup.py clean it's doing more than just cleaning. [11:28] thanks, I'll check that out [11:28] persia: what do you mean? why should they not work together? [11:30] gaspa: The zope3 package only has python2.4 modules, not python2.5 modules. I'm working on my longest outstanding bug (now with 35 duplicates), and I believe everything is now fixed upstream except that it requires python2.4 and builds against python2.5. I'd rather bump zope3 than downgrade, unless there is some reason that zope3 doesn't want to bump. [11:30] RAOF: ping [11:31] Note that I'm very much not a python person, and this package is an abberation, so I have no idea if I would break anything making zope3 use python2.5, and would not be a good person to construct a test case. === \sh_away is now known as \sh [11:32] persia: i had one person that confirms me that zope3 works with python2.5. but he didn't install them from .debs. [11:33] gaspa: Are you someone who knows about things like python and zope? Would you be willing to take a look at the zope3 package to determine if it could be modified to use python2.5? [11:33] <\sh> Kmos, please DO NOT TOUCH MY SYNC BUGS! [11:33] * persia was surprised that the report wasn't ready for todays call, and really, really, really hopes it will be presented before the next call. [11:33] \sh, lol [11:34] yes, i know _somthing_ about zope2... zope3 is completely different. [11:34] <\sh> emgent, nothing to lol about...it's important that kmos know [11:34] gaspa: Hmm. I don't want to break anything else just to make a package that hasn't worked since Breezy work again. Maybe I'll just force python2.4 for now. [11:35] <\sh> persia, afaik zope3 runs with py2.5 very nicely, problem here are the zope products, which are sometimes not know to work with py2.5 [11:35] Kmos: ping [11:36] <\sh> emgent, sorry, i didn't want to sound to harsh... [11:36] \sh: Do you know if we package any of them? [11:36] <\sh> persia, there are some standalone zope product packages...I looked the last time at it, when I installed a zope3 server in my old company === bigon is now known as bigon` [11:37] <\sh> persia, feisty times [11:37] \sh, but i remember that if someone like merge/sync is good talk _FIRST_ with original maintainer (in ubuntu) [11:37] \sh: Hmmm.. I suspect I shouldn't break those. === bigon` is now known as bigon [11:37] Why, oh why, doesn't REVU remember my authenticated session across firefox sessions? [11:38] thoose could maintain dependencies on zope2.x [11:38] <\sh> emgent, well, we are four weeks before UVF and if we don't get all the merges ready for hardy, it will be sad..so therefore I think we can work on those without pinging the last uploader...only when there is something wich the merger doesn't understand [11:38] emgent: Especially at the beginning of a cycle, that is polite. At this point, most syncs / merges come from security, QA, and DCT teams, rather than from the person handling the package for variance. [11:39] persia, ++ [11:39] <\sh> emgent, but this is not the case right now...when I request a sync, I have a reason not to confirm/wishlist them...wishlist is not needed, indeed it's more correct, but annoys when someone else is changing the bug without asking and without testing [11:39] pochu: REVU doesn't remember the authenticated session within a browser session, but it will remember the authentication if you start separate firefox sessions in quick succession. [11:40] \sh, remember to sponsorizze my bugs :P [11:40] <\sh> persia, e.g. zope-sqlrelay (build-deps: Depends: python-sqlrelay (>= 1:0.38-3), zope2.9 | zope2.8 | zope) [11:40] \sh: I suspect most of your syncs are actually medium or high, as the ones I've seen tend to be security related. [11:41] \sh: That's not zope3, so doesn't matter as much. [11:41] <\sh> persia, oh I thought nowadays zope3 is default and is known zope -) [11:41] does the importance of a sync request has any influence when it gets processed? [11:41] geser: None at all. [11:41] <\sh> geser, I don't think so, or I feel like that... [11:42] \sh: Short list of rdepends is http://paste.ubuntu.com/3606/. Do any of those look especially interesting? [11:42] persia: would be nice if it was like, say, launchpad, where I don't have to login every now and then [11:42] pochu: branch, fix, and request a merge :) [11:43] <\sh> but as I said this morning, I don't confirm bugs or I set the importance of the bug without testing or knowing about the issue described in the bugreport, irrelevant if it's a real bug report or just a sync req... [11:43] <\sh> even as a sponsor I'm testbuilding and checking the packages ... so I'm sure that I don't sponsor broken stuff (which doesn't mean, that this never happend ,-)) [11:43] \sh: bug status does matter. Most archive admins only do syncs if they are Confirmed or Triaged (and random people setting this doesn't help keep it accurate) [11:44] persia: isn't zope maintained by Debian/Ubuntu Zope Team ? Don't you think they should know it better and bump python to 2.5 if possible? [11:45] <\sh> persia, python-psycopgda is a postgres zope module...this could be important [11:45] POX_: I thought 2.4 was still default in Debian. Has there been a push to move everything to 2.5? (And yes, generally the Debian Maintainers know best) [11:46] <\sh> persia, today they are important...when we started they weren't :) but ok,but "confirming a bug" means, that the person who confirms the bug knows that nothing really serious breakage is happening [11:46] 2.4 is default, but we have 2.5 as well [11:46] 2.5 is not default yet because few modules are still missing [11:47] <\sh> Kmos, please talk publicly in the channel...not in private this matter needs to be discussed here [11:47] ok [11:47] I only changed your bugs to confirmed (and wishlist, but that on isn't important) [11:48] because I saw slangasek and riddell ignoring your sync requests because of that. [11:48] <\sh> Kmos, did you actually testbuild and testinstalled those packages? [11:48] POX_: Makes sense. On the other hand, the person who originally pushed gaphor (depending on zope3) to 2.5 is an Uploader of zope3, so I'm seeing mixed messages. I think I'll revert in gaphor, as it doesn't work in any supported environment (Well, sarge, but that's not well supported) rather than messing with zope3. Thanks. [11:48] * geser LOLs at the pypy build log where it paints ASCII mandelbrot fractals during build [11:49] all I'm saying: there's "Debian/*Ubuntu* Zope Team" who would probably add python2.5 support if it would be that easy [11:49] <\sh> Kmos, do you really know that I didn't do any crap like your sync req for libgcr410 ? [11:49] persia: ask doko_, he's in that team [11:49] POX_: Which is indeed an excellent point. [11:49] \sh: and I don't say that.. [11:50] POX_: That's the person whose change I will be reverting. Less impact to force gaphor to be 2.4 than to look into zope3, which I don't use, and otherwise know nothing about. [11:50] <\sh> Kmos, no with confirming the bug you know about the package... [11:50] \sh: Bugs subscribed to ubuntu-archive should be set to confirmed [11:50] didn't look at gaphor, but can't it support both (2.4 and 2.5) ? [11:51] Kmos: No. Bugs subscribed to ubuntu-archive should only be confirmed by a member of ~ubuntu-dev who has personally tested the package. [11:51] \sh: you subscribed it, i just try to help, because you said the "confirmed" thing doesn't matter.. but it does matter. for archive-admins to not ignore you [11:51] <\sh> Kmos, dude, I have 50 packages more to check, so don't you think it's easier to file the reports, and after 5 or 6 days to set the status with script to confirmed? [11:51] POX_: gaphor works great with 2.5, except that it can't load modules from zope3 because they aren't provided. [11:51] <\sh> Kmos, in this very moment, it doesn't matter to me, and neither is this a matter for the archive-admins, until I set them to "confirmed" [11:52] oh, so gaphor has a dependency on zope, ok, sorry [11:52] \sh: ok.. sorry for the incovenient [11:52] <\sh> Kmos, anyways,please don't touch those sync reqs and leave the status changed for the guy who created the report... [11:53] \sh: i just saw them to be ignored.. and I see every comment in bug reports, saying the ubuntu changes can be droped, etc. [11:53] POX_: Yep. Cédric has been fighting with a rapidly changing upstream that frequently doesn't work for the past couple years, and finally has something working in sid. Despite the sid modifications including the ability to build with 2.5, I'm reverting to make it sane (as the modifications don't hurt as long as default is still 2.4). [11:53] \sh: one more time, sorry for trying to help. [11:54] like I said yesterday, I think I'm going to didicate myself to phishing [11:54] Kmos: It's not about inconvenience, it's about the fact that just setting it without testing and confirming could lead to a sync that should not have occurred, possibly breaking things. [11:54] <\sh> Kmos, well, if there is a sync req it's obvious that ubuntu changes are dropping...the reasoning is more important...well, I'll add the changelog to the report...I think it's more clear then === asac_ is now known as asac [11:55] POX_: Also, thanks a lot for chiming in. Bouncing this off you has made me much more confident. [11:55] --- Log opened Tue Jan 15 22:05:21 2008 [11:55] 22:05 <\sh> our requests were never ignored, forget the confirmed status...this is just for non-motus who are helping [11:56] \sh: so it needs "confirmed" status, but by the motu's / ubuntu-dev [11:56] <\sh> Kmos, yes, I was wrong nowadays..sorry.../me is old...I heard this morning steve explaining it...and now I confirm those bugs when they are needed [11:56] \sh: ah ok =) [11:57] <\sh> Kmos, but I don't confirm persias reqs or scottks or hobbsees, because I don't anything about their packages they work on... [11:57] * \sh needs some nicotine... [11:58] POX_: zope doesn't support python2.5 and later [11:58] \sh: ok.. i just see archive-admins ignoring them, them I talk to you in irc and you say it doesn't matter.. just tried to help on that, sorry again for that. [11:58] \sh: Given the completely random set of packages I touch, I suspect you are as likely to know about one of my bugs as anyone. [12:00] could someone check this - debian bug 460961 [12:00] Debian bug 460961 in openbios-sparc "Why arch is not only sparc ?" [Wishlist,Open] http://bugs.debian.org/460961 [12:01] <\sh> persia, that doesn't count...the thing is, e.g. yesterday with claws-mail-extra-plugins...I tested them locally and requested a sync for the new claws-mail, but I won't confirm the sync req for claws-mail-extra-plugins now, because I want to wait until the new claws-mail is hitting the archive, to be sure, that my pbuilder is not broken ,-) [12:01] the debian maintainer said it must me arch: all and not arch: sparc, because it only builds on sparc, but can be used on a non-sparc machine. something I don't understand here.. this package FTBFS in hardy. [12:01] Kmos: Because P-a-s takes care it, and this isn't the right forum to ask about Debian bugs, and it doesn't need origin-ubuntu hardy, as P-a-s also takes care of it in Ubuntu. [12:02] What's P-a-s ? [12:02] Kmos: The reason it should be arch-all is that once the bios is built, it can be used by emulators, etc. on other platforms. [12:02] isn't a debian bug, it's also a ubuntu bug, because it FTBFS. [12:02] Kmos: so, how are you following your "i will not modify sync requests of MOTU's" agreement, with \sh's bugs? [12:02] can you not do what you say? [12:03] RAOF: I made some changes to your libdrm package, now it builds a linux-drm-modules package that includes all the modules [12:03] * \sh is brb for 10 mins [12:03] Kmos: or laserjocks [12:03] Kmos: It's not FTBFS in Debian. It's only FTBFS in Ubuntu because we don't support building arch-all on non-i386. It requires manual bootstrapping in Ubuntu by a buildd admin to work. [12:03] persia: ah! i didn't know that.. thanks === doko_ is now known as doko [12:04] Kmos: Might make sense to ask before spamming the BTS, no? [12:04] persia: right [12:05] * persia grumbles that wiki.ubuntu.com is currently slow [12:05] * Kmos http://wiki.ubuntu.com/MOTU/Council/KmosReport [12:06] please write.. [12:06] Kmos: I will. I'm not sure yet if it is '0' or '-'. [12:06] persia: I think about \sh's should be - [12:06] Kmos: were you going to answer my question? [12:07] Hobbsee: you're true. it's done? [12:07] \sh's already has 3 -'s against it, as you would have seen, if you'd looked. [12:07] Kmos: you said you wouldn't do that. then you did. how can any of us trust what you say? [12:07] Hobbsee: if i can open the wiki =) [12:08] Hobbsee: if you see, are different cases.. so you can't apply one rule to all of them. [12:08] Kmos: what did you agree not to do? [12:09] Hobbsee: i agree to not touch bugs, i already know that.. but in that case, I saw archive-admins ignoring it, and it two days.. so I learn the comment and set it to confirmed. [12:09] Kmos: you agreed to not touch sync request bugs from MOTU's. How is your modifying \sh's bugs different? [12:09] Kmos: and why laserjock's? [12:09] Kmos: they were set to confirmed. You agreed not to touch them. you did anyway. [12:09] Hobbsee: laserjock, only changed it to wishlist [12:09] yes, but why? [12:10] to move up in the list [12:10] Kmos: you agreed not to touch them AT ALL! [12:10] archive-admins also don't have done backports in the last two days, and that syncs are in the bottom of the list [12:10] Hobbsee: yes, I know.. i just explained what, and why I did [12:10] Kmos: does it occur to you that not all things are time sensitive, and that in some cases, it's better to leave things alone? [12:10] Hobbsee: it's ok for you now ? [12:11] especially if you've agreed to do so? [12:11] Kmos: no, it's not OK. you said you wouldn't touch them, then you did. [12:11] don't need to remember that [12:11] Kmos: After investigation, openbios-sparc gets a '-' for no Ubuntu bug. [12:11] Kmos: you *do* need to remember it. [12:11] persia: thanks [12:11] Hobbsee: I think I need.. and I like to thank you about that [12:12] Kmos: Please open the corresponding Ubuntu bug, reporting the reason for the FTBFS, and the actions required to resolve it. [12:12] I think I won't forget it.. even if the sync or bug don't never get into archive if it's important or not [12:12] * pochu notices he can now approve and decline universe tasks :) [12:12] <\sh> Kmos, how do you know it's important? [12:12] Kmos: so, which types of sync bugs filed by MOTU's can you touch? [12:12] If a bug has been fixed in Hardy, and is also nominated to be fixed in Hardy, what should I do? Approve it and mark it as fixed, or decline it? [12:13] Hobbsee: none [12:13] Kmos: good. [12:13] Kmos: hwo did you not understand that last time? [12:13] \sh: syncs are generally important.. and only archive-admins can do it. [12:13] pochu: If you approve it, it should automatically have the hardy task marked as "Fix Released" and a note "Status tracked in hardy". I generally approve those to avoid confusion. [12:13] Kmos: still, you said that you wouldn't do this last time. and now you have. [12:13] Kmos: this is not about you. I don't think any MOTU changes other MOTU's syncs without asking him. [12:14] Kmos: what guarentee do we have that your'e not going to do ti again? [12:14] pochu: correct. [12:14] pochu: If it's not yet Fix Released, I usually decline, unless it is considered a release blocker. [12:14] persia: Oh, I didn't know how to make that 'status tracked in ...'. Thank you. [12:14] Hobbsee: I don't do it again.. for sure. Because after my "quarantine" ends, I'll leave for good. [12:14] Kmos: universe syncs tend not to be important. there are also archive admin days, where they all get done. [12:15] persia: if it's not fixed yet and I approve it, will it show in any special list? [12:15] for me, universe is important.. that's a point of view. [12:15] not only main is important [12:15] sure, but all the syncs will get done reasonably quickly. does it particularly matter which order they're done in? [12:15] pochu: I agree. [12:15] <\sh> Kmos, how did you test libgcr410 and pcsd-lite source? [12:15] Hobbsee: no.. but if they're ignored for two days. [12:15] <\sh> Kmos, I'm asking because my cardreader is not working since my update now [12:16] Kmos: 2 days. dude, have some patience. [12:16] pochu: Yes, it will show in the list of bugs tracked for the release, which start appearing in mailing list posts as we get closer to release. Only main is really important, but the default search scripts don't discriminate, so it makes noise for the QA team. [12:16] Hobbsee: that's not my best.. pacience. [12:16] <\sh> Kmos, so I'm fcked now with signing my mails and packages...so what should I do now? [12:16] Kmos: so get better at it, rather than putting in more crap. [12:16] persia: alright. Just to know in case I find a blocker ;) But since I can't touch main nominations I don't think I'll find any real blocker :) [12:17] Kmos: you were told this ages ago. yet you did not do it. [12:17] Kmos: why? [12:17] \sh: I really don't know what to say to you about that.. I can't test every card on earth. [12:17] Hobbsee: because of my pacience *thing* [12:17] maybe.. [12:17] <\sh> Kmos, it's a simple usb smartcard reader which worked since dapper [12:17] wiki is working again [12:17] Kmos: so, fix your patience thing. [12:17] pochu: Likely not. There are sometimes universe blockers, perhaps related to NBS stuff that would otherwise cause massive cannot-start or uninstallible issues, but those need to be also advertised in other ways so we can all hit them soonest. [12:18] Hobbsee: i take medicine pills for that already =) for something like 3/4 years. [12:18] and I smoke.. [12:18] \sh: you're not alone here :) [12:18] Kmos: are they working? [12:18] \sh: that's really strange.. how about to file a bug in debian BTS ? [12:19] Hobbsee: are.. but they can't do everything [12:19] <\sh> Kmos, now it's not a debian bug, it's an ubuntu bug...because the old version did work...so why did you sync it when it's not working? [12:19] <\sh> crap. [12:19] \sh: because testing is overrated, and because we automatically want the newest bling. [12:20] <\sh> Hobbsee, this statement is reserved for wine only *eg* [12:20] \sh: :) [12:20] <\sh> oh wow [12:20] \sh: yes, but we know to blame you if wine doesn't work. [12:20] <\sh> pscs leaves an /etc/init.d/pscsd.dpkg-new [12:21] <\sh> and doesn't overwrite the old one...well it's good but it should have asked me if I want to overwrite [12:21] \sh: I guess pcsc-lite needs a fix as it tries to use /var/run/pcscd/ for it's pidfile [12:21] * Hobbsee adds to KmosReport [12:22] I didn't check yet if it checks that that dir exists on startup [12:22] Kmos: why did you go and modify all the sync requests yourself, instead of notifying the MOTU's that the archive admins do ignore unconfirmed bugs, so they can fix their scripts? [12:24] <\sh> geser, the change was PIDFILE=/var/run/$NAME.pid vs PIDFILE=/var/run/pcscd/$NAME.pid [12:25] \sh: /var/run is a tmpfs so /var/run/pcscd isn't there after a reboot and the init-script doesn't recreate it [12:26] \sh: do you know how start-stop-daemon reacts if it can't write the pidfile? [12:26] <\sh> geser, it doesn't start [12:27] ScottK: ammo added. [12:30] persia: bug 183495 is opened [12:30] Launchpad bug 183495 in openbios-sparc "[FTBFS] openbios-sparc (1.0~alpha2+20070816-1) fails to build in hardy" [Medium,New] https://launchpad.net/bugs/183495 [12:30] * Kmos smoke [12:30] Kmos: Thanks. [12:31] <\sh> geser, the problem I have, the old init file is still in place...but it doesn't start at all... [12:33] <\sh> geser, and the reason is upstream set the new by default : [12:33] <\sh> Statement from upstream changelog: ChangeLog:- default ipcdir is /var/run/pcscd instead of /var/run so the directory [12:33] <\sh> Kmos, there you are [12:35] Kmos: Would you mind updating the description to indicate to the buildd admins how they could resolve the FTBFS? [12:38] <\sh> Kmos, please fix bug #183498 [12:38] Launchpad bug 183498 in pcsc-lite "pscsd doesn't start anymore " [Undecided,New] https://launchpad.net/bugs/183498 [12:40] hi all [12:40] heya mruiz [12:40] <\sh> Kmos, don't worry...I'll fix it [12:41] persia: s/buildd admins/lamont/ [12:41] Hobbsee: I thought infinity looked after sparc [12:42] persia: lamont is back at canonical. it might be infnity who does it, but infinity tends not to look at bugs assigned to him much [12:42] Hobbsee: Ah. In that case, yes. Thanks for the update. [12:42] persia: no problem [12:43] * persia still likes the use of general team identifiers until things are actually in shape to be done and someone is selected to look at them [12:43] persia: this is true, but unfortuantely a lot of the builld admins either will not, or can't, actually fix p-a-s [12:44] Hobbsee: Yes, but this isn't actually really a P-a-s issue. It's a manual bootstrap issue (until someone hacks arch:all arch-specific hinting into P-a-s) [12:45] persia: why is it a manual bootstrap issue? [12:45] We would have the same issue with e.g. bochbios if we selected a different architecture for arch:all [12:45] as in, isn't hte idea just to make it only try to build on sparc? [12:45] persia: it has "It needs manual bootstrapping by an buildd admin." , isn't enough ? [12:45] \sh: thanks [12:45] Hobbsee: Right. it's an arch:all package that can only be built on a sparc, but can be used anywhere. [12:46] persia: oh right. [12:46] Kmos: Not really. Notice that Hobbsee is confused. [12:46] persia: ok [12:46] persia: is there any logical reason why packages would need that? [12:46] <\sh> Hobbsee, qemu e.g. [12:47] Hobbsee: There are three classes of packages that need that: [12:47] \sh: oh, i didn't realise that needed to build on sparc. [12:47] 1) packages that need internet access (for a good reason) [12:47] "It needs manual bootstrapping by an buildd admin, it can only be build on sparc, but can be used anywhere (in other archs)." [12:47] better? [12:47] * Hobbsee thought arch all packagse should be able to build on...any arch? [12:47] Hobbsee: that's "any" [12:47] <\sh> Hobbsee, nope..I mean those packages who can only be build on only but be used by some other archs [12:47] 2) packages that build a bios and need that arch to be real, but can be used in emulators [12:47] \sh: my question is why they can only build on a non-i386 arch.... [12:48] 3) packages that build-depend on packages with interactive license agreements (although there is now a hack in place for sun Java) [12:48] ah, yes, if you're needing to build a bios and that arch needs to be real, yes, i can see your point [12:48] Hobbsee: That's #2. Thinks like openbios-sparc, bochsbios, linuxbios, etc. [12:48] <\sh> Hobbsee, in this special case it's because we don't have any cross compiler on x86/x86_64 to build sparc code... [12:48] s/Thinks/Things/ (wondering why typos are phonetic) [12:48] * Hobbsee knew about 1 & 3 - was asking why it had to be bootstrapped on a particular arch, not i386, not why things need to be bootstrapped in general [12:48] \sh: yummy. [12:49] persia: circular build-depends need also a manual bootstrap on the buildds [12:49] which persia then answered. thanks [12:49] \sh: Well, not quite. Even if we had a cross-compiler, it wouldn't be ideal. [12:49] <\sh> persia, but it would work (as I saw yesterday my experiment with a arm cross compiler ;)) [12:49] geser: Only when bootstrapping. As long as you've already bootstrapped, it's safe (except for versioned circular depends, which is bootstrapping again). But yes, Four reasons. [12:50] \sh: Yes, but you run the risk that your cross-compilation environment doesn't mirror the target properly, and further you have bootstrapping issues (hard to get a cross-compilation environment without a bootstrap bios). [12:51] <\sh> geser, what would you like more, leaving the default and mkdir -p /var/run/pcscd/ during startup if it's not running, or changing ./configure to point to --ipcdir=/var/run and change the debian/pcscd.init to point to /var/run ? [12:51] <\sh> persia, in this special case, yes :) [12:51] <\sh> s/running/exists/ [12:51] \sh: Wouldn't the same apply for e.g. an ARM bios on x86_64? [12:52] <\sh> persia, yepp...but doing kernel and app compilation you don't need to bootstrap, just a gcc with x86/x86_64/arm support [12:52] OK. New python packaging question: I've gotten gaphor to use python2.4, but it now depends on "python (<< 2.5)", which seems to have been inserted automatically during the build process. Any suggestions to avoid this? [12:53] \sh: I'd probably go for mkdir, chmod, chown (if needed) as it's also done in other packages needing a dir below /var/run [12:53] <\sh> .oO(well the gcc needs to be bootstraped yes, but without a bios ;)) [12:53] \sh: Ah. Nifty. Do we compile our qemu ARM bios that way? [12:53] <\sh> persia, hmm...I'm not sure...let's have a look :) [12:54] persia: why do you need it removed? [12:54] persia: just a guess, did you try to change the interpreter to python2.4 if it has any binaries? which you should do anyway if it needs python2.4 [12:54] Hobbsee: because python2.5 is part of the default install, and I don't want to break that just to allow installation of this package. [12:54] persia: oh, point. [12:55] geser: I did that. Do I maybe have to pass a hint to python-support that it doesn't need to block newer python? [12:57] persia: I don't know of any hints. [12:57] <\sh> persia, what is the name of the qemu arm bios? I don't find it [12:58] * persia digs into python-support, wishing that this package could be synced already. This has been way too much effort for an icon in the menu in Hoary. [12:58] \sh: No idea. I keep meaning to get an ARM chroot running, and keep postponing it. [12:58] persia: that might be a crude hack but it could work: specify as supported python version also 2.5 but use python2.4 in the binary. If I'm not mistaken python-support should add a dependency on python2.4 [12:59] persia: if that doesn't work, hardcode a dependency on python2.4 and remove the ${python:Depends} [12:59] geser: So Build-Depends: python-all, pyversions with 2.4, hardcode the interpreters, and hardcode *egg-info/requires.txt ? [13:01] <\sh> persia, I compiled yesterday the whole system of openwrt for arm arch...it bootstraps the toolchain and starts compiling :) very nice...but I didn't have to deal with a bios... [13:01] <\sh> geser, will fix the package this evening.. [13:01] persia: yes, but with pyversions 2.4-2.5, so ${python:Depends} add a dependency on python which doesn't conflict with the one in the archive [13:01] * \sh needs to go to the gym [13:01] \sh: any progress on your fitness? [13:01] <\sh> bbl [13:01] geser: Ah. Maybe that's my only issue. Trying again with "2.4-". Thanks. [13:02] <\sh> geser, yepp...2kgs lost :) [13:02] <\sh> ok cu later ... === \sh is now known as \sh_away [13:19] * persia thinks listing all the upstream changes in debian/changelog and not including the upstream changelog is just duplication of effort [13:21] * rzr updated : http://revu.tauware.de/details.py?package=jaaa [13:22] * rulus updated http://revu.tauware.de/details.py?package=gtkvd too :-) [13:26] persia: FYI ... The problem related to w3c-dtd-xhtml was already reported in Debian about 6 months ago. But looks like the reporter didn't provide enough information to convince the maintainer. I have provided it and the bug is now in 'New' state. So I guess it will get fixed in Debian soon. [13:27] slytherin: Excellent. By the way, would you happen to feel like looking into any upstream version bumps? [13:28] persia: for which package? or general java packages? [13:29] slytherin: specifically I'm looking at libswing-layout-java, javahelp2 , liblucene-java, and junit. [13:30] persia: I will take a look. [13:31] slytherin: Some of these may be syncs from Debian. Also, be interesting to check to see if we can sync libfreemarker-java [13:31] slytherin: Thanks a lot. [13:31] persia: Will let you know by tomorrow end of day. [13:32] slytherin: I have absolutely no idea what that means. If you're up for it, just file upgrade / sync bugs as usual. If you subscribe me, and it's those packages, I'll try to accelerate review. [13:33] persia: Yes, that is what I meant. I will give a general one line report tomorrow. [13:33] slytherin: if you need help from Debian side, just tell me [13:34] man-di: Sure. :-) [13:34] slytherin: It's "tomorrow" which is undefined :) For me that starts in 86 minutes. [13:35] persia: Ok. After 24 hours. :-) [13:36] man-di: Do you happen to know the status of the script-api.jar packaging? [13:37] Good morning all. [13:38] persia: RE: libmilter, I agree for now. Eventually I think we'll need to support it in Main, but I think it's better as a spec for Hardy+1. [13:38] persia: nope, never heard of that [13:39] ScottK: OK. By the way, I'm almost done with the final fix for 30344. Current hack is forcing back to python2.4 for compatibility with zope3, and using the very newest upstream. [13:41] persia: I note that the Debian bug linked to 30344 got marked fix released, but haven't looked at it. [13:41] New upstream finally works in sid. Unfortunately, as zope3 doesn't work with python2.5, we can't do a simple merge (also, the sid package still FTBFS, but that's an easy patch that I'll be submitting as soon as I have something working) [13:44] Ah. OK. [13:45] From what I know about zope3, you definitely want to let someone else worry about making it work with Python 2.5. [13:49] \me plugs his package for revu - http://revu.tauware.de/details.py?package=opensg It's an OSS distributed scenegraph API for doing some cool 3D graphics across multiple machines as well as single machines. Has support for sort first and last. I use's it to drive a 46million pixel display. which is cool, so please revu it. [13:50] Grr. Despite build-depending on python-2.4, and changing all occurrences of /usr/bin/pyhon and /usr/bin/env python to force 2.4, it still compiles for 2.5 if pyversions is "2.4-", Depends on python (<<2.5) with "2.4", and inserts extra #!/usr/bin/python lines in the build process without asking. [13:52] DaveMorris: Although I agree that your package hasn't had a review in overlong, it's best to keep your advertisements to less than once in 24 hours when it's not REVU day so as to avoid annoying potential reviewers. [13:53] persia: elementtree manages to be just for Python 2.4, you might have a look at that package. [13:53] ScottK: Thanks for the hint. [13:53] persia: sure [13:53] I just wanted to make sure people from various timezones saw it [13:57] DaveMorris: I generally recommend staggering advertisements by about 30 hours to achieve that. Makes sure you don't get an extra complaint from somebody who forgot to go to bed as well :) [13:58] thanks for the adice [13:58] *advice [13:58] I'm feeling confident with packaging new things now, is there somewhere I can read on how to do the other tasks MOTU's undertake? [14:02] DaveMorris: ho wabout you fix some build failures. :-) [14:03] DaveMorris: http://qa.ubuntuwire.com/ftbfs/ [14:06] DaveMorris: Most of the work that we do is just looking at packages that need help, and fixing issues. There are some automated tools, like the FTBFS checker, or debcheck (catches issues with dependencies that prevent packages from installing). There's also the bugtracker, which lists all kinds of things that need work, from the simple to the multi-year project. [14:08] got a url for the bug tracker? [14:09] DaveMorris: https://bugs.launchpad.net/ubuntu/+bugs [14:09] ok so bugs in general not bugs with special tags/assigned to a group [14:10] DaveMorris: Right. For a while there was a practice of assigning/subscribing MOTU to bug reports, but that is now deprecated, as there are just too many packages that are handled by MOTU, and it's a waste of the triagers' time. [14:11] DaveMorris: there are some tags, like needs-packaging or bitesize [14:12] DaveMorris: You might try searching for bugs with patches attached, or by tag, or just search for a term that interests you. Also, #ubuntu-bugs is a good place to talk about bugs and finding classes of bugs that need help. [14:12] so when I fix some of the FTBS bugs do I upload them to revu again? [14:12] DaveMorris: Nope. You attach a debdiff. You might want to read https://wiki.ubuntu.com/MOTU/Contributing [14:17] persia: man-di: a quick update on the packages you asked about - http://pastebin.com/m489c582 [14:18] slytherin: I will ping the right people in debian about it [14:19] slytherin: My information reports there being junit-3.8.2.jar floating about. Are you sure about that? [14:20] persia: Let me check once again. [14:20] (Google has 3,860 hits on junit-3.8.2.jar) [14:21] persia: Nothing found on website (sourceforge). So I guess it is better to ignore. [14:21] persia: oops, I was wrong. It is available [14:21] newest junit3 is 3.8.2 [14:21] I packaged it [14:22] heh. I thought so. It is a reverse-dependency for something I'm watching? [14:22] man-di: Safe to sync? [14:22] http://sourceforge.net/project/showfiles.php?group_id=15278&package_id=12472 has it too [14:22] persia: yes, I think so [14:23] persia: I will check debian changelog and log a sync bug accordingly [14:23] man-di: slytherin: Thanks [14:26] persia: done for junit - bug 183529 [14:26] Launchpad bug 183529 in junit "Please sync latest version from Debian" [Undecided,New] https://launchpad.net/bugs/183529 [14:27] slytherin: Great. Thanks. Next target: lucene2 :) [14:28] persia: As soon as w3c-dtd-xhtml is fixed. :-) [14:29] OK. That only leaves libswing-layout-java then. [14:30] man-di: should I file a bug in Debian for libswing-layout-java? [14:31] slytherin: if you like, please do [14:31] slytherin: and if like too, please add a watch file [14:31] Heya gang [14:31] Heya bd. [14:31] to let us get automcatially report new upstream version of it [14:31] bddebian even [14:31] man-di: I neither use that library nor Debian. Just trying to keep things in shape. :-) [14:32] Hi ScottK [14:32] slytherin: you might want to know this page: http://pkg-java.alioth.debian.org/cgi-bin/qareport.cgi [14:32] slytherin: If the Maintainer is "Debian Java Maintainers ", and man-di gives you the go-ahead, you're doing the right thing :) [14:32] hey bddebian [14:33] Heya persia [14:33] Hey bddebian [14:33] and geser :) [14:37] Now I wish there was a separate team for PPC. [14:37] slytherin: There's a PPC interest group. What do you need? [14:38] persia: openoffice.org is missing from standard install. I have filed a bug already but it is in 'New' state. Looks like theer is no one else to confirm. :-) [14:39] slytherin: Have you asked for confirmation in #ubuntu-bugs? [14:39] persia: No, I didn't. [14:39] #ubuntu-bugs [14:39] since Sun is purchasing mysql [14:39] :D [14:39] slytherin: does it actually *build* on ppc? [14:39] -since* [14:40] persia: it is true [14:40] i actually forgot to look into that [14:40] Hobbsee: yes, At least writer, impress and calc. [14:40] last time i looked there was not enough space on the discs for it [14:40] joejaxx: ah, so that's the problem [14:41] joejaxx: I am specifically talking about alternate CD. It is of size around 640 MB IIRC [14:41] * Hobbsee suspects the other problem is that no one actually cares abou tit [14:42] i will take a look into it [14:42] joejaxx: Do you want bug number? [14:42] slytherin: i believe i have it in my email but sure [14:42] :) [14:43] bug 164491 [14:43] Launchpad bug 164491 in ubuntu-meta "[ppc] openoffice.org not present on alternate CD" [Undecided,New] https://launchpad.net/bugs/164491 [14:43] thanks [14:43] hi. who can sync the gpg keys for REVU uploads? [14:45] slytherin: I was just advised that Marek Slama is almost done preparing an update for libswing-layout-java, so just filing the bug is probably sufficient. Should be uploaded this week. [14:45] vud1: I'll start a sync now. [14:46] oks thanks! [14:46] persia: Ok [14:51] persia: I sponsor Marek normally [14:51] slytherin: I will ping you when I uploaded it [14:51] man-di: Ah, so it is you. He was being mysterious about the identity of his Debian contact :) [14:52] man-di: Ok. So I won't file a bug then. [14:53] looks like openoffice.org fails to build on powerpc currently (and also sparc) [14:54] i wish i had a power{5,6} it would make building soo much easier :) [14:54] <_MMA_> joejaxx: Just talk to calc. There's _major_ things happening with OO.o right now. [14:55] _MMA_: yeap i know [14:55] Ng: ping [14:58] Hobbsee: Can I use reportbug on an Ubuntu system to make submissions to Debian's BTS? [14:58] mok0: You can [14:58] mok0: yes, but you need to use the correct distro switch [14:58] reportbut --bts=debian [14:58] but/bug [14:58] Oh, a but report [14:58] OK, thx! [14:58] * StevenK hides [14:58] Wrong channel for that. [14:59] Haha [15:00] Now I'm extra annoyed. I have a python file that hardcodes #!/usr/bin/python2.4, and it somehow gets bumped to 2.5 during the build process. [15:02] persia: You aren't build-depending on a virtual package are you? [15:02] sbuild doesn't support versioned depends on virtual packages. [15:02] ScottK: python2.4, python-support (>= 0.6), debhelper (>= 5) + python-setuptools, xsltproc, docbook-xsl === bigon is now known as bigon` [15:03] ScottK; Yep. I've been encountering that for a while now :) Personally, I think it shouldn't, given how virtual packages are intended to be used, but I understand why it might be nice given how virtual packages are actually used. [15:05] Did you look at elementree? [15:05] Well all I care about on sbuild is that Malone can build packages correctly and it can't always now, so I don't care what you call the fix, it needs to be fixed. [15:06] Hobbsee: hey [15:06] ScottK: Yes, and it was exceedingly helpful. I was able to revert almost all of my hacks and still have a python2.4 package. [15:06] ScottK: Fix the packages [15:07] persia: They aren't fixable. My particular concern is the perl modules conundrum where perl ships modules that are also provided in a later version by another package. [15:07] If sbuild find out it has an insufficient version of a package it ought to see if that problem can be solved and not just fall over and die. [15:08] ScottK: Yes. I know. Ideally upstream wouldn't be quite that annoying. [15:08] But we don't live in an ideal world. [15:08] Right. See my point about there being a difference between how virtual packages were intended to be used and how they are actually used. [15:09] With any luck sbuild will get fixed about the same time as my core-dev application gets approved and so I'll be able to fix the one FTBFS I've caused due to this problem without bugging sponsors. === bigon` is now known as bigon [15:09] ScottK: Soyuz doesn't use Ubuntu sbuild :( === bigon is now known as bigon` === bigon` is now known as bigon [15:09] persia: No, it uses a Canonical fork of it. [15:10] Which also has the bug. [15:10] Of course apparently a lot of the Debian buildd's don't use the Debian sbuild either. [15:10] ScottK: No, it uses a Canonical fork of the sbuild from which the Debian distributed sbuild forked (and which we now sync). [15:11] Right. [15:11] There is a common codebase, but it's fairly historic at this point. GIven the changes, I'm not sure it's easy to backport, but given the advantages of schroot, it might be reasonable to expect Soyuz to migrate when upgrading to hardy. [15:13] infinity promised a fix in short order, but I'm not holding my breath. [15:13] bddebian: You gonna fix Tremulous? [15:14] ScottK: Makes sense. Just wanted to make sure you knew that Soyuz & Ubuntu had different trees. Also, I'm not sure I want it fixed locally only, as it makes it harder to anticipate Soyuz breakage. [15:14] True. There's a patch sitting in Debian BTS that addresses (I think) this problem. [15:15] I cannot help but mention that this situation would be much easier to understand if Launchpad weren't proprietary. [15:15] ScottK: That may well be pending upload waiting for integration with the Debian buildd system as well. [15:16] ScottK: To understand perhaps. On the other hand, the same issue exists in Debian, despite the extreme efforts underway to allow external bootstrapping of DAK and freinds. [15:25] ScottK: I'm not sure of the issue yet [15:26] bddebian: Well given the siretart is a DD and expressed some interest in it getting done you might actually be able to get sponsored on that one. [15:26] Heh [15:31] bddebian: Bascially there's a 3rd party patch available for review and application. Reported to be popular in the tremulous-playing community. === apachelogger__ is now known as apachelogger [15:37] persia: Thanks. Doesn't look like tremulous is a games team package [15:38] bddebian: It's Ubuntu games team, at least. More BTS entries are also nice :) [15:39] bddebian: I'm not sure I understood your latest comment on sdlmame... what do you mean with "whole structure"? [15:39] wallyweek: I mean all of the subdirs are required under .mame, not just .mame/roms ? [15:40] Well files/dirs I should say [15:40] persia: Ubuntu Games Team?? [15:41] bddebian: You know. The people who watch Section: games in Ubuntu and used to have a LP group until Debian Games Team granted all of them SVN access? [15:41] dholbach: I'm confused by your Kmos report. Your "Recommendation" is actually a summary. I don't understand what is to happen from here? [15:41] bddebian: ah! no, they aren't afaik [15:43] ScottK: While I agree with you about that, remember that this is the report to the MC, rather than the report from the MC on the action to take. [15:43] persia: Uhm, OK :-) [15:44] ScottK: Cesare and I summarised the supervision of the last three weeks [15:44] ScottK: right... 'Recommendation' is probably not the right word [15:44] hrm [15:45] dholbach: You also fail to mention him not doing as directed repeatedly and reliably. [15:45] dholbach: If he can't be trusted to do as he's told when he's under close supervision, he can't be trusted at all. [15:46] bddebian: ok, I'll check and address your comment as soon as I'm back home [15:46] wallyweek: NP, thanks [15:47] ScottK: I'll follow up and clarify on that point - I've got a call in a bit, but I'll do it [15:47] dholbach: It reads like a whitewash to me. [15:47] * ScottK is very disappointed. I'll read the mail. [15:48] you're free to reply to that mail [15:48] ugh [15:50] whats uvf? [15:51] nxvl_work: It was UpstreamVersionFreeze. It has now been rolled into FeatureFreeze. [15:51] hi [15:51] persia: thnx [15:51] i just wonder, i the font "lucida" freely redistributable? [15:51] it's in the java package [15:52] i couldn't find anything about licencing fonts, etc. in the sun java license [16:01] hello there [16:02] What about the migration from interdiff to diff.gz (sponsoring new upstream versions) ? [16:02] !hi > emgent [16:03] mruiz: Still under discussion on ubuntu-devel@ (although that seems to have quieted). Feel free to add it to the MOTU Meeting agenda if you like: I was planning to wait until the following meeting to get more involvement from URC+2, where a number of the migration advocates live. [16:03] s/URC/UTC/ [16:04] I'm adding the diff.gz and .dsc to my upgrades bugs [16:04] mruiz: The .dsc is completely useless for upgrade bugs. [16:04] ! [16:06] mruiz: I've expanded on that opinion at https://lists.ubuntu.com/archives/ubuntu-devel/2008-January/024922.html [16:06] mruiz: Also, I believe it's best to follow the current process until we agree on a new one, so I encourage you to post interdiffs [16:08] persia, just interdiffs ? [16:09] mruiz: Yep. The interdiff (appropriately constructed) allows the sponsor to generate the diff.gz. The diff.gz should contain enough information to generate orig.tar.gz, and that contains enough of the package that debuild will generate the .dsc and source.changes [16:09] mruiz: see https://wiki.ubuntu.com/UbuntuDevelopment/Interdiff [16:09] * mruiz is reading there [16:10] I am an advocate of using .diff.gz alone, but it's not yet the approved practice for doing things. [16:12] [off-topic]: does anyone know how can I add fm radio channels in rhythmbox db? There is no ui for this yet and I have to manually edit the db. I just want to know what is format. [16:17] slytherin: I don't use radio channels, but what about right-click on Radio, then Add new radio station? [16:18] pochu: What you are talking about is internet radio. I am talking about FM radio played using tuner cards [16:19] slytherin: oh. No idea then. Maybe try in #rhythmbox in Gimpnet [16:20] pochu: Will try [16:26] need DebootstrapChroot help, reading the wiki and in the 'dchroot' section it says to add a bunch of stuff to fstab ... do i really add those lines to my / fstab? or the chroot fstab? [16:28] Ng: bug #183555. Would you mind subscribing yourself or the terminator team (if it exists) to the Ubuntu terminator bugs? https://launchpad.net/ubuntu/+source/terminator/+subscribe [16:28] Launchpad bug 183555 in terminator "README file has unclear shorcuts" [Undecided,New] https://launchpad.net/bugs/183555 [16:32] Where can I find information to create a man page ? [16:32] pochu: I'll do it now, thanks :) [16:32] Ng: you're welcome [16:33] I've added our team :) [16:33] mruiz: help2man is a nice script if you're creating the man page for an application which supports --help command line option === x-spec-t is now known as Spec [16:37] pochu, I'll find out about it . Thanks! === _nuu is now known as nuu [16:54] dantalizing: which wiki page? what do you want to do with the chroot? [16:57] pochu> From my experience, man pages generated by help2man require a lot of manual tweaking afterwards [16:57] Could anyone help explain install_scripts from distutils to me? It looks to me like this might be the cause of the build-time modifications of the #! line for my python2.4 issue, but I'm not sure I understand well enough to be sure. [16:57] Better than writing from scratch! :) [16:57] bddebian> True :) [17:01] effie_jayx: ping [17:03] geser: https://wiki.ubuntu.com/DebootstrapChroot [17:03] geser: setting up a hardy chroot on gutsy so i can learn to package stuff, or otherwise help out [17:03] Hello, I have a question concerning revu: the mousetweaks module has been submitted for integration into GNOME and in revu. It has been accepted in GNOME and ships with it since GNOME 2.21.5. Does it make any sense to continue with the revu process? [17:04] frafu: Will mousetweaks be a different source package as part of GNOME 2.21.5? [17:04] frafu> If it's accepted into Ubuntu as part of another source package, then there's no need to package it separately [17:05] LucidFox: yes, but creating a man page from the beginning requires more manual tweaking ;) [17:08] dantalizing: I use schroot for accessing my chroots (dchroot is now a wrapper around schroot) [17:08] Lucidfox: The gui of mousetweaks has been integrated into the mouse control panel of the gnome control center. The daemon of mousetweaks comes as a separate package. How can I know if it will also be in Ubuntu 8.04? [17:09] geser: thanks, how is schroot any better than 'sudo chroot ...'? === mathiaz_ is now known as mathiaz [17:09] dantalizing: I use "schroot -p" to get into my dev hardy chroot or "schroot -p -c ubuntu-i386" for my 32bit chroot [17:11] dantalizing: -p is for preserving the environment and I've set up schroot to bind-mount /home (and other dirs) into the chroot to have them available (and schroot unmounts them again when I leave the chroot) [17:11] geser: sounds interesting! What does your schroot.conf look like? [17:11] geser: Why -p? Isn't that dangerous (exposing your environment)? [17:12] Riddell: hi. could you speak to pitti (qtparted)? [17:13] frafu: If the daemon needs a separate package, someone needs to prepare it. REVU is a good place to put the candidate for review. Perhaps adjusting that to only have the daemon, and asking both here and on #ubuntu-desktop for review would be the best way to be sure it gets in for hardy. [17:13] persia: I want X11 access working from my 32bit chroot (e.g. for firefox with flash or skype) and also my usual environment when I use my hardy chroot were I prepare packages (so I don't pollute my normal environment with -dev packages) [17:14] Maybe he should ask seb128 about how he plans to integrate the mousetweaks backend? [17:14] He maintains gnome-control-center, after all [17:15] LucidFox: I think it's better to ask #ubuntu-desktop (maybe during French working hours), than seek a specific person, but I'm a big fan of teams and teamwork. I suspect you've correctly identified the person who would be most likely to have an authoritative answer. [17:16] geser: Ah. I see. You've also separate build chroots then, I guess? [17:16] persia: for building I use normal pbuiler chroots [17:16] pbuilder [17:17] geser: Oh. I thought most schroot users used sbuild over schroot. My mistake. [17:19] Consequently, being shipped with gnome does not necessarily mean that it will be in ubuntu. Moreover, revu is for packages for universe. Would it not be strange to ask to put a package into universe when the same package comes with gnome? I will try to ask seb128... [17:19] persia: when I got my new hardware, I will look on setting up sbuild (and a small local mirror, probably apt-cacher) and will then not use -p for the build chroots [17:20] currently I use schroot only to access my chroots [17:20] frafu: Best practice is to get things into universe, and then push them to main. That gets testing to make sure that it doesn't have issues with the buildds, builds for all architectures, etc. This makes the required Main Inclusion Report much easier to write. === vud1 is now known as tx0pi [17:28] So, I just ran `grep -ri $string /usr/share` and am encountering a lot of "Too many levels of symbolic links". Are these bugs? [17:29] where can I find a good guide on the types of changes needed to add dpatch to a source package? [17:31] it seems like our guides all kinda assume dpatch is already set up... and the examples that ship with dpatch are kinda vague as to what I need to do [17:31] jdong: the patch README? [17:31] dpatch README === tx0pi is now known as txopi [17:32] jdong: man dpatch [17:33] jdong: Better, man dpatch.make [17:33] persia: aha, that's the one! thanks [17:33] geser: I tried that first.... It has no README but it did provide me some fascinating reading about the history of the tool ;-) [17:34] jdong: there are also some examples in the doc/ dir [17:35] yep. [17:35] Wow, Launchpad is really mean nowadays... incomplete bugs self-expire? [17:35] jdong: not for Ubuntu. It's just a preview, but nothing won't change for now. [17:36] jdong: Yes. Generally they get restored if anyone cares (although it does increase the workload for people who take > 4 months to prepare a solution and left it "incomplete" in the meantime) [17:36] yeah I guess if someone cares then it's a non-issue [17:37] * jdong grumbles at that sigsegv bug report filed against the supertux he uploaded yesterday... [17:37] That is at least the philosophy behind the change. I'm not sure how clear that is to the general public when they are advised their incomplete bug became invalid, but that's a different issue. [17:37] as always, unreproducible and I just wasted 30 minutes playing supertux. [17:38] note to self: DO NOT triage game related bugs! [17:38] jdong: Please do. The Games team appreciates any help. [17:38] jdong: Which bug? Does it have a stacktrace? [17:39] jdong: are those 30 minutes really wasted? :) [17:39] persia: bug 183419... I tried toggling fullscreen several times to no luck [17:39] Bug 183419 on http://launchpad.net/bugs/183419 is private [17:39] what? [17:39] err... no it's not? [17:40] https://bugs.edge.launchpad.net/ubuntu/+source/supertux/+bug/183419 [17:40] jdong: you are probably an indirect member of the universe-crash-team so you can see it [17:40] oh yeah it was [17:40] ok it's unprivate now. [17:41] and why is cdrom a nonfree kernel module anyway? :) [17:44] jdong: Try looking in src/gui/menu.cpp around line 750. I suspect the developer forgot to test to make sure the menu pointer was valid before attempting to use it under some special (and likely unreproducible) set of conditions. [17:47] persia: it's probably bad that the this pointer on 0 and 1 are both 0x0, right? :D [17:48] jdong: Usually means that it gets set by a call to some function that returned an error, and the developer didn't check for an error condition. Unfortunately frequent in games, and as you've discovered, best tracked down by source inspection rather than testing. [17:49] jdong: Essentially, a pointer should only be 0 when speficially uninitialised, and should never actually be used in that state. I've seen far too many cases where it gets set somewhere and then used before any validity checking. This might be acceptable in tight loop code, but is generally poor practice. [17:50] persia: hmm I'll investigate then [17:50] Further, in the case of use in tight loop code, it's important to trap the error outside the loop, reset to something sane, and possibly reenter, rather than crashing. Most players would prefer the occasional stutter to a segfault. [17:53] ok I see where the options_menu pointer was used that led to this stack track but not why it's null... [17:53] should've been initialized at this point, but I guess there might be a missed path to this code [17:53] jdong: Just step back up the function. That function was called with a real pointer (although possibly not to the structure the function expected). [17:54] I usually start three or four frames behind the crash to build context, so by the time I get to the crash point, I have some idea what things mean (as otherwise one has to read the entire codebase). [18:15] What package is requestsync in? [18:15] ubuntu-dev-tools. [18:15] Thx persia [18:15] Note that it doesn't work if debian source isn't in sync in all archs, and can be otherwise annoying. [18:16] the new update manager icon is awesome [18:16] I don't use it :-) [18:16] Then why do you want it? [18:16] because of the contrast around the star shape, I managed to mistake the tiny down arrow for something giving me the finger. [18:17] persia: Someone was asking about it on #debian-games :) [18:17] Ah. That's not so dangerous then. Remind them to use the -s parameter. [18:18] * persia isn't sure it would work on a debian system: might need an ubuntu chroot [18:22] POX_: If you're still about, I'm still stuck on pushing gaphor back to defaulting to 2.4. Essentially, setup.py is calling install_scripts to install the /usr/bin entry, and somewhere in this stack, get_script_header is called with executable of "/usr/bin/python2.5", which I'm guessing is the value of sys_executable used by default in get_script_args. [18:22] wow, REVU now has a lintian up-to-date right? :) [18:22] I'm having trouble tracking down where between these two points I can insert a hint that I really do want python2.4, yet still get the various extra bits that install_scripts seems to insert when installing the script. [18:24] pochu: Yep. [18:24] persia: I've a local fix for source-not-in-sync problem in my requestsync-with-python-launchpad-bugs script [18:25] geser: Cool! Do you think you'll be able to push by feature freeze? [18:26] persia: the script is available from bug #147994 for testing [18:26] Launchpad bug 147994 in ubuntu-dev-tools "requestsync should have a command line switch to use python-launchpad-bugs (instead of mailing)" [Wishlist,In progress] https://launchpad.net/bugs/147994 [18:26] it WFM but some broader testing would be nice [18:27] persia: point me to sources, I'm checking RainCT's package right now [18:27] and I'm not sure if it should stay a separate script or be merged with the normal requestsync script [18:28] http://lists.debian.org/debian-mentors/2007/12/msg00151.html "debian/manpages: What is this file for?" [18:29] Seriously... he reviews, cites obscure Debian guidelines on copyright files, and doesn't know what debian/manpages is for? [18:29] is this the proper channel for ppa questions related to packaging...at least i think related to packaging.... [18:29] dantalizing> Ask your questions :) [18:30] LucidFox: uh, I think Patrick Schoenfeld is a known professional troll, though maybe I mixup the name [18:30] Well, his advice does seem to be helpful [18:30] POX_: http://people.ubuntuwire.com/~persia/packages/gaphor_0.12.4-2ubuntu1.dsc builds cleanly in sid and hardy. Works in sid. Works in hardy if one manually adds "2.4" to the shebang line of /usr/bin/gaphor in the binary. [18:31] LucidFox: yeah ok, I mixed up the name [18:31] he maintains a couple of packages [18:31] LucidFox: Some of that is wrong too. At least the debhelper dependency was bumped for a reason. [18:32] Yes, for dh_icons [18:32] but I'm planning to move to cdbs for the next mentors.d.n upload [18:33] gnome.mk takes care of dh_icons both being and not being present, so I'll relax the dependency to just >= 5 anyway [18:34] i keep getting rejected uploads. i'm trying to package latest upstream vzctl, which i was able to do manually on my system. now trying to upload source package to PPA. I was in the Packaging 101 yesterday, and have read Packaging with DebHelper on h.u.c. My .dsc has Files: vzctl_3.0.22.orig.tar.gz and vzctl_3.0.22-0ubuntu2~ppa2.diff.gz. and those files exist in the same dir as the .dsc,... [18:34] ...however PPA is rejecting with: Rejected: [18:34] Unable to find vzctl_3.0.22.orig.tar.gz in upload or distribution. [18:34] Files specified in DSC are broken or missing, skipping package unpack verification. [18:34] Signer has no upload rights at all to this distribution. [18:34] Signer is not permitted to upload to the component 'universe' of file 'vzctl_3.0.22-0ubuntu2~ppa2.dsc' [18:34] and i dont know how to fix it....tia [18:35] dantalizing> First, build with "debuild -S -sa" [18:35] dantalizing: pastebins are good. Check your keys. Use -sa [18:35] rather than just "debuild -S" [18:35] And second, upload the changes file rather than the dsc file... although I think dput rejects dsc files for PPA on the spot [18:36] I did use "debuild -S", and entered my passphrase [18:36] LucidFox: i did upload the changes file. Just mentioning what was in the DSC [18:36] Ah. [18:37] dantalizing: and upload to ppa and not Ubuntu [18:37] Like persia said: use -sa. In addition to -S. Otherwise, dput won't include the orig.tar.gz in the upload. [18:37] dput vzctl_3.0.22-0ubuntu2~ppa2_source.changes [18:37] oh, ok [18:37] ^^ that would go to Ubuntu (unless you changed the default) [18:38] my .dput.cf has ppa.launchpad.net [18:38] as default? [18:38] i only have one entry [18:38] hmm [18:38] the "Signer is not permitted to upload to the component 'universe'" indicates an upload to the official Ubuntu archive [18:40] persia: sorry, pastebin next time [18:40] with bugzilla, can i reply to a comment bugzilla emailed me about and have it add the reply in bugzilla? [18:40] dantalizing> there is also /etc/dput.cf [18:40] the entries in .dput.cf are added rather than replacing hose [18:40] LucidFox: ah [18:41] you need to either set ppa as default, or use "dput ppa [changesfile]" (assuming that the name in square brackets in .dput.cf is [ppa]) [18:41] very nice [18:47] pochu: hello. [18:47] LucidFox,geser,persia: perfect! Working now. Thanks! more noob questions will follow I'm sure.... [18:47] pochu: i have seen the comment about gnomecatalog [18:47] hey josesanch [18:48] To add dh_desktop and dh_icon [18:48] appending this to rules ? [18:48] can works if [18:48] binary-indep: build install [18:48] dh_icons [18:48] dh_desktop [18:48] josesanch: since it's CDBS, it needs to be a bit different. [18:49] install/gnomecatalog:: [18:49] dh_icons [18:49] dh_desktop [18:49] |pastebin | josesanch and pochu [18:49] Argh. [18:49] kitterma: sorry [18:49] !pastebin even [18:49] Sorry, I don't know anything about pastebin even - try searching on http://ubotu.ubuntu-nl.org/factoids.cgi [18:50] !paste [18:50] pastebin is a service to post large texts so you don't flood the channel. The Ubuntu pastebin is at http://paste.ubuntu-nl.org (make sure you give us the URL for your paste - see also the #ubuntu channel topic) === kitterma is now known as ScottK2 [18:50] josesanch: look at https://perso.duckcorp.org/duck/cdbs-doc/cdbs-doc.xhtml [18:52] i could use cleanbuilddir/gnomecatalog:: [18:52] to clean the mo dir. [18:53] pochu: appending this to rules must work then. http://paste.ubuntu-nl.org/52168/ [18:56] josesanch: I think so. You can try a build with that and see if dh_icons and dh_desktop are called at the end of the build, with the debhelper stuff. [18:57] josesanch: regarding the mo dir, better to clean it in setup.py, so other people not using our packaging benefits from it. [18:59] ok.. [19:00] I'll upload this version and wait for comments of anyone else. [19:00] then, if the package is correct i will bump a new version 0.3.4 [19:01] pochu: ok? [19:02] josesanch: sounds good. [19:02] pochu: thanks for all [19:03] yw [19:11] persia: "python setup.py install" line is setting the shebang to python2.5 [19:12] persia: just change this line to "python2.4 setup.py install" and you will have right one [19:12] persia: debian/rules, line 20, 29 and 47 [19:13] dod some knows gOS? [19:14] :q [19:14] err, wronk window :) [19:14] wrong even === thekorn__ is now known as thekorn [19:29] i need some help on building a package with pbuilder [19:29] i am getting an error message [19:29] hostname: unknown host [19:29] the one above [19:30] anyone can help me [19:30] ??? [19:31] Legendario, is your hostname in your hosts file? [19:32] more precisely, is it in your hosts file within the pbuilder tarball? [19:32] i gess i don't know what file is this... [19:32] hmm, perhaps it doesn't need to be in the tarball [19:33] Legendario: what does 'cat /etc/hostname' give you? [19:33] output: P4-Kemel [19:35] so... [19:35] that's my hostname when i type "hostname" on the terminal [19:36] hmm, I would still suspect a problem with /etc/hosts in the pbuilder tarball then, but more context for the error message might help [19:37] I: using fakeroot in build. [19:37] Current time: Wed Jan 16 17:20:58 BRST 2008 [19:37] pbuilder-time-stamp: 1200511258 [19:37] Building the build Environment [19:37] -> extracting base tarball [/var/cache/pbuilder/base.tgz] [19:37] vorian: I thought KTorrent 4 was version 3.x.x? [19:37] -> creating local configuration [19:37] hostname: Unknown host [19:37] !paste | Legendario [19:37] Legendario: pastebin is a service to post large texts so you don't flood the channel. The Ubuntu pastebin is at http://paste.ubuntu-nl.org (make sure you give us the URL for your paste - see also the #ubuntu channel topic) [19:37] jdong: nope, 4.0.0 [19:38] vorian: you sure? [19:38] aye, do you want the tarballs? :) [19:38] Gah, vorian looks too much like vorlon and scares me ;-P [19:38] It should be 127.0.0.1 P4-Kernel [19:38] vorian: woah woah I don't swing that way bro. [19:38] :( [19:38] jdong: liar [19:38] and 127.0.0.1 localhost [19:39] vorian: I was assuming since it was called KTorrent For KDE4 3.0beta1 and the KDE3 series was 2.2.x, it would be 3.0 [19:40] ubotu: thanks, i didn't know how to use it... [19:41] jdong: do you have some time to validate backports please ? :) [19:41] vorian: and btw the watchfile is outdated :) [19:41] jdong: you are hurting my feelings [19:42] slangasek: that's all the output http://paste.ubuntu-nl.org/52178/ [19:42] http://paste.ubuntu-nl.org/52178/plain/ [19:43] #define KT_VERSION_MACRO "3.0beta1" [19:43] vorian: ^^ :) [19:43] don't trust KDE4's FTP version numbers [19:44] Legendario: what does 'grep -i p4-kernel /etc/hosts' give you? [19:44] vorian: so 4.0.0 is the same version as what shipped in 3.97 beta :) [19:44] somerville32: you should never assign hostnames other than localhost to 127.0.0.1, that way lies madness [19:44] jdong: I'll fix it next upstream, hmk? [19:44] :P [19:44] slangasek, Ubuntu does it by default :P [19:44] vorian: yay for epochs :) [19:45] somerville32: I think you'll find that what Ubuntu uses is 127.0.*1*.1 [19:45] 127.0.1.1 P4-Kemel.virtua.com.br P4-Kemel.nau_do_Kemel [19:45] vorian: regex-scrape http://ktorrent.org/downloads/ for ^3.* [19:45] slangasek: that's it. [19:45] Legendario: ok, then I'm out of ideas regarding the source of that error, sorry [19:46] anyone else??? [19:46] I had that problem once [19:46] but I fixed it by correctly setting my hostname stuff [19:47] how? [19:49] somerville32, how did u correct it? [19:52] By having 127.0.0.1 localhost [19:52] and 127.0.0.1 serenity [19:52] serenity being the name of my computer [19:52] which is not a correct fix [19:52] Legendario: I do notice that you don't have 'P4-Kemel' listed as an alias on its own for your host [19:53] Legendario: so try adding P4-Kemel to the end of your 127.0.1.1 line in /etc/hosts [19:56] slangasek, did it. gonna run the pbuilder againg [19:59] ScottK, ScottK2: around? [19:59] Yes [19:59] about filezilla and its changelog [20:00] NEWS is actually the changelog, is it a problem that it is called NEWS and not ChangeLog? [20:00] ChangeLog only contains a link to the svn commit history [20:01] Given policy, I'm inclined to say you ought to install the thing the upstream calls the changelog. [20:02] anyway, as I use dh_installchangelogs, NEWS is renamed to changelog in /usr/share/doc/filezilla/ [20:02] $ cat ChangeLog [20:02] For a full list of changes, please visit http://filezilla-project.org/changelog.php [20:03] I don't think it's really useful for the end user. we do not ship INSTALL files either for example [20:03] Adri2000: So if you install something that upstream doesn't call the changelog and call it the changelog, I think that's definitely wrong. [20:06] policy says "If an upstream changelog is available", not "if a file called 'ChangeLog' is available" [20:06] and it even says "If the upstream changelog files do not already conform to this naming convention, then this may be achieved either by renaming the files, or by adding a symbolic link, at the maintainer's discretion." [20:06] so I really don't see what's wrong there [20:07] I think upstream knows what they believe their changelog is and you are installing a different file and calling it changelog. Installing NEWS also is a good idea. [20:07] slangasek: well, thanks dude. It seems to have worked... but got another error [20:07] I don’t see a reason not to install a file called Poop as changelog.gz if it happens to be the changelog. [20:08] ion_: If there's no upstream changelog, then I'd agree. [20:09] ion_: I see a slight reason not to package something using a file called 'Poop' as its changelog :) [20:09] Legendario: what's the new error? [20:09] scottk: He said NEWS is the changelog. [20:09] ion_: Except that upstream provides a file called Changelog. [20:09] We've already spent more time on this than it's worth. Do whatever you want with the bug. [20:10] ...which doesn’t contain one. :-) [20:10] hello [20:12] slangasek http://paste.ubuntu-nl.org/52191/ [20:14] Legendario: ah, well, that's a bug in your package then [20:14] ok [20:14] thanks [20:14] a lot [20:15] Legendario: the upstream build rules appear to write to /opt/ even when DESTDIR is specified, so you need to sort that out. Perhaps the upstream build rules support some variable other than DESTDIR, or need to be adjusted to support DESTDIR [20:16] slangasek, i think this /opt is set on the make file [20:16] slangasek, do you know how i should correct it [20:17] Legendario: I'd suggest grep'ing to find where it's set and then patch it. [20:18] it is set as a variable at the make file: ROOT=/opt/odfellowship [20:19] the make file inside the souce package [20:20] about bug #174252, do packages that depend on a package that was rebuild against libgif to be also builded again? (eg. aterm agains the rebuilded libafterimage-dev) [20:20] Launchpad bug 174252 in libungif4 "transition to libgif" [Undecided,In progress] https://launchpad.net/bugs/174252 [20:21] RainCT: shouldn't be needed [20:23] how do i match the DESTDIR with the make file ROOT [20:23] ? [20:24] Legendario: with $EDITOR [20:25] geser: can you give me more details [20:25] ? [20:26] geser: oh okay. thanks [20:29] Legendario: simply edit the Makefile and specify the correct value for ROOT [20:29] RainCT: any reason why it should need a rebuild? [20:30] geser: i know that... but what is the right value for avoiding another pbuilder error? [20:30] RainCT: the transition from libungif -> libgif should only affect those packages that link against it [20:31] geser: no, I'm just asking. [20:31] Legendario: which pbuilder error? [20:31] geser: http://paste.ubuntu-nl.org/52191/ [20:32] geser: so motioneye doesn't need a rebuild either? (it has a bug task but it doesn't depend on libgif nor libungif) [20:32] RainCT: also no build-depends? [20:33] geser: no. Build-Depends: debhelper (>=4), imlib11-dev [20:33] Legendario: the best way to patch is that you can overwrite ROOT with a value when you call make from debian/rules === cprov is now known as cprov-out [20:37] RainCT: ask the person who added the task why he believes it needs fixing in mentioneye [20:37] it might be that I missed something === nuu is now known as nu[year] === nu[year] is now known as nuu === bigon is now known as bigon` === bigon` is now known as bigon [21:00] Hi, [21:00] I have a problem with qdevelop package on Hardy: http://revu.tauware.de/details.py?package=qdevelop [21:01] what problem? [21:02] qdevelop depends sqlite.On gutsy, libqt4-dev require libqt4-sql and libqt4-sql require sqlite: [21:02] Depends on: libmysqlclient15off (>= 5.0.27-1), libpq5, libqt4-core (>= 4.3.2), libsqlite0 (>= 2.8.17), libsqlite3-0 (>= 3.4.2) [21:02] It's OK [21:02] On Hardy, qt4-sql don't require sqlite (don't require mysql...) : https://edge.launchpad.net/ubuntu/hardy/i386/libqt4-sql/4.3.3-0ubuntu2 [21:03] but description is:"This package contains the SQL module for Qt. It includes support for PostgreSQL, MySQL, and SQLite databases." [21:03] Depends on: libc6 (>= 2.7-1), libfontconfig1 (>= 2.4.0), libgcc1, libglib2.0-0 (>= 2.14.0), libqt4-core (=4.3.3-0ubuntu2), libstdc++6 (>= 4.1.1-21), zlib1g (>= 1:1.2.3.3.dfsg-1) [21:03] erable: on gutsy does it really link against two version of sqlite? [21:05] erable: try asking the kubuntu people in #kubuntu-devel, they should know better about qt4 than me [21:06] geser: ok. thanks [21:12] hi DktrKranz === bigon is now known as bigon` [21:13] hey somerville32 ;) [21:13] geser: can you approve my mail in motu-council ML ? [21:13] thanks [21:14] DktrKranz, Soren ACKed the ebox-network package upload [21:14] somerville32, ah. nice. I'm after some NMUs in Debian, but I'll have a look later. Thanks for the pointer :) [21:15] DktrKranz: Thanks for doing spamd. You gonna do the *-security uploads too? [21:15] hmm, bug 178152 [21:15] Launchpad bug 178152 in xine-lib "[packaging] patches outside debian dir" [Undecided,Won't fix] https://launchpad.net/bugs/178152 [21:16] ScottK, can we do them? I thought they were for ubuntu-security only. [21:16] Kmos: done [21:16] DktrKranz: We can prepare debdiffs and then they will upload. [21:18] geser: thanks [21:19] ScottK, spamd? === bigon` is now known as bigon [21:20] DktrKranz: Err dspam. Sound better? [21:20] Sorry. [21:20] ScottK, IIRC, it was DarkSun88. [21:20] Sorry again. Got my IRC nicks mixed up. [21:20] Argh. [21:20] heh, it happens, we live near :) [21:20] DarkSun88: Ping ^^^ [21:23] ScottK: Pong. [21:23] DarkSun88: Thanks for merging dspam. You going to do the *-security updates too? [21:24] ScottK: You're welcome, dont' worry. No. [21:25] DarkSun88: Please? [21:26] DktrKranz: hey, I see that you did a fair number of the rebuilds for the libcommoncpp transition? [21:26] No, I don't work to to the *-security updates too. [21:26] to do* [21:27] slangasek, I don't remember exactly, but it is highly probable. [21:28] DktrKranz: did you know that the new libcommoncpp hadn't made it in on sparc when you uploaded? :) [21:28] (but if you don't remember doing it at all, I guess you wouldn't :) [21:28] just mentioning because the packages needed to be reuploaded to get the proper dep on sparc [21:28] damn... [21:28] I'll have a look, thanks for the pointer [21:29] slangasek, in the case, a new rebuild is ok? [21:29] they're all reuploaded now so there's no more work to be done, but it seems inefficient to have had us both uploading them [21:29] Of course. Sorry for that [21:29] slangasek: I'm just curious: there are no binNMUs in ubuntu? [21:30] lucas: nope [21:30] ouch [21:30] yep :-) [21:31] Bin nmus? [21:32] lucas: the advantage is we don't have problems with strict versioned dependencies ({$binary:Versions} vs ${source:Version}) [21:32] TheMuso: binary non-maintainer uploads [21:32] TheMuso: in Debian, wanna-build can be told to rebuild a package for a particular architecture only [21:33] geser: Debian doesn't either, because those bugs have been fixed by now ;) [21:33] slangasek: Right. I must say I'm against havng to upload a binary for an arch, unlike Ubuntu where we only upload source [21:33] geser: sounds a bit like "we haven't invented the wheel yet. the good point is we don't have to care about wheels." [21:33] TheMuso: hrm? who said anything about uploading a binary? [21:34] TheMuso: all we do is tell wanna-build to do a rebuild; so the resulting binaries are handled the same way as any others built on a buildd [21:34] slangasek: Oh right [21:38] Speaking of rebuilds... slangasek: I understand sbuild on soyuz is getting patched today to not mess up the versioned depends on virtual packages anymore. [21:44] spiff === leonel_ is now known as leonel [22:06] Hmm.. unstable has debhelper 6 and now I wanted to request a sync for a package which uses it already. Is it likely that debhelper 6 gets merged for Hardy? (have not found an open bug for this) [22:07] hi all.. i have a .deb of the intel ixgbe driver that i put together (because it's not available in feisty which i'm using)... i have an ixgbe--source.deb now, but i want to make a platform-dependent binary .deb so that i can put it on machines that don't have build tools... what tool should i be using for this? [22:07] blueyed: ask the last merger of debhelper [22:08] What does "ER" stand for? [22:08] somerville32: ENOCONTEXT [22:08] Like, when talking about authentication [22:08] or networking [22:08] somerville32: entity-relation if it's in database context [22:09] somerville32: hmm, I think a more specific context is needed than even that :) [22:09] ER injector? [22:09] doko: Will you merge debhelper 6 for Hardy? I just noticed that it gets used already in a package I'd like to get synced. [22:09] like, a specific protocol or technology maybe? [22:10] somerville32, in italian (well, Roman) slang, "Er" means "The", but I think you are not referring to it ;) [22:10] :) [22:12] blueyed: maybe prepare a patch, if the merge has conflicts? [22:12] somerville32: When you say ER, I think of this: http://www.astronautix.com/lvs/staarder.htm - probably not what you were thinking [22:12] ER is an overloaded acronym [22:13] doko: there are two conflicting files only: man/po4a/po/{debhelper.pot,es.po} [22:26] let me ask you guys something... can anyone download a .deb package from REVU? [22:27] Legendario: which deb? there are no debs on REVU [22:28] geser: and how does the .deb gets to repository? [22:29] Legendario: you mean the official repository? [22:29] there the buildds build them [22:29] but there is no buildd for REVU [22:30] geser: yes, the universe. Isn't the REVU a way to the repos? [22:30] !rev [22:30] Sorry, I don't know anything about rev - try searching on http://ubotu.ubuntu-nl.org/factoids.cgi [22:30] !revu [22:30] REVU is a web-based tool to give people who have worked on Ubuntu packages a chance to "put their packages out there" for other people to look at and comment on in a structured manner. See https://wiki.ubuntu.com/MOTU/Packages/REVU [22:31] and if a package is seen by two MOTUs acceptable it get uploaded to universe where it gets build by the buildds [22:32] geser: what are the buildds? [22:32] !revu [22:32] REVU is a web-based tool to give people who have worked on Ubuntu packages a chance to "put their packages out there" for other people to look at and comment on in a structured manner. See https://wiki.ubuntu.com/MOTU/Packages/REVU [22:34] Legendario: buildd = build daemon, software which does the automatic building of binary debs [22:36] geser, so the REVU builds it againg and place it on the repo, right? [22:36] Legendario: REVU is only for reviewing of source packages [22:37] heya [22:37] REVU is independent from the official repository [22:38] hello is there a way to apply a default gonf key that dosnt require a logout login to apply it to the current user? [22:38] help !paste [22:39] KillerKiwi2005: Set it in the schema? [22:40] geser, so let's see if i got it... what i send to REVU is not the .deb package i've built with pbuilder which is under my pbuilder/result directory... [22:41] geser, it's only all the preparation for it. The building is done againg by the buildd when it reaches the repository [22:41] geser, is that right? [22:41] for gnome people: https://edge.launchpad.net/~ubuntu-gnome-enthusiasts [22:42] hmm [22:42] the ZOMG kde posts on the planet are bad enough [22:42] do we really need one for GNOME? [22:43] soren: you mean set it for the user ? [22:44] soren: that seems wrong... [22:44] KillerKiwi2005: No. In the gconf schema. [22:45] soren: I do gconftool-2 `$GCONF_CONFIG_SOURCE` --makefile-install-rule, and that sets the defaults but the current user is not updated [22:46] KillerKiwi2005: You need to SIGHUP gconfd-2. dh_gconf should take care of all of this. [22:46] KillerKiwi2005: It might take up to 30 seconds before gconfd-2 will feel like looking at the new schema, though. If you're in a hurry, SIGINT it. [22:47] soren: so kill gonf and it will do its thing on restart? [22:48] KillerKiwi2005: dh_gconf calls gconf-schemas, which in turn sends a sighup to all running instances of gconfd-2 (afair). [22:48] KillerKiwi2005: Well... That's not entirely true. [22:48] KillerKiwi2005: dh_gconf instruments the postinst with calls to gconf-schemas. The rest is accurate. [22:49] KillerKiwi2005: Considering that you're asking in this channel, I'm assuming this is in an Ubuntu package. Otherwise, much of what I'm saying doesn't apply. [22:50] soren: lol, actaully no, although I will need to know for ubuntu as well [22:52] is this the place to ask for help with the PPA? [22:52] <_MMA_> the_belgain: #launchpad [22:53] thanks [23:00] Legendario: yes, that's right. To REVU you only upload the source package (.dsc, .diff.gz and .orig.tar.gz) but you should test-build it in your pbuilder so you know if it builds it hits universe [23:04] POX_: Indeed. Thank you very much. Our most duplicated outstanding bug for the last few releases may now be closed. [23:11] anyone know where i can find a tutorial on creating kernel module packages? i would like to package the intel ixgbe drivers for use on feisty... i have tried using dh_build and selecting "kernel module", then configuring debian/rules, then dpkg-buildpackage to build the source .deb, then module-assistant to build the binary... module-assistant fails on me [23:12] ceekay: I'd recommend looking at linux-ubuntu-modules source, searching the wiki, and asking in #ubuntu-kernel, as we don't really support other separate kernel modules (with m-a) very well in Ubuntu. [23:13] heh, I sent him here from -kernel ... [23:13] I would study how other packages handle it, like how the alsa-driver source package generates the alsa-source binary package, which is then usable by module-assistant [23:14] Right. That makes it tricky :) === xhaker is now known as xhaker_ [23:19] somerville32: Thanks. I forgot that it was actually Friday morning my time. :p [23:20] :) [23:23] thanks crimsun and persia ... i know this is pretty non-standard.. i just want to do it the debian/ubuntu way if possible :) [23:24] !bot | Legendario [23:45] persia: The soyuz fix for versioned depends on virtual packages is released. [23:48] ScottK: Saw that. Excellent news for your perl issue. [23:48] Yeah. Saves me having to repack half the perl modules. [23:48] Atually 2, not half [23:49] * persia hopes someone with local sbuild and time will consider testing the patch from Debian bug #456934 for inclusion in our sbuild to match this. [23:49] Debian bug 456934 in sbuild "sbuild: Wrong handling of or'ed build-dependencies" [Important,Open] http://bugs.debian.org/456934 [23:49] tjaalton: Ok. Any particular reason for building all the modules? [23:49] Atually/Actually [23:49] tjaalton: Apart from raw greed for bleeding edge drivers :) [23:49] ScottK: gaphor up too. Thanks for all your help with that bug over three releases :) [23:52] persia: so you got the python dependency fixed? [23:52] persia: You're welcome. Thanks for fixing. [23:52] persia: Can it be backported? === bigon is now known as bigon`