[00:10] slangasek: so for the libpam-smbpass seed change, I need to bzr branch ~ubuntu-core-dev/ubuntu-seeds/ubuntu.hardy/, modify samba-server, ci and push. [00:10] slangasek: and then I also need to upload a new version of ubuntu-meta [00:11] yep [00:22] I'm curious. During a upgrade, a lot of the time, a daemon will be stopped, before popping up the configuration file thing. For instance, in this case... upgrading has stopped apache, and then presented me with a choice to deal with /etc/syslog.conf conflicts. [00:22] cjwatson: I'm assuming the package on revu you pointed to is needed... While I'm not up on font packages, I can certainly have a look in terms of general packaging. [00:22] Does anybody think it should be possible, and preferable, to ask these questions before stopping a service in some fashion? [00:23] preferable, sure, but there's no good way to predict which packages are going to generate conffile prompts [00:23] I personally would prefer upgrades to keep services stopped and in a non functional state for a very short time. [00:23] Why doesn't the service restart happen near the end of the process? [00:23] For instance, postinst. [00:23] fundamentally, to keep the package prerm script as simple as possible [00:24] Oh, ttf-ubuntu-title is already in the archive. [00:25] cjwatson: I found a better fix for the edit dialogue sizing issue, setting word wrap on text labels seems to change their properties somewhat so I set it to have a minimum sizehint policy rather than the dialogue [00:26] TheMuso: it's not absolutely critical, but we've been working with the (new) upstream to get everything merged and it would be nice to have the final result in hardy; a not-installed-by-default font should be fairly non-risky [00:26] Riddell: have you tried it with various languages? [00:26] cjwatson: Right, it looks like the version in the archive and the version in revu are the same, in terms of version number, minus a few 0s. [00:27] I suspect they're actually very different inside [00:27] cjwatson: the messagebox in question_dialog has a question icon already, that's set by QMessageBox.Question [00:27] the version number is bizarre, but it does collapse to 2.0 rather than 0.2 [00:27] Riddell: ah, ok [00:27] cjwatson: yes that Next text is wrong, how do I find out what the debconf translation is called? [00:28] Riddell: it's the ubiquity/imported/go-forward template [00:28] (likewise go-back, etc.; see translate_widgets in gtk_ui.py) [00:29] Riddell: in fact, if you just do self.get_string('next') that should do it [00:29] Riddell: so your usual translation code ought to take care of it if you just leave it alone, unless you've also set that button to read something else [00:30] cjwatson: Right I understand what you're getting at, I'll have a look at that today. [00:30] * TheMuso gets breakfast. [00:45] is msgid usually the english US version of something? [00:45] i don't see a po/en US [00:47] that's customary, yes, in part because historically using English msgids meant you were assured the source string was ascii [00:47] and therefore upwards-compatible with any other national encodings [00:47] ahh [00:47] slangasek: are you sure a new version of ubuntu-meta needs to be uploaded to take into account the samba-server task change ? [00:47] mathiaz: hmm, now that you ask the question, no :) [00:47] so fixing an english problem means changing all the po/ files [00:47] how convenient lol. [00:47] slangasek: I've built a new version of ubuntu-meta, and there isn't change between .99 and .100 [00:48] xtknight: preferably done with a global find-and-replace, yes :) [00:48] mathiaz: then I guess you don't need to upload it \o/ :) [00:48] slangasek: ok - but I think a new version of tasksel should be uploaded [00:48] possibly correct [00:56] cjwatson: the edit dialogue doesn't seem to be tranlated at all, a job for tomorrow i think along with the buttons [01:08] Riddell: are you using bzr? [01:08] Riddell: 'cos that was a long-standing bug up to 1.8.2, but I fixed it in bzr [01:09] mathiaz: I'll do tasksel now [01:09] cjwatson: I'm not, I'll try that [01:14] mathiaz: tasksel only needs to be uploaded when *structural* changes are made to the seeds. Simply adding a package to a seed never requires a tasksel upload. [01:14] cjwatson: right - that's what I was thinking. [01:15] when you said "but I think a new version of tasksel should be uploaded"? :-) [01:15] cjwatson: so I don't need to upload any new package ? [01:15] no, you do not [01:15] samba-server doesn't have an associated metapackage; Launchpad will update Task fields by itself, which is all you need [01:15] cjwatson: oh... awesome :) [01:16] hi mathiaz :) [01:16] cjwatson: I've just pushed my commit to the ubuntu-core-dev branch and all the rest is taken care automatically [01:16] hi emgent [01:17] mathiaz: yep [01:17] uploaded tasksel anyway for mythbuntu's benefit [02:15] Hello [02:15] * slangasek waves [02:17] Riddell: bcm freeze> can you please create a bug report about that? It's probaly not something I can fix in jockey itself, but we can at least investigate it [02:17] cody-somerville: ping [02:17] hi pitti [02:18] LaserJock: looking at -proposed [02:18] emgent, pong [02:23] LaserJock: I don't see flashplugin in -proposed [02:32] pitti: seb128 got if for me [02:33] pitti: however, the verification is done so can we move it to -updates? [02:33] LaserJock: oh, is it? great, yes, I can do that; what was the bug#? [02:34] pitti: bug #173890 [02:34] Launchpad bug 173890 in flashplugin-nonfree "flashplugin-nonfree fails to install due to md5sum mismatch" [Undecided,Confirmed] https://launchpad.net/bugs/173890 [02:34] next time I'm going to use a new bug, that one is a mess [02:34] apparently you did not specify a bug# in the changelo? [02:34] that's baad :/ [02:34] and my clever SRU script doesn't see it that way :) [02:34] well, with as messy as it is I didn't want it [02:35] pitti: ah, sorry about that [02:35] LaserJock: btw, is someone looking at the previous releases? [02:35] not yet [02:35] we can probably stop caring about edgy, but dapper/feisty would be nice [02:35] I'll try to rope somebody into it [02:35] it's a trivial thing [02:36] copied [02:37] awesome, thank you [02:37] pitti, You approved the Abiword FFe, correct? [02:37] thanks belong to you :) [02:37] cody-somerville: no, I said that it's worth considering if there is a tested package for it [02:37] but it's getting really tight [02:38] how does the package in the PPA perform? [02:38] any good/back feedback about it? [02:38] (sorry, I'm pretty much out of the loop at the conf here) [02:39] I've heard good things but I haven't tested it myself yet. [02:39] I just wonder about why nobody seemed to care/notice for 5 months, and suddenly it's generating so much fuss [02:40] pitti: 5 months? [02:40] 2.6 was just released a couple weeks ago [02:41] I mean tracking 2.5 in hardy, etc. [02:41] if it's really that fresh, it's quite natural that we didn't get it? [02:42] of course it's our fault for not tracking all our upstreams :-) [02:43] LaserJock: sorry, I put that badly: I meant, I was interested why it suddenly created so much fuss: whether we ignored 2.6 for so long, or because it is actually so fresh [02:43] it was actually fresh [02:43] I think lazyness [02:44] The split up the source packages in the next release from whats in the archive now [02:44] *They [02:44] cody-somerville: huh? [02:44] LaserJock, They split up the source package [02:44] And change how it is packaged all about [02:45] pitti: upstream also had a blog post on Planet Gnome titled "Ubuntu Sclerosis" which has something to do with it as well [02:45] squeaky wheel and all that [02:45] tsk [02:46] but 2.4 is now long longer maintained, is 2 years old [02:46] and 2.6 fixes something like 7 or so LP bugs and has lots of goodies [02:46] yes, apparently asking for a description of the upstream changes in this new version that no one in Ubuntu has seen prior to 2 weeks from the RC is "OMG bureaucracy" [02:47] I'm pretty sure the changes are listed in the description [02:47] slangasek: yeah, I had a talk with #abiword about that [02:47] cody-somerville: they don't maintain a changelog, which is also helpful :-) [02:47] Ah. [02:48] slangasek, I'm going to upload some experimental changes for the xubuntu seeds in regards to language packages, fyi. [02:48] slangasek: I especially like that 2.6 didn't even exist when you made your first comment [02:49] slangasek: I'm glad we have mind-reading release managers ;-) [02:49] heh [02:49] I hate the ff3 type-ahead matching so much. isn't there anything to be done to make it ONLY search in urls? geeh [02:49] cody-somerville: feel free, just be sure to make them un-experimental by the beginning of next week :-) [02:50] :) [02:50] pitti: Could I bug you to do a sync while you're around? [02:50] * cody-somerville needs several syncs done. :) [02:50] StevenK: well, try :) [02:51] keescook: really? I think it's sooo great [02:51] pitti: Bug #214434 . Puuhhhhlease :-) [02:51] Launchpad bug 214434 in sapwood "Please sync sapwood 3.0.0.debian.1-2 (universe) from Debian unstable (main)" [Undecided,Confirmed] https://launchpad.net/bugs/214434 [02:52] pitti: I'm sure it's good for many people, but I don't type page titles into my urlbar. I type URLs, so I want it to match URLs. :P [02:53] * pitti turns the sync crank [02:54] and there doesn't seem to be any config options for it. :( [02:54] keescook: And it tries to match the URL against page titles? [02:55] StevenK: if I type "f", I have "facebook.com" listed first, and "mattroloff.com" listed second even though clearly it does not start with an "f". [02:56] I want completion, not searching, basically. [02:56] ah [02:56] cody-somerville: if you need an urgent sync, toss it to me [02:56] I was gonna say, urls work fine here [02:56] but yeah, searching, not completion [02:56] mattroloff.com starts with a silent f. So silent it isn't written. :-P [02:56] heh [02:58] bug #214306 is important and bug #214302 would be nice [02:58] Launchpad bug 214306 in ristretto "Sync ristretto 0.18-1 from debian unstable (main)" [Medium,Confirmed] https://launchpad.net/bugs/214306 [02:58] Launchpad bug 214302 in xfce4-mpc-plugin "Sync xfce4-mpc-plugin 0.3.3-1 from debian unstable (main)." [Wishlist,Confirmed] https://launchpad.net/bugs/214302 [02:59] keescook: Maybe you have to set the config option in that cesspool of files firefox calls a configuration system. [03:00] StevenK: according to upstream bug reports, there is no such option. :( [03:00] Those nasty people [03:00] Also, if a core-dev could take care of https://code.edge.launchpad.net/~xubuntu-dev/ubuntu-seeds/xubuntu.hardy/+merge/202 :) [03:01] Are those the experimental changes you warned slangasek about? [03:01] Yes. [03:01] Why? [03:01] Just curious [03:03] The code browser doesn't show the blue lines for some reason, weird. [03:04] it`s 4.04 AM i go to sleep. night people :) [03:04] * cody-somerville waves. [03:05] cody-somerville: You sent me a contentless ping. This is a contentless pong. Please provide a bit of information about what you want and I will respond when I am around. [03:05] Hobbsee, I'm good now, thanks anyhow ;] [03:06] cody-somerville: You do know that's a script that replies? [03:07] StevenK, I'm pretty sure she uses @msg or w/e it is called [03:10] cody-somerville: too bad that xfce.mk will go away; it made things soo nice and consistent [03:10] * Hobbsee looks in [03:10] pitti! [03:10] * Hobbsee sends pitti to bed. [03:10] cody-somerville: 214302 needs ack from ~ubuntu-release [03:10] since it's a new upstream version [03:11] Hobbsee: it's only 9pm for him. ;) [03:11] Hobbsee: huh? it's ten past nine [03:11] pitti, I think I'm a delegate or w/e it is called :S [03:11] and it is mostly a bugfix release anyhow [03:11] cody-somerville: ah, that makes sense [03:15] cody-somerville: zzzzynced [03:15] :) [03:15] * cody-somerville hugs pitti. [03:15] * pitti hugs cody back [03:15] cjwatson: universe is stricter in a lot of cases, yes. I've also found this rather backwards. However, the degree of clue, thoughtfulness, and general responsibility for core-dev and motu are very different. :( [03:15] Also, I was wondering, why isn't it possible to force syncs where upstream has a different original tarball? [03:16] cjwatson: it's one of the reasons i'm quite inactive. [03:16] cody-somerville: because we can't replace files in the archive, only add new ones and obsolete older ones [03:16] cody-somerville: otherwise you'd potentially modify sources in stable releases, etc. [03:16] Okay. [03:17] cody-somerville: the fact that we have to use strictness to make motu's actually think about what we're doing says a lot about the types of people we're giving upload rights to, and that that probably neesd a rethink. [03:18] pitti, As for xfce.mk, I'm hoping I can get debian to adopt it :) [03:18] In the mean time, best to reduce delta due to lack of Xubuntu manpower. [03:19] StevenK, Were you going to do that merge or were you just curious? [03:19] LaserJock: at some point, relevant, clueful people, would do well to sit down and hash this all out, before MOTU falls to the ground. What are your thoughts on using the infrastructure from UDS to do so? [03:19] cody-somerville: The latter. [03:20] Hobbsee: I don't much like UDSs for this sort of thing unless theres good pre- and post-UDS work [03:20] Hobbsee: james_w and I had some thoughts in -motu [03:20] * cody-somerville needs a core dev to do merge #202 on lp :) [03:21] keescook: oh, timezone screwage again. I thought it was 2am or something. [03:21] Hobbsee: also I think he sent an email to ubuntu-bugsquad about patch/debdiff handling [03:22] LaserJock: neither, but i'm meaning the infrastructure - ie, the VOIP and possibly gobby. [03:22] ugg [03:22] gobby is useful, VOIP is not [03:22] LaserJock: then again, if the powers that be decide that there is no problem, then the issue is moot. [03:22] real time chat tends to help. [03:22] Ubuntustudio reports great success with VOIP [03:22] <_MMA_> :P [03:23] _MMA_, has gotten me into it ;] [03:23] <_MMA_> Just for the team anyway. [03:23] the VOIP at UDSs is really nasty [03:23] you can't hear what half the people are saying, people often switch rooms, etc. [03:23] LaserJock: i'm assuming it would be after it. Or before it. [03:23] LaserJock: everyone on headsets, around the world. [03:24] I personally don't like any type of voice communication, but I might be alone in that [03:24] :-) [03:24] * cody-somerville would be up for trying atleast. :) [03:24] LaserJock: heh. [03:24] LaserJock: you're not alone - i hate it too - voice is too high. But i know it tends to work better. [03:25] * LaserJock works in lab all day with no windows, voices scare me :-) [03:25] heh [03:25] <_MMA_> cody-somerville: You'll see UDS VOIP nastiness soon enough. ;) Not as good as Skype unfortunately. [03:25] :) [03:27] each time it seems to get better. So i'll live in hope :P [03:27] like I said in -motu earlier, I think we need to 1) list our developer bug processes 2) figure out what the developer, contributor, and triager should do in each case 3) design processes and policies correspondingly [03:28] 1) can be pre-UDS 2) at UDS and 3) post-UDS [03:28] * _MMA_ remembers Hobbsee in Boston. Wasn't too bad. [03:28] LaserJock: first, i think you need to ascertain if people actually care - particularly those in authority. [03:29] Hobbsee: I don't actually [03:29] LaserJock: no? [03:29] not particularly [03:29] MOTU decide there policies for the most part [03:30] LaserJock: if you've got no traction with those in authority, then you can either fight to get it fixed with them, or raise people up from underneath. You have to see if thsoe people care, too. [03:30] Core devs, well good ideas are usually pretty straightfoward there [03:30] core dev, sure. I believe our problem is with MOTU, is it not? [03:30] not sure [03:30] but honestly there isn't much "authority" per se [03:30] we just do what we decided to do, not much to it [03:31] if you attempt to raise people up, and stir them up into some form of change, you can pay a pretty high personal price for that - so people think you're overreacting, all the time, for eg. [03:31] I'm more concerned with getting enough developers involved [03:31] there's the MC. [03:31] the MC isn't really all that relavent here [03:31] afaik, they're still the governing body - they have no pwoer, but you can't actually bypass them for new processes either. [03:31] the MC doesn't determine MOTU Policy [03:31] i thougth they could still veto it? [03:32] mmm, doubtful [03:32] and even if they could I doubt they would [03:32] LaserJock: i hope you suceed where i haven't, and don't crash and burn. [03:32] well, there's a difference though, honestly [03:32] it's not about what *I'm* doing, it's about what MOTU (and developers in general) is doing [03:33] 12:32 autofocus_new_items = OFF [03:33] oops [03:33] true. But it's not anarchy. That's what they keep telling me, anyway. [03:33] no [03:33] that's why we have MOTU meetings, the TB, etc. [03:34] what I'm saying is that we are not dependent on the MC here [03:34] so policy can change according to the whims of whoever turns up. Yeah :( [03:34] they are there to make sure we "fight fair" more or less and to give useful suggestions [03:34] yes, so people should show up, right? :-) [03:35] indeed. [03:35] The fact that they don't is a sign that it may not be the most effective place for policy decisions. Of course, that's a catch 22 [03:35] people whine on mailing lists that it should have been at a MOTU meeting, yet don't show up to the meeting. [03:35] that's why I'm saying to have pre-UDS,UDS, and post-UDS involvement [03:35] * Hobbsee nods [03:36] LaserJock: so, your thoughts for hwo to proceed are to discuss and vote on things at MOTU meetings, and hope that no one tells you that there is no problem? [03:36] people can tell me there is no problem all they want [03:37] if enough people feel otherwise it doesn't matter [03:37] in generally, for obvious stuff we can just do it and if people complain figure it out [03:37] feel otherwise and turn up to do something, yeah. [03:37] for stuff that needs voting or more discussion, use MOTU Meeting [03:38] and if there's a conflict we can't resolve as MC/TB for guidance [03:38] *ask [03:38] LaserJock: i'm interested in solving the core problems. Patching the outside bits, while temporarily helpful, is not the long turn solution, and tends to just lead to more beuocracy, and procedures changing each time. [03:39] sure [03:39] of course, it's hard to do tha twhen at least 2 members of the MC tell me that there *is* no problem. [03:39] we need a comprehensive look at this stuff [03:39] well, why should that stop you? [03:39] so presumably it's better to talk to those who are enlightened on the problem, assuming there is one, and i'm not mad, [03:39] sure [03:40] and get them to show up and deal with it. [03:40] I often come across problems I see that in reality are not a big deal [03:40] well, i migth attribute it to that too, but the fact that cjwatson and slangasek and such are complaining about the end results of the problem suggest that i'm really not mad, and the problem does exist. [03:40] you gotta be realistic, open to the other side, but if you think you have a good basis for what you're thinking go for it [03:41] and presumably that it's big enough it needs to actually be sorted out, rather than swept under the carpet. [03:41] yep [03:41] LaserJock: hwo many people do you think will speak, if members of the perceived "authority" shoot the idea down? [03:41] ok [03:42] first, you're making this a "us and them" deal [03:42] and second, you're assuming "authority" can shoot us down [03:42] LaserJock: as soon as they say "$person is overreacting", the discussion dies, and no one looks at what person actually said. [03:42] umm, not really [03:42] I've seen more often than not encouragement from the MC to get things done [03:42] at least, the mailing list posts tend to suggest that. [03:42] globally, yes. [03:43] I've talked to I think all members of the MC and beyond about issues I'm seeing [03:43] and pretty much everybody agrees that there are issues we should look at [03:43] certainly, the autority can't shoot things down - but by saying "i don't think this is a problem", then most people will be inclined to believe them. [03:43] heh, you got all of the MC to agree to that? wow [03:43] not if those people also see the issue [03:45] LaserJock: right, so all the issues stuff needs to come from you and others, then. [03:45] that works. [03:45] so, in general, keep it open, keep it civil, and if it's something that needs to be done it will get done [03:46] i fail to see hwo my last mailing list post wasn't, and i still got called as overreacting. [03:46] but, whatever. [03:46] so i clearly need to go thru you guys, to avoid that. [03:47] well, to be brutally honest, I think you're tending to exagerate stuff a bit to make sure it's getting attention [03:47] there are clearly issues, but the sky is not falling and MOTU will survive [03:47] despite what I sometimes think myself [03:47] i'd hope MOTU will survive. [03:48] and i think that it will survive in some form. [03:48] But i don't think the form will be overly useful, as it'll be filled with beuarocracy, and it will be very hard to get things done. [03:48] nobody likes beuarocracy [03:49] we just don't know always how to fix it [03:49] whether that actually translates to that the sky has fallen, if MOTU can't effectively get stuff done...that's an exercise for the reader. [03:50] I worry far more about apathy than I do authority [03:50] i find that due to the apathy, the majority of the decisions get left to the authority. [03:50] right [03:51] Which is usually bad and wrong, but it happens. [03:51] though authority is often a nice thing to have, it does tend to create "us and them" scenarios that aren't all that good for a community like ours [03:52] so, we have no one to blame but ourselves, no? [03:52] if our apathy leads down a road we'd rather not go [03:53] indeed. [03:53] but all is not lost certainly :-) [03:53] although i don't think the blame game is helpful [03:53] well, all is lost if we keep not caring. [03:53] or at least, close to lost [03:53] no, but what it *does* mean is that we are also the solution, right? [03:54] you guys are the solution, sure. [03:54] and you, if you like [03:55] hah [03:56] Hobbsee: Yes, you too if you want to be. [03:57] This _feels_ like a scene out of the movie when the bad guys corner the good guy and try to convince the good guy to join them, ;] [03:58] Yeah but there is no evil intent here. [03:58] cody-somerville: heh [03:58] "Come to the dark side Sarah" [03:59] perhaps we need some sort of MOTU or Developer CoC [03:59] a bit more technical like [04:00] "you shall employ thought before uploading"? [04:00] something like that [04:00] though perhaps better phrasing ;-) [04:00] haha yeah. [04:00] "You shall not peddle crack at your fellow developers" [04:01] * StevenK looks at jtisme [04:01] Doh [04:01] * StevenK looks at jdong [04:02] haha [04:02] once upon a time I started writing a MOTU Manifesto, perhaps I should finish that [04:04] I think a first good step would be for us to write down the perceived problems in a wiki problem with solid examples. [04:04] * cody-somerville is having a hard time identifying what exactly the issue is. :/ [04:05] yikes, that can get a bit messy [04:05] "problems" are generally general :-) [04:05] it's hard to pull out specifics a lot of the time, especially if it's social issues [04:06] cody-somerville: but I do agree we need to probably do some of that [04:06] Maybe we need to employ CPS? [04:06] I just don't think a "Everything that's wrong with MOTU" wiki page is going to end well ;-) [04:07] I think we can get a long ways by not asking "what are we doing wrong" but rather "how do we want to work" and "what are we doing right?" [04:07] I don't necessarily disagree with you. [04:09] I'm just saying that because I think I've been involved in at least like 2 or 3 "what's wrong with MOTU" wiki pages ;-) [04:09] Right. [04:09] and by the time your done hashing out the problems you don't have any energy for the solution and you feel depressed :-) [04:09] at least that's how it is for me [04:10] I think maybe the problem is the lack of application of creative problem solving skills (TM). [04:10] There was a book I started reading once that described a number of CPS techniques that I think would be helpful to us. [04:10] I'll make a note to stop at the library for it. [04:10] I think the two top social issues are apathy and distrust [04:11] So the objective is absolving that apathy and distrust? [04:11] which leads to sloppiness and beaurocracy [04:11] * cody-somerville nods. [04:12] well, somewhat [04:12] So, maybe some team building exercises would be helpful? [04:12] we certainly don't want technical things like workflows, processes, policies to make things worse [04:12] But how do you plan to measure success? [04:12] but rather better [04:13] well, in some ways it's a bit more fundamental, IMO [04:13] how can you do a team building exercise when 3/4 of the team doesn't show up? [04:14] So the problem is getting people to show up? [04:14] I tried to get some MOTU Hug Days going so we could do a last minute team QA blitz and we got 0 responses to the -motu email [04:15] hmm [04:15] as far as people signing up, nixternal did reply with some suggestions [04:16] * cody-somerville wonders why he didn't reply :/ [04:16] you were probably too busy [04:17] I kinda feel like we have too much going on [04:17] we have a bazillon teams and "programs", etc. [04:17] * cody-somerville is pretty heavily loaded with work + leading Xubuntu. [04:18] * TheMuso would have likely been one who replied to that, if he was still only a community member. [04:19] right, lots of reasons not to [04:19] so my thinking is perhaps, using a sports phrase, we need to get back to fundamentals [04:20] look at what *has* to be done and work out good workflows/policies/processes for them [04:20] in particular looking at developers [04:21] look at things that are working and things that aren't [04:21] if they aren't then we can drop them until we can make them work [04:21] What do you feel currently isn't working? [04:22] mentoring [04:22] sponsorship to some degree [04:22] and the "red tape" part is a lot of it [04:23] developers don't want to be paper pushers, right? [04:23] they want to work on fun stuff [04:23] I've never had a problem having to do red tape, as it keeps me on my toes, thinking about why I am doing what I'm doing. [04:24] And it makes me look more carefully at anything I sponsor, and makes me think about how careful I must be with my own work. [04:24] I agree [04:24] For me, anything to keep me aware of how careful we as developers need to be is a good thing. [04:24] but there's a line [04:24] we don't want more than necessary [04:25] Yes I agree, but we need to keep all developers, whether MOTU or not, to be thinking about just what they are doing, and what damage could result as a result. [04:25] for sure [04:25] gah [04:25] but it's kinda like those popup dialog boxes [04:25] I don't find that my work is inhibited by the current processes. [04:25] people can either give up or just click through [04:26] cody-somerville: you were just complaining about seed changes ;-) [04:26] Well, that'll be changed soon [04:26] I do need someone to still do that merge for me [04:27] * TheMuso will. [04:28] Thanks TheMuso :) [04:29] LaserJock: is the issue more about not having enough contributors, or about contributions not being high enough quality? [04:30] well, both [04:31] my more immediate concern is consistency [04:32] I need to know that the developers around me are working to the same standards and are being held to the same standards [04:32] I need to know that the bug triagers are going to do things the way I expect them to [04:33] I need to know that what I do matters [04:33] I'm guessing those are key elements [04:33] * cody-somerville nods. [04:34] so I think the problem is in a lot of ways, not about the specific processes (they by-and-large work well) but that we're all on the same page [04:34] cody-somerville: I have a text conflict in supported... [04:34] one thing I've thought about doing is having cheatsheets [04:34] well, it's good that you don't feel lack of contributors is a big issue - you have some muscle (maybe not as much as you want), but it needs better training to be effective [04:34] a developer cheatsheet, a triager cheatsheet, etc. [04:35] where you can see at a glance what you need to do [04:35] * bryce nods [04:35] because frankly I spend more time looking up the current wiki page standards than I do actually doing the work [04:35] I have had similar problems with both X bug reporting and triaging, so have been adopting the documentation/training approach, which seems to have had some good results so far [04:36] TheMuso, It won't merge cleanly? [04:36] bryce: yeah, we've done quite a bit of that and it's helped [04:36] I put some triager docs and bug reporter docs at http://wiki.ubuntu.com/X, and refer people to those regularly. I notice people have started catching on [04:36] but I think we're maybe trying to bite of more than we can chew a lot of times [04:36] Universe especially is tricky [04:36] cody-somerville: No. I did a bzr pull from the core-dev tree and then merged yours, and there is a conflict. [04:37] because of the shear numbers of packages and upstreams [04:37] I'm trying to turn "When reporting an X bug, *always* attach your Xorg.0.log" into an adage that everyone knows without thinking. ;-) That alone would save me *soo* much time. [04:37] TheMuso, I imagine it has to do with emacs? [04:37] LaserJock: yeah I got the impression from the discussion scrollback that "needs to be more fun" and "needs to be less overwhelming" are also issues [04:37] cody-somerville: Yes. [04:38] bryce: I agree completely [04:38] TheMuso, One moment please. [04:38] cody-somerville: Sure. [04:39] yeah, I can sympathize with the overwelming amounts of work - I think it's true for us all. Certainly with X alone there seems to be no end of bugs. But I look at firefox or openoffice and things don't seem quite so bad. ;-) [04:39] bryce: there are something around 200 pages on the wiki for development [04:40] between MOTU and UbuntuDevelopment related pages [04:40] the launchpad integration for bug reporting should make all of that automatic. [04:41] there are a lot of helpful books and so for dealing with overwhelming workloads, that generally recommend some combination of prioritizing, limiting scope of commitments/responsibilities, etc. [04:41] bryce: that's basically what I'm getting at [04:41] I'd like to be able to go to the control panel area, and hit Report a Bug or something... and have it automatically attach all files that might even be remotely relevent. [04:41] LaserJock: information overload? [04:41] figure out what we *have* to do, prioritize it, and then have fun working on it [04:41] wasabi: I would love to see that too [04:41] right [04:42] wasabi: I was JUST thinking the same thing! [04:42] LaserJock: for volunteer projects especially, morale and fun are the most precious resource [04:42] one thing I think we can do is to develop more pushing of certain things to Debian [04:42] like the low priority stuff [04:42] Well, the crash report tool does most of it. Just need a way to file arbitrary bugs using it... and then include the X config/log. [04:42] I think a bug reporting feature should be added to every distro; it will collect relevant info based on the error you're reporting [04:42] And maybe dmesg output, the last hours logs, except security stuff. [04:42] LaserJock: a good plan, lots of people, useful docs, and plenty of morale can do just about anything ;-) [04:43] apport perhaps can be extended to accomplish it. [04:43] Heck, I'd like to be able to hit Report a Bug in a running application, and have a core dump attached, even if it's not crashed. [04:43] minimizing divergence from Debian I think is also a key for MOTU [04:44] Oh hey, apport is launched for Report a Problem now. Didn't know that. [04:44] the more we have Debian maintain stuff the better off it is for both Debian and Ubuntu [04:44] Has anyone else experienced bug #214898? lol [04:44] Launchpad bug 214898 in ubuntu "latest batch of updates deleted all user docs, pics, music and vids?" [Undecided,New] https://launchpad.net/bugs/214898 [04:44] Heck, somebody just needs to add X logs to apport files. [04:45] TheMuso, Try now :) [04:45] bryce: what do you think of the idea of having like cheatsheets for developers and bug workers? [04:45] cody-somerville: Ok jus a sec [04:45] Apport is nice for bugs related to crashes... but there's a lot of bugs NOT related to crashes that require the same info [04:45] LaserJock: I think it's a *great* idea [04:45] temugen: hit Report a Problem, from any launchpad integrated application. [04:46] bryce: I think we tend to use the same documentation for learning as for reference, which is tiresome [04:46] LaserJock: I did what I guess could be described as a cheatsheet here - https://wiki.ubuntu.com/X/Reporting [04:46] I don't want to wade through how to make a debdiff, I want to just remember which flags/tags/subs to flip :-) [04:47] cody-somerville: Much better, committed. [04:47] The Gnome Main Menu, under System, needs Report a Problem as a link. [04:47] Has this been brought up? [04:47] LaserJock: I don't know how many bug reporters actually read that (I wish I could make them all read it), but it's proven useful for newbie / occasional Xorg bug triagers. Gives them a paint-by-numbers way of making sure the bug report has the needed info. [04:47] i'm sure there's something relevant about pitti's mega-killall script. [04:48] Dude. apport already does all this. This problem is already solved. [04:48] LaserJock: yeah I also have my own packaging "cheatsheet" for how to do everything from rolling packages, to doing SRU's, sponsoring other people's uploads, etc. I refer to it a lot. [04:48] X11 needs to put a file in /usr/share/apport/general-hooks that attachs the X logs. [04:48] That's all. [04:49] bryce: in wiki page format? [04:49] LaserJock: nah, just a text file [04:49] I was wondering if I could grab one of those LaTeX cheatsheets of the net and make PDFs [04:50] LaserJock: it started as just my personal notes when learning packaging, and grew... [04:50] bryce: ... in fact you already have a hook for displayconfig. You know this right? :) [04:50] wasabi: indeed [04:50] Move it to general-hooks. [04:50] =) [04:51] And we can just add a Report a Bug option to the main menu, which should be there anyways. Use that to promote bug filing. [04:51] Tada. [04:51] wasabi: I think the issue is that I stuck it against xorg-server but needs to be against xorg-server-core or something. pitti gave me a suggestion, I just haven't had time to do it [04:51] well, there was concern about using apport for general bug reporting in stable releases [04:51] (and yes, I've probably spent 10x that time asking people to attach stuff in the meantime...) [04:51] getting flooded, etc. [04:51] true [04:52] LaserJock: Is that a real concern? Does that really happen? [04:52] yes [04:52] that's why it was turned off for Gutsy [04:52] to be honest, I kind of wonder if the bug flood from apport is part of what did in displayconfig-gtk [04:52] Hmm. [04:52] we use it while in development, but turned it off right before release [04:52] When we ship such buggy apps... [04:52] lol [04:52] just staying on top of the bug reports was consuming most of the development energy, so other important and necessary coding never got done [04:52] Somehow I think that needs to be solveable in other ways. [04:52] * cody-somerville coughs something about displayconfig-gtk. [04:52] cody-somerville: obviously we just nee perfect code [04:52] I'm not sure what those other ways might be. [04:53] MS probably has a department handling the automatic error reporting in Vista/XP [04:54] !register [04:54] By default, only registered users can send private messages - Information about registering your nickname: http://freenode.net/faq.shtml#userregistration - Type « /nick » to select your nickname. [04:54] If you're worried about crash reports, do we not have anyway to do automatic duplicate detection? [04:54] If you'r worried about everybody leaving messages in easy to make bug reports... that's different. [04:55] sometimes I find with bugs reported via apport, the user is kind of doing a "fire and forget", and it can be hard to get followup for any additional info, or to test fixes or whatever [04:55] crashes should be easily duplicate checkable. [04:55] (I would sort of like it if apport-reported bugs could be kept separate from "regular" bugs. But anyway.) [04:55] They are tagged, no? [04:56] bryce: the fire-and-forget issue would be huge if we had apport on for stable releases [04:56] that's one thing I don't like about it [04:56] I have a heck of a time getting responses from reporters [04:56] you dont even need a LP acct to report with apport do you? [04:57] if they could be auto-duped they might be more useful, since it'd give you a measure of the "intensity" of a given crash bug. [04:57] TheMuso, What do you think of me adding ps3pf-utils to the desktop seeds for PowerPC archs (since it is commonly used on the PS3)? [04:57] I thought so, but I've never actually used it so I don't know [04:57] I'd say just ignore no responses. Is it that big of a deal? [04:58] "Sorry, need more info." Forget about it. [04:58] wasabi: well, it certainly can be [04:58] For the 50 you do that for, maybe 3 actually are obvious from the stack trace. [04:58] cody-somerville: Its up to you, but I don't see a problem in doing so. [04:58] and that's 3 bugs solved. [04:58] we are drowning in bugs already [04:58] TheMuso, Will you merge my seeds again once I push it? [04:58] Well, I'd rather have the information recorded than not at all... I guess. [04:58] yeah, it's not hard to deal with non-responders, but it's a time thing... time I could spend doing other stuff [04:59] wasabi: well, ideally I would to [04:59] but when 90% of my bugs are usless apport reports and I can't get a response from anybody (worst case scenario) then it is a problem [04:59] ABout the dupe detection, is there any solution for that? I mean, if 10,000 people hit the same exact crash, the stack trace is going to be pretty much exactly the same. [04:59] At least among some very small number of them [04:59] slangasek, would you please merge debian-cd w/ http://bazaar.launchpad.net/~mythbuntu/debian-cd/mythbuntu-debiancd/ ? It contains updated artwork. Thnaks [05:00] wasabi: I believe that is being worked on [05:00] Thanks even [05:00] wasabi: like once a bug has so many dupes then it gets turned into a real bug report [05:00] Ahh, yeah. That seems reasonable. [05:00] rather than getting a report for any little crash [05:00] Maybe if the dupe detection just automatically works, it' sjust a matter of sort order. [05:01] TheMuso, Actually, I'll get back to you on it. [05:01] Sort by dupe count desc/apport desc [05:01] wasabi: I think bdmurray uses bug-buddy to help identify apport dupes [05:01] And work from the top, maybe never getting to the bottom. ;) [05:02] Could automatically mark apport bugs with superceeded packages as 'obsolete', in a way that htey drop off the default list when a new version is released. [05:03] cody-somerville: ok [05:09] Hobbsee: what end result am I complaining about? [05:16] superm1: merged, thanks [05:17] thanks [05:19] hrm... so I rebooted and got compiz... and metacity is no longer my windowmanager??? [05:21] hrm.. how do I tell what my window manager is, I wonder? [05:27] cd [05:27] woops wrong tab [05:27] bryce: how about something like http://laserjock.us/files/ubuntu/uqrc.pdf [05:28] but you know, filled in with stuff :-) [05:28] ** (ck-list-sessions:12271): WARNING **: Failed to get list of seats: Failed to execute program /usr/lib/dbus-1.0/dbus-daemon-launch-helper: Success [05:28] how very, um, annoying [05:31] Failed to execute program <>: Success ? How very curious [05:39] that just means it's trying to print errno [05:39] and expecting it to be non-zero when, evidently, it isn't === fabbione is now known as fabbione-away [06:03] humm, noticed a rather disturbing bug while resizing my disk using the livecd.. somehow sda1 got mounted during the gparted operations, which obviously meant that resize failed [06:04] I guess there's no way for gparted to prevent that? === asac_ is now known as asac [06:07] perhaps the same workarounds employed in ubiquity need to be in a gparted wrapper [06:08] right, also the swap partitions had to be swapoff'ed [06:25] Can someone give me a bit of sed magic that will append a string after a certain string in a group of files? [06:25] or better yet, just one file :) [06:27] cody-somerville: SOmething like 's/orig string/orig string and new string/g' [06:28] The g at the end may not be needed. [06:28] Depending on whether the original string occurs more than once, and whether you want to replace more than one occurance. [06:31] would this work?: sed -i 's//\n\n\n' [06:31] hehe [06:32] cody-somerville: You'd need to litter that with escapes [06:32] sed -e 's^\)^\1\n\n\n^' [06:34] * jdong notes to self: ask StevenK all sed questions [06:34] I'm not sure at all if sed will respect and expand the \n's. It made add literal \n's. [06:34] s/made/may/ [06:35] * cody-somerville will test. [06:36] woot woot [06:49] good morning [07:08] good morning [07:14] Good morning [07:38] what package initially creates the Documents, Pictures, Music folders? [07:38] i'm seeing a Bug 214898 stating that all his default folders contain no files anymore [07:38] https://bugs.launchpad.net/ubuntu/+bug/214898 [07:39] xtknight: xdg-user-dirs I think./ [07:45] TheMuso, thanks [07:45] that's it im pretty sure [08:33] the writing support for the danish language is broken due to the following bug 214969 ... any ideas on the best solution for this? ... probably either the depend on myspell-da or hunspell-da should be removed; is any of those dictionary-types more used than the other? [08:33] Launchpad bug 214969 in language-support-writing-da "language-support-writing-da depends on conflicting packages" [Undecided,New] https://launchpad.net/bugs/214969 [08:48] hi [08:49] I was wondering if it would be possible to go pass feature freeze and update support in 8.04 to Unicode 5.1 that came out last week? [08:52] moyogo, have the changes in 5.1 been reflected in code yet? [08:53] neftune: update to 5.1 would just mean changing the unicode data files in general [08:55] moyogo, which ones are you talking about specifically? [08:56] gucharmap, glib and probably data files perl or python use [09:03] tjaalton: there? can you explain me why you decided to make a hard depends out of libflashsupport? because of the bug? [09:03] bug 183943 [09:03] Launchpad bug 183943 in flashplugin-nonfree "flashplugin-nonfree should include libflashsupport as a dependency" [Medium,Fix released] https://launchpad.net/bugs/183943 [09:08] asac: yes, so it should work without that now? [09:08] tjaalton: w definitly cannot keep it anymore ... it causes crashes like hell :( [09:08] tjaalton: i am just trying to figure for whom it breaks. [09:08] tjaalton: so far it works for everyone [09:08] i asked [09:09] tjaalton: could you try what i asked on the bug? [09:09] asac: ok.. I had some problems with it back then, so I'll try without the lib now [09:09] tjaalton: thanks [09:10] moyogo, it looks like glib 2.16.3 just added 5.1 support and is already in hardy. python's "unicodedata" module has lagged considerably in the past [09:13] neftune: yeah I just realized glib has :-), so I guess it leaves only some apps and libraries [09:13] bonjour [09:13] neftune: gucharmap would get 5.1 if the unicode data was regenerated when build [09:15] seb128: Are you aware that your last totem upload breaks youtube playing (I get the no plugin for youtube warning) where it used to work just before the update ? [09:16] stgraber: yes, there is several comments about that on launchpad, I doubt I'll have time to look at it before hardy though, youtube plugin require universe plugins anyway, would be nice if somebody checked upstream if they know about the issue [09:21] moyogo, http://bugzilla.gnome.org/show_bug.cgi?id=527238 [09:24] <\sh> asac, the problem is just: without libflashsupport flashplugin will only play sound via alsa , so someone need to setup .asoundrc correctly...and stop pulseaudio, when it's enabled by default (via esd sound mixing)...or with flashsupport you need to adjust your PA stuff to let the flash plugin stream play on your default device...which is hard to accomplish without padevchooser installed by default [09:25] \sh: yeah, but libflashsupport causes crashes ... and those are not rare [09:25] <\sh> asac, I'm running now with flashsupport and pulse since I found out how pulse actually worked...and since then no problems anymore [09:25] \sh: all the people i asked today had no problems with keeping pulseaudio enabled [09:26] \sh: go to youtube video ... wait till sound plays, hit browser back ... hit browser forward, wait till sound plays and repeat [09:26] you will see a crash [09:26] <\sh> (on i386 that is...I have to recheck the situation on amd64 still) [09:26] <\sh> asac, ahhhh [09:26] often it happens after the second time ... sometimes it needs 10 [09:27] <\sh> asac, yes...this i see with latest flashplugin too...I wondered if that was because of the new flashplugin [09:27] no thats flashsupport [09:27] <\sh> groovy [09:27] \sh: can you test if moving away libflashsupport.so really breaks sound for you again [09:27] <\sh> asac, give me a sec [09:27] everyone i asked had problems earlier in this cycle but don't have sissues now [09:29] \sh: i think fixing asoundrc for some users is ok, requiring to disable pulseaudio would be a bad trade :( [09:30] <\sh> well, moving flashsupport out of the way, restarted firefox, gives me no sound in the moment, because .asoundrc is set to use pulse [09:31] <\sh> I would need to set asoundconf back to alsa device != pulse [09:31] <\sh> which I will do now [09:33] \sh: and then you need hardware mixing [09:33] <\sh> Amaranth, yepp. [09:33] <\sh> and this is not working, I'll restart the session to be sure... [09:33] no one has hardware mixing :) [09:33] unless they have an X-Fi or something [09:33] <\sh> Amaranth, nope it worked before without pulse..so it should work now, too [09:33] <\sh> brb [09:34] mvo around ? [09:34] you have a high-end sound card then [09:34] \sh: try plain default setup and at best without your funky bluetooth thing ;) [09:34] (for now) [09:35] kagou: yes [09:35] Amaranth: straber could play multiple streams in flash without flashsupport ... and has no hardware mixing according to him [09:35] stgraber that is [09:35] in flash, sure [09:36] because pulseaudio will release the device after being idle for a bit [09:36] then dmix takes over [09:36] mvo, can i assign you to Bug #208419 specially the nautilus-share problem. We need to add/or modify a patch to install libpam-smbpass in the same time that samba [09:36] Launchpad bug 208419 in nautilus-share "Integrate samba password in PAM" [Medium,Confirmed] https://launchpad.net/bugs/208419 [09:36] then where is the problem? [09:36] and ... is that problem worse than crashing firefox :/ [09:37] <\sh> asac, layer 8;) [09:37] <\sh> asac, asoundconf is not enough...I had the change the default sound card in gnome settings [09:37] <\sh> to [09:37] <\sh> o [09:37] <\sh> now I have sound [09:37] asac: you can't play a sound in anything else at the same time as flash [09:37] Amaranth: stgraber could use mplayer at the same time iirc [09:37] <\sh> pressing back button, pressing forward...no crash and sound [09:37] <\sh> but only via alsa [09:38] because if flash is playing pulseaudio can't get its exclusive grab of the card and if pulseaudio is playing flash is blocked [09:38] * \sh can use rhythmbox at the same time too :) via usb headset mixer [09:38] http://www.stgraber.org/download/pulse.png, that's firefox without libflashsupport + mplayer using esd + mplayer using pulse [09:38] \sh: that's a different device, cheater :P [09:38] <\sh> Amaranth, nope...I have both sounds on my headset [09:39] kagou: ok, I will check the bug out [09:39] stgraber: err, that says flash is going through pulse [09:39] so either asoundrc or libflashsupport [09:39] Amaranth: I don't have a .asoundrc [09:39] perhaps it is the default then [09:39] and he doesn't have flashsupport. [09:40] maybe latest flash update uses pulse now? [09:40] but flash is going through pulse which means either flash grew pulse support without telling about it or alsa is routing things through pulse [09:40] <\sh> anyways...i have alsa now, and I can listen to two different sounds from two different apps at the same time [09:40] * asac digging in the dark [09:40] <\sh> no PA daemon running, esd totally shut down in sound settings [09:40] thanks mvo [09:40] \sh: can you see if you can get it working without soundrc? [09:40] \sh: then you are not using pulse at all [09:41] but with pulse enabled? [09:41] \sh: so that doesn't count either :P [09:41] like what stgraber has? [09:41] maybe flash can use ESD and Pulse has its ESD compatibility stuff ? [09:41] asac: uh, my ff3-session got busted somehow, it just keeps on loading new windows.. is this something new with beta5?-) [09:41] <\sh> ok now pulse started... [09:41] <\sh> asoundrc is set to use my usb headset [09:41] <\sh> (which I need, I don#t have speakers on my pc here) [09:42] tjaalton: not sure what you describe :) [09:42] <\sh> so...ff3+flash is using still alsa... [09:43] <\sh> now setting back to pulse output [09:43] asac: started ff, it first loads the session ok (~10 windows with a lot of tabs), then a second later it loads it again, and again until the machine is so slow that the process slows down :) [09:43] \sh: close firefox and rhythmbox, set to pulse, open rhythmbox, open firefox [09:44] <\sh> Amaranth, I did :) [09:44] either flash will magically be going through pulse now or you get no sound [09:44] stgraber: can you test this as well ^^ [09:44] ugh, last penguin.swf post was in december [09:45] <\sh> WTF? [09:45] <\sh> my usbheadset device disappeared now while switching to pulse [09:45] <\sh> grmpf..restarting session [09:45] would starting firefox with esd wrapper help if we detect pulse? [09:45] and no changelog [09:46] asac: isn't flash a separate process now? [09:46] i thought it used XEmbed [09:46] isn't wrapper inherited to child processes? [09:47] slangasek: I'd like to get and ACK on bug 213991 before I upload. [09:47] Launchpad bug 213991 in libvirt "libvirt should allow for 'model' of nic" [Wishlist,New] https://launchpad.net/bugs/213991 [09:47] asac: working [09:47] * soren goes to bed [09:47] asac: the esd wrapper is for things using /dev/dsp, not alsa [09:47] asac: rhythmbox + firefox, both going through pulse [09:47] right ;) [09:47] * Amaranth wishes adobe published a changelog somewhere [09:47] (but without flashsupport i guess) [09:48] no flashsupport indeed [09:48] actually i think if we find that flash goes through pulse i don't see why flashsupport would be required [09:48] <\sh> Amaranth, ok...back to pulse...flash tries to connect to pulse but is not establishing the connection or crashes...you can see that with padevchooser -> volume control [09:48] maybe you have a system config to route alsa through pulse and flash is less stupid now [09:48] <\sh> (without flashsupport in /usr/lib) [09:49] if you don't have a system config to route alsa through pulse i'd say hardy is pretty broken, actually :P [09:49] \sh: please be sure that everything is clean and you can still reproduce. maybe pulse server was busted after playing around that much. [09:49] Amaranth: mplayer -ao alsa fails so I guess I don't [09:49] (i don't trust pulse that much) [09:50] I guess we'd rather preconfigure everything with support to use pulse and leave everything else nonworking [09:53] Amaranth: if that means no sound in flash its a pretty tough decision [09:56] tjaalton: do you use a special extension to restore your session? [09:57] <\sh> asac, nope...I restarted pulse [09:57] <\sh> asac, rhythmbox goes through pulse now [09:58] <\sh> and flash tries to establish a connection on pulse...but doesn't succeed...and flashsupport is now in /root and not in /usr/lib anymore [09:58] asac: no, it's built-in [09:59] asac: I think ff2 already had the feature, before you needed an extension for it [09:59] \sh: but if flash tries to establish a connetion to pulse it isn't it a pulse bug? [09:59] <\sh> asac, nope [10:00] "before that.." [10:00] \sh: why? i mean it tries to use pulse [10:00] <\sh> asac, I wonder what magic adobe implemented now to try at least a pulse connection [10:01] <\sh> asac, yes...but I think somethings changed in flashplugin without knowing it...I didn't see that behaviour with 115 but now with 128 [10:01] <\sh> aeh 127 [10:01] yeah. but why doesn't pulse accept the connection. thats what bothers me [10:01] stgraber: does it work for you? [10:01] <\sh> asac, because something's wrong in flashplugin...it's not pulse [10:01] \sh: maybe you tweaked pulse in some way? [10:02] <\sh> asac, nope [10:02] asac: yep, that's why I get the "Adobe Flash: Flash Animation" in pavucontrol [10:02] <\sh> asac, I need flash at work :) so I'm not a friend of breaking my working environment [10:02] \sh: i386 or amd64 ? [10:02] <\sh> stgraber, i386 right now...I test with amd64 this evening [10:02] I'm running 64 [10:02] asac: I know what it's doing.. it opened all the windows that I had opened at some time during the last session. it loaded 87 windows in total :) [10:03] <\sh> stgraber, do you see this flashing connection thingy in volume-control of padevchooser? [10:03] tjaalton: ah. so not a bug? [10:03] <\sh> stgraber, for application streams? [10:03] tjaalton: ah now i read what you mean [10:03] asac: it is, it should load what I had when I closed firefox [10:03] tjaalton: can you reproduce? [10:03] asac: I'll try to reproduce with a clean profile [10:04] tjaalton: maybe that happened when the upgrade was pushed underneath while running firefox? [10:04] \sh: no, I have "Adobe Flash: Flash Animation" there while watching something on youtube [10:05] asac: well, then I had to kill it since "file - quit" didn't work. it loaded fine with beta5 then, but this is probably the first time I've restarted since the update [10:05] <\sh> stgraber, well, not here..I moved libflashsupport.so out of the way, as discussed, and now I have a flashy thingy... [10:05] <\sh> I'm trying to record that now [10:05] <\sh> give me some mins [10:05] tjaalton, did your vdr upload fix the xine frontend ? (i recently got a dvb-t stick, when i tried vdr it didnt wnat to work with xine) [10:05] tjaalton: ok. i'd file it under the "all kind of obscure things can happen if you don't close firefox while upgrading" [10:05] \sh: the only difference I see between your flash and mine is that I'm using nspluginwrapper [10:07] ogra_cmpc: I don't know, never used that myself. It probably needs some fiddling because 64bit FTBFS'd [10:07] ogra_cmpc: I'll check what debian has now [10:08] ah, k, i'll wait then, no pressure or something :) [10:08] asac: I wonder why that is [10:08] Does firefox not keep its own resources in memory? [10:09] Amaranth: it loads things lazily [10:09] Amaranth: so if you don't use a module (mostly chrome files), then upgrade and then try a previously not used feature it will either pull in incompatible modules (like in firefox2) or nothing (like now) [10:10] <\sh> stgraber, asac : http://archive.linux-server.org/flashproblem.ogg [10:10] note: this is not really about shared libs, but about the chrome/*.jar files [10:12] \sh: strange. but i don't see why this is not a pulse bug [10:12] <\sh> asac, because with the old flashplugin this didn't happen... [10:12] well ... the old plash plugin apparnetly didn't try pulse at all [10:12] <\sh> asac, so something changed during the update from 115 to 124 [10:12] <\sh> asac, right :) [10:13] \sh: yours is "alsa playback" mine isn't ... [10:13] that still doesn't tell me that its a flash bug. maybe flash could workaround a bug in pulse. [10:13] <\sh> stgraber, again...everthing is setback to pulse.. [10:14] asac: I'm be much more likely to blame flash than pulse [10:14] <\sh> stgraber, only libflashsupport.so is moved away [10:14] flash doesn't do _anything_ right [10:14] Amaranth: why? [10:14] but pulse does? [10:14] "hmm, lets see how many alsa connections we can open" [10:14] asac: seems that sessionstore.js is reducing in size now that I let firefox start without killing it, so no urgent need to do anything :) [10:14] hehe [10:14] <\sh> stgraber, the alsa playback is triggerd from flash, not from anything else :) [10:15] here I have no mention of alsa in pavucontrol, like if flash is trying to connect using ESD or Pulse and not alsa [10:15] tjaalton: ok thanks. [10:15] (reducing when I close windows) [10:15] <\sh> stgraber, yepp...esd is not running, only PA [10:15] <\sh> stgraber, which is somehow the default when you click on sound settings, esd mixing [10:16] <\sh> stgraber, and this evening I'll have to see how my amd64 is behaving [10:16] Amaranth: if we have an idea what flash does wrong, we could raise that with adobe. but i can't go there claiming that they are broken if we are not sure whats going on on pulse side. [10:16] it seems that flash is using ESD here and my pulse is listening to ESD connection (first time I use pulse, so I assume it's the default) [10:17] (well ill do anyway, but i want to know as much details as possible) [10:17] stgraber: ah, right [10:17] asac: looks like flash can use ESD instead of alsa, then it works with pulse using the ESD compatibility of pulse [10:17] it tries alsa, fails, tries oss, fails, tries esd, gets pulse [10:17] stgraber: ok. so how does it work? [10:17] asac: as Amaranth said [10:17] stgraber: would we need to change anything to make everything work? [10:18] _but_ if pulse is configured to release the device on idle then flash can snag it and break pulse [10:18] asac: I never changed my pulse config, so I'd assume listening on ESD port/socket is default [10:18] Amaranth: ah. so you say flash would end up in alsa if pulse releases it? [10:18] <\sh> grmpf [10:18] yes [10:18] \sh: tweaked that? [10:19] I'm not sure if we have pulse configured to hold on to the device, the default is to release it after 10 seconds or so [10:19] \sh: can you try : mplayer -ao esd something-here ? [10:19] <\sh> asac, no...this alsa playback is bugging me...why does pulse catch that... [10:19] <\sh> stgraber, yepp [10:19] \sh: this means that esd released your device (like Amaranth said) [10:19] lets figure if thats the default [10:20] <\sh> stgraber, it's connecting to pulse [10:20] <\sh> stgraber, I can see the stream normally...without flashing [10:20] we have: [10:20] ### Automatically suspend sinks/sources that become idle for too long [10:20] load-module module-suspend-on-idle [10:20] is that the config? [10:20] asac: I don't think having pulse locking the device is a good idea (most games won't work if we do so) but flash shouldn't try alsa/oss first and start with pulse (if we can change this behaviour) [10:20] stgraber: ok. but can we reproduce \sh flashing? [10:21] asac: without the ESD part of PA I think so [10:21] asac: Right, we either need to not load that module or somehow make flash use esd first [10:21] stgraber: what do you mean? how did you enable the ESD part? [10:21] I just tried on a newly installed computer and PA is listening for ESD connection by default (ie. mplayer -ao esd /usr/share/sounds/login.wav works) [10:22] he has pulse active and has that module installed :) [10:22] Amaranth: seems to be the default ... [10:22] \sh: ^^ ?? [10:22] pulseaudio-esound-compat [10:22] <\sh> mplayer -ao esd /usr/share/sounds/login.wav <- works :) via pulse [10:23] ah, ubuntu-desktop depends on that package :) [10:23] <\sh> ii pulseaudio-esound-compat 0.9.10-1ubuntu1 PulseAudio ESD compatibility layer [10:23] <\sh> yes.. [10:23] sorry, i don't get why \sh doesn't have that default even though he didn't change any pulse config. what am i missing? [10:23] but flash can't connect to pulse using the esd compat stuff [10:23] that seems to be the problem for \sh [10:23] but not a problem for stgraber [10:23] Amaranth: why does he use compat stuff and stgraber not? [10:24] They both do [10:24] <\sh> asac, diff in amd64 and i386? [10:24] <\sh> asac, nspluginwrapper e.g.? [10:24] ok, so its nspluginwraper vs. not nspluginwrapper? [10:24] i can't imagine why [10:24] They both have the compat stuff and it works with mplayer but doesn't seem to work for \sh with flash [10:24] I can't imagine nspluginwrapper breaking a wire protocol [10:24] evand, cjwatson: small fix to gtk_ui.py committed, please check for sanity [10:25] But I have seen stranger things [10:25] Amaranth: actually in this case it would not be "breaking", but "fixing" [10:25] stgraber: uses nspluginwrapper [10:25] Oh, \sh is the one with x86? [10:25] yeah [10:25] thats even more crazy [10:25] <\sh> Amaranth, right [10:25] \sh: can you run "pacmd" [10:25] <\sh> yes [10:25] \sh: then : list-modules [10:26] \sh: and look for : module-esound-protocol-unix [10:26] <\sh> index: 6 [10:26] <\sh> name: [10:26] <\sh> argument: <> [10:26] <\sh> used: -1 [10:26] <\sh> auto unload: no [10:26] that's what i get too [10:26] \sh: same here ... [10:26] used is weird but it is -1 for everything :P [10:26] <\sh> groovy [10:27] crap don't run 'exit' in there [10:27] it doesn't exit the shell, it exits pulseaudio [10:27] Amaranth: list-modules is : List loaded modules so ... [10:27] use ctrl+d to exit the shell [10:30] Amaranth: maybe i missed it, but does it work for you? (in ephy obviously :)) [10:30] i don't have ephy installed :P [10:30] i was just about to test this [10:30] * asac bows down [10:31] i still have flash 124 :/ [10:31] ephy will be cool again when it uses webkit, right now i don't see a use :) [10:31] TheMuso: then feel free to have a look at bug #203429 [10:31] ok, so how should I test this pulse/ff/flash -madness? [10:31] Launchpad bug 203429 in initramfs-tools "resume script missing functions" [Undecided,Fix released] https://launchpad.net/bugs/203429 [10:31] whoops [10:31] 11538 jldugger 20 0 687m 72m 22m S 190 7.2 13:21.50 pluginappletvie [10:32] ok, what version of flash should i have? [10:32] doesn't matter, works with the one i have [10:32] oh, i need to move libflashsupport aside [10:33] duh [10:33] Amaranth: 124 [10:33] tjaalton: use the latest, use default setting for everything pulse related ... then see if rhythmbox works together with flash sound [10:34] ALSA lib pcm_dmix.c:874:(snd_pcm_dmix_open) unable to open slave [10:34] about a bazillion times [10:34] no sound, obviously... [10:34] you probably already have an alsa software running (rhythmbox ?) [10:34] Amaranth, thats exactly the error libflsashsupport works around :) [10:34] ogra_cmpc: I know :P [10:35] stgraber: No, I have pulse running [10:35] The whole point is to get flash talking to pulse :P [10:35] wont work without libflashsupport [10:35] ogra_cmpc: yesterday you said that flashssuport just cares for open file descriptors ;) [10:35] ogra_cmpc: stgraber's system apparently says otherwise [10:35] asac, exactly ... and the message above is one of the symptoms [10:36] and libflashsupport causes crashes [10:36] so 3 systems, 3 different outcomes [10:36] this sucks :P [10:36] well, flashsupport crashes for everyone ;) [10:36] ogra_cmpc: here flash just uses ESD and everything works fine (tm) [10:36] Amaranth, libflashsupport worked fine for three releases in ltsp [10:37] stgraber, yeah, thst indeed different [10:37] \sh gets a pulse connection error, I get an alsa connection error, stgraber gets everything working [10:37] ogra_cmpc: that doesn't matter. and its likely that its due to a race/deadlock so you might just not have seen this [10:37] i think there is an env variable to make it choose which to use by default [10:37] asac: ok, not using any .asoundrc tricks, "software sound mixing" is on, and the rest should be default as well. If I start firefox/flash first all is well, but if I leave RB running and restart ff, the flash clips stay grey [10:37] asac, how about setting the good old DSP environment variable in /etc ? [10:37] and poiint that to esd [10:38] ogra_cmpc: the wrapper only works for oss according to Amaranth [10:38] iwj took that out a while ago, how about trying if it works now :) [10:38] i thought, anyway :/ [10:38] well, how did stgraber then get flash talk to esd ? [10:38] you can test it i guess, by manually running firefox wrapped [10:39] if that works id consider to wrap firefox i ncase i can detect pulse enabled [10:39] padsp certainly doesn't help [10:39] what is the esd one? [10:39] esddsp? [10:39] thats what i don't know and nobody could tell me :( ... maybe just esd /usr/lib/firefox-3.0b5/firefox ? [10:39] right [10:40] rhythmbox is playing 'mad' music :P [10:40] ogra_cmpc: i have no esddsp [10:40] esddsp is corret [10:40] fits the mood [10:40] do i need a package for that? [10:40] asac: esound-clients [10:40] oh, I just noticed, I'm not running PA using the init.d script but it's gnome that's starting it ... not sure it makes any difference though [10:40] ok ... \sh does it work to run firefox as above? [10:41] asac, esound-clients [10:41] tjaalton: and you? [10:41] <\sh> hmm...via esd? [10:41] but i dont know how trhe dependency chain is here, it might conflict [10:41] so it's running as myself and not as "pulse" [10:41] but esddsp is broken, it tries to preload non-existent files [10:41] asac: sorry, what? [10:41] ERROR: ld.so: object '/usr/lib/libesddsp.so.0' from LD_PRELOAD cannot be preloaded: ignored. [10:41] thats what i get [10:41] eek, indeed [10:42] tjaalton: nevermind ... esddsp is broken too :( [10:42] that lib doesn't exist at all [10:42] damn [10:42] where is it? [10:42] it moved to /usr/lib/esound/libesddsp.so.0 [10:42] change the script [10:42] <\sh> asac, what do I need to do? sry..didn't catch it..doing some real life work too here;) [10:43] run esddsp /usr/lib/firefox-3.0b5/firefox [10:43] \sh: ^^ [10:43] mv .asoundrc.asoundconf .asoundrc.asoundconf.old && asoundconf set-pulseaudio [10:43] no help [10:43] does that chzage something ^^^^ [10:43] * \sh needs to install that first [10:43] ogra_cmpc: that's not what we want :P [10:43] Amaranth, that should be exactly what you want [10:43] can't we fix human beings not caring about sound? [10:43] <\sh> asac, sound and linux is evil...and I wonder when we can fix that [10:44] cant we just fix libflashsupport ? [10:44] ogra_cmpc: feel free. i tried, but didn't succeed [10:44] and warren didn't want to take a look apparently [10:44] ogra_cmpc: no, that breaks pulse [10:44] <\sh> asac, no change with flash thingy on padevchooser volume control [10:45] because it doesn't handle flash opening 1024 connections [10:45] <\sh> same with padsk [10:45] <\sh> padsp even [10:45] Amaranth: can't pulse close idle connections :) ? [10:45] that being asoundconf change [10:46] Amaranth, indeed you need libflashsupport for the handles [10:46] but in that setup it works stable since feisty for ltsp [10:46] <\sh> ogra_cmpc, libflashsupport is not nearly alpha stage regarding adobes comments ,-> [10:46] ok. so our bug is that flash doesn't close asound handles? [10:46] \sh, libflashsupport was abandoned by adobe over a year ago [10:46] \sh: i doubt that adobe really considers flashsupport a solution at all [10:47] its maintained in the pulseaudio source since 6monthjs [10:47] <\sh> well, what I don't understand is, what flash is doing since new version...the old 115 didn't try to connect to pulse somehow...and never showed this behaviour [10:47] * \sh tries to find the damn changelog for flash linux [10:47] \sh: new flash doesn't try for me either [10:48] there is some flash specific env variable to make it prefer esd, i swear [10:48] stgraber: can you dump your envs? [10:48] Amaranth: can't you block the other options somehow? so it ends up trying esd? [10:49] that's what i'm saying [10:49] no i mean to test, block them outside of flsah ;) [10:49] although the more i think about it the more i think that might be been a feature of the original (horrible) libflashsupport [10:49] * asac runs strings on the binary [10:49] <\sh> Amaranth, no env var for me here... [10:49] no, there is no way to block [10:49] asac, you didnt ask in #pulseaudio about http://www.pulseaudio.org/ticket/225 yesterday, right ? [10:50] because it is already blocked, it tries until it isn't blocked :P [10:50] <\sh> and firefoxrc tells me that dsp is none [10:50] asac, i pinged the channel with a question bout it, lets see what the actual maintainer say :) [10:50] \sh: firefoxrc is not considered [10:50] http://ubuntu.pastebin.com/f1f783113 [10:50] should be removed [10:50] asac: ^ [10:51] <\sh> asac, I'm just checking what could lead to something which makes sense [10:51] \sh, all we need is a fix for http://www.pulseaudio.org/ticket/225 [10:52] yeah. what firefoxrc did was to tell firefox start script to start with a wrapper [10:52] fiddling around with esd and other stuff doesnt really get us anywhere [10:52] no luck parsing the strings output of flashplugin binary so far (searching for ESD env) [10:53] asac, forget it, flash alsways used libesd directly, i doubt they fixed that when they switched to alsa support [10:53] FLASH_ALSA_DEVICE [10:54] asac: apparently we are the first to find that, whatever it may be [10:54] thekorn: does launchpadbugs.commentsbase.attachments work for you? [10:55] google knows nothing [10:55] thekorn: I tried it on https://bugs.launchpad.net/ubuntu/+source/acidbase/+bug/135425/comments/1 and the set is empty [10:55] Launchpad bug 135425 in acidbase "Attack names not shown on default install" [Undecided,New] [10:55] Amaranth, not really, else there wouldnt be a bug abo8ut it upstream [10:56] ogra_cmpc: err, i meant FLASH_ALSA_DEVICE [10:56] Amaranth, ah [10:56] Amaranth: the reason why nobody saw that is because the backtrace gives no info about that [10:56] we found the way to reproduce by accident [10:56] and maybe its only uncovered in ffox 3 (which is not yet that wide spread) [10:56] asac: err, i meant FLASH_ALSA_DEVICE [10:56] :P [10:56] oh ;) [10:57] <\sh> what could be the value...export FLASH_ALSA_DEVICE=pulse doesn't fix it ,-> [10:57] there is also a string: SOUND_COMPLETE :) [10:57] thekorn: and it crashes with the HTML connector: http://paste.ubuntu.com/6692/ :-( [10:58] \sh, padsp ? [10:58] \sh: i imagine it is something like 'hw:0' [10:58] <\sh> Amaranth, I wonder if we can work around it when introducing a monitor device or something :) [10:59] <\sh> I mean I can reroute a PA monitor device to a pcm.monitor device :) [10:59] <\sh> which then can be used for recording software not capable of pulse [10:59] \sh, did you try with padsp as device ? [10:59] <\sh> ogra_cmpc, padsp /usr/lib/ff-3.0b5/firefox doesn't change anything.... [10:59] if you route it through various monitors it very likely that yous sound gets async over time [11:00] <\sh> ogra_cmpc, or you mean export FLASH_ALSA_DEVICE=padsp ? [11:00] yes [11:00] <\sh> no change [11:00] <\sh> same behaviour...flashing in pulse but no sound [11:01] <\sh> lunch [11:01] <\sh> bbl [11:05] I get "unsupported version 0 of Verneed record" for svn's /usr/lib/libsvn_wc-1.so.1 [11:05] Anybody with a clue what that means? [11:09] ogra_cmpc: i think i joined the pa channel late. what was the outcome? [11:12] dholbach: bug.comments[x].attachments is always empty in the text mode, because LP's +text does not link an attachment to a comment [11:12] thekorn: ah ok - I just wondered if there was a way to figure out how old a patch is [11:12] cjwatson: how come partition_button_undo is a special case for translations? just because there was an existing string for it and not for the others? [11:13] dholbach: can you please file bug bugreport on the crash in the html-mode, this should work, [11:13] kagou: nautilus-share should be fixed [11:13] thekorn: will do [11:13] you should be able to workaround this by calling b.attachments before b.comments[x].attachemnts [11:15] cjwatson: the partitioner has lost the default mountpoints and partition types, is that deliberate? http://kubuntu.org/~jriddell/tmp/partition-buttons.png [11:23] asac: heh, seems that enabling libflashsupport again didn't change my situation. the flash-videos get hung almost like they did without the lib (when RB is playing in the background) [11:23] asac: hum, refresh sorted that out [11:29] asac, that pulse bug sounds suspiciously like we could just fix it by dropping the second pthread_mutex_lock() from FPI_SoundOutput_FillBuffer [11:29] ogra_cmpc: is it guarded by a !same_thread NOR ui-thread ? [11:30] hmm [11:31] could we drop the FF side here ? [11:31] ogra_cmpc: firefox side? what do yo umean? [11:31] well, there is a pthread_mutex_lock() as well as one issued by libflashsupport [11:32] we only want one, right (at least to teh same mutex) [11:33] ogra_cmpc: which line do you refer to in flashsupport? [11:33] i didnt look at the code, just followed the logic in the bug [11:35] oh [11:35] i was unbder the impression FPI_SoundOutput_FillBuffer comes from libflashsupport [11:35] seems its not [11:37] asac, did anyone try with libflashsupport connected directly to alsa ? [11:38] (needs a recompile of libflashsupport with alsa dep added) [11:38] ogra_cmpc: i tried it [11:38] ah, k [11:38] it crashes even easier [11:38] but in general the same [11:38] yeah, one layer less :( [11:41] <\sh> back [12:02] ogra_cmpc: had a chance to try the kernel I posted yesterday? [12:09] hmm, I just applied the last day or two's worth of hardy updates and my theme (darklooks) started showing black on black tooltips. the colour options in Appearances let me fix it, but what could have reset that? (or at they options that I couldn' previously override?) [12:12] cjwatson: is it a convenient moment to chat about ps3 port stuff? [12:30] amitk, looks good so far ... (beyond tha fact that the kernel takes 5 min to boot (guessiong thats through debug stuff you enabled tohough) and usplash being totally broken) [12:30] * ogra_cmpc hugs amitk [12:31] oh, wait, now it got stuck [12:32] * ogra_cmpc tries agign [12:35] ogra_cmpc: don't enable wireless for now, when n-m asks for access, deny it [12:35] ah [12:35] right, i had the keyring question [12:37] slangasek: i thought it was you, anyway [12:39] amitk, bah, nm as a whole seems broken ... even making a new connection breaks [12:43] thank you mvo :) [12:43] amitk, beyond that its just beautiful :D [12:44] ogra_cmpc: also try the -13 kernel that I am uploading to rookery, it should allow wireless to work too. [12:44] amitk, any idea what causes the NM breakage ? anything i can work on to take load off you ? [12:44] ah, ok [12:46] its funny, your kernel seems to become fster with every boot [12:46] ogra_cmpc: it learns its way around ;) [12:46] the first one took nearly 10 mins ... the subsequent ones about 5 ... now its less than one [12:46] i.e. as fzast as mine [12:47] ogra_cmpc: could be garbage collection on the flash FS? [12:48] <\sh> are the buildds already on manual? [12:49] should'nt be [12:49] <\sh> good [12:52] amitk, hmm, right [13:04] amitk, i'm at my 20th successfull suspend cycle now, i guess we can consider that part working :) [13:05] ogra_cmpc: heh, now to figure out which one of our 1000 patches is breaking it in the Ubuntu kernel ;) [13:06] ouch [13:06] amitk: "git bisect" :-) [13:06] amitk: git-bisect or the equivalent for whatever VCS you're using should help a lot. [13:07] ogra_cmpc: the -13 kernel i sent is a patch on top of stock 2.6.24.2, so atleast we have it working on 2.6.24 [13:07] Mithrandir: yeah... already prepping for git bisect [13:24] slangasek, Once my push mirrors on launchpad, will you merge the xubuntu seeds and rebuild please (with a cherry on top)? :) [13:24] https://code.edge.launchpad.net/~xubuntu-dev/ubuntu-seeds/xubuntu.hardy/+merge/203 [13:26] cody-somerville: hello [13:27] Hello jeromeg [13:27] i was thinking of something the other day [13:27] * cody-somerville nods. [13:27] shouldn't the xfce about dialog in the menu display the version of xfce and ubuntu the system is using ? [13:29] cjwatson: i hate to be wasting your time, but the mythbuntu.hardy seeds don't work with germinate: http://www.pastebin.ca/979505 maybe you can take a look? http://bazaar.launchpad.net/%7Emythbuntu/ubuntu-seeds/mythbuntu.hardy/ is current [13:30] jeromeg, Good point. It already displays the version of Xfce but not Xubuntu. [13:30] cody-somerville: oh ok [13:31] jeromeg, Would you be kind enough to file a bug against xfce4-utils (source package)? [13:31] mvo still around ?! :) [13:31] cody-somerville: i'll [13:32] cody-somerville: i could propose a patch, but i've no idea on how to localize the string [13:33] jeromeg, We only use the numbers after release. [13:34] cody-somerville: sorry, by localize i meant, make it display in french in a french environment, in english, in german in ... [13:34] so that it fits all languages [13:34] "Xubuntu 8.04" is pretty universal :) [13:34] oh ok :) [13:35] i thought something like "Your system runs Xubuntu 8.04." [13:35] kagou: why [13:35] but yours is fine and does not need translating :) [13:35] kagou: yes :) [13:35] jeromeg, Well, how is the current text translated? If it isn't already translated, I wouldn't worry about it. [13:35] i've no idea [13:35] jeromeg, and it is too late to add new strings anyhow even if we wanted it translated [13:35] mvo, as you know nautilus-share, we need to solve Bug #212098 and may be you can create a patch [13:35] Launchpad bug 212098 in nautilus-share ""easy" file sharing not notifying about logout/restart" [High,Confirmed] https://launchpad.net/bugs/212098 [13:36] cody-somerville: i think it was translated in launchpad, but now that xfce moved to universe... [13:36] kagou: I will have a look [13:36] mvo, great. I was thinking about a popup notification (dbus ?!!) [13:36] cody-somerville: for example the logout dialog is not translated anymore, because there is xubuntu patch to add a suspend button [13:37] jeromeg, We should see about tackling that issue for Intrepid. [13:37] cody-somerville: yep [13:38] mvo, your last patch for nautilus-share working great :) Thanks you [13:38] kagou: great, thanks [13:40] amitk, ok, nm working again ... sound is MIA now :) [13:48] http://svcs.cs.pdx.edu/gitweb?p=dolt.git [13:48] ^ slangasek will explode with joy [13:48] jeromeg, As for now, you can just pull the info from /etc/lsb-release and slap it on the top of the paragraph :) [14:08] \sh: http://paste.ubuntu.com/6697/ [14:08] i guess thats what you are seeing [14:09] ogra_cmpc: stgraber: Amaranth: maybe read this as well ^^ [14:09] yeah [14:09] sounds familiar [14:10] so \sh probably had the plugin loaded [14:13] ogra_cmpc: which lock did you suspect? [14:45] evand: I've implemented the resize widget in kde_ui, see also 215131 [14:45] bug 215131 [14:45] Launchpad bug 215131 in ubiquity "resize widget changes for clarity" [Undecided,New] https://launchpad.net/bugs/215131 === Sebast1an is now known as Sebastian === thekorn_ is now known as thekorn [15:00] Riddell: fantastic! Thanks a bunch. Do you think it is ready for Hardy, or should it wait for Intrepid? [15:02] <\sh> asac, what plugin? [15:02] <\sh> ag... [15:02] \sh: from what i understand thats about the PA plugin for alsa [15:03] \sh: "Opened 3 months ago [15:03] oops [15:03] \sh: http://paste.ubuntu.com/6697/ [15:03] <\sh> asac, when it means: pcm.!default { type pulse } [15:03] <\sh> ctl.!default { type pulse } [15:03] <\sh> then yes [15:03] thats what we use in ltsp [15:04] (which works fine apparently) [15:04] <\sh> you need to, when you have apps which are not running with pulse natively [15:04] <\sh> (as I need to run, e.g. recordmydesktop ;)) [15:05] <\sh> asac, will it be fixed, when removing this alsa plugin, or will flash+flashsupport still crash? [15:05] \sh: thats about how flash tries to work without flashsupport [15:05] apparently it works for some though ... which is interesting at least [15:06] \sh: flashsupport is rather inheritantly broken [15:06] <\sh> as I said, sound + linux is sometimes really crappy ;) [15:06] yeah, thats no news :) [15:06] for all of us i guess [15:06] asac, thats very likely HW specific [15:07] cards that can do dmix might provide a free channel for direct flash access aside of pulse [15:07] <\sh> ogra_cmpc, no..I think the HW problems are somewhat solved, the problems are the different sound architectures introduced over time [15:07] \sh, i meant the fact that it works for some [15:07] <\sh> ogra_cmpc,oh, ok :) [15:08] if alsa can do dmix because you have HW that can, it will be possible to run pulse and flash native on alsa alongside [15:08] sadly the cheap cards cant do dmix [15:08] and indeed cheap == majority :P [15:09] if all cards would have dmix support we could get rid of soundservers altogether [15:09] <\sh> well, I would say: integrated cards == majority... [15:09] dmix == software mixing? [15:10] at least that's how I've understood it.. [15:10] * asac ETOOMUCHSOUNDCONFUSION [15:10] :) [15:11] hey slangasek [15:11] http://alsa.opensrc.org/home/w/org/opensrc/alsa/index.php?title=DmixPlugin [15:12] asac, come on, its only three layers [15:13] hum [15:13] <\sh> ogra_cmpc, two layers too much :) [15:14] mvo: pling [15:14] \sh, fully agreed [15:14] <\sh> ogra_cmpc, let's fix it for intrepid ;) [15:14] even though you wont get around kernel vs userspace [15:16] <\sh> ogra_cmpc, of course...use the technique we used when we were young...two tins with a nylon fiber...to transmit soundwaves from a to b ;) [15:17] evand: yeah, it's mostly a copy of the gtk code [15:17] \sh, nowadays you need four cans and two strings ... stereo, you know ... :) [15:18] Riddell: ah, indeed. Just pulled. Mind if I subscribe ubuntu-release to the bug, or can you approve it? [15:18] evand: approve? [15:18] subscribe away [15:18] Riddell: It needs a UI freeze exception, no? [15:18] why does alsa need such cryptic things like asoundrc ... it should just work, shouldn't it? [15:18] <\sh> ogra_cmpc, oh damn...yes :) [15:18] heya [15:19] asac, its should, yes ... asoundrc is a wrokaround and fine tuning tool [15:19] yeah ... still everywhere i look i see "use this or that asoundrc" ... and i don't understand any of those [15:19] but you have working sound ... you jusy introduce an unstable layer through adding pulse [15:20] i still don't get why we did that without checking for regressions in all major apps [15:20] i mean esd was not enabled by default in gutsy? [15:20] <\sh> asac, the problem starts, when you use e.g. a headset via usb...it's add a new soundcard actually...and then you need to point alsa to the correct soundcard to use [15:21] <\sh> asac, with pulse enabled it's a bit easier...now you click and point ;) [15:21] asac, i dont think anyone actually had assigned a task for sound [15:21] which is odd [15:21] ogra_cmpc: so where did this come from? [15:22] upstream [15:22] afaik gnome defaults to pulse [15:22] so we did the switch too [15:22] yes, but why is it enabled by default now? [15:22] it always was [15:22] login and logout sound depend on it [15:22] oh. so things have just improved that much that we now notice that somethign like firefox is broken with esd :) [15:23] asac: do you know if it's deliberate that network manager doesn't do the equivalent of "iwconfig wlan0 ap off" when it switches from wireless to wired? [15:23] and thus weadded flashssupport ) [15:23] Ng: the driver should do the same for wlan0 essid off [15:23] no [15:23] upstream added pulse [15:23] Ng: i think network manager does that [15:23] we followed [15:24] asac: ah interesting, because it's not doing that here on 4965 and so after I switch to wired, the wireless chip is in a loop trying to get on the AP [15:24] ogra_cmpc: but before it was esd. and with esd everything was broken for me all the time [15:24] Ng: yeah. try the linux-backports module for iwl [15:24] asac: I've tried lbm and lum [15:25] hmm, "iwconfig wlan0 essid off" has associated me with a nearby Belkin AP [15:25] I've definitely never associated with that before [15:25] Ng: thats a driver issue for sure [15:25] Ng: check if the module is set to "auto-associate" ... if there is such a parm at all [15:26] Ng: is the bheaviour any different if you turn of wireless completely in applet? [15:26] no such parameter i can see for iwl4965 [15:26] I'll test shortly [15:28] Ng: parm: disable_hw_scan:disable hardware scanning (default 0) (int) [15:28] no idea if that will make you loopy [15:34] asac: I'd rather give up wired than have to specify all my wireless networks by hand ;) [15:42] Ng: he? does that really stop network manager from finding networks? [15:43] asac: oh, I don't actually know, I was guessing based on the description [15:48] Ng: try. for me these HW_XXX things read more like: let the driver do some magic that nobody wants anyway ;) [15:50] asac: hey, I want lots of magic! ;) [15:50] but I will explore more options === cody-somerville_ is now known as cody-somerville [15:59] slangasek, I got disconnected. Did you see my request? [16:00] <\sh> heading for home...cu later [16:02] bye \sh [16:02] mvo: bug 206866 is ubuntu-only since the crash was introduced in compiz' patch 037_fullscreen_stacking_fixes.patch. I heard you're packaging a new upstream release, could you please add a check to the call in plugins/move.c whether "w" is actually valid? This should fix this bug. [16:02] Launchpad bug 206866 in compiz "compiz.real crashed with SIGSEGV in updateWindowAttributes()" [Medium,Triaged] https://launchpad.net/bugs/206866 [16:04] afflux: let me have a look [16:28] mvo: another thing is bug 209216, do you think we should set it to wontfix? [16:28] Launchpad bug 209216 in compiz "Attempted to unregister path (path[0] = org path[1] = freedesktop) which isn't registered" [Low,New] https://launchpad.net/bugs/209216 [16:42] lamont: any idea why firefox 3 translation tarball didn't appear in http://people.ubuntu.com/~lamont/translations/ ... though its in .changes? [16:42] lamont: http://launchpadlibrarian.net/13136097/firefox-3.0_3.0~b5%2Bnobinonly-0ubuntu1_i386.changes [16:42] asac: are translations still being pushed through my people account? [16:43] and no, no clue without going to look. [16:43] lamont: according to carlos yes. [16:43] lamont: xulrunner-1.9 worked, firefox-3.0 didn't for whatever reason. [16:43] lamont: as far as I know, yes [16:43] ok [16:43] lamont: would be really cool to get this sorted before i upload the next update ( so we can test if everything works now) [16:44] lamont: or at least an idea why this isn't working ;) [16:44] pitti is unfortunately away this week [16:46] seb128, where does get gnome the info about screensize from ? my panels seem to think the screen is only 1024x768 while i have 1280x800 atm [16:47] ogra_cmpc: it has no guessed value [16:47] weird [16:47] ogra_cmpc: that usually happens when you have multiple screens [16:47] ogra_cmpc: ie a projector connected to a laptop [16:47] hmm, i only have one LCD panel [16:47] the panel will adapt to the projector resolution [16:48] lol [16:48] what? [16:48] seb128, thanks, lots of hugs for you [16:48] np ;-) [16:49] it hs an external output and obviously changing the resolution for that from 1024x768 to off fixes it [16:49] even though there is nothing connected to it ... weird === davmor2 is now known as davmor2_away [16:55] carlos: Err, we've been uploading translation tarballs with package uploads for years now... Are you serious that you (or someone) still fetches stuff from lamont's home on rookery? [16:56] infinity: we keep them for debugging purposes [16:56] infinity: that's why we only keep last 9-10 days [16:56] slangasek: Hm. Can one fix to hotkey-setup be made? [16:56] slangasek: The init script should be "echo -n 7", not "echo -n 4" [16:56] slangasek: (breaks lid closing on various HP machines) [16:57] mjg59: "lid closing"? resulting in crashes? [16:57] Yes [16:58] bitter [16:58] hum [16:58] carlos: Kay, well I can look into why broke in this case, but I assume that the actual translation workflow (which is, I hope, done via the soyuz upload queue) has not been harmed by this? [16:58] i guess i should try that, because i'm getting these crashes. there's also a bug report somewhere. [16:58] infinity: no, we had a bug there [16:58] and we are trying to confirm that we got those translations [16:59] infinity: if you did any change that may explain the lack of firefox tarball [16:59] you don't need to do anything [17:00] carlos: I can't explain the lack of tarball, no. I'm going to look into it shortlyish. [17:00] infinity: ok, thanks [17:02] mjg59: are you referring to this bug? https://bugs.launchpad.net/ubuntu/+source/acpid/+bug/157691/ [17:02] Launchpad bug 157691 in acpid "Hardy/Gutsy crashes when the lid is closed on a HP 6710b, HP 6510b and HP 2510p" [Undecided,Confirmed] [17:03] Yeah, that'd be it [17:04] No clue why it's filed against acpid [17:05] no clue either. i had a workaround on my normal install, but it'd always freeze when testing LTSP. glad that it gets fixed, though :) [17:07] ?!! [17:07] half of vmware is missing from my disk [17:07] slangasek: #157691 has a debdiff attached now [17:07] an important half? :-) [17:08] LaserJock: the init script, and stuff in /etc [17:08] A sign you should start using virtual manager :) [17:09] ted1: is that that thing that requires processor features I don't have? :p [17:09] Oh, it still runs (mostly) without those :) [17:10] vmware works fine for me [17:10] :) [17:10] since I have a real, genuine, licence :p [17:10] VMWare never worked for me, and virt-manager has worked well once soren sat down and made me actually read the documentation and follow ALL of the steps instead of the ones that looked like fun. [17:11] I've had the most luck with vmware but I'd love to get something faster (as always) [17:11] ted1: what does virt-manager use? kvm? [17:11] laga: Yeah, kvm. [17:11] i usually use virtualbox, but sometimes i have odd problems with it.. mostly pebkac, though ;) [17:19] evand: I think that's my ubiquity hacking for the day done, I made a small change to gtk_ui you might want to check for sanity [17:19] vmware is so tedious and proprietary. It might become slightly more uninteresting if they get their modules into mainline. :-) [17:25] soren: I'm not sure I understand what's being proposed in 213991 - just getting libvirt to support passing an additional option to kvm? [17:26] Wondering what the deal is with having some of the boost packages in universe, specifically libboost-regex-dev and libboost-date-time-dev [17:26] slangasek: Pretty much. To be perfectly honest, it's slightly more involved than that, but essentially you're right. There'll be a small change to virtinst as well to actually make use of it. [17:27] Riddell: looks ok, though I modified your fix for the crash when .disk/info doesn't exist slightly. [17:29] evand: ta [17:29] slangasek, , i'v talk to mvo about Bug #212098 [17:29] Launchpad bug 212098 in nautilus-share ""easy" file sharing not notifying about logout/login" [High,Confirmed] https://launchpad.net/bugs/212098 [17:29] megabyte405: No one has requested to move them? [17:30] megabyte405: But, I'm pretty sure with the ASIO MIR approved on getting the Abiword one finalized, they'd come into main too as they're dependencies of ASIO [17:30] Riddell: thanks for the fixes, much appreciated. [17:30] ted1: OK - we got a tentative "ok" on libasio-dev (Depends on abiword getting updated, which seems pretty likely at this point) which depends on those, does that mean tey'd come in? [17:30] i don't know if we should patch nautilus-share or if we should patch samba+libpam-smbpass for nautification [17:30] ted1: ah, great, you answered my question before I finished typing it :) [17:31] megabyte405: Heh, yeah. How are the Abiword packages coming along? [17:31] ted1: great, I am about to upload a new one that has the tcp backend (read: libasio-dev) enabled, which should put them basically in release form [17:32] megabyte405: what about that scrolling issue? [17:32] Just put up the MIR for libwv-1.2 which I'm hoping should be straightforward, since it's just a library we stopped copying and pasting into the source tarball [17:32] I get something quite similar in Firefox soemtimes [17:32] LaserJock: I am thinking that is abiword showing a bug in Ubuntu (uneducated guess is X), since Martin, another Abi dev, doesn't get it in Fedora, I didn't patch that part of the code in the package, and I don't see it in Windows either [17:33] cody-somerville: merged, and rebuild started [17:33] LaserJock: and your comment adds credence to my suspicion. For what it's worth, I had to scroll quite fast to trigger it - certainly annoying, but not a show stopper. [17:34] Keybuk: yes, I was tickled to see the dolt announcement :) [17:34] Keybuk: unfortunately, so far it only handles "compile", not "link", so the better part of the work is still to be done :) [17:35] kagou: that wasn't me, just the network bouncing my connection, but hello :) [17:35] hehe [17:37] :) [17:38] slangasek: heh [17:39] erm [17:39] You might want to run `apt-get -f install' to correct these. [17:39] The following packages have unmet dependencies: [17:39] udev: Conflicts: volumeid but 117-4ubuntu2 is installed [17:39] E: Unmet dependencies. Try using -f. [17:40] * ogra just tried to set up a pbuilder [17:40] slangasek: Did you see my poke? [17:40] ogra: hmm, I saw that in my early testing [17:41] if you were setting up a pbuilder, why would volumeid be installed? [17:41] no idea [17:42] i installed pbuilder and ran sudo pbuilder create [17:42] no fancy settings or so [17:42] it's Priority: important [17:42] and Task: minimal [17:42] I wonder whether that means debootstrap drags it in? [17:42] that would be it [17:43] mjg59: patch in #157691> this is definitely correct for all intel/ati/radeon systems? [17:44] If you don't pass --variant=buildd to debootstrap, yes, it grabs Priority: important. [17:44] soren: yeah, the libvirt change looks ok then [17:44] slangasek: It's what I actually meant to do the first time, I misread the spec [17:44] * ogra wonders how to configure this funny touchpad on teh new lappie [17:44] slangasek: Thanks very much. [17:44] ogra: should fix itself next time the publisher thuds then [17:45] Is Martin Pitt here? [17:45] Keybuk, i have ülenty other pbuilders so i personally dont mind ... [17:45] mjg59: oh, well, if it's the /spec/, then no possible harm can come from following it, right? ;) [17:45] megabyte405: He's at a conference today [17:45] megabyte405, he's at a conference [17:45] slangasek: Oh, potential for failure and misery, but in principle it's the safest of the options [17:45] ah - well he just commented on an MIR we need for abiword :) [17:46] mjg59: ok, let's get it in and keep an eye out for regressions [17:46] slangasek: If you could upload that, that would be great [17:47] right, grabbing [17:48] geez vbox on this new laptop is faster than native ubuntu on my old one === Mez is now known as Floodbot5 [17:51] mjg59: uploaded === Floodbot5 is now known as Mez [17:53] slangasek: Thanks! [17:57] <\sh> asac, well, I could reproduce the crash of flash this morning on the i386 desktop, but I tried it now with amd64 .. and this works like a charm, using pulse [17:58] asac / carlos : translation mirror stuff fixed-ish. Should be able to find firefox-3.0 in 20080410 now. [17:59] infinity: cool, thanks [18:00] carlos: So, to be clear, that's not used by rosetta anywhere, right? You just use it as a backup/verification source? [18:00] carlos: Actual imports are done via soyuz uploads, I hope...? [18:00] infinity: right [18:00] infinity: yeah [18:00] carlos: Kay. Is there still value in having the backup/verification? [18:00] well, in this case was useful ;-) [18:00] carlos: If there is, we should get it moved to ~ubuntu-archive, so poor lamont can stop owning it. :) [18:01] but I think that I'm the only one using it [18:02] carlos: Well, I'm happy to either ask the distro guys to take ownership of it in ~ubuntu-archive, or kill it off completely, I just don't want lamont maintaining it for another 3 years. :) === ogasawara__ is now known as ogasawara [18:02] infinity: I don't maintain it. :-) [18:02] <\sh> infinity, rock...thx for the wine p-a-s entry...looks like that YokoZar and I will get more bug reports for wine on lpia now :) [18:03] infinity: let me check with jtv and danilo and I will tell you what to do [18:03] lamont: That was kinda my point. :) [18:03] heh [18:03] it's currently maintained through the process of "malevolent neglect" [18:04] haha [18:04] hello everyone I didn't say hi to :-) [18:12] slangasek, It appears the cds are still obese! === slangasek changed the topic of #ubuntu-devel to: Archive: frozen like a penguin paradise | Ubuntu 8.04 LTS Beta released | Development of Ubuntu (not support, not application development on Ubuntu) | #ubuntu for support and general discussion for dapper/edgy/feisty/gutsy, #ubuntu+1 for hardy | #ubuntu-motu for getting involved in development | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs [18:13] Ah, the alternative cds are fine but the live cds are not. [18:13] cody-somerville: due to the same package set as before? I may have failed to account for propagation delays [18:18] slangasek, let me check [18:19] slangasek, right. No change between manifests. [18:32] mjg59: Help me out. I got /sys/devices/virtual/backlight/acpi_video{0,1} [18:33] mjg59: i guess there was supposed to be only one there. I am investigating why my dell changes brightness 2 steps at a time [18:47] Should fixes for bugs like Bug #184084 be left to Intrepid seeing as we are in Final Freeze now? [18:47] Launchpad bug 184084 in checky "Extension description mentions Iceweasel/Icedove/Iceape" [Undecided,Confirmed] https://launchpad.net/bugs/184084 [18:49] jdstrand/kees: Apr 10 08:15:39 mix kernel: [36454.520112] audit(1207836939.319:6): type=1503 operation="inode_permission" requested_mask="::rw" denied_mask="::rw" name="/dev/tty" pid=3491 profile="/usr/sbin/cupsd" namespace="default" [18:49] oh, wait. my bad. [18:49] I think that means it's not asking for mmap anymore === davmor2_away is now known as davmor2 [18:54] lamont: righto! it just wants to spew to the tty. [18:54] denied [18:54] * lamont looks at the actual logfile to make sure that was all [18:55] hi, is the 'gnome panel freezes when clock applet is clicked' occurring again for anyone? [18:55] lamont: hey, has anyone prodded you about helping to get more info for bug #81242? [18:55] Launchpad bug 81242 in postfix "postfix-ldap is linked against gnuTLS" [Medium,Triaged] https://launchpad.net/bugs/81242 [18:55] keescook: yep. [18:56] slangasek: I've been waiting for someone to tell me what to do with it... [18:56] I just link against libldap2 [18:57] lamont: how about what I wrote in the last comment? :) [18:57] happy to switch to another lib... [18:57] (since turning off ldap support would be more likely to get me shot...) [18:57] infinity: thanks a lot! [18:57] lamont: i.e., "please give me stderr from this failing command" [18:57] lamont: actually, a) you link against libldap-2.4-2, b) there is no other lib kthx [18:58] having postfix be the only thing in main linking to a different LDAP implementation would be more likely to get you shot by *me* ;) [18:58] heh [18:58] the comment from upstream is more upstream bitching about the exit()s in the code existing at all - the code should really report an error back to the caller. [18:59] yes [18:59] and, as you say, there are very few exit() calls in the library today, so triggering such a failure could be problematic [18:59] but I need to know /which/ one postfix is triggering, because I don't have time to convert all the uses of exit() before release ... :) [19:00] depending on the user, I expect any of those fatal logs could be it... [19:00] that is, I rather doubt that there is a test case, and that upstream would say "all" [19:00] * Caesar wishes Adobe would learn something about release mangement and versioning for their Linux Flash plugin [19:00] OTOH, what about making gnutls and ssl just be friends? [19:00] gnutls and ldap. [19:01] sort of like making it so that db4.2 and db4.3 can be linked in the same app..... [19:02] um [19:02] this is *not* an issue with gnutls and ssl being loaded at the same time [19:02] or if it is, we need a test case to show why [19:02] other apps are able to get by with both loaded; c.f. apache [19:03] it's definitely not a symbol conflict error - it may be a resource contention error over entropy, but in that case I still need to see the stderr to figure out where the contention is happening [19:08] evand: really my last commit to ubiquity today, might be time to ask slangasek about that map patch you had earlier and upload :) [19:11] Riddell: thanks, and whoops. I hadn't noticed that the change needed to be made in the KDE ui. I actually have one last thing I'd like to try to get in :/, but I should be done with that within the hour. [19:12] I've been a bit disconnected (being at the LF summit, traveling, etc.)... What's the current policy for uploading? Do we ask permission for even critical bugfixes or how does it work right now? [19:12] xhaker: The patch to fix that breaks other things. I'll investigate it at some point, but it's likely to require major restucturing [19:12] soren: 1) read ubuntu-devel-announce, 2) profit? :) [19:13] mjg59: ok, so it is known to behave like that. thanks. [19:14] slangasek: Reading e-mail over this sorry excuse for an internet connection makes me cry, but ok, I'll see what I can squeeze throguh. :) [19:15] soren: maybe just https://lists.ubuntu.com/archives/ubuntu-devel-announce/2008-April/ then? :) [19:15] That's what I'm doing :) [19:18] So I need to "get in touch with ubuntu-release" or do I need to milestone a bug? [19:19] in regards to what? [19:19] you don't need to do either of those things to be able to upload [19:19] True. [19:19] though presumably if you're uploading, the bug should be milestoned in its own right? [19:20] That sounds reasonable. :) [19:20] soren: I find that bribing slangasek helps [19:20] "If you approve this upload, This milestoned bug will be fixed" [19:20] Does it count if I have to milestone a bug first? [19:20] :) [19:20] * slangasek hehs [19:21] I think that vorlon knows about "Activity Log" [19:21] yes...:) [19:21] I was afraid that might be the case. [19:21] slangasek: I will obviously have a pending udev upload soon [19:21] now that I've got vmware working [19:22] and that I've beaten evand for various issues, I can test my fix properly [19:22] Keybuk: great! :) [19:22] * evand rubs the lumps on his head from the abuse. [19:22] (the 117-5 fixed just about everything else, I decided not to wait since volumeid had to go) [19:22] evand: you love it [19:22] hahah [19:24] slangasek: In this particular case, I have an upload that fixes some things, that *should* have been reported and milestoned, but weren't. You want me to file the bug first to have some sort of virtual paper trail? [19:25] soren: if the bug isn't filed yet, don't go to the extra effort until the release team has had a chance to look at it in the queue [19:25] soren: if we need a paperwork trail we can come back to you [19:25] slangasek: Cool. That's what I wanted to know, I suppose. Thanks. [19:34] How long would it approximately take for Ubuntu to have the latest versions of libxml2 & libxslt that were released this week available somewhere (like universe maybe)? I'm a Gnome Planner developer and I am wondering if making those versions required deps would cause much delay in availability of Planner on Ubuntu (because of testing and such). [19:37] Griffon26: at this point, we're pretty much entirely frozen for our release [19:37] they'll be in 8.10 for sure [19:38] Griffon26: until then you might consider having a team PPA for Planner [19:38] * Griffon26 grabs a dictionary [19:38] Griffon26: basically a seperate repository. I can help you go through it if you want. [19:38] I'm an upstream dev. I don't know much about Ubuntu =) [19:39] I was just looking for a ballpark figure. I'm not running Ubuntu myself. [19:39] Griffon26: It's my job to help upstream devs like yourself get squared away on ubuntu [19:40] Griffon26: I'm on my way to lugradio live tomorrow but if you send me a mail (jorge@ubuntu.com) I can answer any questions you might have about ubuntu. I can give you the quick version of everything you might need to know about ubuntu [19:41] slangasek: if I smack a package upload through the queue myself, is that naughty? :p [19:41] Keybuk: probably :) [19:41] ie. readahead-list === cprov is now known as cprov-out [19:42] jcastro: I may only need an answer to the following question, otherwise I'll send you mail. Is there a reasonably easy way for users to get upgrades to packages in between releases? [19:43] Griffon26: a PPA would be the best bet (personal package archive), otherwise between releases you have -updates which is for security and major bugfixes, and -backports. [19:43] I see [19:44] Griffon26: does the user have to build something from source if he wants to use the new version of planner? [19:44] Griffon26: the downside to PPAs is users need to know about them and add them manually. [19:45] laga: I'm assuming they will until you have a binary version of the release that I'll do this weekend. [19:45] well, -security is for security. :P [19:45] -updates will have security and SRU (stable release updates) [19:45] hi keescook! [19:45] we confused ourselves. ;) [19:45] -updates doesn't have security [19:45] * keescook hugs jdstrand [19:46] -security has security [19:46] Griffon26: well, that binary version will happen in 8.10 then which will also have the dependencies. so i guess your new version of planner will also have to go to a PPA [19:46] laga: it's just that if I dep on the new libxml2/libxslt versions, people who build from source on Ubuntu may suddenly need to upgrade libxml2/libxslt as well [19:46] -updates does have security, whenever there is a sync [19:46] Griffon26: unless people build it themselves [19:46] LaserJock: ^ [19:46] Griffon26: yeah, you can also put those deps in the same PPA if you want. [19:46] mjg59, do you happen to know anything about "IDEACO IDC 6680" touchscreens ? [19:46] jdstrand: well, that's kinda iffy [19:46] LaserJock: we ask for a sync of -security to -updates for large packages like the kernel, firefox, etc [19:47] LaserJock: this is to reduce the load on security.ubuntu.com [19:47] Griffon26: i understand the problem.. i guess the PPA is the best thing you can do. i assume that libxml2 and libxslt are rather important packages, so building them from source might not be a good idea [19:47] jdstrand: so there are exceptions ;-) [19:47] LaserJock: IIRC, the admins will periodically sync -security to -updates without us explicitly requesting it also [19:47] ogra: Nope [19:47] ogra: What does it present as? [19:48] keescook: -security also gets new hardware support sometimes ;) [19:48] mjg59, simple input event device, it even works just not in sync with my finger or the stylus [19:48] this just happened recently actually-- I requested a sync for firefox, but they just sync'd it all [19:48] /dev/input/event9 atm [19:49] jdstrand: all of what? [19:49] jcastro: laga: I think I have a good idea of what's possible now. I'll recommend on our list that whoever builds it in Ubuntu also publish a PPA with libxml2/libxslt. [19:49] whatever we uploaded to -security that wasn't in -updates yet (eg mysql) [19:49] Griffon26: ideally, you'd want someone who us a planner enthusiast who is willing to maintain your package in ubuntu, then he/she goes through our developer process. [19:49] but that is rare. :D [19:50] LaserJock: rmadison mysql-server-5.0 | grep gutsy [19:50] yeah, people building Planner from source are rare. Somehow the cross-section between project managers and software engineers is not that large. I wonder why ;-) [19:50] LaserJock: I didn't ask for mysql to be sync'd, yet, there it is [19:50] :) [19:51] Griffon26: if someone is interested send them my way and I'll go through the process with them [19:51] jcastro: will do. Thanks for your help. Both of you. [19:51] jdstrand: odd, I don't think that's policy but they probably had a reason [19:51] LaserJock: pretty sure it all gets back to the load on security.u.c [19:52] sure, but that would be only exceptions to the rul, not the rul [19:52] this is something that has probably evolved over time, I don't make those decisions ;) [19:53] in any case, security goes to -security and SRUs go to -updates [19:53] whatever the archive admins do aftward I don't much care :-) [19:55] slangasek: do universe things ever make it into the release notes? it might be nice to mention the selinux work. [19:57] keescook: I think a "tour of the universe" might be useful come to think of it [19:57] jcastro: I like the sound of that [19:57] "tour of the universe" [19:59] http://fridge.ubuntu.com/node/51 [19:59] that was pre-dapper when I did that. [20:01] cool [20:01] I'll ping corey and/or mgunes, see if they're interested in doing another tour [20:02] right, if we have a "tour" type page, that would be suitable; obviously selinux wouldn't fall under errata though, and I don't imagine it would go in the press release [20:03] slangasek: yeah, I'd like to find a place for it since it's been a long-standing gripe ("but Ubuntu can't use SELinux") that's fixed now. [20:04] maybe if the release notes linked to the tour page? [20:05] burgundavia once did a universe tour, one thing a day for a week, which is also a pretty good idea to get sustained interest [20:08] <_MMA_> jcastro: mgunes has been MIA. He's had some personal issues. An email sould get a responce. Though it might be delayed. [21:01] slangasek: ok to accept readahead-list? ok, great! [21:04] Keybuk: bwuh? has it even been uploaded yet? === Jonbo_ is now known as Jonbo [21:05] slangasek: yes :-) [21:05] I haven't seen it in the unapproved queue... [21:05] it didn't spend very long there [21:05] heh [21:06] haha [21:29] bryce: is bug #137234 covered by your projector work? [21:29] Launchpad bug 137234 in xserver-xorg-video-intel "[gutsy] Second display not enabled with "intel" video driver" [High,Confirmed] https://launchpad.net/bugs/137234 [21:30] Is there a place to get an old debian package? [21:31] It doesn't seem to be in the pool anymore. [21:32] slangasek: btw, there is a sync request that's been in for -amd, that didn't make it in before the cut off (we've been testing the fix, but unfortunately whomever does syncs didn't get to it in time). [21:33] bryce: I'll have a look at all those goodies today/tomorrow [21:35] slangasek: cool thanks. It apparently is very strongly desired by the LTSP guys, and upstream. [21:36] (and ogra for edubuntu) [21:36] ah, that one :) [21:36] hello [21:36] * slangasek waves to calc === IR3404 is now known as Xyzzy === Xyzzy is now known as Xyzyyz [21:39] bryce: so I didn't see that you answered wrt bug #137234? [21:39] Launchpad bug 137234 in xserver-xorg-video-intel "[gutsy] Second display not enabled with "intel" video driver" [High,Confirmed] https://launchpad.net/bugs/137234 [21:40] oh sorry, looking [21:41] bryce, btw, i have a new laptop with intel card ... apparently the external output ws enabled by default which caused gnome to think it is on a 1024x768 display [21:41] (screen size is 1280x800, i had a floating bottom panel) [21:43] slangasek: I'll reply - that bug seems to be a mix of S-Video not working, plus VGA-projector troubles; sounds like things have improved in Hardy for him, but still having some trouble. [21:43] bryce: ok, thanks [21:48] slangasek: a lot has improved in the couple months since the last comment, so it's quite possible the issue got fixed with one of our patches; I've asked him to retest [21:49] bryce: ok. it came to my attention because he brought it up (on u-d-discuss) in response to my call for bugs that should be critical; but yes, this only proves that he /thinks/ the bug is still there :) [21:50] people find X crashes/corruption so scary that I hear that a lot [21:53] of course, they also extrapolate that the crash they see must affect all Ubuntu users (sometimes they're right, but usually not) [21:53] hehe. I wonder why :-P [21:53] then I ask for a ''full backtrace'' and I think that completely pushes them over the edge of sanity [21:54] :-) [21:55] well, yes, this particular bug submitter also believes that "the most likely environment to use ubuntu is academia" (?) :) [22:00] slangasek: :-) [22:01] slangasek: btw the sync request is bug #211385 [22:01] Launchpad bug 211385 in xserver-xorg-video-amd "please sync xserver-xorg-video-geode (main) from Debian (main)" [Undecided,New] https://launchpad.net/bugs/211385 [22:03] i'm so tired, i'm going to be dead tomorrow due to this crappy jetlag :( [22:04] calc: changed to summertime recently? :-) [22:05] Nafallo: change from -5 to +1 today [22:05] er -5 to +2 maybe [22:06] 7 hours difference [22:06] calc: where are you? :-) [22:06] Prague [22:06] I was in Houston yesterday [22:06] sounds more like +2 indeed :-) [22:07] got on the plane at Wed 3:30PM got to the hotel at Thu 12:00PM, heh [22:07] us Londoners stole +1 ;-) [22:07] ah yea [22:07] :-) [22:09] ArneGoetje, ping. I was checking to see if you were aware of breakage in terms of language packs? [22:09] ArneGoetje, see http://cdimages.ubuntu.com/dvd/20080410/report.html or http://people.ubuntu.com/~ubuntu-archive/livefs-build-logs/hardy/ubuntu-dvd/20080410/livecd-20080410-i386.out for information [22:11] slangasek, it looks like the issue is actually that openoffice.org-hyphenation-en-us got moved to universe for some reason or another? [22:12] slangasek, were you aware of whether that was intended? [22:12] mario_limonciell: it wasn't moved, it was never /in/ main [22:12] we've worked this out, the dependency has been moved back to openoffice.org-hyphenation [22:12] in the very near past you worked it out? that dvd was just queued a few hours ago [22:13] yes [22:14] So, seriously, is there no Debian archive somewhere? [22:15] ted1: sounds like you're looking for snapshot.debian.net? === ogra_ is now known as ogra [22:17] slangasek, so should all pending language issues be fixed at this point if another DVD is made (we've been trying to get a good one the last few days) [22:18] slangasek: Yes, yes I am. Thank you. [22:19] hmm, f-spot segfaults on a fresh install [22:19] ugh [22:19] mario_limonciell: looks like it, yes [22:19] mario_limonciell: want me to fire one off? [22:19] http://paste.ubuntu.com/6719/ [22:21] ogra: I get a SIGSEGV after I Quit via the menu :-) [22:22] with f-spot ? [22:22] ogra: yes. not fresh install though. [22:22] slangasek, yeah if you can [22:22] ogra: upgraded gutsy. [22:22] it looks like that are two errors ... [22:23] ogra: at least :-) [22:23] mario_limonciell: running a livefs build only [22:23] one is a missing sqlite db and the other is that it cant connect to dbus [22:23] slangasek, yeah probably a good idea (waste of space to make dvds when that livefs is what keeps failing) [22:24] ogra: the missing sqlite db is well known issue [22:25] stratus, any idea if thats fixed in debian ? [22:25] mario_limonciell: can I just leave it running and wait for feedback from you if there are still failures? [22:25] yeah [22:25] i'll check on it in an hour or two [22:25] (or if you want me to follow through with an ISO build) [22:25] hmm, takes longer than an hour to finish, doesn't it? :) [22:26] if it succeeds anyway [22:26] ogra: see 81905 in lp [22:26] well it will fail within an hour at least :) [22:26] ok :) [22:26] ogra: it seems that f-spot have no idea where the sqlite db is and guesses that the table tags is missing [22:29] ogra: did f-spot setup .gnome2/f-spot/photos.db ? [22:29] yup [22:30] is it a really fresh install, right? I assume you didn't copy your home directory there. [22:30] right [22:30] !bug 206149 [22:30] Launchpad bug 206149 in ubuntu "The system doesn't recognize my Realtek RTL8185 wireless card." [Undecided,New] https://launchpad.net/bugs/206149 [22:30] then sqlite .gnome2/f-spot/photos.db and check if the missing table is there [22:30] and -b ./ solves it [22:31] oh cool [22:31] probably a bug during the build, f-spot has no base dir set or a wrong one [22:31] yeah [22:31] smells like that [22:32] i still have a segfault on exit though [22:32] but it starts and runs fine [22:32] oic [22:32] any idea what's up with that bug? [22:32] at least no d-bus related bug too [22:38] slangasek: is it ok if I upload a translation update of edubuntu-docs? [22:40] LaserJock: yes [22:40] thanks [23:37] slangasek: ping? [23:39] Keybuk: pong [23:41] slangasek: proposed udev patch [23:41] http://people.ubuntu.com/~scott/fix-persistent-net.pl.txt [23:41] well, patch/er/ [23:41] slangasek: uploaded new lrm for the latest ABI. it also fixed faulty dh_strip which was supposed to skip stripping nvidia, but didn't since the versioning was changed [23:47] Keybuk: looks sane to me... :) [23:47] tjaalton: thanks, will process it [23:47] slangasek: thanks [23:47] slangasek: I've tested it in every crazy situation I can replicate [23:49] Keybuk: let's go for it then :) [23:50] tjaalton: I note there's an open bug report for the nvidia stripping thing, and it's not mentioned in the changelog, so please close by hand [23:51] slangasek: uploaded [23:51] slangasek: right, will do