[00:03] Riddell or ScottK: mind reviewing if you have a second? http://dl.dropbox.com/u/69524/q4wine_0.114-r1-0ubuntu1.debian.tar.gz [00:05] Quintasan: What is that? [00:06] ScottK: https://bugs.launchpad.net/ubuntu/+bug/360942 [00:06] Ubuntu bug 360942 in ubuntu "[needs packaging] Q4wine - binary" [Wishlist,New] [00:07] oh great, I'm missing description [00:07] Quintasan: Since it's for wine, please ask Yokozar to look at it. [00:08] ScottK: okay [00:15] ooh, it's Quintasan http://fridge.ubuntu.com/node/1958 [00:16] Riddell: :D [00:20] Well, I'm going to bed. Good Night [00:21] nini Quintasan :) [00:21] * Mamarok should do that, too [00:38] Riddell: live CD build worked on i386, so I think we're actually in good shape. === dendrobates is now known as dendro-afk === dendro-afk is now known as dendrobates [03:36] There is good news and there is bad news: [03:37] 1. Good news: Software installation is easy enough that my totally non-technical step daughter got Skype installed with no help. [03:37] 2. Bad news: I know because her mike didn't work out of the box and I had to show her about kmix and mixer settings to get it working. [04:03] that *could* be a bug (WRT #2) [04:03] OTOH, we've fixed a crackton of those sorts in current ALSA [04:04] meaning, we now autoselect the appropriate mic based on jack presence, so if you have an external mic plugged in, the internal mic is muted [04:04] IIRC I had to tell it to use analog input instead of digital. [04:04] on karmic, that would be apt-get install linux-backports-modules-alsa-karmic-generic (and a reboot) [04:09] I'm actually thinking I need to take a break from Ubuntu. [04:09] it's healthy [04:10] I'm looking forward to the day when I can actually use it without having to drag myself away from fixing stuff [04:10] I'm disproportionately upset about the whole MoM discussion. [04:10] yeah, that was a bit nasty [04:11] It really brings into question Canonical's reliabilty as a partner to the community IMO. [04:11] "Sorry, it's broken, but we've pitched it over the fence to fix. Good luck." [04:11] well, there certainly are commercial forces driving the company's motives [04:11] Certainly. [04:11] I can understand that. [04:12] But ditching standard tools the people have used for years because one person is too busy and recommending unreliable, incomplete replacements is not a confidence builder. [04:12] I've largely avoided this particular tidbit by having merged manually from the onset, but I understand how, if led to rely on its continued presence, its recent downtime would be annoying. [04:13] unfortunately, I believe the core plumbers are utterly overworked [04:13] I understand the downtime (even if I'm not happy about it). It's the "and we're not going to fix it, but we didn't tell anyone" part I dislike [04:13] No doubt, but MoM doesn't take a core plumber. [04:14] So the fact that it was the sole responsibility of such a person is a management failure. [04:15] I don't even profess to understand how it was developed; my inclination is that Scott hacked away in a cave and emerged "victorious" after a while [04:15] I took a brief look at the code and it's not that horrible. [04:16] ScottK: didnt other-scott say he's been trying to find someone to take it over for years? [04:16] maco: Certainly. That's why it's a management failure at Canonical. [04:17] While I don't doubt there's a bit of ego involved all around, I agree that it could have been announced more widely. [04:18] (Then again, all around we seem to have this announce problem. None of us is immune.) [04:25] what is MoM an acronym for? [04:25] merge-o-matic [04:25] aah, sounds technical [04:25] It was. [04:27] hi seele [04:27] maco: yo [04:58] so whats the whole MoM battle? [04:58] I only really get a gist of it from the mailing lists [04:58] but don't know the backstory [05:00] Intially it was a closed source tool provided by Canonical. [05:00] Probably written by Keybuk in his spare time. [05:01] Then it stopped getting updated during one release cycle, a few people in the community got fed up and made DaD. [05:01] DaD was a FOSS version of MoM. [05:01] It's merging algorithm was never quite as good, but it had comments and was updated more frequently. [05:02] So for a while, MOTU used DaD and Canonical people used MoM. [05:02] (non-Canonical core-dev were rather less common then) [05:02] Then two years ago, Canonical agreed to open source MoM and the DaD devs agreed to merge their improvements. [05:03] It took some time, but it finally got done. [05:03] So DaD went away. [05:03] All is ~ well for a while. [05:03] k [05:03] There are a few community contributions, but it's not a lot of work for Keybuk to keep up. [05:03] Which is good, since he has no time. [05:04] Then Debian package source format V3 shows up and everytime MoM hits one, it falls over dead. [05:05] Today, Keybuk got pressed to fix it and washed his hands of it. [05:05] i thoguht he'd been blacklisting problematic packages for a months...longer than v3's been around [05:05] maco: That was in the archive. [05:05] As long as that was done it hit MoM irregularly. [05:05] Now that v3 can go in the archive, it hits every run. [05:06] The consensus from the Canonical people in the conversation was that it was unreasonable for the community to expect Canonical to continue to provide this service and we should use bzr merges instead. [05:06] Which might be OK except: [05:07] 1. All through the distributed development disucssions whenever it's been brought up, there have been repeated assurances that no one would be forced to switch. [05:07] 2. There's a very long list of packages that won't import yet and so can't be done this way. [05:08] 3. It's complex and not completely documented. Between Laserjock and I we tried 5 merges tonight and managed only one upload. [05:08] #2: there is? [05:08] i thought james said all the packages were in bzr now [05:08] ? [05:09] He may have said that, but the instructions on merging using bzr say to consult a list of packages that won't import and that list is still there and not short. [05:09] hmm [05:09] The one upload I managed, I only managed becuase I asked how to get stuff done. [05:09] hm ok then... <100? [05:09] I was always under the impression that merging will all happen in bzr now.. [05:09] I dodn't count. [05:09] dod/did [05:09] and that the deb sources would also be in bzr [05:09] so that it wouldn't be a problem [05:10] shtylman: It will because there is nothing other than a completely manual merge as an option. [05:10] shtylman: That's an interesting theory. [05:10] 4. Even once you know the workflow, it has more steps, more things to go wrong, and is pretty generally inferior to the old way of doing things. [05:11] The added granularity of history that using a VCS would normally give you isn't relevant for merges. [05:12] I see [05:12] what type of merges are we talking about here? [05:12] I honestly have no idea [05:12] as I have never done package management [05:13] It's for when we have local changes from Debian and they update their package. [05:13] I see [05:14] So the Debian changes and the Ubuntu changes get merged from a common ancestor. Then you have to figure out how much of the Ubuntu difference is still relvant and needs to be preserved [05:14] makes sense [05:14] seems kinda like a rebase [05:14] I guess it kinda is [05:14] In a sense it is. [05:15] shtylman: What VCS do you use the most? [05:15] mostly git [05:16] obviously bzr for ubuntu stuff [05:16] git...i should learn more git [05:17] I like bzr [05:17] shtylman: Right, so say you'd only used git, had used it for years, and someone took it away and gave you an alpha release of bzr that you had to use instead (or you could just not use a VCS) and they did it with no warning. [05:17] I just really like some git features more [05:17] I'm not opposed to bzr. I generally like it. I've promoted it's use on $work projects. [05:17] ... yea... not good [05:18] This particular use case is just not yet fully baked. [05:18] I can see where that would be a problem [05:18] didn't they try to do that once already? [05:18] I guess the migration is just hard [05:18] back when bzr sucked? [05:18] So that's kinda the current situation from my perspective. [05:18] I see [05:18] hey i have a merging question. how should my debian/changelog look at the end? [05:19] JontheEchidna: We tried it during Gutsy in Kubuntu like we are now (for the Debian dirs), but that was a team choice to try it, it wasn't required. [05:19] combine all the changelog entries that are ubuntu-specific into one new changelog that comes after the last debian one? [05:19] I combine all changelog entries, sticking the new debian entries in chronological order [05:19] maco: All the previous Ubuntu entries should stil be there in order. [05:19] One can argue chronological or version order. [05:20] chronological or version order? [05:20] version order, right? [05:20] oh [05:20] heh [05:20] didnt read scott's second line ;) [05:20] * ScottK uses vesion order generally. [05:21] There are times when it doesn't quite make sense. [05:21] Then the new entry at the top should document all the remaining changes. [05:31] would you be willing to sponsor a mergey upload for me? [05:33] just to point, none of the packages I oversee are imported. [05:33] crimsun: arent they all git anyway? [05:33] crimsun: hey wanna upload hunspell? http://people.ubuntu.com/~maco.m/needs_sponsorship/ [05:33] pkg-alsa is entirely SVN [05:34] maco: maybe in 25 minutes; I need to walk home [05:34] ok [05:34] crimsun: hey did you read email from me? [05:34] no [05:34] crimsun: pm real quick? [05:34] I'm really stretched thin ATM === dendrobates is now known as dendro-afk [06:49] maco: your changelog is pretty curt. Which changes did you retain/drop? [06:49] crimsun: kept all. it was a clean merge, just the changelog had a conflict [06:50] can you respin with a more verbose changelog entry for the next merger? [06:50] the ubuntu changes are just a list of extra words to add to the dictionary [06:50] yeah [06:50] no need to list them all; you can use "refer to version X for extra words" [06:52] * Merge with Debian, keeping all Ubuntu changes (extrawords.txt and its support in debian/rules) [06:52] that work? [06:52] yep [07:03] um, by the way...i cant test this build in pbuilder...that whole "aptitude is fubar" thing [07:03] aptitude just spits a stacktrace at me while trying to install the build-deps [07:04] but since it built in debian and given what the ubuntu changes are..it should theoretically build fine on the buildd [07:04] maco: There's more than one way to do it. [07:04] ScottK: im listening [07:04] Look in the pbuilder config file in /ect. [07:05] IIRC there's a dependency resolver option that doesn't use aptitude. [07:06] ah i see [07:07] PBUILDERSATISFYDEPENDSCMD= should i use gdebi? manpage says thats faster than classic [07:14] ScottK: thanks! the build is running now [07:23] crimsun: ok, reupload with more verbose changelog and with having successfully built in pbuilder [07:23] oh my its 230am? [07:29] Now it is. [07:30] *snort* [07:30] * crimsun points maco toward -v20070829-2ubuntu4 [07:30] whatd i do? [07:30] you forgot to include the new Debian changelog entry [07:31] I rerolled the source package for you this time; just keep it in mind for next time, please. [07:31] whatd i forget what? [07:31] -3ubuntu1 is the new entry [07:32] -2ubuntu4 was the previous version in ubuntu [07:32] -3 is in debian [07:32] so -3ubuntu1 is the new merged one [07:33] crimsun: what> [07:33] *? [07:33] passing -v20070829-2ubuntu4 to debuild -S means "include all changelog entries *since* this version in the _source.changes" [07:34] without it, you only include the most recent, which omits -3 [07:34] ooooooo i see [07:34] im sorry. thank you [07:34] np. Z. [07:35] Unless of course you're doing it the new fashioned way and using bzr builddeb where -v means be verbose. [07:49] um actually, yes i was using bzr builddeb [07:50] i assume thisd be: bzr builddeb --old=lp:debian/blahblah then? [07:51] oh, probably -rREVNO [07:52] hmm nope. not -rREVNO [07:53] helps if i read all of --help's output... [07:57] To pass -v like in dpkg-buildpackage it's -- -v. [07:57] Because -S -- -v is so intuitive [08:06] ScottK: bzr merging not your friend? [08:16] Riddell: for the record, cjwatson talked to doko about a glibc backport. It's in doko's hands now [08:17] the number of duplicates of https://bugs.kde.org/show_bug.cgi?id=196207 is piling up every day... [08:18] KDE bug 196207 in general "GTK/Glib applications crash when running under KDE4 3x (malloc_printerr, free(): invalid pointer)" [Crash,Resolved: upstream] [08:19] gah, the bug description is not really accurate, extends to Qt applications, too [09:16] agateau: ping [09:19] jussi01: pong [09:20] agateau: I understand you are notifications goto man? [09:20] jussi01: sort of :) [09:20] :D [09:24] agateau: Im using 4.4 beta, notifications seem well screwed. they seem to stick around forever, unless you click on the cross. ie. we get a huge accumulation of little numbers that never go down... [09:24] oh that, I am sort of the "alternative notifications goto man", then [09:25] I created Colibri just because I didn't like the Plasma notifications [09:25] agateau: ahh :D [09:25] Not sure I can help you here, you want to discuss this with notmart on #plasma [09:25] what, pray tell, is colibri? [09:26] http://kde-apps.org/content/show.php/Colibri?content=117147 [09:26] jussi01: you may find this post interesting as well: http://agateau.wordpress.com/2009/12/16/introducing-colibri-an-alternative-to-plasma-notifications/ [09:28] * jussi01 reads [09:36] agateau: hmmm i will try this passive thingy [09:36] cause i dont like this monumental plasma notifies :) [09:37] :) [09:37] as jussi01 sayed a lot of them if you dont cancel them [09:37] o.o [09:37] hmm is there already a package or can i ask on getdeb for packaging it o.o [09:38] ghostcube: I haven't packaged it, but it was (I think) nixternal who commented on the post that someone was packaging it [09:38] ah ok [09:38] nixternal: winke winke [09:42] office party time! [09:42] wth this looks so borked even if its wanted [09:43] https://code.launchpad.net/projects [09:43] * Riddell hangs tinsel up around the place [09:43] Riddell: ?? [09:43] xmas party? [09:44] Riddell: is this an Edimburgh-only office party? the Chapelle-Rablais office is hard at work! [09:44] seems hes busy doing nothing :D [09:44] oh the Edinburgh office is hard at work, just with Slade playing in the background [09:45] oh ok :) === Quintasan1 is now known as Quintasan [10:01] good morning [10:02] morning [10:14] 11 am is not morning:) [10:55] :P === yofel_ is now known as yofel [11:01] ulysses__: depends on when you got up :P [11:01] at 7:30 :/ [11:02] what a shame, I went to sleep at about 5 o clock in the morning and got up at 10 [11:02] :D [11:02] can't sleep at night?:P [11:02] yeah, pretty much [11:02] during the day too :P [11:04] what did I miss about virtuoso? [11:04] Mamarok: nothing, 6.0.1 is still not working or debian guy is crazy [11:05] Mamarok: the problem is probably with virtuoso or nepomuk itself [11:05] I've tried to poke in #nepomuk-kde but there was noone answering [11:29] Quintasan|Szel: write a mail directly to Sebastian Trueg, or poke him in @nepomuk-kde [11:46] here comes a packaging problem: [11:47] gah, /me hates bash when it doesn't allow Ctrl+C *grrr* [11:48] dpkg: error processing /var/cache/apt/archives/kde-window-manager_4%3a4.3.85-0ubuntu1~karmic1~ppa3_amd64.deb (--unpack): [11:48] trying to overwrite '/usr/share/kde4/config/aurorae.knsrc', which is also in package kwin-style-aurorae 0:0.2.1-0ubuntu1 [11:48] here you go ^^ [11:50] and btw, there is a typo in the website announcement: [11:50] udo apt-get dist-uprade [11:51] * Mamarok is not good at copy-paste today [11:51] it reads dist-uprade, missing the g === ScottK changed the topic of #kubuntu-devel to: Lucid Alpha 1 Released! | Kubuntu has the Doctor on the brain | https://wiki.kubuntu.org/Kubuntu/Todo | Oustanding merges: http://people.ubuntuwire.com/~lucas/merges.html |Congratulations to Quintasan for becoming a MOTU [13:57] Hello [14:08] ehlo [14:12] pleasingly I have no merges with my name next to it [14:14] oh actually it helps if I search by my launchpad id [14:15] :D [14:16] * Quintasan is off to help his dad, then back to backtracing crashes [14:17] * jussi01 hugs Riddell [14:21] ScottK: arounds? [14:21] freeflying: Yes. [15:31] Riddell: We're really close to having things about as built as we are going to. How would you feel about doing your qt4-x11 upload to fix powerpc today? [15:33] hmm, I havn't started looking at it, I don't know the right syntax for arch specific patches [15:33] or it may need to be done in the rules file specially [15:33] let me look into it [15:34] Whatver you did for armel last cycle (IIRC nepomuk stuff) worked out OK. [15:55] Sparc is fully built now. [16:34] ScottK: you have a sparc station? === seele_ is now known as seele [16:34] seele: No. Just tracking it in Launchpad. [16:35] ah [16:35] IA64 is done now too (don't have that either) [16:53] JontheEchidna: when you got a second can you ack colibri on revu? thanks === santiago-ve is now known as Guest72071 === santiago-ve is now known as Guest56540 [18:07] hi [18:07] updated to beta 2 worked fine :) [18:08] anyone knows who is doing the ppa for nvidia vdpau drivers ? [18:10] ah got it njsnider [18:10] bjsnider [18:10] o.o [18:48] ok Qt 4.6 and kde 4.4 beta 2 in combination with new nvidia driver definetly rox [18:48] teh fast [18:51] well, Qt 4.6 sux for me, lot of bug reports refering to it === Guest56540 is now known as foursixnine === blizzzek is now known as blizzz === yofel_ is now known as yofel === santiago-ve is now known as Guest9969 [19:29] anyone have any idea about why kded4 eats 100% cpu in 4.4 beta2? [19:34] no, try unloading modules or using gdb [20:21] * Daskreech waves [20:23] Quintasan|Szel: ping === santiago-ve is now known as Guest52774 === Guest52774 is now known as foursixnine [21:05] just installed that Beta 2 stuff on Karmic, I have all my desktop settings gone (activities, etc.), 2 panels (which luckily at least kept the settings, but twice, one smaller and hard to get rid of) and plasma crashes galore. I guess this is to be expected? [21:05] Oh and Akonadi doesn't run anymore [21:07] +: it's fast, desktop effects work like double speed now [21:07] back up ~/.kde before upgrades === blizzz is now known as Blizzz [21:48] nixternal, hi :D I looked at the lucid todo for kubuntu and are there any developmental tasks I can take part in [22:28] Mamarok: https://launchpad.net/ubuntu/+source/eglibc/2.10.2-2ubuntu4 [22:29] Riddell: kdepim finally finished on armel, so that's fully built too. [22:30] http://www.techradar.com/news/software/operating-systems/8-of-the-best-kde-distributions-657523?artc_pg=3 [22:30] * ScottK now waits for Riddell brilliant debian/rules hack for powerpc only patches [22:33] txwikinger: You have been busy. We all chatted about that ages ago. [22:40] ScottK, hi :D how are you [22:40] dhillon-v10: Busy [22:41] ScottK, do you have any idea when merge-o-matic is going to be back, I heard its down right now [22:42] Never unless someone fixes it. [22:42] ScottK, alright that's good enough for an answer :D I just continue my work as it then [22:43] dhillon-v10: The answer wasn't a joke. It's completely maintainerless right now, so unless someone volunteers to work on it, it won't get fixed. [22:43] ScottK: what about the backport to Karmic? [22:44] Mamarok: Rules say it has to be fixed in the development release first, so this would be step one. [22:44] ScottK, I am really good at web designing, much better than packaging, can I get working on it, please [22:44] All I know is I say it building. [22:44] OK, nice to know :) [22:44] dhillon-v10: Source is in the merge-o-matic project on Launchpad. [22:45] ScottK, alright so what's the main problem behind it, is it not updating the packages [22:45] dhillon-v10: It needs to deal with v3 format source packages. [22:46] ScottK, what's v3 format package [22:46] dhillon-v10: Remember the start of the conversation where I said I was busy? Google can help you with that one. [22:47] ScottK, I am sorry I didn't mean to bother you, I'll google that [23:21] Is the Akonadi kcm module being transitioned out in 4.4? I see it's still in kdepim-runtime but it doesn't have a .desktop service entry. [23:22] Zorael: It's in a separate binary now. [23:22] (I think) [23:33] Zorael: upstream decided to remove the module, since basically the only thing you could do with it is break Akonadi (unless you really knew what you were doing) [23:34] JontheEchidna: Ah, okay, makes sense. [23:39] * Riddell puts up some mistletoe here too and stand underneith it [23:49] * claydoh dcc's smooches to Riddell :) [23:49] yay! [23:51] * claydoh is enjoying his beta2 presents , many thanks to all! [23:52] ScottK: what do you think of this? http://bazaar.launchpad.net/%7Ekubuntu-members/qt/ubuntu/revision/50 [23:52] I've really no idea if that's the right syntax in rules [23:52] Looking [23:54] unix:!mac:*-g++*:QMAKE_LFLAGS += -Wl,--gc-sections <- is this common for removing this line? [23:54] back in the day, I would add that line when building on my old mac [23:55] the linker explodes on ppc with --gc-sections [23:55] meaning 10.04's current linker explodes on ppc [23:57] Riddell: Looks right. Looks similar to what you did in http://bazaar.launchpad.net/~kubuntu-members/kdebase-runtime/ubuntu/revision/69 [23:59] Of course I left the #ENDIF of of my IA64 build fix on the first try, so don't go by me.