[03:04] Riddell, ping [03:04] Riddell,you looked at amarok build? or should I file an upgrade bug for amarok in LP> [07:48] * c2tarun phew finally dsl connection worked on kubuntu :D [07:50] morning [07:51] Riddell: ping [09:14] Riddell: by the way I uploaded koffice-l10n to revu (I'm not sure, but we can do that for the main archives ?) [09:16] hey folks [09:16] Riddell: apachelogger: this link on Kubuntu.org is 404: http://releases.ubuntu.com/kubuntu/lucid/kubuntu-10.04.1-alternate-amd64.iso [09:16] found this out yesterday, when we wanted to download it [09:17] thanks bambee, will take a look shortly [09:19] markey_nokia: hmm, we're on .2 now not .2 [09:19] markey_nokia: hmm, we're on .2 now not .1 [09:19] where did you find that link? [09:19] Riddell: typo then? [09:19] Riddell: it's the download link... bottom of "direct download" page [09:20] this page? http://www.kubuntu.org/getkubuntu/download [09:21] http://www.kubuntu.org/getkubuntu/download#download-block [09:21] ^ [09:21] Riddell: you're welcome. There are other things to do ? (directly or indirectly for kubuntu) [09:30] bambee: http://stefan.derkits.at/files/libmygpo-qt/ needs packaged for amarok [09:31] markey_nokia: hmm well kubuntu.org seems just to have broken [09:32] Riddell: Ok, I will package it [09:45] Riddell: hmm ok. looks like a simple typo in the URL, no? [09:45] 1 instead of 2 [09:46] yes, something not updated since .2 was released [09:46] but now ubuntu.com is broken for me, tsk [09:48] markey_nokia: pressing the Begin Download button gets the .2 for me [09:48] lemme try again... [09:48] ah hah, I see it [09:49] Riddell: definitely gets the .1 here, and we tried it on several computers [09:50] yep, onto it [09:55] markey_nokia: fixed, thanks === debfx_ is now known as debfx [12:16] Binary only demotions to universe o debconf-kde-helper {debconf-kde} [12:16] do we want that? [12:17] dantti: does kpackagekit need that or is the library enough? [12:22] Riddell: uh .. you *can* disable them [12:22] it's under settings [12:22] shadeslayer: disable what? [12:22] derp [12:22] Riddell: google suggestions in rekonq [12:22] pardon? [12:23] http://identi.ca/notice/67301633 [12:23] oh right [12:23] tell it to the microblogosphere :) [12:23] i see the issue being resolved :P [12:24] you can't exactly disable them [12:24] because it would break the inherent nature of the urlbar thingy [12:24] well make up your mind, you just said they could be [12:25] yeah ... you can't .. i spoke too soon [12:29] apachelogger: Riddell lolz http://www.wikipeetia.org/Ubuntu_(operateng_sytem) [12:37] ScottK: why do we want python bindings for an obsolete library (qt assistant)? [12:39] Riddell: Because they're there, people might be using them, so we might as well. Also it will enable us to sync python-qt4 from Debian. [12:40] In fact filing the sync request is on my TODO for this morning. [12:41] syncing that from debian is good [12:41] adding support for obsolete libraries isn't good though [12:41] shadeslayer: very odd [12:41] but the sync wins out in goodness [12:41] apachelogger: whaddup [12:41] shadeslayer: must you be highlighting me though? [12:41] I am reverse engineering a ray tracer [12:41] this is madness! [12:42] ah [12:42] apachelogger: leave it ... it's madness [12:42] half the lines of my description start with "assumption:" the other half with "educated guess:" [12:42] Riddell: Filed. [12:43] though I have an evil plan to conclude all assumptions and guesses in the last sentence, when the big picture emerged from the mess of shitty code [12:54] Riddell: hey, where can I get the image for the Kubuntu stickers you gave out? [12:56] the one with the broken typography? [12:56] lol, yeah [12:56] http://people.canonical.com/~jriddell/tmp/sticker.png [12:56] but you'd be better recreating it [12:57] Riddell: I heard you have Project Neon stickers apparently [12:57] Riddell: oh, in that case, do you have the svg? [12:57] Quintasan: I have a few :D [12:57] nigelb: no I made it in Krita [12:57] Quintasan: come to UDS if you want them [12:57] Riddell: ah. I suck at anything design, I might just use the same thing again. [12:58] Riddell: Make Canonical give me sponsorship for sure :D [12:58] Quintasan: ok, since Riddell is using the UDS clause, I won't offer to mail to you. [12:58] FFFFFFFFFFFF [12:58] lol [12:58] I only have like 5 to 6 which I grabbed from shadeslayer :P [12:58] hehe [12:59] show me em [12:59] i have a bunch of them as well [13:00] I'm planning to take 6 sheets of Ubuntu and Kubuntu stickers so I have some at every conf to give out :D [13:01] Kewl :D [13:18] Riddell: the lib is enough [13:19] Nightrose: on the one hand you want non-technical submission and on the other you do not like me proposing talks about vodka or wine, make up your mind :P [13:19] * Nightrose pokes apachelogger hard [13:19] Nightrose: how about ... "how to take over a dying project and make it sexy again" [13:19] sounds much better ;-) [13:20] Nightrose: is that non-technical enough? [13:20] yes [13:20] splendid [13:20] :P [13:40] whaa [13:41] apachelogger: that sounds like Project Neon [13:41] shadeslayer: lets do a talk together [13:41] where? [13:41] Desktop Summit? :D [13:42] Your room? [13:42] DS [13:42] * apachelogger writes some brainfuck [13:43] apachelogger: dude ... i'm all for doing it, only constraint being, i don't want to be on a stage :P [13:44] fair enough, you get a mic and sit in the audience [13:44] sure [13:45] i get to be the guy who takes the mike to everyone in the audience [13:45] when they have questions ofcourse [13:48] no questions allowed [13:48] * apachelogger had bad experiences with questions [13:49] hmm, gtk-recordMyDesktop fails here [13:49] some Python error [13:49] is there a working alternative? [13:49] need a screencast recorder [13:50] "AttributeError: TrayPopupMenu instance has no attribute 'popupmenu_continueitem'" [13:50] What's the error? [13:50] see above :) [13:50] Ah. [13:51] any ideas? [13:52] apachelogger: haha :D [13:52] markey_nokia: How about RecordItNow? [13:52] lemme check [13:52] Google claims that's a Qt front end to it. [13:52] there is [13:53] "The following packages have unmet dependencies: [13:53] recorditnow: Depends: libpolkit-qt-1-0 but it is not going to be installed" [13:53] can't install it [13:53] sudo apt-get install python-appindicator may also help. [13:53] ok [13:53] (with the GTK one) [13:54] that also gives an error... [13:54] (using KPackageKit) [13:54] is that PackageKit fail? [13:54] markey_nokia: you do not want to use recorditnow [13:55] http://verb3k.wordpress.com/2010/01/26/how-to-do-proper-screencasts-on-linux/ [13:55] ScottK: I do not want to use recorditnow [13:55] haha LD [13:55] :D [13:55] markey_nokia: It's not the droid you are looking for. [13:56] ScottK: indeed. I want an easy solution, just pressing a button or so [13:56] hmm [13:56] or I could run Linux in a VM, and use a Windows tool, I guess [13:56] might work for now [13:57] but still... I thought this is easy to do with Linux :) [13:57] maybe it's not [14:08] debfx: why is kubuntu/10_soundmenu_support.diff not needed in Amarok? === seaLne_ is now known as seaLne [14:14] Riddell: I guess it's all handled by mpris, agateau told me it's not necessary anymore [14:25] debfx: ok thanks [14:37] Riddell: ping amarok failed to build too :( [14:40] * c2tarun it seems packages are failing just because I build them :( koffice, kdeedu, amarok .... [14:43] c2tarun: last few lines of FTBFS please [14:43] or the whole log if you can pastebinit [14:43] shadeslayer: here is whole log http://launchpadlibrarian.net/66654622/buildlog_ubuntu-natty-i386.amarok_2%3A2.4.0.90-0ubuntu1_FAILEDTOBUILD.txt.gz [14:44] c2tarun: that's bug #737137 [14:44] Launchpad bug 737137 in cmake (Ubuntu Natty) "find_library fails to locate multiarch libraries" [High,New] https://launchpad.net/bugs/737137 [14:44] everything is failing, presumably libz has had the multiarch treatment [14:45] who is working on it? [14:45] nobody as far as I know [14:46] Importance: High and still unassigned. [14:48] somebody uploaded a file on that bug, anyone looked at it yet? [14:50] that might be the best chance for a quick fix, just hardcoding in the new paths [14:50] it'll need to be done once for each platform of course [15:01] Riddell: Are you around to take care of the release meeting? [15:03] ScottK: I am, anything you think I should mention? [15:03] besides the world being broken due to multiarch [15:03] I'd like to know where we are on the gles transition. [15:04] mm, yes === hunger_ is now known as hunger [15:36] apachelogger: why didn't you get this http://wetab.mobi [15:47] Riddell: I might have a fix for the cmake issue [15:48] debfx: oh? [15:48] do tell [15:50] Riddell: adding /lib/${CMAKE_SYSTEM_PROCESSOR}-${CMAKE_SYSTEM_NAME}-${CMAKE_C_COMPILER_ID} to CMAKE_PLATFORM_IMPLICIT_LINK_DIRECTORIES [15:50] and the same for /usr/lib [15:50] that sounds promising [15:52] debfx: in Modules/Platform/UnixPaths.cmake ? [15:54] Riddell: yes, http://paste.kde.org/7590/ [15:58] debfx: doesn't work on i386, it ends up as i686-linux-gnu [16:00] grr why is it i386 when we are building for i686 [16:05] shadeslayer: idunno [16:06] shadeslayer: >++++++++++[<++++++>-]<.> [16:06] what [16:06] what is that [16:06] turing ftw [16:06] brainfuck programming language? [16:06] shadeslayer: brainfuck [16:06] * apachelogger is currently writing a svg builder in brainfck :D [16:07] * shadeslayer looks [16:07] debfx: CMAKE_PLATFORM_IMPLICIT_LINK_DIRECTORIES doesn't work if regardless it seems, I need to set CMAKE_SYSTEM_LIBRARY_PATH [16:07] not using more than 6 cells [16:07] sorta madness [16:08] though originally I wanted to go with 3, turns out you spend half your time calculating perfect arithmetic manipulation ^^ [16:09] compiling it with gcc? [16:09] I can imagine that the only thing harder than using that language is writing a compiler for it [16:10] nah, using bf [16:10] !info bf [16:10] bf (source: bf): a fast Brainfuck interpreter. In component universe, is optional. Version 20041219ubuntu3 (maverick), package size 13 kB, installed size 104 kB [16:10] !language | ubottu You should know better! [16:10] Daskreech: Please watch your language and topic to help keep this channel family-friendly, polite, and professional. [16:11] JontheEchidna: actually writing a compiler is pretty easy as there is no complexity in the grammar [16:11] Stupid bot [16:11] debfx: now I have http://paste.kde.org/7592/ [16:11] JontheEchidna: if one knows how to build a parse tree that is ^^ [16:11] lol [16:13] Hmm, somehow I like USC's recommendations better than mine (in Konsole): http://i.imgur.com/fGIvC.png [16:13] Riddell: I think we should add it to CMAKE_PLATFORM_IMPLICIT_LINK_DIRECTORIES as well [16:14] JontheEchidna: you will be eatien by appstream and slow package managers anyway [16:14] debfx: ok [16:14] but but but... they're slow [16:14] I know [16:14] slow is the new fast [16:14] AppStream is an example of horrible technology being forced across the entire ecosystem [16:14] JontheEchidna: your software is totally 1980's [16:15] well, maybe not horrible, but not great [16:15] appstream would not need to be so slow if they first forced a common stack across the ecosystem [16:16] It's the cross-distro package management abstraction that is doing that [16:16] nonono [16:16] it is abstracting [16:16] ...abstraction comes at a price... [16:17] the stronger the abstraction the more you need to pay [16:17] ... [16:17] Eventually it's just abstractions all the way down. [16:17] I don't see a cross-distro package manager as being a big need anyways, since most people will only ever experience on linux distro [16:17] nah, it is layers :P [16:18] abstraction is not necessarily capsulation [16:18] APT is what makes Debian-based distros shine in the first place, so I think not utilizing it to its fullest extent is diminishing what makes Debian-family great [16:18] JontheEchidna: i've read that twice. Did you mean one linux distro? [16:18] Daskreech: He did. [16:18] JontheEchidna: Agreed. [16:19] Daskreech: yes [16:19] JontheEchidna: So is it Kpk/Muon shootout at the next UDS? [16:19] I'd disagree with that but ok [16:19] People are too fickle for that [16:19] ScottK: Yes, by 10.10 I expect there to be a "critical mass" of the Muon Suite having more features than {K}PK to justify switching [16:20] does it do appstream? [16:20] 10.10? [16:20] er [16:20] 11.10 [16:20] typo [16:20] cause even with slowness and stuff, appstream is the important community factor [16:21] (especially if you take project brezn into account) [16:22] why do we need to compromise on inferior technology when the average user will only ever use one distro? [16:23] JontheEchidna: I agree with your conclusion, although not necessarily the premise. [16:23] The average user will use Windows because that's what came on their computer. [16:24] it is not about the user [16:24] it is about the software [16:24] all them distros put steam behind appstream [16:24] making it naturally more attractive for deployment as more people are concerned with improvement and maintenance and whatnot [16:25] also from a social POV it encourages more inter-distro knowledge exchange and stuff [16:25] It's an interesting theory. Not a new one though. [16:26] I could see Muon using certain components of AppStream, such as the application metadata rather than app-install-data [16:26] Inter-distro knowledge exchange is not all good. If I'm off inter-distro knowledge exchanging then I'm not working on my distro, so I need to get enough out of it to be worth the time. [16:27] So far all I needed to know about OpenSUSE I learned from their patches in their BTS (don't use them). [16:28] and it seems that I already support the OCS reviews stuff since I support the Ubuntu reviews server (and could easily support another if the Ubuntu API isn't used) [16:28] so yeah, it could be said that Muon will support most of the AppStream bits, expect that the installation doesn't use PackageKit [16:29] supposedly appstream's spec is abstract enough from an implementation that it should not matter whether one uses pk at all? [16:29] (didn't read anything TBH, all I know comes from people telling me how awesome it is :P) [16:29] right, the usage of packagekit is totally in the realm of the client implementation [16:29] http://distributions.freedesktop.org/wiki/AppStream/Implementation [16:30] I am all for a fast appstream ;) [16:30] that said [16:30] there is fragmentation [16:30] as the opensuse guys apparently did not consider building their initial impl based on kpk [16:30] making me go ewwwww [16:31] oh actually [16:31] right, it's not like Bretzn took any consideration to existing utilities anyways [16:31] JontheEchidna: IIRC their KDE client looks very sexy [16:31] you should steal some things [16:31] I've not been able to find code [16:31] ah it is a stealh project [16:31] I always knew opensuse had something to do with the brotherhood of nod [16:31] now it becomes all too clear :P === zkriesse_ is now known as Guest56267 [16:34] I am not entirely convinced it is not vapourware :P [16:34] (j/k) [16:34] JontheEchidna: fregl will know [16:34] IIRC he attended the meeting of the brotherhood and other distros regarding appstream [16:34] * apachelogger wonders how much code he should produce for his brainfuck program [16:35] http://i.imgur.com/EcCbj.png [16:35] ^I hate it when that happens [16:35] the trick is to not have food [16:35] I do not have any food === Guest56267 is now known as zkriesse__ [16:35] so I need to start drinking at 3pm [16:36] works out pretty well === zkriesse__ is now known as zkriesse_ [16:36] there is an article on the wiki about kubuntu patches specs ? (mostly about the number) [16:37] bambee: are you asking? [16:37] Riddell: yup [16:38] bambee: I don't understand what you mean by kubuntu patches specs [16:38] why is debconf ncurses now pink [16:38] has someone ported fluffy to debconf? [16:38] * apachelogger hides [16:38] Riddell: See kirkland's mail to ubuntu-devel last night. [16:39] debian is pink too, dont you know [16:39] Riddell: patches are apparently named "kubuntu_NUMBER_what_does_my_patch.diff" => why a number ? [16:39] numbers are used to maintain order, in case two patches modify the same file [16:39] JontheEchidna: Except they don't maintain order. [16:39] debian/patches/series does that. [16:39] (they used to) [16:40] they maintain order with humans ;) [16:40] Meh. [16:40] I guess that was back in the dpatch days [16:40] JontheEchidna: ok thanks [16:40] :) [16:40] of course people at times forget how quilt work and mess up the series [16:40] JontheEchidna: Actually simple-patchsys. dpatch has 00list. [16:40] so quilt is not always in line with number thus humans [16:41] fortunately enough patches in kde rarely depend on each other ^^ [16:41] that is because most KDE software is holy and needs not patching ^^ [16:42] yeah [16:42] like phonon [16:42] you people actually patch my mighty phonon [16:42] very distrubing I must say [16:42] technically debian does [16:43] debian-- [16:43] and we just pick up the changes [16:43] ohhhhhhh [16:43] Riddell: talking about colorz [16:43] I have a file somewhere that overloads the aubergine grub to be blue [16:43] probably should go in kds [16:44] ^yes plz [16:44] apachelogger: I've in love with qt, finally :) [16:44] I don't want to ride the aubertruck [16:44] you nevar blogged the pic! :P [16:44] nigelb: you are doing a talk on Qt development at UADW then? :D [16:44] JontheEchidna: are you doing KDE? [16:44] I dented it, though [16:44] pretty plz [16:44] apachelogger: too many talks on qt at uadw [16:44] apachelogger: Qt haz taken over. [16:45] well [16:45] for obvious reasons [16:45] https://wiki.ubuntu.com/UbuntuAppDeveloperWeek [16:45] apachelogger: very obvious, once I started getting a hang of it. [16:45] I'm still stumbling but very impressed too [16:45] * apachelogger pokes JontheEchidna really hard so he adds a KDE talk [16:46] apachelogger: you mean something like how to use KAppTemplate to kickstart app development? [16:46] apachelogger: oh? does that mess anything up? [16:46] debfx: on arm this patch ends up with /lib/armv7l-linux-gnu [16:46] Riddell: no [16:46] Riddell: very easy with grub2 [16:46] debfx: and it needs to be /lib/arm-linux-gnueabi [16:47] it basically just has a higher number than the ubuntu thing and sets the color stuff to something else [16:47] debfx: so I think we're best just to do it hardcoded by platform [16:47] has like 10 sloc or so [16:47] JontheEchidna: that'd be good [16:47] JontheEchidna: for example [16:47] JontheEchidna: go and give instructions on a real world app [16:47] * JontheEchidna loves this runner: http://i.imgur.com/UkpCP.png [16:47] JontheEchidna: write a micro blogging app so we can get rid of choqok :P [16:48] apachelogger: except for grub you need to remove the aubergine settings first before you set your own ones, I tried to override it in a higher number which looks horrible since I still got a aubergine background for about a second before mine came up [16:48] * apachelogger really wonders why text is broken in qml on android [16:49] JontheEchidna: I second apachelogger [16:49] I did file a bug a while ago for it not being easily changable [16:49] yofel: get a faster machine :P [16:49] apachelogger: that's on my thinkpad! maybe I should blame nvidia [16:49] maybe [16:49] I did not see any funny color business on my netbook [16:49] * apachelogger is out of booze :O [16:50] zomg [16:50] Riddell: You might talk to lool about armhf as it's going to make it a triplet, IIRC. [16:50] maybe I'll do something neat with QZeitgeist [16:52] apachelogger: https://wiki.ubuntu.com/UbuntuAppDeveloperWeek/Timetable [16:53] JontheEchidna++ [16:53] oops, meant to put that in friday's slot [16:53] <3 wiki tables [16:55] how time we have left to push BlueDevil 1.1 final? [16:55] apachelogger: actually tuesday is good, I miscalculated the UTC [16:56] afiestas: Are there new features between what we have now and final? [16:56] Before next Friday would be nice, in any case. [16:56] ScottK: nope [16:56] If we could get it next week, then it would be no trouble, after that we may have to justify it. [16:57] okz [16:57] I will start to work like a crazy then [16:57] Great. [16:57] I have been kinda offline the last 3 weeks because of personal stuff so I have to catchup [16:57] Welcome back. [16:58] I'm going to downlad latest iso and install Kubuntu again [16:58] apachelogger: 7 sessions from kde/qt [16:58] Riddell, ScottK, debfx: If you look for the multiarch path, that might be subtly different; I think you want DEB_HOST_MULTIARCH [16:58] The triplet is in DEB_HOST_GNU_TYPE [16:59] for instance on i386, DEB_HOST_MULTIARCH=i386-linux-gnu and DEB_HOST_GNU_TYPE=i686-linux-gnu [16:59] lool: Thanks. [17:00] lool: well we need something that can be used outside of package building [17:00] cmake is used by non packagers too [17:01] Riddell: You need the multiarch library path? [17:01] lool: yes, to find libraries in multiarch directories [17:01] So slangasek has proposed a lsb-architecture helper I think [17:01] but I'm not sure we're using that [17:02] apparently, this is not used in dpkg; instead, gnutriplet_to_multiarch() basically returns the triplet except on i386 [17:03] Riddell: I would talk to slangasek and ask about the plans there [17:03] he just came online some minutes ago [17:03] http://paste.ubuntu.com/582163/ is the dpkg-dev triplet [17:03] best thing he came up with was $(gcc -print-search-dirs | cut -f2- -d'=' |sed -e's/:/\n/g'); do readlink -f $dir; done | uniq [17:05] 18:04 < slangasek> yes, even if we were using lsb-architecture, this is not an agreed lsb interface yet [17:05] 18:04 < slangasek> dpkg-architecture -qDEB_HOST_MULTIARCH is the only standard interface so far [17:06] ah, that seems more promising [17:06] But that's dpkg specific [17:06] good enough for now I think [17:06] what kdepim is kubuntu going to ship? [17:06] DEB_HOST_MULTIARCH is what I was mentioning earlier, not sure whether you got it [17:06] It was added to dpkg-architecture in natty recently [17:23] debfx: how about this? http://paste.kde.org/7604/ [17:55] shadeslayer: the tab bar in rekonq doesn't like to redraw itself properly today, have you seen that? [17:55] yep [17:55] known regression [17:55] lemme check logs if it was fixed [17:56] commit ae58104b91693b0b1d7404b5fe7600d6c157a5ce [17:57] Riddell: rekonq should have a 0.6.90 release soonish [17:57] Riddell: we should hardcode the paths in that cmake files (i.e. calling dpkg-architecture only while building cmake) [17:58] debfx: why? [18:00] parley : Depends: kdeedu-kvtml-data (= 4:4.6.0-0ubuntu2) but 4:4.6.1a-0ubuntu1 is to be installed [18:01] that's why kubuntu mobile isn't building on ARM [18:01] which I guess is due to the cmake issue plus the gles issue [18:02] shadeslayer: with spell check? :P:P [18:02]