[01:19] ::qt-bugs:: [1290155] package libqtcore4 failed to install @ https://bugs.launchpad.net/bugs/1290155 (by Raphael Silveira) [04:55] shadeslayer: Please have a look at the kde4libs version requirement in your last kwalletmanager upload and then fix. It's stuck dep-wait ATM. [05:50] hi all, I see this https://trello.com/c/lXnsdJw1/23-kde-sc-4-13-beta1 but no more notes on it -- the Dot article is going out in a few hours and there is nothing about us..... [05:55] this cold is horrid, so I'm turning off my computer now [07:07] good morning [09:43] apachelogger: any update on bug 1281036? [09:43] bug 1281036 in kde-workspace (Ubuntu) "Why is kdm still depending on consolekit?" [Undecided,New] https://launchpad.net/bugs/1281036 [09:56] yofel: ck pointless, logind patch somewhere and intended to land upstream, whether or not that is actually going to happen I do not know [09:57] there was something peculiar though [09:57] like a lot of stuff depends on ck [09:58] I know that ximion did a bunch of things to get rid of ck in tanglu, including patching. So I guess we'll postpone that to 14.10? [10:00] Good morning. [10:01] yofel: well the kdm level dep can be removed, there is no point in us having it [10:01] but for example polkit-qt also deps ck [10:01] morning [10:01] so even if the dep goes away from kdm you won't get rid of ck unless polkit-qt is fixed [10:01] Riddell: yo [10:02] moin Riddell [10:02] true [10:05] shadeslayer: did we mis-communicate over 4.12.80? [10:14] ooh interesting, dolphin got a dropbox plugin fileviewdropboxplugin.so [10:17] oh [10:17] yofel: I totally fixed the bug in workspace bzr... :P [10:17] oh oops ^^ [10:19] someone is fixig this errors http://qa.kubuntu.co.uk/ninjas-status/build_status_4.12.80_trusty.html ? :) [10:21] soee: yes, but any help appreciated [10:24] yofel: after looking at it for a bit I think the polkit dep could also simply be removed [10:24] though it does direct dbus stuff, so I am not sure if that would not introduce unintended lockups due to stuff not being there (qtdbus likes to do that) [10:40] shadeslayer: added libqmobipocket-dev to kfilemetadata, unless there's a reason you missed it out [10:45] newest bug reports make me cringe again [10:46] so very pontless [10:47] Riddell: sure if i would only know hot to do it :) i have to many projects (web development) to find time to get into packaging :/ [10:48] you know, I really need web developers btw :P [10:48] apachelogger: if you need help feel free to contact me:) [10:49] there's a card or two having to do with web development ^^ [10:49] card or two ? [11:08] ::workspace-bugs:: [1281036] Why is kdm still depending on consolekit? @ https://bugs.launchpad.net/bugs/1281036 (by Simon Rettberg) [11:10] why is kdm still existing is a better question :) [11:20] oh kdm was a life savior few times last week :D [11:21] soee: why's that? [11:23] Riddell: i had some problems when lightdm didn't start and kdm worked for me [11:23] but after fresh install 2-3 weeks ago all seems to be just fine [11:25] btm im impressed how quick sometimes bugs get fixed in ubuntu :) one day after my report there was fix released https://bugs.launchpad.net/ubuntu/+source/ubuntu-drivers-common/+bug/1289420 :) [11:25] Launchpad bug 1289420 in ubuntu-drivers-common (Ubuntu) "Regression in the gpu-manager: bbswitch quirks erroneously applied" [High,Fix released] [11:28] 'Morning all [11:53] small review http://www.theregister.co.uk/2014/03/10/ubuntu_flavours_review/?page=2 [11:53] "While OpenSUSE remains my favorite KDE distro, Kubuntu 14.04 is looking like a great way to run KDE atop a Debian-based system." hmm [11:54] I tried opensuse and still really dislike the re-themeing and yast stuff [11:54] used it a long time ago, than like a year ago in VM [11:55] but Kubuntu is the one [11:55] i remember one feature i liked about opensuse i think [11:55] it was one click installation button [11:56] ihave open suse 13.1 on a separate hdd, since I heard good things , but 2days in and i have repos problems [11:57] don't think I'll be playing around with it much longer === soee_ is now known as soee [12:16] morning shadeslayer :) [12:17] good morning :) [12:17] my computer is so slow [12:18] shadeslayer: it'll be that god awful operating system you use get windows 8.1 it'll cure all that for you ;) [12:18] yeah, I hear it has boxes and rectangly stuff [12:18] sounds all cool and futuristic [12:19] shadeslayer: Haha [12:20] Yeah the box is what holds the dvd and the rectangles are the things that leak from your wallet to pay for the sticker that accompanies the dvd :D [12:22] Riddell: mind fixing kwalletmanager? [12:22] 4.12.3 in Trusty [12:22] shadeslayer: ok [12:22] I don't have upload rights :( [12:22] W8.1 is a tablet OS , doesn't belong on a regular desktop or lappy [12:24] most of my friends are windows users, and they're switching back to W7 on any new machines they buy [12:28] * Riddell doesn't see much of interest on http://summit.ubuntu.com/uds-1403/ except 1 qt session [12:28] speaking of which, how's the 5.2 upload going Mirv? [12:30] shadeslayer: can you remember why you made the baloo package replace kde-runtime ? [12:55] trying to overwrite '/usr/lib/kde4/kio_timeline.so', which is also in package kde-runtime 4:4.12.3-0ubuntu1 [12:55] ah hah [12:55] baloo does overlap [13:06] Riddell: powers that be waiting for bug #1207270 fix to be perfect plus at the same time some autopilot test regressions hunting. the landing itself would be ready, everything is at https://launchpad.net/~ci-train-ppa-service/+archive/landing-006/+packages [13:06] bug 1207270 in Ubuntu UI Toolkit "Font size too small in browser with Qt 5.1 / DPR rework" [Critical,In progress] https://launchpad.net/bugs/1207270 [13:06] I'm now entertaining myself with arm64 builds so that those could also be binary copied [13:06] powerpc/ppc64el is in pretty good shape [13:07] ahoneybun_: about the navigation, it's too much, just make it an actual item at the top of the page, not magically appearing [13:08] it makes the site too busy [13:08] also it makes for random visual nonesense [13:10] ahoneybun_: http://i.imgur.com/hQUnhEc.png like here I have a menu ontop of a menu, and the overlay menu doesn't even have a search, so it's a 50% less useful menu than the menu it is layed over [13:17] Riddell: did you get my email with all of the ready packages? [13:19] sgclark: yes I should have uploaded them all [13:20] although do let me know if I didn't [13:22] Riddell: kdoctools, kio, kservice need upload [13:23] ok onto it [13:24] sgclark: kdoctools is the latest [13:25] hmm, ok let me check [13:25] sgclark: kio and kservice should also be the latest [13:26] Riddell: I have no idea why they are red then :( I fixed not-installed [13:28] sgclark: mm yes you may be right looking at the contents of the packages in the PPA, will upload again [13:29] Riddell: now you know :) [13:30] Riddell: also I noticed krunner has the 4 version thing going on [13:30] sgclark: uploaded again, let's see how that does [13:30] Riddell: thanks! [13:30] shadeslayer: know what? [13:30] Riddell: why baloo replaces kde-runtime [13:31] ah yes, I do [13:32] Riddell: I think all that was left for baloo was license stuff [13:34] shadeslayer: yep, onto it [13:34] shadeslayer: I also split out the libraries [13:34] Riddell: cool :) [13:35] and just removing nepomuk from kde-runtime, ah nepomuk, we knew thee well atrocious name and all [13:35] why remove it? [13:35] lol crashtastic [13:35] no longer builds with it [13:36] Riddell: huh? Amarok still depends on Nepomuk [13:36] well [13:36] 'depends' [13:36] so if you have nepomuk installed, you get meta data integration with amarok [13:36] same for digikam I believe [13:36] kde-runtime doesn't build with nepomuk any more by default [13:36] Riddell: I'd consider that a problem [13:37] we could turn it on if those parts are still needed [13:37] vhanda around yet to ask him? [13:37] nope [13:37] poke Alex / tsdgeos for the correct course of action here IMHO [13:39] apachelogger: looks like no one cares about app-install-data or knows how to fix it :( [13:40] shadeslayer: Did you get my ping? [13:40] * Riddell adds an KDERUNTIME_BUILD_NEPOMUK=true [13:40] ScottK: kwalletmanager? [13:40] shadeslayer: Yes. [13:41] ScottK: I can't upload a fixed version, Riddell should have fixed it [13:42] shadeslayer: yeah, I saw, much fancy [13:42] still on the todo list, will get to it shortly [13:42] what needs doing? [13:42] Riddell: build depends is the wrong version [13:42] shadeslayer: supposedly one could do that stuff outside canonical tho :P [13:42] apachelogger: "supposedly" [13:42] just requires an archive mirror and a bit of python haxxoring to make it run [13:42] it be not documented afaik [13:42] archive mirror ought to be the most exhausting thing here ^^ [13:43] mm [13:43] shadeslayer: good code is documentation :P [13:43] really, the problem is the mirror part [13:43] you need a full local mirror from what I have seen [13:43] oh [13:43] then the crawler will walk through all packages and extract desktop files [13:44] I /could/ potentially do it [13:44] shadeslayer: just send a bump mail to the thread :P [13:44] every day [13:44] until someone does something ^^ [13:44] apachelogger: https://trello.com/c/3qlMcV3D/87-evaluate-apport-kde [13:44] we do need apport-kde [13:45] to submit stuff to daisy [13:45] no we do not [13:45] we need python-apport to submit stuff to daisy [13:45] we need apport-kde to handle !kde apps [13:46] oh hm [13:46] ScottK: shadeslayer: kwalletmanager uploaded! [13:46] true true [13:46] that being said *we* don't need it because we almost exclusively deal with kde apps [13:46] so it is not mission critical for us [13:46] it is however sensible to have it anyway [13:46] right [13:46] apachelogger: https://trello.com/c/f3ax0qf7 < can I start working on that? [13:46] of course it is not maintained, so it is all a bit fishy [13:47] shadeslayer: not yet [13:47] ok [13:47] https://trello.com/c/jddah0bu [13:47] https://trello.com/c/ai7BGDWe [13:47] https://trello.com/c/q4sdntKb [13:47] actually [13:47] shadeslayer: https://trello.com/c/LXHZgmZR [13:48] other than that the QA boards still have stuff [13:48] apachelogger: did you finish refactoring? [13:48] alas, most of it is low profile, but still nicer than other things [13:48] shadeslayer: no, that doesn't block the bug fixes you have pending there tho [13:48] apachelogger: okay, new driver manager release then [13:51] shadeslayer: also FWIW, as far as I know there is no dep3 parser at this time so while the tool would be simple, the parser would need writing from scratch [13:51] yep [13:51] or well, if someone wrote a parser it's likely in perl :P [13:51] *nod* [13:51] we shall write one in the awesomest language known to man [13:52] assembly [13:52] :3 [13:52] what's wrong with assembly? :O [13:53] c is the nicest management container for asm, and asm is the nicest language to use in a c program :P [13:54] the perl fanboys should port their stuff to ruby, they should feel right at home while not preventing everyone else from reading the code ... [13:55] Riddell: kdoctools needs another push, mis-named manpages in not-installed fixed [13:58] sgclark: uploaded! [13:59] * apachelogger uploads muon with a bazillion patches [14:00] :P [14:01] muon worked pretty well for me so far [14:08] you clearly haven't used the test case document then xD [14:08] apachelogger: there be bugs in kubuntu-driver-manager [14:09] apachelogger: http://i.imgur.com/NjUFucz.png [14:09] Apply button should be enabled [14:09] but it's not [14:10] oh la la [14:10] one wonders how that can break with the changes I made [14:11] I don't think it was broken before, but just need to confirm [14:12] yeah works with 0ubuntu2 [14:12] oh wait [14:12] aha [14:12] apachelogger: yeah logic is busted :/ [14:12] werx for me [14:12] maybe I did not push [14:12] lemme check [14:13] apachelogger: test case : click on broadcom driver -> click on different radeon driver -> click on currently active radeon driver [14:13] goes kaput [14:13] and I can imagine why [14:14] well, I have no broadcom now do I? and you did not take my suggestions that you should introduce a fake mode :P [14:14] shadeslayer: how does that work though? [14:14] emit changed(true) -> emit changed(true) -> emit changed(false) [14:14] well but that is bugged to begin with [14:14] last one causes button disabling [14:14] so that should have been broken with ubuntu2 as well, no? [14:15] or did you only introduce diff handling later? [14:15] it is [14:15] this would need state tracking in the kcm :/ [14:15] no it wouldn't necessarily [14:16] oh? [14:16] in the slot of the module you go foreach widget: if (widget->didYouChange()) { emit changed(); return } [14:16] hm, yeah makes sense [14:17] it's cheap enough for this case [14:17] the other option would be simple state checking a la qmap [14:17] that's what I was thinking [14:17] or actually [14:17] you just need qset [14:18] if a widget is in the set, it has changed [14:18] all in all it's probably some 10 sloc :P [14:18] *nod* [14:18] next iteration [14:18] actually, you should do that [14:18] probably cheaper than asking the widget [14:18] since I want to get current bug fixes out [14:18] or not [14:18] it's probably all the same xD [14:19] shadeslayer: personally I'd go with the hasChanged() function on widget [14:19] it separates the data somewhat more from the decision of what to do with the data [14:24] Riddell: Thanks. [14:27] Riddell: http://paste.ubuntu.com/7067871/ [14:27] sgclark: it's a plugin, it's allowed to have weird dependency information [14:28] you can add a libkf5khtml5.lintian-override if you want [14:28] Riddell: ok thanks [14:30] Riddell: apachelogger plz upload http://people.ubuntu.com/~rohangarg/upload/kubuntu-driver-manager_14.04ubuntu3.dsc [14:33] shadeslayer: that should be in the set? [14:34] or not [14:35] shadeslayer: please check with cj why it's not I guess [14:35] Uploading kubuntu-driver-manager_14.04ubuntu3_source.changes: done. [14:35] Successfully uploaded packages. [14:36] cheerio [14:38] agateau: no nepomuk for gwenview now? [14:38] Riddell: it is all baloo powered, now! [14:41] * Riddell looks to gwenview for the bare necessities [14:46] Riddell: monday nag. I can has review for Ubiquity fix? https://code.launchpad.net/~agateau/ubiquity/kde-fix-cropped-greeter-buttons/+merge/208628 [14:49] Riddell: btw for arch specific install files I just do foo.install.armhf right? [14:49] hm [14:49] though that won't scale for when I want to build for all arm ach's [14:49] *arch's [14:49] shadeslayer: no great answer that I've found [14:49] you can do foo.install.armhf and foo.install.arm64 [14:49] yeah [14:49] but then that's just duplication :( [14:49] but then you have to remember to keep them all up to date [14:49] yeah [14:49] or [14:49] symlinks [14:49] ln -s foo.install.arm foo.install.armhf [14:49] ln -s foo.install.arm foo.install.arm64 [14:49] you can also have the failed individually installed in debian/rules depending on the arch [14:49] voila [14:49] but then you have to live with --list-missing complaining [14:50] Riddell: how so? [14:50] some binaries will only be built on armhf/arm64 [14:50] ScottK: ^^ any opinions on symlinks inside packaging [14:51] pft [14:51] ScottK: ^^ any opinions on symlinks inside packaging [14:51] I haven't been following, so no opinion on specifics, but generically, it happens all the time. [14:52] agateau: yes, sorry, will get to it [14:53] ScottK: basically, kwin_gles should only be built for armhf and arm64 , and that would mean writing arch specific install files [14:53] Right. We've had those before. [14:53] No problem there. [14:53] cool, so I just write kde-window-manager.install.arm and symlink that to kde-window-manager.install.arm64 and armhf [14:54] I believe you can. [14:54] shadeslayer: please talk with MG first, to make sure that is still what he wants [14:54] thx :) [14:54] Or just symlink one to the other, no need for three. [14:54] apachelogger_: oh okay [14:55] shadeslayer: what are the files that differ? [14:56] Riddell: at the very least kwin_gles [14:56] I haven't built the package yet without gles to see what else changes [14:58] apachelogger_: sent him an email [14:58] oh, he is off again [14:58] he starts work way too early I say [14:59] ^^ [14:59] apachelogger_: I think he starts at 8 [15:00] some days it looks more like 7 :P [15:01] Q: is it just me or does debdelta look like a bloody big workaround [15:04] any idea why ark always neds a lot of time to open archive, even very small one ? [15:05] no [15:05] it largely depends on the compression format used though, some have more sensible access methods than others AFAIK [15:05] maybe im missing some dependency, not sure === apachelogger_ is now known as apachelogger [15:10] apachelogger: btw which cache rescans do you talk about here https://trello.com/c/q4sdntKb [15:13] shadeslayer: see description [15:14] did you just add that [15:34] shadeslayer: yes [15:34] * shadeslayer is trying to reproduce on a live session [15:38] apachelogger: http://people.ubuntu.com/~rohangarg/vids/Kubuntu%2014.04.webm [15:39] so what part of that is wrong? it scans for languages -> installs -> scans for languages again? [15:42] so why does that not play in dragon [15:42] if stuff was only broken half as much it would still be broken beyond repair [15:44] shadeslayer: see my comment that progress bar is not rescanning the languages, it is rescanning the xapian index [15:44] well, updating really [15:44] PHONON-GST BEGIN: void Phonon::Gstreamer::MediaObject::setError(const QString&, Phonon::ErrorType) [15:44] PHONON-GST "Internal data flow error." [15:44] :( [15:45] Riddell: khtml ready [15:45] apachelogger: I don't get a cache rescan when I move a language back [15:45] shadeslayer: I think gstreamer doesn't speak webm [15:46] shadeslayer: oh, well, even after install it shouldn't update [15:46] it does a qapt cache update, but that ought not cause the progressbar [15:46] no, wrong again [15:46] http://people.ubuntu.com/~rohangarg/vids/Kubuntu%2012.04.webm [15:46] it reopens the qapt/apt cache xD [15:46] the madness [15:47] ignore the name :P [15:47] its the same video [15:47] is it not [15:47] no, this one removes [15:48] no [15:48] it's the same video :P [15:48] now it's 403 [15:48] apachelogger: yeah try again [15:48] ah wrong commnd [15:48] shadeslayer: don't use apply at all [15:48] just move stuff around [15:49] yeah, I don't see a cache rescan [15:49] boom crash [15:49] shadeslayer: so what I see is -> start -> scan -> engb to selection -> scan -> engb out of selection -> scan ->... [15:49] nope, don't see that on the live session [15:49] basically whenver something moves I get a xapian rescan for some reason [15:49] also, i made it go kaboom [15:50] well that's not healthy [15:50] also [15:50] why do you test shit in the live session again [15:51] * apachelogger sighs at all the open tabs that now made him forget what he was looking at -.- [15:53] hmm, closing the lid disconnects from irc, didn't set it to do so. Must be a bug :/ [15:53] apachelogger: http://paste.ubuntu.com/7068347/ [15:54] * apachelogger sighs at qapt [15:55] apachelogger: http://open.spotify.com/track/4WamMWxFi6ts3xkIhivF5c [15:55] ? [15:56] apachelogger: https://www.youtube.com/watch?v=uKjeTadGPgY listen to this [15:56] starts off like really bad punk [15:56] is it really bad punk? [15:57] might easily be worse [15:57] * apachelogger throws a keyboard at shadeslayer [15:57] * shadeslayer catches [15:58] apachelogger: granted this is alot better https://www.youtube.com/watch?v=VvcohzJvviQ [15:58] same band [15:58] * apachelogger aint clicking youtube links from shadeslayer anymore [15:58] heh [15:59] the locale kcm is also very shitty [15:59] sgclark: uploaded! [15:59] it internally calls itself locale kcm, it really is called language though [15:59] ... [15:59] yeah [15:59] it's weird [15:59] ehm [15:59] wait what [15:59] whatwhatwhat [16:00] shadeslayer: that backtrace cannot be from a recent kcm locale [16:00] whatx4 [16:00] that backtrace should contain libkubuntu somewhere [16:00] the kcm has no direct access to the qapt classes anymore [16:00] ah, possibly old iso then [16:01] ........ [16:01] * shadeslayer is upgrading his install [16:01] shadeslayer: get yourself an autostart script that triggers a zsync after 15 minutes into the session [16:01] xD [16:07] Riddell: i386 sybol fix plasma-framework ready [16:11] shadeslayer: about that gst card, if you tag something blocked, at the very least paste the relevant IRC discussion :S [16:12] Riddell: can you push your baloo changes? [16:12] say you get hit by a bus and I need to go on rehab, someone might have a hard time finding out why exactly that card is being blocked by app-install-data [16:12] Riddell: alternatively, can I make gwenview depend on libbaloo-dev [16:14] Riddell: knewstuff, plasma-framework ready [16:15] apachelogger: the poor bus [16:16] shadeslayer: hang on, I'm looking at baloo and gwenview [16:16] oh [16:16] I was looking at gwenview [16:16] shadeslayer: although I'm a little stuck so maybe you can help [16:16] sure, ssup [16:16] Riddell: also I don't see khtml in que nor rebuilt recently so add that to the list [16:19] Riddell: also, KDE SC 4.12.3 copied for Precise [16:20] shadeslayer: awooga, lovely [16:20] shadeslayer: pushed baloo packaging [16:20] shadeslayer: I split it 1 package for each library as debian likes it [16:20] shadeslayer: kept 1 package baloo for everything else [16:21] shadeslayer: but this gives a runtime package that stuff like gwenview I expect ought to depend upon and the baloo package depends on the libraries [16:21] shadeslayer: so circular dependency [16:22] shadeslayer: so I altered the .symbols files to make them depend on baloo in the same way that libkdecore5 add a depends on kde-runtime in the .symbols files [16:22] shadeslayer: but it doesn't seem to work, at least gwenview doesn't depend on baloo [16:22] shadeslayer: any ideas? [16:22] you can add it manually I think, via debian/rules [16:23] shadeslayer: for gwenview? === PaulW2U is now known as G4MBY [16:23] moment, looking at packaging [16:24] Riddell: so gwenview depends on libbaloowhatever and baloo, baloo depends on libbaloo* .. how is that circular? [16:25] shadeslayer: gwenview depends on libbaloofiles4, libbaloofiles4 depends on baloo, baloo depends on libbaloofiles4 [16:27] but libbaloofiles4 depends on baloo only as a runtime dep [16:27] remove baloo from libbaloofiles4 dep [16:27] lemme check how to do that [16:28] so yes we could manually add baloo as a runtime dep to gwenview but I'd rather not have to remember to do that for every package which uses baloo [16:28] I'd rather libbaloofiles4 magically added a depend on baloo for anything that needed it [16:28] which my packaging should make happen [16:28] but doesn't === rudyismydog is now known as mydogsnameisrudy [16:32] Riddell: https://wiki.debian.org/Projects/ImprovedDpkgShlibdeps , looks like it should work [16:33] * shadeslayer looks [16:35] Riddell: building, give me a moment [16:40]