[00:20] Sput: no windows backend? other than the vlc backend you mean? :P [00:43] apachelogger: I wonder if we should turn off the locale missing notificaton for the live image [07:17] apachelogger: yeah, bundling vlc seems a bit overkillish for playing audio notifications, that's why I'm wondering if I should just use QtMultimedia for that [09:40] Riddell: gstreamer1.0-pulseaudio was not installed on update [09:41] hmm [09:41] also: pgst when built against gst1 doesn't support device control without pulseaudio apparently [09:41] still doesn't play [09:41] Riddell: does preview in dolphin work for you? [09:43] oh, it locks up, how nice [09:43] oh wait [09:43] wrong phonon xD [09:47] oh [09:47] dolphin loads webkit [09:47] who'd have though [09:47] t [09:48] for html preview maybe? [09:48] new webkit should be in archive now, update again? [09:49] but no dolphin preview doesn't work :( [09:51] nah, I think dolphin doesn't do video anyway [09:51] yes it does [09:51] Sput: vlc for you use case is probably like 1mib of compressed data [09:52] you just have to throw away the plugins you do not need [09:52] dolphin says "(dolphin:32349): GStreamer-CRITICAL **: gst_mini_object_unref: assertion 'GST_IS_MINI_OBJECT (mini_object)' failed [09:52] ** Message: don't know how to handle audio/mpeg" [09:52] which is weird since dragon plays it fine [09:53] Sput: anyway, you could also use windows media foundations directly I do not care, it's your call [09:54] (both phonon4 and qtmm are not really good at notification sounds btw, due to lack of pcm caching) [10:06] apachelogger: well, I'm mostly interested if QtMultimedia has downsides compared to Phonon, for our simple use case (also in Linux; for example, does QtMM work properly in a KDE desktop?) [10:07] qtmm sucks! phonon rocks! [10:07] Sput: what Riddell said [10:08] is that political, ethical or technical opinion? :) [10:08] Sput: I have no clue whether it works correctly in plasma, though it should as there's not much you can do wrong as long as sound gets piped through pa [10:08] qtmm has shittier API for your simple use case though :P [10:08] yes, but as Phonon is no longer a part of Qt5, people start to complain that it's an extra dependency... [10:09] well, I'll talk to TheOneRing (he asked me to figure this out) [10:09] Sput: an extra dependency? [10:09] extra on top of what? [10:10] "a separate package" :) [10:10] Sput: do it kde style, give options :D (add cmake options :D ) [10:10] windows has packages? :O [10:11] jussi: yeah, abstract the abstraction libraries ^^ [10:11] apachelogger: :D :D :D [10:11] jussi: yes, it's all optional already now; but I don't want to ship yet another audio notification backend [10:11] obviously for me it would be easiest to just continue with Phonon as the code is all there :P [10:12] (and works fine in Qt5 for that matter; thanks for creating proper CMake stuff!) [10:12] cmake omnomnom [10:13] well, I'll talk to TheOneRing. he's tempted to port the ds9 backend to phonon-qt5 because he thinks that shipping vlc just for notifications is a bit overkill [10:13] Sput: I think someone else was actually using the ds9 backend somewhere [10:13] Sput: you might want to tell him to consult with j-b though [10:13] hmm [10:14] I really doubt libvlc+libvlccore+fileio+vorbis/pcmdecoder&demuxer would be big [10:14] ok [10:14] Sput: just remembered, I think that someone else is also using ds9 because they thought vlc was too big for notifications [10:14] can't remember who though [10:14] for Qt5? [10:15] meh, the proper solution would be to just use knotify on all platforms! [10:15] Sput: for qt4 [10:16] just saying, there may be other interested parties [10:18] oh, we use the ds9 backend for qt4 [10:19] anyone uses the kwallet gpg backend here? [10:19] because I'm going to disable it [10:20] Sput: :O [10:21] Sput: oh random additional thought: windows media foundation may in general be more appropriate than ds9, so a backend rewrite may be handy [10:21] btw I will not be available till around 2 [10:22] shadeslayer: I am going to get drunk until then [10:22] I'm here for another 20 minutes or so though [10:22] apachelogger: have fun [10:22] good for me, I can sneak in patches [10:23] :O [10:23] outragous [10:24] yeah [10:26] like this one for example http://paste.ubuntu.com/7161940/ [10:28] 'Morning folks [10:38] apachelogger: since when did the mailing list become a bug tracker :O [10:38] shadeslayer: where? [10:38] Fwd: [Apport KDE] [2014-03-27] Automatic crash report generated by DrKonqi [10:39] I dunno [10:39] it's hard to file a bug if apport is broken, don't you think :P [10:39] a bit of a chicken-egg situation right there [10:39] sure, but a bit of googling would have landed you on the bug report ;) [10:40] shadeslayer: me? [10:40] no, the guy posted to the mailing list [10:40] s/to/who/ [10:40] shadeslayer meant: "no, the guy posted who the mailing list" [10:40] well yeah, you know who people are :P [10:40] s/who/how [10:41] :) [10:41] gtg [10:41] cya at around 2 I guess [10:41] o/ [11:03] Sput: hmm if people complain about extra dependencies that kindae puts the whole kf5 idea as being a complaint target [11:06] Riddell: have you had time to look at calligra? [11:24] sgclark: yep, uploading it now [11:32] Riddell: KF5 will only catch on as planned if deployment is easy also on other platforms than Linux [11:32] (which is why I keep telling people they should keep this in mind, but I'm not sure if that's actually happening) [11:33] I'm not going to make KF5 a requirement for Quassel (much as I'd love to) if it's not really easy to package in Windows [11:34] preferably, KDE provides some CMake stuff that can automagically download and build a list of KD5 modules and their dependencies for Win and Mac [11:35] (and makes sure that it's setup in a way where a normal CMake buildsystem can then pick up the components; I have no clue if CMake config mode actually works, as there's no standard location to install cmake configs to) [11:41] sgclark: up for some backporting? [11:41] sgclark: also calligra-l10n to be done with all its fiddle [11:43] Riddell: still trying to make a saucy pbuild env, not going well. Sure I can attempt i10n, which is top priority? [11:45] sudo pbuilder-dist saucy create [11:51] sgclark: either way around [11:51] sudo debootstrap saucy saucy when you can't be bothered with pbuilder [11:58] sgclark: calligra-l10n-package in kubuntu-automation may well help with calligra-l10n [11:58] sgclark: it's a weird multi source package [12:28] apachelogger: your command was successful, thank you [12:28] Riddell: starting with backport [12:32] Riddell: the calligra-l10n internal script didn't do the job? That was pretty easy to use last time I worked on it [12:33] yofel: I've always got confused by it and had to fix stuff up by hand [12:33] Riddell: since you don't like pbuilder, I can recommend lxc rather than regular chroots very much [12:33] ::workspace-bugs:: [1283820] content of windows (and desktop) appear black when using kwin, but not with kwin_gles - in... @ https://bugs.launchpad.net/bugs/1283820 (by avlas) [12:34] apachelogger: do you have an easy pbuilder-like lxc setup? I do like using it for manual building, but for simple testbuilds lxc is just a hassle [12:34] Riddell: good point, the usage documentation of it was... lacking [12:35] nope, I only looked into lxc yesterdya for like an hour, but it should not be hard to create lbuilder xD [12:35] lx-create -> then overlayfs snapshot per build instance [12:35] it's possible sure, and a overlayfs setup like auto-upgrade-tester uses would be sweet [12:36] all you need then is a way to conviently exchange the src package and binary results [12:37] which you could do via a mount (which may be a bit eww) or the host script would simply copy the data into the clone before calling execute and copying the data out before calling destroy [12:37] well, you can just copy that to /var/lib/lxc/*/rootfs/tmp/, the actual command executing needs to happen over ssh though I think [12:37] it really maps very well to how pbuilde rworks anyway [12:37] true [12:37] yofel: nope, you could stdin pipe [12:37] BUT [12:38] pbuilder doesn't do that either [12:38] instaed you'd deploy a tiny harness script into the clone and lxc-execute that from the host [12:38] if you say so, auto-upgrade-tester fiddels with ssh IIRC, which is why I said that [12:38] which I think is what pbuilder does as well [12:38] oh, good point [12:38] yofel: I think that's to mimic actual user input [12:38] you could still stdin pipe though I guess [12:40] I may change the neon5 builder to use lxc some time, quite possibly results of that can be salvaged into a possible lbuilder [12:42] is the freeze over? [12:42] package building is really the only thing that I still use pbuilder for. When I need a simple chroot I go with lxc usually (mostly because that has the debootstrap stuff cached) [12:43] shadeslayer: freeze is over when beta2 announcement is out [12:43] pft okay [12:44] Riddell: are you finishing off http://qa.kubuntu.co.uk/ninjas-status/build_status_4.12.95_saucy.html [12:44] * yofel just retried a bunch on that page [12:45] okay [12:54] shadeslayer: hmm I'd quite forgotten about it :( [12:54] pykde4 needs install file mangling (s/3.4/3.3/), otherwise we're good so far [12:55] *nod* [12:55] will get to it post lunch [12:57] yofel: oh, what I don't know yet, how do you enable non-root user execution of containers? [13:03] apachelogger: I've never actually tried it: https://www.stgraber.org/2014/01/17/lxc-1-0-unprivileged-containers/ [13:04] uhu, very hacky [13:04] let's not think about that for now ^^ [13:39] is there a way to see, which binary application registered which applications at the dbus-service? [14:19] yofel: placing data in the overlay is a bit of a hack xD [14:20] hmm, system settings>multimedia>audio&video settings doesn't open in 14.04. it just hangs with the cursor turning over [14:21] same goes for my desktop 14/04 install, altho it has pulseaudio installed [14:23] because Riddell broke all of multimedia [14:23] * apachelogger ^5's Riddell and welcomes him to the club :P [14:25] well alsamixer is still available [14:27] much interesting [14:27] kio_sftp is no more? [14:29] ah [14:29] nooooooo [14:29] needs 0.6.0 :( [14:30] shadeslayer: sftp works for me, what's up? [14:30] https://launchpadlibrarian.net/170854895/buildlog_ubuntu-saucy-i386.kde-runtime_4%3A4.12.95-0ubuntu1~ubuntu13.10~ppa1_FAILEDTOBUILD.txt.gz [14:30] needs newer libssh in saucy [14:30] backport it? [14:30] aye [14:32] yofel: if you want to write lbuilder... http://paste.ubuntu.com/7163019/ should have all the bits needed in terms of actual lxc interaction [15:27] hi, i want to debug ksmserver, but when i attach to ksmserver with gdb, it always says that is already debugged (which it is clearly not) and if i kill the other program, the session exits. i suppose, that is, because kinit is attached to ksmserver and when kinit is killed, the session ends. how can i debug ksmserver? [15:28] menace: please don't cross post :) [15:28] and I think you'll get your answer in #kde-devel [15:30] shadeslayer: let's see ;-) [15:31] but i thought, perhaps it is a distribution specific issue... [15:31] doubt it [15:34] since forking and attaching from kinit/session-beginning is sometimes handled differently, i was not sure... only reason [16:01] shadeslayer: does not look like it, that anyone's helping, does it? ;) [16:01] ask on the mailing list? [16:02] yeah, I'll subscribe to it [16:03] oh, i think, i found a possibility [16:10] Riddell: I got the GPG thing working [16:10] Riddell: your key must be fully trusted === shadeslayer is now known as shadeslayer_ === shadeslayer_ is now known as shadeslayer [17:02] Riddell: how's the release going [17:02] need any help? [17:27] shadeslayer: which GPG thing? [17:27] Riddell: kwallet gpg [17:27] shadeslayer: oh weren't we going to disable that? [17:27] Riddell: I did [17:27] but I also got it to work [17:28] afiestas *really* recommended that we disable it [17:28] what's the default? [17:28] GPG for me [17:29] and it says "for better protection" [17:29] but it doesn't work at all for a new user [17:29] yep [17:29] so very unfriendly [17:29] IMO, if upstream doesn't recommend it, they should be disabling it [17:29] you want to connect to wifi and it says use gpg and it breaks [17:30] or at least not-enable-by-default [17:30] shadeslayer: release in about 4 hours, needs a story on kubuntu.org and this checked over https://wiki.kubuntu.org/TrustyTahr/Beta2/Kubuntu === shadeslayer_ is now known as shadeslayer [18:09] Riddell: calligra saucy backport in my ubuntu one. Runs through saucy pbuilder with only one dep change. [18:13] sgclark: is it in a PPA? then we could just copy it [18:13] PPA? [18:14] yofel: I don't know what you mean sorry [18:14] sgclark: ah ok, PPA's are our launchpad archives, like kubuntu-ninjas/ppa or kubuntu-ppa/experimental [18:15] every launchpad user can have some so you might want to look into it [18:15] sgclark: https://help.launchpad.net/Packaging/PPA [18:19] sgclark: in the meanwhile, where's the package? Then I could upload it too [18:19] yofel: in my ubuntu one, I don't have your email to give access [18:19] yofel@gmx.net [18:33] :) [18:33] hello [18:33] hi Peace- [18:34] hi yofel [18:34] yofel: want to see something neat? [18:34] always ^^ [18:34] yofel: http://wstaw.org/m/2014/03/27/plasma-desktopUA2242.png [18:34] :3 [18:35] NERD o.O? [18:35] * Peace- hheheheh nerd [18:35] silly plugin name [18:35] for vim [18:35] hehe [18:35] yeah, I recognised vim, but that's cool indeed [18:35] yofel: http://wstaw.org/m/2014/03/27/plasma-desktoppX2242.png [18:36] notice the +'s on the left [18:36] :O [18:37] mah i love kate :P [18:37] yofel: oh, even more fun : http://wstaw.org/m/2014/03/27/plasma-desktopwH2242.png [18:38] :O [18:38] yofel: also, that's qvim [18:38] or Qt Vim :P [18:38] wait what, that exists? [18:38] gimme [18:39] :D [18:39] yofel: https://bitbucket.org/equalsraf/vim-qt/wiki/Home [18:41] yofel: ./configure --enable-gui=qt --prefix=/opt/vim-qt --with-vim-name=qvim --enable-rubyinterp --enable-pythoninterp --enable-python3interp [18:41] for extra pythonness [18:42] sgclark: did you use a different tarball to build calligra for saucy? Trying to apply your files to the trusty tarball gives: [18:42] dpkg-source: error: file ./calligra_2.8.1.orig.tar.xz has size 138470468 instead of expected 138471552 [18:43] yofel: I still haven't found a good way to find files inside a project [18:43] or jumping around into functions [18:43] ctags just seem very 90's [19:04] yofel: sorry having internet issues, will re-upload the orig file [19:08] sgclark: I would rather find out why they're mismatching, otherwise I'll have to do a full upload when a diff upload should be sufficient [19:08] yofel: my internet cut out possibly while it was uploading [19:09] sgclark: then it would be mine, as I downloaded the tarball from launchpad. The .dsc has the filesize in it though and that doesn't match [19:13] yofel: did you get it ? I am building off tarball retrieved from depot [19:18] for some reason your tarball doesn't match with depot, MD5: f35d3b792875b31576804090872a9771 != 85629dd9ed72a0b57ed279528b0879d7 [19:19] but now I can rebuild the package easily [19:19] what ? how did that happen? have I just wasted days of work? I am confused [19:23] sgclark: it should be fine, after all the archive upload to trusty used the correct tarball [19:23] I'll just take your changes and rebase them on the right file [19:24] ok [19:26] sgclark: what's fixMsooXmlTableStyleCase.diff ? [19:27] sgclark: aah, the tarball was respun upstream to fix the error that this patch fixes [19:27] hence the different files [19:27] yofel: ahh [19:28] that makes sense then [19:35] sgclark: what exactly did you change? Your packaging was a bit outdated so the diff is a bit large and I don't really see what you changed for the backport http://paste.ubuntu.com/7164538/ [19:39] yofel: my packaging was a bit outdated? This is my first backport and all I did was run it against pbuilder saucy and had to change okular depend version, as newest does not exist in saucy. Your paste looks like all of the work I did for trusty build... [19:40] sgclark: that's the diff between the trusty archive package and yours [19:41] backporting usually means taking the package from trusty, and adjusting it for saucy. Here I guess jr changed the package before upload so they don't match [19:41] Oh I think I know what happened, looks like I failed to bzr add patches [19:42] though we no longer need so I should rebuild calligra with new tarball? [19:46] more going on there though. I don't know what all he changed [19:48] I'll look at things in detail in a moment, a bit busy right now [19:49] yofel: no there is something very wrong here, patches being added back in that should have been removed. Disregard this being ready. [19:49] the patches are there because the tarball is different [19:50] yofel: I am thinking I want to grab the new tarball and rebuild. What do you think? [19:51] sgclark: probably best, run this: [19:51] dget -xu https://launchpad.net/ubuntu/trusty/+source/calligra/1:2.8.1-0ubuntu1/+files/calligra_2.8.1-0ubuntu1.dsc [19:51] then you'll start from the current trusty package [19:52] ok [19:57] sgclark: do you have a gpg key? I could explain you how to setup a PPA if you want. Having the package in an archive makes it easier to a) test it as we have the binaries b) review it as it's in a public location c) one can copy packages between archives easily [19:59] yofel: yeah, I set up a persoanl ppa. will upload this to that when ready [20:00] yay [21:41] beta 2 feels nearly out [21:42] yofel: I grabbed the tarball from depot and that error (the one I created patch for) is still there. Where did you get your tarball that was fixed? [21:42] sgclark: I ran the dget line I gave you [21:43] which downloads the tarball from launchpad [21:43] and that has the same md5 as depot [21:45] sgclark: wait, which patch are we talking about exactly? [21:45] well the depot was not fixed so I am back to havving no idea why the mdsums would be off [21:46] BorDer needs to be Border in MsooXmlTableStyle.cpp [21:46] let me verify this [21:48] sgclark: ok, seems like upstream messed up again? [21:48] 85629dd9ed72a0b57ed279528b0879d7 is indeed still broken, and I just realized that raymond complained about that today [21:48] I don't know but this is getting frustrating. [21:49] sgclark: in any case, we'll have to fix the trusty package first, as that's broken too [21:50] sgclark: the tarball was respun though, so the md5sums were still off [21:50] yofel: I am going to apply my original fix to this latest depot I downloaded ok? [21:50] it's just that the new tarball is broken too [21:51] will that resolve mdsums? [21:51] sure, as long as yours is 85629dd9ed72a0b57ed279528b0879d7 [21:52] yofel: verified, moving forward [22:16] agateau: are there any plans to do a release of libdbusmenu-qt with the cmake config stuff and Qt5 support in? [22:16] 0.9.2 is getting a bit stale :P [23:36] Sput: IIRC he stopped dealing with it when he left Canonical. [23:39] Subject: [ubuntu-release] Ubuntu 14.04 LTS (Trusty Tahr) Final Beta released [23:40] whee [23:44] btw, our https://wiki.kubuntu.org/TrustyTahr/Beta2/Kubuntu shows down at the Software Common to all Ubuntu flavours section: [23:44] Include: Nothing found for "^##StartTrustyReleaseBugs"! [23:44] Include: Nothing found for "##EndTrustyReleaseBugs"! [23:44] in red [23:50] hmm, some breakage there, wasn't like that this afternoon [23:51] * Riddell removes [23:52] * Riddell publishes http://www.kubuntu.org/news/kubuntu-1404-lts-beta-2 === mhall119 is now known as mhall119|offline [23:59] weeeee [23:59] slightly concerned we'll now get lots of bugs about amarok crashing from people who need to update to latest qtwebkit