[19:58:11] * dilfridge has changed topic for #gentoo-meetings to: "Gentoo Meetings | FOSDEM talk -> #gentoo-fosdem ! | KDE team meeting 17 Jan 1900 UTC http://wiki.gentoo.org/wiki/Project:KDE/Meeting/2013-1" [19:58:46] --> creffett (~creffett@gentoo/developer/creffett) hat #gentoo-meetings betreten [20:04:33] (kde) abcd, ago, alexxy, creffett, dastergon, dilfridge, jmbsvicetto, johu, kensington, mschiff, patrick, reavertm, scarabeus, thev00d00 [20:04:38] 1. roll call [20:04:46] -*- dastergon is here [20:04:51] -*- mschiff here [20:04:52] -*- dilfridge here [20:05:02] -*- creffett is here [20:05:50] -*- alexxy partialy here [20:06:05] wow, we actually reached our minimum for a meeting [20:06:11] I may faint. [20:06:35] I'll give a couple more minutes for any stragglers, then we'll get started [20:08:41] -*- alexxy need some sleep, group theory is cool but... [20:08:49] I'm here and late :-) [20:09:04] also cooking ;-) [20:09:11] good enough for me, let's get this show on the road [20:09:32] 2. Keywords [20:09:59] <-- EM3RY (~quassel@204.12.167.10) hat das Netzwerk verlassen (Ping timeout: 248 seconds) [20:10:07] we dropped ppc64, but do we want to add any stable/testing keywords? [20:10:36] -*- creffett would be interested in getting arm stabilized at some point [20:11:38] --> dMaggot (~user@201.227.175.3) hat #gentoo-meetings betreten [20:13:17] creffett: +1 [20:15:23] 3. CMake minimum requirement [20:15:35] kdelibs-4.10 will require CMake 2.8.8 as a minimum. Should we bother enforcing this (given that 2.8.9 is already the lowest version in-tree)? If so, where - kdelibs, KDE eclass, or even CMake eclass? [20:16:12] err sorry got distracted [20:16:23] arm is great [20:16:51] re 3) I'd say kde eclass [20:16:54] just to make sure [20:17:21] +1 for 3 [20:17:48] cmake-utils.eclass says CMAKE_MIN_VERSION="${CMAKE_MIN_VERSION:-2.8.4}" [20:17:58] maybe it's reallly easiest there [20:18:09] (change the default min version to 2.8.8) [20:19:01] sounds good to me, I'll go make that change [20:19:03] +1 [20:19:24] +1 [20:19:44] 4. Bugs [20:19:56] --> lmiphay (~lmiphay@86-45-44-228-dynamic.b-ras2.prp.dublin.eircom.net) hat #gentoo-meetings betreten [20:19:58] x11-themes/oxygen-gtk: Rename x11-themes/oxygen-gtk to x11-themes/oxygen-gtk2. Import x11-themes/oxygen-gtk3 from the overlay or we keep x11-themes/oxygen-gtk with 2 slot and rename x11-themes/oxygen-gtk3 to x11-themes/oxygen-gtk. [20:20:21] basically, split it or slot it? [20:21:07] what does gnome team say? [20:21:21] I don't know if we've asked gnome team [20:21:35] they should know best [20:21:55] okay, will ask them [20:21:55] x11-libs/gtk+ itself uses slots.. [20:22:40] I know that some packages have weird methods of handling gtk+ 2 vs. 3 [20:22:57] *shrug* [20:23:12] we'll ask GNOME team about it [20:23:17] media-libs/qt-gstreamer - split out QtGlib (bug #439198) discuss, vote [20:23:38] I wrote upstream about that [20:23:53] and another question I had (basically gstreamer 1.0 support) [20:24:02] unfortunately haven't got a reply about it [20:24:09] okay [20:24:19] I actually have a similar problem [20:24:29] I'm using gstreamer 1.0 from the gnome overlay [20:24:37] and haven't been able to update telepathy ever since [20:25:00] I personally am not in favor, my comments are on the bug but basically I think that it's making things needlessly complex and if upstream eventually does split it out we'll write an ebuild then [20:25:08] can we differ this decision at least for a coumple of weeks? [20:25:18] I'll try to ping upstream about this again in the weekend [20:25:32] sounds good to me [20:25:41] any further comments on this one? [20:25:50] I'll take that item and will inform of any updates [20:26:11] okay [20:26:19] kde-misc/polkit-kde-kcmodules: Store polkit configuration changes to /etc instead of /usr (bug #438790) Discuss/vote: long-term solution [20:26:24] -*- creffett shudders at this one [20:26:53] here, the main question is "what does upstream think" [20:27:14] there was a thread about it on the release list some time ago [20:27:18] mhm [20:27:33] there was no response on the upstream bug itself [20:27:51] imo nothing but the package manager should ever change anything in /usr [20:28:33] I think I summarized the responses from the list a couple meetings ago, but the general idea was that a bunch of distros have their own custom patches to change things, and upstream basically said "uh, why are we still doing this if everyone patches it out" [20:29:41] and sometimes they learn from it and fix it [20:30:16] and they said that in Frameworks 5 you'll be able to export XDG_CONFIG_DIRS to specify these things [20:30:32] there are actually two issues here, since kdm does the same thing [20:31:33] and for that, a number of distros just symlink /usr/share/config/kdm to /etc/kde/kdm [20:32:33] that's because kdm doesn't use kconfig for its settings [20:33:00] --> toralf (~toralf@d227155.adsl.hansenet.de) hat #gentoo-meetings betreten [20:34:16] <-- toralf (~toralf@d227155.adsl.hansenet.de) hat das Netzwerk verlassen (Quit: Konversation terminated!) [20:34:35] note that all of the proposed solutions were for main KDE, not polkit-kde-kcmodules, the general consensus on that one was "we're not shipping it" [20:34:41] so...opinions on what to do? [20:36:30] does it fix a bug? [20:36:48] https://bugs.gentoo.org/show_bug.cgi?id=438790 [20:37:50] it has a temporary fix of CONFIG_PROTECT, but no long-term fix [20:38:36] how about symlinking this to somewhere in /etc? [20:38:49] could work [20:38:55] I'll experiment with that and report back [20:39:19] Change the default DB backend for app-office/akonadi-server (bug #441596) Discuss/vote: there are some comments that the mysql backend brings improved stability/performance [20:40:03] any input here? this seems like one that needs reavertm's input [20:40:36] I've found that mysql is much better [20:40:49] that is on a high power desktop [20:41:01] but stability wise much nicer [20:41:29] I am using postgres as backend... but I would not make that the default ;) [20:41:43] you all know the kde wiki page about this? [20:41:50] I don't [20:41:53] moment [20:42:20] http://techbase.kde.org/Projects/PIM/Akonadi/Database [20:43:50] okay [20:44:00] hmm well I would say make mysql default [20:44:01] only real problem with the "full databases" as mysql or postgresql is that e.g. a laptop crash (dead battery) may leave you with a broken database [20:44:14] agreed [20:44:17] but then, you basically only lose a cache, so why bother [20:44:24] vote: make mysql the default backend? [20:44:25] +1 [20:44:27] +1 [20:44:35] not with goto settings, no [20:44:42] s/goto/good/ [20:44:56] mschiff: what do you mean? [20:45:06] YOu can configure postgres to be safe [20:45:25] --> tampakrap (~tampakrap@gentoo/developer/tampakrap) hat #gentoo-meetings betreten [20:45:26] to be protected agains broken database [20:45:33] hi [20:45:37] ah ok [20:45:40] hi dad [20:45:43] because of battery problem etc [20:46:02] hi tampakrap [20:46:36] mschiff: ok so you say, the crash problem is not fundamental [20:46:43] I had mysql running for quite a long time with akonadi on my laptop and never had an issue with it [20:46:46] ok [20:46:48] right [20:47:23] and I *had* cases where it crashed or I had to switch it off the hard way [20:47:26] so again [20:47:31] vote: make mysql the default backend? [20:47:37] +1 [20:47:40] +1 [20:47:50] +1 [20:47:57] +1 [20:48:00] cool [20:48:16] remember changing the default useflag settings in the ebuild... [20:48:23] I'll note it on the bug [20:48:37] -*- dilfridge thinks this will improve our relations to the kdepim guys [20:49:05] is it that bad? 8-) [20:49:06] dilfridge: and maybe with the improved relations you can do something about kmail? ;) [20:49:14] sigh [20:49:20] what do you want, a miracle? [20:49:39] app-office/akonadi-server-1.9.0: add Qt 5 support (bug #450412) How do we want to start rolling out qt5 support in general? [20:49:47] I already thought of porting kmail-4.4 to kdepim-4.10 [20:50:17] I am looking at the git logs of kdepim quite often... they are really doing a lot in fixing bugs since some time... [20:50:53] our overlay has a kmail-4.4.11.1-r100.ebuild, that is my personal playground [20:51:03] (does not build right now) [20:51:05] as I understand it, qt4 versus qt5 has to be global, you can't have both at once [20:51:29] bastards [20:51:52] haven't tried qt5 yet (I masked it) but I don't think one blocks the other (in terms of ebuilds, don't know about builds) [20:52:17] there are some comments in the bug about how you can't have, say, qt5 libs and a qt4 consumer [20:52:45] question number 1: should we start adding qt5 support? [20:52:49] right, that's the type of issues I would expect: you probably would have to rebuild everything to use one or the other [20:53:00] <-- ABCD (~quassel@gentoo/developer/abcd) hat das Netzwerk verlassen (Ping timeout: 245 seconds) [20:56:02] any comments? [20:56:18] --> ABCD (~quassel@gentoo/developer/abcd) hat #gentoo-meetings betreten [20:56:38] how? [20:57:04] adding the USE flag to packages that have support? [20:57:05] what? [20:57:23] but KDE SC itself will not support qt5 right? [20:57:33] (4.xx) [20:57:53] did qt team say anything how they want the useflags? eg. either qt4 or qt5, or only a qt5 useflag that disables qt4 support, or... ? [20:58:09] -*- creffett adds this one to the "talk to other teams" list [21:00:00] 5. Open floor [21:00:05] afterwards kde will need akonadi[qt4] or akonadi[-qt5] [21:00:07] yeah [21:00:18] someone please tell the qt guys not to be ridiculous [21:00:29] a category "qt" is a joke [21:00:30] -*- creffett nominates kensington in his absence [21:00:43] "qt-base" if anything :D [21:00:50] no, the joke is suggesting dropping the qt-prefix [21:01:00] noooo at is much better [21:01:05] bug 450818 is goint to bite our... well, you know, our feet [21:01:10] *qt [21:01:23] qt/core, qt/dbus, etc., etc....sure, let's just drop a bunch of name collisions into the tree [21:01:33] there's still no nswer fro teh qt team, from qt upstream, etc [21:03:43] dMaggot: pesa's question should be answered (re: where is it from, has it been upstreamed) [21:04:47] https://bugreports.qt-project.org/browse/QTBUG-29082 [21:05:05] <-- ABCD (~quassel@gentoo/developer/abcd) hat das Netzwerk verlassen (Ping timeout: 245 seconds) [21:06:08] works for me [21:06:42] creffett: right, I should have updated the bug with that info [21:06:50] dMaggot: just done [21:07:00] what else do we have for open floor business? [21:07:26] cookie distribution [21:07:36] fine [21:07:40] cookies to all who attended [21:07:56] -*- dilfridge munching [21:07:57] and some cookies for tampakrap's baby [21:08:10] plasma-active is about half packaged in my local overlay so far [21:09:03] --> ABCD (~quassel@gentoo/developer/abcd) hat #gentoo-meetings betreten [21:09:27] I'm still trying to figure out how to package "maliit", which is a dep of one of the packages which provides the input method framework (if anyone has a good example of a package that is configured by way of qmake, please let me know) [21:10:39] eww [21:10:42] but other than that it's nearly ready to hit the overlay, though I'm not going to package the patches until the next Active release because they don't apply to latest KDE and I really don't want to respin a few hundred line long patch [21:10:56] I'll probably be putting it in category plasma-active [21:11:13] push it! and then write to -dev and request kde-active category... [21:11:40] it can wait until we've actually tested the packages :) [21:11:58] plasma-active is less easier understood than kde-active I'd say... [21:12:09] yeah, but the official name upstream is plasma active [21:12:20] :| [21:12:27] eh, we can bikeshed it once I'm finished packaging [21:12:28] they are really branding specialists [21:12:48] mhm [21:13:09] any other business for open floor? we still have time [21:13:19] kde-plasma-active? [21:14:08] Is anybody else using the 49.9999 packges? [21:15:15] mschiff: no but I'm already on .98 [21:15:25] k [21:15:34] just for the record [21:15:58] I had a good time using 4.9.49.9999 and will keep to do so with 4.10 (already do) [21:15:58] mschiff: kensington: btw the pykde4 patch that I added does not help, I still get a sandbox violation [21:16:08] in pykde4-4.9.98 [21:16:20] there is one more commit that could cause it [21:16:28] -> weekend or anyone else [21:16:59] kensington is working on porting all python dependencies to python-r1.eclass btw [21:17:09] nice [21:19:39] I don't envy him that [21:20:58] nothing else? /me finds a gavel [21:21:17] *bam* meeting adjourned [21:21:27] yeah yeah yeah :) [21:21:46] ;) [21:21:53] * dilfridge has changed topic for #gentoo-meetings to: "Gentoo Meetings | FOSDEM talk -> #gentoo-fosdem !"