[17:04] KDE SIG meeting start, who's present today? [17:04] present [17:04] ping: ltinkl, Kevin_Kofler, jreznik , kdefoo* [17:05] * jreznik is here [17:05] Present. [17:08] looks like agenda is still empty, topics? [17:08] [10:55] Updates: cmake 2.6.3, kde-plasma-networkmanagement (new package), Qt 4.5 [17:08] [10:59] https://bugzilla.redhat.com/show_bug.cgi?id=qtscriptgenerator​ [17:08] Bug 489550: medium, medium, ---, jreznik, ASSIGNED, Review Request: qtscriptgenerator - Qt bindings for Qt Script [17:08] [11:01] taglib-extras: https://bugzilla.redhat.com/show_bug.cgi?id=491647​ [17:08] Bug 491647: medium, medium, ---, nobody, ASSIGNED, Review Request: taglib-extras - Taglib support for other formats [17:08] * SMParrish wakes up [17:09] than is sick today, won't be here. Send him good moji/karma to get better. [17:09] *** rdieter sets the channel topic to "KDE SIG Meeting -- https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2009-03-24​ -- cmake-2.6.3". [17:09] first, cmake-2.6.3 update [17:10] this probably should land soonish, before or with kde-4.2.2. it's in updates-testing now [17:11] there's one bug open for a pkg failing to build with it though. [17:11] (A package which is not yet in Fedora, but currently under review.) [17:11] --> lfoppiano_ has joined this channel (n=lfoppian@host148-131-static.28-87-b.business.telecomitalia.it). [17:12] <-- thomasj has left this server ("leaving"). [17:12] oh, ok [17:12] CMake crashes on it, which is a CMake bug, there's a quick hack patch we could apply, but I think it leaks memory. [17:12] <-- yunustj has left this channel ("Leaving"). [17:12] But even if we fix that, it'd still not build because it expects CMAKE_CURRENT_LIST_FILE to contain something other than what it contains with 2.6.3. [17:13] And I'm not sure what's at fault there. [17:13] <-- thomasj__ has left this server ("leaving"). [17:13] --> thomasj has joined this channel (n=thomasj@e180149179.adsl.alicedsl.de). [17:13] <-- thomasj has left this server (Client Quit). [17:13] is cmake upstream involved or notified yet? [17:14] if not, probably should [17:15] <-- SMParrish has left this server (Remote closed the connection). [17:15] I think it hasn't been reported upstream yet. [17:15] I agree it should. [17:15] shrug, I'm on the fence either way... I guess I'd lean toward keeping it in testing a little while longer [17:16] it should be reported upstream! [17:16] There should probably also be a bug filed at the upstream for the package which fails to build. [17:16] but... I'll tag it for the f9/f10 buildroots, so that it's used for subsequent builds [17:16] *** oget_zzz is now known as oget. [17:17] any other opinions or comments? [17:17] --> thomasj has joined this channel (n=thomasj@fedora/thomasj). [17:17] I think we should push it to stable if we're going to use it for builds. [17:18] If KDE is built with 2.6.3, it installs files to a location where only 2.6.3 will find it. [17:18] So if users then try to build KDE apps with 2.6.2 it won't work. [17:18] and broken buildroot could be problem - maybe some other apps do not build, or only that one? [17:19] Rawhide has had 2.6.3 for a while. [17:19] jreznik: that's the only one we're aware of... all rawhide builds have been fine so far that I recall [17:20] --> SMParrish has joined this channel (n=quassel@cpe-069-134-255-095.nc.res.rr.com). [17:20] That app (Zorba XQuery) is the only 2.6.3 issue I'm aware of. [17:20] There's the kdepimlibs issue since at least 2.6.2, but that update fixes it (and the 2.6.2 currently in F9 and F10 has the bug). [17:21] ok, I'm convinced, stable +1 [17:22] --> MostafaDaneshvar has joined this channel (n=MostafaD@unaffiliated/mostafadaneshvar). [17:23] me too +1, we'll see :) [17:23] *** rdieter sets the channel topic to "KDE SIG Meeting -- https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2009-03-24​ -- kde-plasma-networkmanagement (new package)". [17:23] move on... kde-plasma-networkmanagement (new package) to stable ? [17:23] MathStuf isn't here, I'd rather get his input first [17:24] but I don't see any reason not to -> stable this too [17:25] let's table this until we here from MathStuf [17:25] hear even [17:25] OK [17:25] *** rdieter sets the channel topic to "KDE SIG Meeting -- https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2009-03-24​ -- Qt 4.5". [17:25] <-- MostafaDaneshvar has left this channel. [17:25] it's not in state to be called stable [17:25] Qt 4.5 is similar to cmake, it should go near or before kde-4.2.2. [17:26] jreznik: But it's a new package, which doesn't obsolete any existing package. [17:26] --> petreu has joined this channel (n=peter@fedora/petreu). [17:26] jreznik: networkmanagement or Qt? (if networkmanagement, I'd tend to agree with you). [17:26] FYI, I queued cmake-2.6.3 for stable. [17:26] <-- valente has left this server ("Leaving."). [17:27] Kevin_Kofler: yes, I know... but people will see it as new update and try to use it and it fails... etc... rdieter: netman [17:27] perhaps add to networkmanagement's %description about it's testing/unstable nature? [17:27] but I'm ok with leaving in testing longer too [17:27] rdieter: I wanted to propose it (but who reads descriptions?) [17:28] jreznik: it's better than nothing, and a note to the bodhi update as well (probably should be done regardless) [17:28] regardless of whether we push to stable soonish, that is [17:28] but that's all stuff MathStuf should have input on [17:29] +1 [17:29] <-- nicubunu has left this server ("Fading into sunset"). [17:29] yes, we should do it - explicit warning, may kill children :) [17:30] About Qt 4.5: I guess we should start planning it. [17:30] There's a bunch of packages which will need to be updated with it. [17:30] --> _daemon_ has joined this channel (n=gorillaJ@fedora/hitboxx). [17:30] ok, Qt, since some of kde workarounds for Qt'isms are buildtime, I'd also lean toward this landing near kde.4.2.2 [17:31] not necessarily in the same update, but pushed near or at the same time [17:31] Either because they have workarounds #ifdef-ed by Qt version detected at build time or because the Qt 4.5 patches don't work with 4.4 (kde-plasma-weather is one of those). [17:31] I think if we want to bundle this with 4.2.2 it should be in one grouped update. [17:31] or... crazy idea... do 4.2.2 first, then qt shortly after? [17:31] Otherwise we need to bundle some rebuilt KDE stuff with it. [17:32] but maybe all at once is the simpler/saner path here [17:32] all at once, starting from Qt 4.5 [17:32] * ltinkl wakes up [17:33] I'd have suggested doing Qt 4.5 first, but with 4.2.2 getting tagged tomorrow, I think all at once is probably the best solution. [17:33] And that way we also get the KDE stuff rebuilt automatically. [17:33] We still need to track down other stuff to update at the same time. [17:33] kde-plasma-weather definitely. [17:33] ltinkl do you have your script to find qt deps? [17:33] umm, qt deps_ [17:33] But there may be a couple of other packages which got patched for Qt 4.5 in Rawhide. Maybe send a heads-up to the fedora-devel-list? [17:33] ? [17:34] Or even fedora-devel-announce? [17:34] Kevin_Kofler: probably, good idea. mind taking care of that? bonus points for making a wiki page to document which pkgs need to be batched with qt-4.5 too [17:35] arora is waiting for it :D [17:35] Where should I send to? fedora-devel-announce? [17:35] I think so... [17:36] esp due to the License change too [17:36] (almost forgot) [17:36] are we going to update Qt 4.5 also in F10/F9? [17:36] ltinkl: yup, that's what we're talking about. :) [17:36] doing qt-4.5 along with kde-4.2.2 [17:37] rdieter: ok, sry, I came late :) [17:38] <-- drago01 has left this server (Remote closed the connection). [17:38] We should probably also refresh the qt-copy patches. [17:39] Kevin_Kofler: I did that, a few days ago [17:39] 0319, but we could recheck to see if anything has changed since [17:39] --> wsinw has joined this channel (n=IceChat7@125.25.7.79.adsl.dynamic.totbb.net). [17:39] The most recent change is March 21. [17:40] "add missing included to 0275-qtconcurrent-threadcount.diff that would stop amarok 2 compiling" [17:40] <-- wsinw has left this server (Client Quit). [17:40] ok, our builds already include a workaround for that. but using fixed/pristine qt-copy patches is preferable [17:41] alright, I think we're all in agreement about qt, anything else? [17:41] --> J5 has joined this channel (n=quintice@66.187.234.199). [17:42] let's move on then... https://bugzilla.redhat.com/show_bug.cgi?id=qtscriptgenerator​ [17:42] Bug 489550: medium, medium, ---, jreznik, ASSIGNED, Review Request: qtscriptgenerator - Qt bindings for Qt Script [17:42] *** rdieter sets the channel topic to "KDE SIG Meeting -- https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2009-03-24​ -- recent pkg reviews". [17:42] --> drago01 has joined this channel (n=linux@chello062178124130.3.13.univie.teleweb.at). [17:43] my opinion - package doesn't contain generator but generated bindings... so is it correct to name it qtscriptgenerator? [17:43] Hmmm, about Qt, there's this bug which affects merkaartor: http://www.qtsoftware.com/developer/task-tracker/index_html?method=entry&id=248604​ - unfortunately, we don't have access to the fix which went into 4.5.1. :-( [17:43] The merkaartor maintainer has been talking to me about that on #fedora-kde. [17:44] Kevin_Kofler: ok, let's keep an eye on that, hopefully something we can address before qt-4.5/kde-4.2.2 goes stable [17:44] I think it'll have to be worked around in merkaartor unless we can get the Qt Software folks to send us the patch. [17:44] <-- JSchmitt has left this server (Remote closed the connection). [17:44] I guess we should try contacting Thiago or one of the other folks working for them. [17:45] jreznik: I agree, though I'm concerned that we're straying from upstream name and what folks could expect wrt naming. .... I'll ask around, other distros... and on amarok list and see what folk think [17:46] I wonder if it makes sense to ship the generator itself as well. [17:46] jreznik: rough approach I'm working on now is making a subpkg qtscriptbindings, Required by parent qtscriptgenerator, but I'm having trouble finding content for the parent pkg... :) [17:47] rdieter: generator itself? [17:47] Kevin_Kofler: I thought about that too, but shipping a generic binary named /usr/bin/generator may be unwise too [17:47] Rename it to /usr/bin/qtscriptgenerator? [17:47] * rdieter is worried about straying from upstream behavior/expectations again... :) [17:47] But if all it works on is the Qt headers, I'm not sure how useful it is to ship it. [17:48] it's only Qt specific, isn't it? [17:48] Kevin_Kofler: but you're right. not sure. how about /usr/bin/qtscriptgenerator and %_qt4_bindir/generator [17:48] <-- fab has left this server ("Leaving"). [17:48] or maybe let it only in qt4 bindir? [17:49] as it is for qt... [17:49] someday, we really need to set %_qt4_bindir = %{_bindir} ... why haven't we done that already? maybe there's a good reason, but I don't recall.... [17:49] <-- giallu has left this server (Read error: 110 (Connection timed out)). [17:50] *** stickster is now known as stickster_food. [17:50] I checked, it can take a header file and a typesystem file. [17:50] So it should be usable to generate bindings for other stuff as well. [17:51] <-- mbacovsk has left this server (Read error: 110 (Connection timed out)). [17:51] ok, I'll include it... let's go with only %_qt4_bindir for now... [17:52] other review under scrutiny: https://bugzilla.redhat.com/show_bug.cgi?id=taglib-extras​ [17:52] Bug taglib: was not found. [17:52] rdieter: Well, there are 2 reasons why they're separate: Qt 3 stuff getting confused when finding unprefixed Qt 4 executables in /usr/bin and broken apps expecting to find everything in _qt4_prefix. [17:52] Kevin_Kofler: ok, I knew there was something. :) [17:52] https://bugzilla.redhat.com/show_bug.cgi?id=491647​ [17:52] Bug 491647: medium, medium, ---, nobody, ASSIGNED, Review Request: taglib-extras - Taglib support for other formats [17:53] this already has a pre-review done by a potential new contributor, so finishing this up should be pretty easy [17:54] these 2 should pave the way for amarok-2.1.x (qtscript* can be used with amarok-2.0.x too) [17:56] <-- mcepl has left this channel. [17:56] that's it for the agenda, got a few minutes left, anything else for today? [17:57] I've started work on Leonidas theme [17:57] and there's now SVN repo on fedorahosted [17:58] --> sdziallas has joined this channel (n=sebastia@p57A2AA2C.dip.t-dialin.net). [17:58] what's the status of the artwork? is it finished already? [17:58] <-- nphilipp has left this server ("Leaving"). [18:00] --> neverho0d has joined this channel (n=psv@vpn-pool-78-139-211-133.tomtel.ru). [18:00] jreznik: how about a brief status update onlist (fedora-kde?), and what still needs work, and where help is needed? [18:00] ltinkl: no, some artwork team mockups but I don't have even sources yet (I was asking for it, so I'll take my own way again probably :D ) [18:00] *** mmcgrath is now known as mmcgrath_pto. [18:01] I'm starting to really get fed up of this "new artwork at every release" drill. [18:01] I realize our artwork team wants to have something to do. ;-) [18:01] rdieter: ok, I'll check it with artwork team... [18:01] But for us it only means work. :-( And Solar already looks great. [18:02] Kevin_Kofler: +1 [18:02] don't know why they can't just gradually improve and build on existing theme [18:02] Kevin_Kofler: one problem is - they start tooooooo late... [18:02] That too. [18:02] Kevin_Kofler: I wouldn't mind so much... but it's taking too long... (jreznik +1 too) [18:03] alright, times up for today, thanks all. [18:03] so instead we can have for beta all done we haven't start yet [18:03] ok, thanks [18:04] I can post the log, anyone up for posting a meeting summary? [18:04] jreznik? Or should I? [18:05] svahl: I'll do it today, take a rest :D [18:05] The beta will just ship with Solar artwork on the KDE spin. [18:05] thx :) [18:05] There's no way we can change it now. [18:06] Kevin_Kofler, right. That's IMHO the best way now. All leonidas-backgrounds package is also to big for the current live images