File:Fedora-kde-sig-2008-06-10.txt

[18:05] *** Kevin_Kofler nastavil téma kanálu na "KDE SIG Meeting - https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2008-06-10 - automoc: What's the right way of packaging it? Upstream is confusing about that". [18:05] --> JSchmitt has joined this channel (n=s4504kr@p4FDD1FE8.dip0.t-ipconnect.de). [18:05]  First topic: automoc: What's the right way of packaging it? Upstream is confusing about that [18:05]  lvillani put this on the agenda. [18:05] --> kital has joined this channel (n=Joerg_Si@fedora/kital). [18:05] yep [18:06] I sent an email to upstream developer regarding automoc.files.in [18:06] if it's needed or not [18:06] <-- drago01 has left this server (Read error: 104 (Connection reset by peer)). [18:07] --> drago01 has joined this channel (n=linux@chello062178124130.3.13.univie.teleweb.at). [18:07] but it seems (after reading Automoc4Config.cmake) that is not really needed [18:07] or at least, we don't need it to be packaged in %{_libdir} [18:07] lvillani: don't include it if it's not needed [18:07] --> rnorwood has joined this channel (n=rnorwood@nat/redhat/x-abac00903191ff15). [18:08]  But is it really not needed? [18:08] but upstream says that it's a must-have and that it must be in it's installed location (%{_libdir}) [18:08] I forwarded his email to rdieter, he agrees that this is confusing [18:09] anyone willing to do a test build with this automoc package? [18:10]  Just build it in dist-f10 and I'll try to build akonadi against it. [18:10]  Is the multilib issue fixed yet? [18:11] lvillani: we just remove the file in rawhide. If it causess build problem we always can add it back [18:11] it should be [18:12] <-- kital has left this server (Remote closed the connection). [18:12]  OK, then as than said, go ahead and build it in dist-f10. As I said, I'll build akonadi against it, we'll see if it builds. [18:12] than, sounds good [18:12]  Anything else on this or can we move on? [18:13] fine, I think we can move on [18:13] *** Kevin_Kofler nastavil téma kanálu na "KDE SIG Meeting - https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2008-06-10 - KDE 4.0.5 update - status update". [18:13]  Next topic: KDE 4.0.5 update - status update [18:13]  ltinkl issues a kdelibs build with some more fixes, I'm waiting for the build to complete to update the update. [18:13]  Then the plan is to move it to stable, unless there are any objections? [18:14]  Just one thing: the crystalsvg changes: should I add a versioned kdelibs3 Requires or a versioned crystalsvg-icon-them Conflicts to prevent updating only part of these and running into file conflicts or is it not worth it? [18:15] Kevin_Kofler: it's not worth it [18:15]  OK. [18:15]  Then I think that's all about 4.0.5, move on? [18:15] *** Kevin_Kofler nastavil téma kanálu na "KDE SIG Meeting - https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2008-06-10 - https://bugzilla.redhat.com/show_bug.cgi?id=446093". [18:16]  OK, now we're in the recent bugs discussion. [18:16]  First one: https://bugzilla.redhat.com/show_bug.cgi?id=446093 [18:16]  Bug 446093: low, low, ---, Ngo Than, ASSIGNED, split kdeedu RFE [18:16] it's bad idea to seperate package for each app. [18:16] -math should be fine with me [18:16] <Kevin_Kofler> But then how do we accomodate the SIGs which want to have only parts of kdeedu? [18:17] <Kevin_Kofler> But what is -math? For example, is Step Math? [18:17] <Kevin_Kofler> It's a Physics tool, but of course there's lots of (applied) Mathematics in it. [18:18] calgebra, kalgebra, kbruch, kmplot, kpercentage [18:18] <Kevin_Kofler> Maybe we should rediscuss this next week when rdieter is back from vacation? [18:18] it could be packages in -math [18:18] <Kevin_Kofler> I'm OK with a -math subpackage, though I anticipate there will be requests for some more subpackages for other topics. [18:19] <SMParrish> I think you can justify going in either direction. Personally I think -math is the solution. Can address other SIG requests as we get them [18:19] is it really worth splitting it because there's no space on live CD? [18:20] how about removing other packages ? [18:22] <Kevin_Kofler> I'd say we go with kdeedu-math, but we should discuss this with rdieter when he comes back before implementing it. [18:22] <MathStuf> what else is taking up the space on the CD that kdeedu is looked at for trimming "fat"? [18:23] <Kevin_Kofler> There are tons of Mathematics-related tools in Fedora. [18:23] it's fine with me with kdeedu-math, but no go with seperate package for each app! [18:23] <MathStuf> ah, kdeedu is 79MB [18:23] <Kevin_Kofler> *maxima, octave, several plotting tools, just to mention some. [18:24] <Kevin_Kofler> And soon there might be even a huge mastodont of a program called "SAGE", but that's a PITA to package. [18:24] <Kevin_Kofler> Mainly because it tries to be its own distribution. [18:24] <Kevin_Kofler> But that's not really a topic for KDE SIG. ;-) [18:25] <Kevin_Kofler> In any case, I can definitely see how a Mathematics spin is heavily size-constrained. [18:28] ok, we should move on [18:29] <Kevin_Kofler> Right. I'll ask rdieter when he comes back whether he's OK with this, but right now the conclusion is to go for kdeedu-math. [18:29] *** Kevin_Kofler nastavil téma kanálu na "KDE SIG Meeting - https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2008-06-10 - https://bugzilla.redhat.com/show_bug.cgi?id=447852". [18:29] <Kevin_Kofler> https://bugzilla.redhat.com/show_bug.cgi?id=447852 [18:29] <bugbot_> Bug 447852: low, low, ---, Ngo Than, CLOSED UPSTREAM, system:/ kio broken [18:29] --> rnorwood_ has joined this channel (n=rnorwood@nat/redhat/x-60fd88bb99c1a48e). [18:29] <-- rnorwood has left this server (Read error: 110 (Connection timed out)). [18:29] <Kevin_Kofler> MathStuf found out that the offending dialog actually comes from KDE 3. [18:29] <Kevin_Kofler> That's why it tries to use ioslaves which don't exist anymore in KDE 4. [18:30] <Kevin_Kofler> We have to disable that KDE 3 dialog. [18:30] +1 [18:31] Matstuff: is there a testcase how to reproduce the problem? [18:32] <Kevin_Kofler> I guess installing kdebase3 is part of it? [18:32] <Kevin_Kofler> AFAIK that code is in kdebase3. [18:33] <MathStuf> yeah, just put in a removable media [18:33] <MathStuf> the KDE3 dialog and the plasmoid show up [18:34] ah i can reproduce it [18:35] the KDE3 dialog should be disable in thi scase [18:35] is the system ioslave part of kdebase3? then just remove it from there [18:35] ltinkl: yes [18:36] I don't think it's really needed outside of old kdebase, KDE 4 doesn't use it [18:36] ltinkl: could you please take care of it? [18:36] sure [18:36] ltinkl: thanks [18:36] * pusling takes notes [18:36] <Kevin_Kofler> The system:/ ioslave is harmless though. [18:36] <Kevin_Kofler> What really needs to be disabled is the offending dialog. [18:37] hmm, right, that would be better [18:37] I'll try to come up with something that just disables the dialog [18:37] it's more probably kded_mediamanager or sth like that [18:37] [Chyba] checks: neznámý příkaz. [18:38] *** stickster_afk je nyní znám jako stickster. [18:39] oh, and the previously mentioned bug http://bugs.kde.org/show_bug.cgi?id=163717 is related to that [18:39] <bugbot_> Bug 163717: medium, medium, ---, Petr Machata, CLOSED CURRENTRELEASE, Need ltrace 0.3.36 to debug x32 binaries on x64 [18:39] ltinkl: yes, it's in kded_mediamanager [18:39] stupid bugbot [18:39] <Kevin_Kofler> OK. ltinkl: please take care of this. [18:40] <Kevin_Kofler> And let's move on. [18:40] *** Kevin_Kofler nastavil téma kanálu na "KDE SIG Meeting - https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2008-06-10 - https://bugzilla.redhat.com/show_bug.cgi?id=439587". [18:40] <Kevin_Kofler> https://bugzilla.redhat.com/show_bug.cgi?id=439587 [18:40] <bugbot_> Bug 439587: medium, low, ---, Kevin Kofler, ASSIGNED, Unable to move Icons in panel [18:40] the KDE bug crashed in [18:40] #11 0x00002aaab063bd58 in libhal_device_get_property_QString from /usr/lib64/kde3/kded_mediamanager.so [18:40] same stuff probably [18:40] <Kevin_Kofler> Right. [18:41] <Kevin_Kofler> #439587 (unable to move icons in panel) has been reopened because the patch doesn't apply as is for KDE 4.1. [18:41] the feature "moving icons on panel" is not included in beta1? [18:41] <Kevin_Kofler> We had hoped that it would get fixed upstream, but unfortunately they don't want to implement this before 4.2 because they consider it a "new feature". [18:41] i thougt it's iin beta1 [18:42] <Kevin_Kofler> No, users confirmed it isn't. [18:42] <-- rnorwood_ has left this server (Remote closed the connection). [18:42] it seems we have to adapt the patch for 4.1 [18:43] <Kevin_Kofler> Right. Now the question is who does it. [18:43] unless we screwed up in debian packaging, I can also confirm - whats in panel is impossible to move from where you drag+dropped it [18:43] <Kevin_Kofler> I could have some fun with it, but I don't have Rawhide installed anywhere (I have F8 here and F9 on my new laptop). [18:44] <Kevin_Kofler> It's probably easier to work on for somebody who actually has Rawhide or some other KDE 4.1 installed somewhere. [18:44] --> rnorwood has joined this channel (n=rnorwood@nat/redhat/x-18ccf6fd7663d216). [18:45] <Kevin_Kofler> Patch against 4.0.3 here: http://cvs.fedoraproject.org/viewcvs/rpms/kdebase-workspace/F-9/kdebase-workspace-4.0.3-kde%23158301.patch?rev=1.2&view=markup [18:45] <Kevin_Kofler> But that uses the KDE 4.0 libplasma API. [18:45] <Kevin_Kofler> So it won't work as is in 4.1. (It doesn't even apply, because the relevant files have also changed.) [18:45] --> huff has joined this channel (n=dhuff@nat/redhat/x-59858684a982d493). [18:46] <Kevin_Kofler> (The patch comes from Stefan Binner from OpenSUSE.) [18:47] * spstarr_work is sitting out today's meeting busy @ work [18:48] <Kevin_Kofler> So, do we have any volunteer to work on porting that patch? [18:50] it seems noone has interest to work on this. [18:51] <Kevin_Kofler> We can't really count on the OpenSUSE folks to do it for us, because they'll ship 11.0 with KDE 4.0. [18:51] I can at least try having trunk installed locally from sources [18:51] no promise though, the API is completely different [18:52] yes. there's a lot of change there [18:52] i will take a look at this [18:52] some stuff got better in trunk (wrt plasma and desktop) but I still considers this very unfinished [18:53] -s [18:53] <Kevin_Kofler> Yeah, a lot of things got postponed to 4.2. :-( [18:53] cf my systray adventures in http://bugs.kde.org/show_bug.cgi?id=155381 [18:53] <bugbot_> Bug 155381: medium, medium, ---, Dave Jones, CLOSED RAWHIDE, drivers/block/cfq-iosched.c:1070: spin_is_locked on uninitialized spinlock d9d133b8. [18:53] <Kevin_Kofler> IMHO they're way too strict with their freezes. [18:53] <Kevin_Kofler> Plasma should have been exempted for 4.1 like it was for 4.0. [18:54] <Kevin_Kofler> And some other stuff should get freeze exemptions too. [18:56] k, anything else on the agenda? [18:56] <Kevin_Kofler> No. [18:56] *** Kevin_Kofler nastavil téma kanálu na "KDE SIG Meeting - https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2008-06-10 - Open discussion". [18:57] <Kevin_Kofler> We have 3 minutes left for anything else. [18:57] since svahl is not here, I'll post the minutes if noone objects [18:57] <-- danishka has left this server ("Leaving"). [18:57] <Kevin_Kofler> ltinkl: OK [18:58] <Kevin_Kofler> There's also the minutes of the previous 2 weeks, I wanted to take care of those and didn't get around to it. :-( [18:58] it seems that Automoc4Config.cmake does not require automoc4.files.in but Akonadi still tries to include it, no objections about that but... [18:59] the point is that FindAutomoc4.cmake looks for Automoc4Config.cmake in %{_libdir}/automoc4 which, imho, is wrong [18:59] suggestions? [19:00] i think that %{_datadir}/cmake/Modules is a better place for Automoc4Config.cmake, but this will probably involves patching FindAutomoc4.cmake for packages shipping it [19:00] (ie: akonadi) [19:00] (I just did a couple of local test builds, so I'm not 100% sure about that) [19:02] --> iWolf_ has joined this channel (n=jft@oh-71-2-108-50.dyn.embarqhsd.net). [19:02] <Kevin_Kofler> Well, our time is up. [19:02] <Kevin_Kofler> lvillani: We can discuss this on #fedora-kde later. [19:03] ok