Jul 10 18:04:32 If anyone present has anything to discuss in today's KDE SIG meeting which is not yet in the agenda: http://fedoraproject.org/wiki/SIGs/KDE/Meetings/2007-07-10 please add it now. Jul 10 18:07:19 Waiting for rdieter (and possibly than and svahl too, anyone knows if they'll show up?) because there's no way to hold a meeting with all the key people missing. ;-) Jul 10 18:21:34 here now (sorry) Jul 10 18:22:18 had another meeting Jul 10 18:23:50 I can give a summary of akademy, I suppose. Jul 10 18:24:06 Great, let's start with that! Jul 10 18:24:25 first off, had a great time, met lots of great folks. Jul 10 18:24:58 my talk, referenced on http://akademy.kde.org/conference/programme.php Jul 10 18:25:09 you can download an .ogg video ... Jul 10 18:25:22 I watched it, don't think I made too much of a fool out of myself. Jul 10 18:25:57 I watched it already. You did well, much better than some of the other speakers. Jul 10 18:26:03 :) Jul 10 18:26:21 that's ok, some folks are better coders that speakers, clearly. Jul 10 18:26:28 Obviously. Jul 10 18:27:34 We (fedora) seem to be the only distro to be comitted to shipping kde4 (as the default kde desktop anyway). Jul 10 18:28:27 So much for all those "KDE-centric" distros... :-/ Jul 10 18:28:32 the distro folks were all in agreement, that stuff should be installable to the the same prefix, ie, /usr Jul 10 18:28:35 * dgilmore is half here Jul 10 18:28:52 Great. Jul 10 18:28:55 problem being that there's still many problems with that. Jul 10 18:29:10 Do they have a solution for the -devel stuff already? Jul 10 18:29:14 that's ok, for us, since we don't plan on shipping a full kde3. Jul 10 18:29:52 -devel ? no, that wasn't even on the radar yet, but I think we'll have no option but explicit Conflicts. Jul 10 18:30:05 I think the only real solution to avoid the conflict would be to put all the unversioned symlinks into /usr/lib/kde4 (and possibly move the plugins to something like /usr/lib/kde4/plugins to avoid the confusion). Jul 10 18:30:21 That, or use something like /usr/lib/kde3.5/lib/. Jul 10 18:30:22 rdieter: maybe we can put kde3 headers in /usr/include/kde3 Jul 10 18:30:36 it's not the headers, but library symlinks that are the problem. Jul 10 18:30:41 ahh ok Jul 10 18:30:42 That's not the biggest problem (and some distros do that already), the libraries are. Jul 10 18:30:49 And some devel-only executables. Jul 10 18:31:21 I'll keep pushing Than to give us a status report on compat-kde3* progress. Jul 10 18:31:58 worst case, if we don't hear from him within X days, we can start on it ourselves. Jul 10 18:32:01 Too bad Than is not present this week. He was present last week, but he didn't have the compat packages ready yet. Jul 10 18:32:14 <-- vpv (i=vpv@kapsi.fi) hat #fedora-meeting verlassen Jul 10 18:32:25 Have you seen last week's summary? http://fedoraproject.org/wiki/SIGs/KDE/Meetings/2007-07-03 Jul 10 18:32:34 So, Than's batting 1/4 (present 1 out of the past 4 weeks). Jul 10 18:33:02 :( Jul 10 18:33:11 no (sorry), I'll take a look asap. Jul 10 18:33:23 i really want to see his compat packages so we can get a kde4 build in rawhide Jul 10 18:33:44 aseigo made a point at akademy that distros need to work harder to push patches upstream. Jul 10 18:34:01 I think we're all in agreement here. Jul 10 18:34:07 yeah. Jul 10 18:34:32 I think his comment was directed toward a few other distros who historically are bad at that. Jul 10 18:34:44 I talked about this with Than last week, we need to sift through our KDE 3 patches to see which still apply to KDE 4, so we can try to get the applicable ones upstream too while we are at it. Jul 10 18:35:18 Some of the KDE 3 patches are really needed for KDE 4 (and should go upstream), e.g. the ConsoleKit support. Jul 10 18:35:46 I took a first stab at updating it to KDE 4, but haven't had the time to test it (I don't even know if it compiles). Jul 10 18:36:04 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=228111 Jul 10 18:36:19 absolutely, documenting patches is essential. Jul 10 18:37:15 Kevin_Kofler: I'll snag it an incorporate into the next kdebase4 build(s). Jul 10 18:37:57 since we're a "team" now, Than needs to do a better job wrt cvs commit messages, changelogs, patch comments/documentation. Jul 10 18:38:01 Do you have any of the remaining modules packaged already or just the 3 core ones? Jul 10 18:38:14 or the rest of us simply need to practice ESP/clarivoyance. Jul 10 18:38:28 --> Rathann (n=rathann@debi.pekin.waw.pl) hat #fedora-meeting betreten Jul 10 18:38:37 no, would have liked to worked on that... but alas... Jul 10 18:39:09 Not that I'm picking on Than alone, obviously Jul 10 18:39:22 I'm fairly busy these days too unfortunately (trying to finally get my diploma thesis in Mathematics completed). Jul 10 18:39:29 we all need to use better documention, practices... Jul 10 18:39:42 math is hard. :) Jul 10 18:40:02 The problem is that the more we document, the more time this is going to take. :-( Jul 10 18:40:21 true, but it's for posterity, 5 minutes now, saves loads more later. Jul 10 18:40:51 even for my *own* packages, I forget what some patches are for, etc... Jul 10 18:41:01 much less looking at someone else's work. Jul 10 18:41:13 it's a necessary evil. Jul 10 18:41:28 True. Jul 10 18:41:32 Next topic: redland/soprano Jul 10 18:42:03 unless anyone has other ?'s wrt akademy? Jul 10 18:42:57 --- rdieter hat das Thema geändert auf: http://fedoraproject.org/wiki/Communicate/FedoraMeetingChannel -- KDE SIG meeting -- redland/soprano Jul 10 18:43:01 ok, moving on... Jul 10 18:43:07 Redland has finally been approved yet, unfortunately, thomasvs isn't importing it now. I'm starting to get really fed up of waiting. :-( Jul 10 18:44:21 And do we have any Soprano packages yet? Because redland alone isn't enough, the dependency chain is kdelibs4 (Nepomuk) Requires soprano Requires redland. Jul 10 18:44:39 well, we can continue to offer to help/comaintain. Jul 10 18:45:34 if things drag out too long, we could just assume comaintainership, import the stuff ourselves to get things started. Jul 10 18:46:18 Kevin_Kofler: you interested in comaintaining those? or anyone else here? Jul 10 18:46:32 <-- Rathann hat sich getrennt ("Gone.") Jul 10 18:46:43 I can, if no one else can/will. Jul 10 18:47:05 I'm willing to comaintain redland to get things done. Jul 10 18:47:19 But Soprano needs a specfile first. Jul 10 18:47:45 wrt soprano, I'm not aware of any pending submissions... lemme go checkif I have anything on that... Jul 10 18:48:00 yes. Jul 10 18:49:06 *very* quick-n-dirty mind you... http://kdeforge.unl.edu/apt/kde-redhat/SOURCES/soprano/ Jul 10 18:50:04 I'll submit that for review later this week... if no one beats me to it. Jul 10 18:50:22 I'm using that in the latest kde-redhat kde4 builds. Jul 10 18:50:51 OK, I'll review this for you once redland is in. Jul 10 18:52:09 I don't see soprano in the current kde-redhat unstable though. Jul 10 18:53:04 oops. Jul 10 18:53:18 I'll go double-check that after the meeting. Jul 10 18:53:42 One last topic I wanted to bring up is the infamous ICEauthority patch. Jul 10 18:53:50 FYI, this patch apparently originates from Kubuntu. Jul 10 18:54:03 They're using almost the same patch with a comment about what it's supposed to fix. Jul 10 18:54:06 --- rdieter hat das Thema geändert auf: http://fedoraproject.org/wiki/Communicate/FedoraMeetingChannel -- KDE SIG meeting -- Fallout of ICEauthority/sudo patch (#247455) Jul 10 18:54:09 --> racor (n=rc040203@HSI-KBW-082-212-056-027.hsi.kabelbw.de) hat #fedora-meeting betreten Jul 10 18:54:21 yeah, 2 problems here. Jul 10 18:54:45 It's supposed to fix using KDE apps with sudo. Jul 10 18:54:45 1. it wasn't documented (here anyway) what it addressed, and why it was needed. Jul 10 18:55:01 2. upstream was not consulted Jul 10 18:55:11 Indeed, I shouldn't have to look up what the patch does in the Kubuntu diff.gz! Jul 10 18:55:29 kubuntu makes heavy use of sudo, so I can understand their motivations... Jul 10 18:56:05 fedora doesn't, so I think we can likely conclude the patch is more bad than good. Jul 10 18:56:46 The "memory corruption" error could probably be zapped by ignoring getenv("HOME") entirely and using only savhome, but that wouldn't fix the XDMCP breakage, and it's just piling another hack above the previous one. Jul 10 18:57:27 yeah, it's nice to have XDMCP again. :) Jul 10 18:57:28 For reference, the KDM code producing the error: Jul 10 18:57:28 --> scop (n=scop@cs181043142.pp.htv.fi) hat #fedora-meeting betreten Jul 10 18:57:39 if (strcmp( savhome, getenv( "HOME" ) ) || memcmp( savhome, "/tmp/", 5 )) Jul 10 18:57:39 LogError( "Internal error: memory corruption detected\n" ); /* no panic: recursion */ Jul 10 18:57:39 else { Jul 10 18:57:39 sprintf( buf, "rm -rf %s", savhome ); Jul 10 18:57:39 system( buf ); Jul 10 18:57:39 } Jul 10 18:57:51 We can be glad it does the sanity check! Jul 10 18:58:05 rm -rf /root could be pretty bad on some systems! Jul 10 18:58:57 eew. Jul 10 18:59:37 Though I believe savhome would be good anyway, it's the strcmp(savhome,getenv("HOME")) which fails because getenv("HOME") has been changed by the hack. Jul 10 19:00:02 * rdieter nods Jul 10 19:00:38 I'd feel a whole lot better using only upstream-sanctioned/approved solutions here. anything else is unacceptable. Jul 10 19:01:43 I agree that the hack needs to go given the breakage, so let's do this on F7 too (and FC6, but that has to be done by Than). Jul 10 19:02:27 ok, I think we're out of time... FPC meeting. Jul 10 19:02:33 Right.