14:01:54 <rdieter> #startmeeting KDE SIG Meeting -- https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2009-10-06
14:01:54 <zodbot> Meeting started Tue Oct  6 14:01:54 2009 UTC.  The chair is rdieter. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:01:54 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:02:07 <rdieter> #topic roll call
14:02:10 <Kevin_Kofler> Present.
14:02:12 <mathstuf> here
14:02:15 <SMParrish> here
14:02:25 * than is present
14:02:27 <svahl> here
14:04:08 <rdieter> #topic  F-12 kde spin status
14:04:27 <rdieter> ok, any reports, comments on F12-related business?
14:04:54 <Kevin_Kofler> I think we're good to go. svahl?
14:05:08 <rdieter> oh, one thing we'd been talking over the past week, was considering including both phonon backends to make switching easier
14:05:28 <svahl> the x86_64 spin is slighlty oversized (703 megs) due do newer wallpaper. But so far no complaints
14:05:41 <Kevin_Kofler> There's this bizarre bug files about mislabeled liveuser files on the KDE spin: https://bugzilla.redhat.com/show_bug.cgi?id=520022
14:05:42 <buggbot> Bug 520022: medium, low, ---, fedora, NEW, [KDE Live] setroubleshoot:      SELinux is preventing loadkeys (loadkeys_t) "write" /home/liveuser/.xsession-errors (user_home_t).
14:05:45 <Kevin_Kofler> *filed
14:06:37 <Kevin_Kofler> Hmmm, if we're already oversized, adding another Phonon backend might not be that great.
14:06:52 <Kevin_Kofler> Though I guess it's just a few kB, as we have the backend libs anyway.
14:07:24 <jreznik> svahl: I hope they won't use this wallpaper and use again the old one... this one looks really horrible...
14:07:33 <svahl> for the bug: what I'm supposed to do there? adding a restorecon in the ks?
14:07:42 <rdieter> -gstreamer backend is ~229k
14:08:29 <rdieter> and there shouldn't be any new deps from it
14:09:03 <rdieter> I guess that bug report didn't mention any bad effects of the denial (ie, login failed or such)
14:09:14 <rdieter> so, not sure if it's worth considering it as a blocker yet
14:09:45 <rdieter> but once we have a beta, I'm planning installing it on another box or 2
14:09:59 <rdieter> so I can help test, debug here better.
14:10:41 <rdieter> anyway, any objection to adding both phonon backends to kde-desktop in comps?
14:10:41 <jreznik> today it started with knotify crash... but I have on installation on SSD card, I hope I can get backtraces
14:11:37 <svahl> both as default?
14:11:38 <Kevin_Kofler> svahl: I have no idea why the files are mislabeled in the first place.
14:11:46 <Kevin_Kofler> So I don't know how to fix it.
14:11:55 <SMParrish> svahl: xine as default
14:12:13 <rdieter> svahl: both installed by default, but -xine will be preferred/used by default
14:12:25 <svahl> I meant both in default section of @kde-desktop
14:12:34 <rdieter> svahl: yes, right
14:12:54 <than> i prefer to add both in comps
14:13:41 <rdieter> ok, I'll add it after meeting, if anyone has other comments, objections in the meantime, yell.
14:13:47 <rdieter> any other F-12 items?
14:14:48 <jreznik> rdieter: qtwebkit
14:15:02 <svahl> Kevin_Kofler, I'll see if I can reproduce it. Maybe adding a restorecon at the end of the kde ks won't hurt and fixes it
14:15:19 <Kevin_Kofler> jreznik: We have that under "recent bugs".
14:15:26 <Kevin_Kofler> Can we move on now?
14:15:35 <jreznik> Kevin_Kofler: ah, ok
14:15:40 <rdieter> #topic Tracking ongoing development/projects
14:16:12 <rdieter> just a quick comment, we've got a of cool development going on in many areas, but no way to track status, progress.
14:16:44 <rdieter> I thought we could possibly leverage bugzilla to help out there (but I'm not picky on the tool used, as long as we do something)
14:17:11 <rdieter> stuff I had in mind here, includes: kdm fprint support, ooo kde integration, PolKitOne-qt
14:17:38 <jreznik> wiki is probably better - I don't have idea how to track progress in bz
14:17:56 <Kevin_Kofler> F13 feature pages? :-)
14:18:06 <rdieter> well, progress, as in progress reports, bz is good at tracking targets, and interdependencies
14:18:12 <jreznik> problem is that Fedora Feature page is more - to be 100% in some time not depending on state
14:18:35 <Kevin_Kofler> A feature can be postponed to F14 if it isn't ready in time for F13.
14:18:36 <rdieter> wiki/bz could complement each other here too
14:19:11 <rdieter> anyway, I don't have a specific proposal, just think about it, and publicize current activity somehow
14:19:36 <jreznik> rdieter: I like it! I was thinking about it some time
14:19:55 <jreznik> as many times someone started work on something and in time it dissapeared
14:20:07 * jreznik has to try taskjuggler :D
14:21:01 <rdieter> jreznik: so your job at least is to wiki/bz-ize your status/progress on polkit1-qt stuff, ok?
14:21:19 <rdieter> and if we can poke those working on the other stuff I mentioned, would be nice.
14:21:44 <Kevin_Kofler> I think we should really use feature pages for that stuff.
14:21:54 <Kevin_Kofler> All the other subprojects in Fedora advertise their cool work that way.
14:21:55 <jreznik> rdieter: ok, but I see one probleme - it should be probably tracked upstream
14:21:57 <Kevin_Kofler> We shouldn't be left out.
14:21:57 <rdieter> Kevin_Kofler: it could evolve into that shure.
14:22:10 <jreznik> Kevin_Kofler: of course, this should be F13 feature
14:22:12 <Kevin_Kofler> People always complain about us not filing enough feature pages.
14:22:29 <rdieter> jreznik: I don't care honestly, just as long as it's tracked somewhere, somehow.  I'm not picky right now
14:22:41 <jreznik> Kevin_Kofler: like one feature per project?
14:22:52 <jreznik> sound good
14:22:58 <Kevin_Kofler> Yes, one per project.
14:24:40 <jreznik> so what we are going to have - kde 4.4, pk-1-qt, kde fingerprint, kde abrt
14:24:59 <rdieter> looks like a good start
14:25:05 <Kevin_Kofler> OO.o-kde would be great too.
14:25:39 <Kevin_Kofler> BTW, for pk-1-qt, we really need this and especially the KAuth backend in time for KDE 4.4.
14:25:41 <jreznik> ok, so I try to prepare Feature pages for these projects
14:26:09 <Kevin_Kofler> We want the KDE 4.4 update to F12 to be complete, not ship a fake KAuth backend or none at all.
14:26:26 <jreznik> Kevin_Kofler: I want it too :) high priority - there's no problem to port kauth already
14:26:48 <jreznik> actually I was about to do it today, but fighting with some gnome code for auth. agent...
14:26:51 <rdieter> ok, let's move on (for now, can discuss further details later on irc/ml)
14:27:11 <jreznik> just note action item for me :D
14:27:48 <rdieter> #action jreznik to work on Feature pages - kde 4.4, pk-1-qt, kde fingerprint, kde abrt
14:28:17 <rdieter> #topic qt-4.5.3 builds broken, missing translations
14:28:52 <Kevin_Kofler> Staying in the tradition of Qt 4.5.x being broken all over the place here. :-(
14:28:54 <rdieter> our qt-4.5.3 builds broke, seems related to translations, but that's about the extent of my qmake-fu got me when I tried debugging
14:29:19 <Kevin_Kofler> 4.5.0 had many regressions, 4.5.1 a few (even compared to 4.5.0), 4.5.2 was decent, now we're up in breakage land again. :-(
14:29:40 <rdieter> in short, need someone to take on the task of debugging, and contacting relevant upstreams to find out what's going on
14:30:06 <jreznik> rdieter: I talked with than, he will take care
14:30:26 <rdieter> cool
14:30:48 <than> yes, i will take care
14:30:55 <rdieter> #action than to look into qt-4.5.3 build (translation-related?) breakage
14:31:15 <rdieter> #topic kde-4.3.2 status
14:31:26 <rdieter> than: how goes kde-4.3.2 import/building?
14:32:08 <than> the kde-4.3.2 build should be done today
14:32:08 <jreznik> rdieter: ltinkl is working on it, I saw lot of commits so probably it's moving
14:32:32 <rdieter> woo, good news, thx
14:33:09 <rdieter> #topic qt/selinux, https://bugzilla.redhat.com/show_bug.cgi?id=527079
14:33:10 <buggbot> Bug 527079: medium, medium, ---, than, ASSIGNED, setroubleshoot:      SELinux is preventing /usr/bin/arora from changing a writable memory segment executable.
14:33:21 <rdieter> this one looks fun.
14:34:05 <Kevin_Kofler> This is QtWebKit's JIT triggering execmem.
14:34:15 <rdieter> looks like disabling the jit is the plan for now?  should we wait on 4.5.3 builds getting fixed, or consider back-porting the workaround to our 4.5.2 builds too?
14:34:23 <Kevin_Kofler> jreznik disabled the JIT in devel, as is already done in webkitgtk in F-12 and devel.
14:34:37 <jreznik> we should have it for F12
14:34:40 <SMParrish> FYI the base webkit packages are also causing selinux issues
14:34:56 <Kevin_Kofler> BTW, the JIT is also still 32-bit x86 only. :-/
14:34:59 <Kevin_Kofler> That really sucks.
14:35:05 <Kevin_Kofler> No x86_64. :-/
14:35:26 <jreznik> only i386 is enabled by default
14:35:27 <jreznik> and even it's causing problems...
14:35:30 <Kevin_Kofler> So I guess we can just disable it for i?86 too, it's how things always are on x86_64 anyway.
14:35:33 <rdieter> sounds bad-ish, so I'd recommend backporting the workaround for 4.5.2 (shouldn't be hard).
14:35:45 <jreznik> np
14:36:02 <rdieter> I can do it, unless someone else wants the fun.
14:36:04 <jreznik> gentoo has the same workaround as JIT USE tag
14:36:34 * thomasj way too late, but here
14:36:35 <jreznik> just few lines from pri away
14:37:18 <Kevin_Kofler> Of course, we could also try to argue with the SELinux folks that libQtWebKit.so.* should be allowed execmem.
14:37:25 <Kevin_Kofler> But they're going to kill us for even suggesting it.
14:37:30 <jreznik> :D
14:37:33 <rdieter> #action backport JIT disabler patch to qt-4.5.2 in F-12/ branch (rdieter, or whoever beats him to it)
14:37:53 <rdieter> next...
14:37:55 <rdieter> #topic FUDCon Toronto 2009
14:38:12 <rdieter> For posterity, who's going to FUDCon Toronto 2009?
14:38:14 <Kevin_Kofler> There are some contrived ways to do JITs which don't trip over SELinux's execmem protection.
14:38:25 <Kevin_Kofler> I guess it could be done somehow for QtWebKit.
14:38:26 <mathstuf> looks like a go for me
14:38:28 * rdieter raises hand
14:38:31 <mathstuf> no finals
14:38:33 <Kevin_Kofler> (But someone needs to code that.)
14:38:37 <SMParrish> I'll be there
14:38:49 <jreznik> sounds good but... thanks our gypsies we need visas again :(
14:39:27 <rdieter> invited aseigo , and he's accepted.  so far, so good (provided we get funding for his costs)
14:40:07 <Kevin_Kofler> I won't be there. Toronto is 6920 km from here according to Marble...
14:40:07 <rdieter> jreznik: is that a yes from you?
14:40:30 <SMParrish> I think funding will work out.  Paul is working out the final details of that this week
14:40:37 * jreznik does not have passport, would need visas (in Vienna :D) and money (12000 CZK)
14:41:10 <rdieter> ok, wasn't sure how to interpret your previous visa comment. :)
14:41:11 <jreznik> rdieter: it's I'd like to visit Toronto but not sure I can make it in time (to have passport, visas...)
14:41:14 <mathstuf> rdieter SMParrish: i may be able to meet in buffalo at the bus station
14:41:20 <mathstuf> its probably how I'll be travelling
14:41:31 * mathstuf doesn't see himself driving
14:41:45 <rdieter> mathstuf: sweet, let's coordinate that.  party bus.
14:41:48 <mathstuf> :)
14:41:50 <jreznik> Kevin_Kofler: but would be fun, I hope there are directl flights from Vienna
14:42:10 <SMParrish> bus also picksup at the airport there as well
14:42:16 <mathstuf> ah
14:42:23 <mathstuf> maybe they're close
14:42:25 <Kevin_Kofler> jreznik: I think there are, but I'm not sure.
14:42:35 <rdieter> other than the logistics, we need to brainstorm for stuff to do, kde-related talks/hack-sessions
14:42:49 <Kevin_Kofler> Check Austrian Airlines's website.
14:44:23 <mathstuf> :(
14:44:31 <mathstuf> google maps broke in konq recently
14:44:39 <Kevin_Kofler> mathstuf: Use Marble!
14:44:52 <rdieter> #topic Open Discussion
14:44:53 * jreznik is checking online system for passport requests
14:44:55 <mathstuf> was hopiong for search for the airport
14:45:10 <rdieter> that's it for the agenda for today, anything else to discuss today?
14:45:18 <Kevin_Kofler> You can enter a search into Marble's location combobox and hit ENTER.
14:45:32 <mathstuf> for buffalo; don't think it does landmarks yet
14:45:34 <Kevin_Kofler> It'll send it to OpenStreetMap's search server and return hits.
14:46:11 <Kevin_Kofler> That is, it will if the server is responding.
14:46:11 <mathstuf> off to class
14:46:15 <Kevin_Kofler> It has been really overloaded lately.
14:47:42 <Kevin_Kofler> Wow, you really need to check which airport you're landing at. Buffalo has tons of airports.
14:48:24 <svahl> If someone imports the extragear packages I'll do the builds then (still need access to ktown)
14:48:46 <SMParrish> Kevin_Kofler: airport code is BUF
14:48:55 <rdieter> svahl: send me a public ssh key, and I'll add you to ktown
14:49:13 <svahl> rdieter, ok, will do it after the meeting, thx
14:49:23 <jreznik> I can't promise anything but I'd like to come
14:49:52 <rdieter> jreznik: would be swell, keep us informed.
14:50:19 <jreznik> even it's not cheap, means travelling (I hate travelling)... and  visa :( dammed gypsies :(
14:51:20 <rdieter> well, seems we've exhausted meeting topics, let's wrap things up (yay, for ending early)
14:51:22 <rdieter> #endmeeting