14:06:30 <rdieter> #startmeeting KDE SIG Meeting -- https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2009-10-27
14:06:30 <zodbot> Meeting started Tue Oct 27 14:06:30 2009 UTC.  The chair is rdieter. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:06:30 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:06:36 <rdieter> #topic roll call
14:06:42 <rdieter> who's present today?
14:06:54 * mefoster raises hand ...
14:07:03 <rdieter> SMParrish, than: ping
14:07:18 * SMParrish here
14:08:35 <rdieter> mefoster: hi
14:09:33 <rdieter> well, let's get started a bit slowly,
14:09:33 * jreznik is here, sorry I'm late, dst changes
14:09:40 <rdieter> #topic agenda
14:09:46 <rdieter> currently, we have, https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2009-10-27#Agenda
14:09:59 <rdieter> any other agenda items to consider today?
14:10:17 <mefoster> We could talk about the sesame/soprano thing ...
14:10:31 <rdieter> mefoster: ok, thanks, I'll append that
14:13:57 <rdieter> [09:13] <rdieter> ok, let's move on then
14:13:59 <rdieter> [09:13] <rdieter> #topic Fedora 12 artwork
14:14:02 <rdieter> meh
14:14:05 <rdieter> #topic Fedora 12 artwork
14:14:09 <rdieter> jreznik: ?
14:14:15 <jreznik> ok, news - http://rezza.hofyland.cz/fedora/artwork/f12/constantine-ksplash-rain-80.png
14:14:26 <jreznik> I called it Constantine Smile theme
14:14:28 <ltinkl> neat
14:14:46 <rdieter> that's very different than before alright
14:14:46 <jreznik> it's already in our SVN
14:14:49 <rdieter> cool
14:14:53 <Sho_> there are some odd glitches on that pic
14:15:01 <Sho_> around the grey circles
14:15:10 <jreznik> I know
14:15:16 <jreznik> I have to fix it
14:15:28 <jreznik> I hope it's one of final wallpapers
14:15:35 <rdieter> it's not final... yet?
14:15:59 <jreznik> actually I don't know
14:16:00 <rdieter> well, it's going to happen today sometime, anyway
14:16:07 <jreznik> I hope
14:16:18 <rdieter> we can ping the #fedora-design folks, today was the deadline set by FESCo
14:16:21 <ltinkl> when was the feature and artwork freeze supposed to happen? :)
14:16:30 <jreznik> and from iteration 60 to 80 there are only few changes
14:16:38 <Sho_> aside from the glitches it's fine
14:16:47 <rdieter> the freeze was long past, but the design folks asked for an exception
14:17:12 <jreznik> rdieter: after we complained...
14:17:40 <ltinkl> they ask for exceptions every release
14:17:55 <jreznik> now I have to do something with KDM theme, I can do it today I hope - similar to default Air
14:18:34 <rdieter> jreznik: thanks for the update, I can help out too after meeting (and find out about wallpaper status from the design team)
14:18:55 <jreznik> rdieter: this is latest one from artwork mailing list
14:18:59 <rdieter> cool
14:19:02 <rdieter> for posterity
14:19:10 <rdieter> #link http://rezza.hofyland.cz/fedora/artwork/f12/constantine-ksplash-rain-80.png
14:19:18 <rdieter> #topic eigen2 status
14:20:02 <rdieter> a quick update, per last week's meeting, I reverted to eigen 2.0.6 for F-12, and rebuilt dependent pkgs, all tagged.
14:20:22 <rdieter> anything else here?
14:21:34 <rdieter> ok... :)
14:21:39 <rdieter> #topic qt-4.5.3: updates, F-12
14:22:47 <rdieter> how do we feel about qt-4.5.3 currently in updates-testing?  and how to handle F-12 (release in 0-day updates-testing or include in F-12 final)?
14:23:08 <rdieter> seems to be a couple of minor'ish outstanding issues
14:23:36 <rdieter> who's currently using 4.5.3, and how's it working for you?
14:24:18 * rdieter is, on f12
14:24:27 <mefoster> I have 4.5.3 and haven't noticed anything, but I don't know what I should be noticing
14:24:30 <SMParrish> I'm having no issues so far
14:24:44 <ltinkl> no issues either
14:25:12 <rdieter> 2 known issues: qt-4.5.3 + opera's qt4 build = a bit of flakiness,   and 3star echo mode causes auth dialogs to pause,freeze
14:25:55 <rdieter> regarding F-12, I'm of a mind to get it in earlier rather than later, for more testing, feedback
14:26:25 <SMParrish> I would prefer it be in F12-final so to get into more hands faster
14:27:02 <rdieter> ok, let's lean towards getting it into updates and f12-final, but I'd like to hear from than first
14:27:40 <rdieter> #info: leaning towards getting qt-4.5.3 into updates and f12-final asap, awaiting feedback from than
14:27:54 <rdieter> #topic soprano status: sesame, virtuoso
14:28:03 <nucleo> but what to do with opera issue?
14:28:27 <rdieter> nucleo: not much we can do, that's really between opera and nokia/qt
14:29:23 <Kevin_Kofler> I don't give a darn about Opera, but the 3 stars issue is IMHO a blocker.
14:29:24 <nucleo> may be it is possible wait some time when this issue whould be fixed?
14:29:29 <rdieter> nucleo: I'm very much against holding the update hostage due to unknown bugs from some closed-source application
14:29:43 <Kevin_Kofler> Password dialogs crashing with an option which appears to be quite commonly used = big problem.
14:29:49 <ltinkl> nucleo, rdieter: opera and qt are in the same building, I'm sure they will handle this :)
14:30:03 <rdieter> ltinkl: interesting. :)
14:30:26 <rdieter> Kevin_Kofler: ok, we can hash out options after meeting
14:30:41 <rdieter> mefoster: soprano/sesame status ?
14:30:45 <ltinkl> Kevin_Kofler: what is the 3 star prob?
14:30:58 <mefoster> Regarding the soprano-sesame backend: my initial investigations came up with this: https://fedoraproject.org/wiki/MaryEllenFoster/SopranoSesame
14:31:06 <mefoster> It's not actually as bad as it looks ...
14:31:16 <rdieter> ok, we're not in a hurry, let's back up and finish discussing qt (sorry mefoster, hold a sec)
14:31:30 <mefoster> rdieter: holding ...
14:31:51 <rdieter> ltinkl: https://bugs.kde.org/show_bug.cgi?id=211250
14:31:55 <buggbot> Bug 211250: normal, NOR, ---, kdelibs-bugs kde org, NEW, Password dialogs get stuck
14:31:57 <Kevin_Kofler> ltinkl: If you use the option to echo passwords with 3 stars, the password KLineEdits crash.
14:32:14 <Kevin_Kofler> With 4.5.3 and 4.6 beta, but not 100% reproducibly.
14:32:22 <Kevin_Kofler> With 4.5.2, it works.
14:33:06 <ltinkl> Kevin_Kofler: didn't KPasswordLineEdit(?) change as well?
14:33:18 <ltinkl> just to know we are not hunting the wrong bug
14:33:37 <Kevin_Kofler> The relevant code in KLineEdit (there's no separate class for passwords) didn't change on the 4.3 branch.
14:33:39 <rdieter> ltinkl: also seen on debian, archlinux, maybe you can poke at it, or some find some kde-ninja's to do so too?
14:34:00 <ltinkl> rdieter: ok, I'll look into it
14:34:03 <Kevin_Kofler> And downgrading Qt while keeping KDE at 4.3.2 is reported to "fix" it.
14:34:30 <rdieter> cool, we'll continue to consider that a blocker, until further notice
14:34:33 * thomasj again very late, sorry, work
14:34:38 <Kevin_Kofler> The relevant code is a hack in klineedit.cpp, searching for threestars will find it.
14:34:49 <Kevin_Kofler> thomasj: Don't worry, I was also late today. ^^
14:35:03 <ltinkl> Kevin_Kofler: I still think the fix would be in KLineEdit, not in Qt
14:35:18 <Kevin_Kofler> Well, the hack is fairly straightforward.
14:35:22 <Kevin_Kofler> We need to know why it breaks.
14:35:25 <Kevin_Kofler> Then we know what to fix.
14:35:51 <Kevin_Kofler> It's some sort of race condition or event-based pseudo race condition.
14:36:01 * rdieter noticed notting just tagged qt-4.5.3 for dist-f12
14:36:51 <rdieter> just as well
14:36:53 <rdieter> move on now?
14:37:15 <Kevin_Kofler> OK, move on.
14:37:21 <rdieter> [09:31] <mefoster> Regarding the soprano-sesame backend: my initial investigations came up with this: https://fedoraproject.org/wiki/MaryEllenFoster/SopranoSesame
14:37:22 <rdieter> [09:31] <mefoster> It's not actually as bad as it looks ...
14:37:27 <rdieter> mefoster: your floor
14:37:32 <Kevin_Kofler> That's "not bad"?
14:37:39 <mefoster> Yes
14:37:45 <Kevin_Kofler> It looks as bad as SAGE to me...
14:37:59 <mefoster> It should be able to build it without maven, in which case all of the packages in the top two categories are fine
14:38:11 <Kevin_Kofler> Well, I wish you good luck.
14:38:12 <mefoster> There's a bunch of jpackage packages which should be straightforward to import
14:38:17 <Kevin_Kofler> I still think Virtuoso is the way to go.
14:38:45 <rdieter> Kevin_Kofler: ideally, we'll have both (eventually)
14:38:47 <mefoster> Creating a new package for a small Java library amounts essentially to cut and paste
14:39:09 <mefoster> I'll try to get the jpackage things in next (already done one)
14:39:14 * rdieter likes the sound of that
14:39:31 <Kevin_Kofler> rdieter: What for? To have another series of meetings wasted arguing over which one should be the default like for Phonon? ^^
14:39:50 <rdieter> Kevin_Kofler: we'll use whatever upstream uses by default
14:40:01 <mefoster> If continue with this it would be nice to know that it'll be used somehow :)
14:40:05 <Kevin_Kofler> (IMHO, Virtuoso should definitely be the default. We don't have room for all that Java stack on the live image.)
14:40:48 <rdieter> we'll discuss virtuoso in a minute. :)
14:41:28 <Kevin_Kofler> soprano-sesame2 is definitely something we can ship if you manage to get it packaged properly. :-)
14:41:31 <mefoster> akurtakov has said he's willing to do some reviews for me
14:41:32 <rdieter> mefoster: in the least, it'll be valuable to offer as a short term solution, as well as providing a proper upgrade path when/if we do use virtuoso
14:41:47 <Kevin_Kofler> I just don't think it's a good idea to ship by default.
14:42:18 <mefoster> I'll keep that wiki page up to date with thecurrent status (I may reformat it to make it clearer what's going on)
14:42:20 <rdieter> mefoster: cool, any estimate on time-frame ?  a week, a month?
14:42:21 <killefiz> mefoster: can you let the list know when there is anything that needs to be reviewed?
14:42:23 <Kevin_Kofler> Also because of RAM efficiency concerns (Java stuff, and Sesame2 in particular, eats a lot of RAM).
14:42:54 <mefoster> I'd say a month is realistic ... (may be famous last words)
14:43:13 <rdieter> Kevin_Kofler: from feedback I've seen, the runtime footprint of virtuoso vs sesame2 is approx the same
14:43:26 <mefoster> It mostly boils down to getting a bunch of Java packages reviewed or, in some cases, updated
14:43:32 <Kevin_Kofler> I think we can have Virtuoso shipping in less than that (but I'm the kind of person who doesn't have qualms about shipping alpha stuff, so maybe don't listen to me ;-) ).
14:43:34 <rdieter> I think virutoso may win by a small margin
14:43:49 <rdieter> mefoster: cool, thanks
14:44:17 <rdieter> Kevin_Kofler: it's looking like getting nepomuk/virtuoso working properly will have to wait for kde-4.4
14:44:38 <Kevin_Kofler> I can backport required kdelibs and kdebase* changes...
14:44:48 <Kevin_Kofler> But if we wait for stable stuff, we'll have to wait for 4.4 indeed.
14:44:57 <Kevin_Kofler> So it might be possible to get Sesame2 in faster.
14:45:17 <Kevin_Kofler> (if mefoster is motivated enough and if she finds reviewers in reasonable time)
14:45:21 <mefoster> So I'll concentrate first on pushing the jpackage-only dependencies through the review process
14:45:58 <mefoster> Will try to get them all processed by the end of the week
14:46:07 <mefoster> (as in, review bugs opened)
14:47:14 <rdieter> awesome.
14:47:29 <rdieter> #topic open discussion
14:47:34 <rdieter> anything else for today?
14:47:54 <Kevin_Kofler> rdieter: Virtuoso status?
14:48:00 <Kevin_Kofler> Anything new?
14:48:03 <rdieter> oh, right
14:48:23 <rdieter> requisite virtuoso-opensource-5.0.12 pkgs are built as updates, in f12
14:48:39 <Kevin_Kofler> Good.
14:48:50 <Kevin_Kofler> So that leaves Soprano, and possibly kde* patches?
14:48:54 <rdieter> soprano-2.3.65 is built in devel/ branch, with virtuoso support, avail in kde-unstable repos too
14:48:56 <mefoster> (p.s. -- Kevin_Kofler -- a lot of those jar files are build dependencies, not runtime dependencies)
14:49:22 <Kevin_Kofler> There was some patch on Trüg's blog to fix an issue with Virtuoso itself, is that one already in the 5.0.12 release?
14:50:23 <rdieter> in my limited testing, tagging files seems to work.  I'm still at a loss for what the strigi indexer is for exactly, or how to use or take advantage of it (maybe it's just not working, dunno)
14:50:41 <rdieter> but I do see abrt crashes regularly from nepomuk too
14:51:07 <Kevin_Kofler> http://trueg.wordpress.com/2009/10/14/and-yet-another-post-about-virtuoso/
14:51:19 <Kevin_Kofler> Can you please check that the 5.0.12 release has this fix already?
14:51:36 <Kevin_Kofler> (From the comments, looks like it does.)
14:51:37 <rdieter> Kevin_Kofler: it does
14:52:01 <Kevin_Kofler> OK, great.
14:52:13 <Kevin_Kofler> Re backtraces, did you report them upstream yet?
14:52:23 <svahl> sry for beeing late. missed the time change (again)
14:52:48 <Kevin_Kofler> Re strigi, not sure.
14:52:49 <rdieter> Kevin_Kofler: no (kinda waiting on kde-4.4 alpha/beta before getting hardcore about it)
14:53:05 <Kevin_Kofler> I think the quicker we get things reported, the better.
14:53:20 <rdieter> we've got some incremental improvement over the status quo, but nothing seriously considering shipping yet
14:53:24 <Kevin_Kofler> Of course I also have a hidden agenda of wanting to ship that Virtuoso stuff ASAP (like "F12 0-day update" soon ;-) ).
14:54:22 * rdieter is shocked and appalled.  :)
14:54:49 <rdieter> before we run out of time, svahl, do you have anything worthy of mention ?
14:55:21 <rdieter> and fyi, yesterday, added kcm-gtk to kde-desktop (default) comps, removed gtk-qt-engine
14:55:33 <svahl> not really. no known bugs of the beta live images (filed against fedora livecd kde)
14:55:47 <svahl> ok, thx
14:56:09 <rdieter> svahl: it's ~90k, hopefully doesn't hurt us space-wise
14:57:04 <nucleo> rdieter: did you see that kcm-gtk didn't work when it installed on LiveCD?
14:57:12 <svahl> don't think so, gtk-qt-enginge was also quite small
14:57:24 <rdieter> nucleo: needs more testing, true
14:57:43 <rdieter> seems to work for some, not others
14:57:59 <Kevin_Kofler> The problem is that it writes into a .gtkrc-2.0-kde4 file which is not registered in the GTK2_RC_FILES search path.
14:58:16 <Kevin_Kofler> Maybe it fights with kdebase-workspace over that environment variable?
14:58:28 <Kevin_Kofler> kdebase-workspace adds its own .kde/share/config/gtkrc-2.0 to that.
14:58:31 <nucleo> gtk-qt-engine writes in .gtkrc-2.0-kde4 too
14:58:47 <rdieter> Kevin_Kofler: maybe, we need to look at that, make it integrate better somehow
14:59:14 <rdieter> kcm-gtk's origins came form gtk-qt-engine originally
14:59:48 <rdieter> then maybe even use qtcurve-gtk2, to make Sho_ happy
14:59:54 <Sho_> :)
15:00:09 <Kevin_Kofler> For Strigi with Nepomuk: you should get your files indexed if they're in the right directory, then nepomuksearch:/keyword for some keyword which appears in one of your files should find that file.
15:00:10 <rdieter> err, make all fedora-kde users happy.
15:00:13 <Sho_> both opensuse and kubuntu do it now
15:00:19 <nucleo> but gtk-qt-engine creates ~/.kde/env/gtk-qt-engine.rc.sh
15:00:23 <Kevin_Kofler> If it doesn't, it's a bug and should be reported upstream (like the crashes).
15:00:26 <Sho_> not that i would normaly use kubuntu as an example about what to do
15:00:38 <Sho_> s/about/of/
15:01:10 <rdieter> nucleo: that's less than ideal, as it requires a logout/login on first use
15:01:36 <Kevin_Kofler> I think there's no better way.
15:01:42 <Kevin_Kofler> You can't inject things into the toplevel environment.
15:01:53 <rdieter> Kevin_Kofler: can't share use of .kde/share/config/gtkrc-2.0 somehow ?
15:02:09 <Kevin_Kofler> ksmserver would have to support some setEnv D-Bus call, I know the GNOME equivalent supports that now, but ksmserver doesn't AFAIK.
15:02:49 <rdieter> well, barring details, it warrants some love and attention
15:02:57 <Kevin_Kofler> Well, that file is written at each session startup by code in kdebase-workspace.
15:03:08 <Kevin_Kofler> And I think also on color setting changes.
15:03:35 <Kevin_Kofler> I don't think clobbering it is a good idea.
15:03:51 <Kevin_Kofler> What I could do is hack the code which sets GTK2_RC_FILES to include the kcm-gtk one.
15:03:58 <Kevin_Kofler> It'll just be ignored if it's not there.
15:04:07 <rdieter> sounds better to me, yeah, something like that
15:04:24 * mefoster has to leave soon to reboot her computer to Windows to give a demo :(
15:04:29 <rdieter> well, looks like we're out of time
15:04:36 <rdieter> mefoster: thanks, enjoy (I think)
15:04:42 <XulWork> hey guys, can we package k3b alpha, because old k3b doesnt work with f11 transcode
15:05:05 <rdieter> XulWork: kde-unstable repo @ kde-redhat.sourceforge.net
15:05:14 <Sho_> fwiw: very bad experience with k3b alpha here
15:05:23 <Sho_> couldn't burn an iso the other day
15:05:27 <XulWork> hrm
15:05:29 <Kevin_Kofler> We've already had K3b alpha in F12, we reverted it.
15:05:32 <rdieter> XulWork: the transcode issue is known, I just need to get off my arse, and patch it (but I think k3b alpha has the same problemj)
15:05:52 <Kevin_Kofler> Pushing it as an F11 update is definitely a bad idea.
15:06:01 <Kevin_Kofler> Patching the actual bug is better.
15:06:12 <XulWork> ya if you can patch it
15:06:24 <Kevin_Kofler> I'm sure it can be fixed.
15:06:27 <XulWork> k
15:06:45 <rdieter> XulWork: https://bugzilla.redhat.com/show_bug.cgi?id=524348
15:06:47 <buggbot> Bug 524348: medium, low, ---, rrakus, ASSIGNED, transcode --progress_rate 0
15:07:22 <rdieter> alright, let's wrap up the meeting, thanks everyone
15:07:25 <rdieter> #endmeeting