14:07:32 <rdieter> #startmeeting KDE-SIG Meeting -- https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2009-08-11
14:07:54 <rdieter> #topic Init
14:07:57 <rdieter> who's present today?
14:08:01 <svahl> present
14:08:02 <Kevin_Kofler> Present.
14:08:05 <jreznik> present
14:08:12 <SMParrish> Present
14:09:41 <rdieter> alrighty,
14:09:51 <rdieter> #topic KDE 4.3.0 blockers (for F11/F10)
14:10:29 <rdieter> this should help, https://bugzilla.redhat.com/showdependencytree.cgi?id=515075&hide_resolved=1
14:11:23 <rdieter> 2 open items: fish://  and cntl-f12 dashboard shortcut
14:11:48 <Kevin_Kofler> https://bugzilla.redhat.com/show_bug.cgi?id=516416
14:11:49 <buggbot> Bug 516416: medium, low, ---, than, ASSIGNED, fish protocol broken in kde 4.3
14:12:07 <Kevin_Kofler> There's an upstream commit being suspected for the regression.
14:12:17 <Kevin_Kofler> We need to try reverting it and testing whether that fixes it.
14:12:32 <ltinkl> +1, revert and see
14:12:40 <rdieter> seems upstream borked fish:// but good, yeah, who wants to work on testing out reverting ?
14:13:12 <Kevin_Kofler> The offending commit: http://websvn.kde.org/?view=rev&revision=946444
14:13:25 <ltinkl> rdieter: I'll give it a shot
14:13:41 <Kevin_Kofler> Also suspicious: the fish.cpp change in: http://websvn.kde.org/?view=rev&revision=933202
14:13:42 * thomasj late present
14:13:47 <rdieter> ok, #action ltinkl to try reverting suspect upstream commit(s), and test things out
14:14:02 <rdieter> thomasj: hiya
14:14:03 <Kevin_Kofler> If reverting 946444 doesn't fix it, we should try reverting 933202 too.
14:14:09 <thomasj> Hi :)
14:14:16 <Kevin_Kofler> I didn't see any other relevant change between 4.2 and 4.3.
14:14:51 <Kevin_Kofler> https://bugzilla.redhat.com/show_bug.cgi?id=516445
14:14:52 <buggbot> Bug 516445: low, low, ---, than, ASSIGNED, Show Dashboard Shortcut (ctrl-f12) doesn't work after upgrade to kde 4.3.0
14:14:58 <Kevin_Kofler> That's the second one.
14:15:01 <ltinkl> that one is funny :)
14:15:03 <Kevin_Kofler> We don't seem to have much of a lead there.
14:15:34 <jreznik> ctrl+f12 - I can't reproduce it anymore... I was trying to reinstall kdesettings, deleting users globalshortcuts... for new user it's ok
14:15:36 <ltinkl> somehow the default shortcut got lost during the update; system setting thinks the shortcut is conflicting
14:15:47 <rdieter> there's a lot of mysterious "shortcuts not working" bugs on upgrades.  we saw that a bit with 4.2 too.
14:15:54 <ltinkl> but the other shortcut it reports is the very same one :)
14:16:09 <SMParrish> ctrl+f12 works for me.  I can't reproduce
14:16:17 <ltinkl> it doesn't for me
14:16:25 <ltinkl> ie I can reproduce that
14:16:36 <rdieter> yeah, it either "just works" or mysteriously fails.
14:16:42 <rdieter> fun bug
14:17:08 <svahl> I'm not sure but maybe I've seen it fail when desktop effects were enabled
14:17:26 <svahl> normally I don't use it and it works then
14:17:30 <jreznik> svahl: I don't have DE enabled and it failed
14:18:06 <ltinkl> svahl: same as jreznik
14:18:12 * thomasj cant reproduce it
14:18:13 <svahl> ok, was just a try. :)
14:18:49 * jreznik has 2 other boxes to upgrade, will check it
14:18:55 <rdieter> is that one reported upstream (yet)?
14:19:00 <rdieter> if not, shouldn't we?
14:19:02 <SMParrish> My opinion since its not 100% reproducable is bounce it upstream
14:19:09 <Kevin_Kofler> Of course we should.
14:19:16 <Kevin_Kofler> There's no upstream reference in our bug.
14:19:17 <ltinkl> yup (it hasn't been reported upstream afaik)
14:19:17 <Southern_Gentlem> what are the video cards of who it works vs dont
14:19:24 <Kevin_Kofler> So I guess it hasn't been reported upstream yet.
14:19:34 <SMParrish> just triaged it to request upstream filing
14:19:48 <ltinkl> Southern_Gentlem: wrong window? :)
14:19:48 <rdieter> Southern_Gentlem: I don't think video cards have anything to do with this. ??
14:19:48 <Kevin_Kofler> Southern_Gentlem: I doubt it's a graphics driver bug.
14:19:51 <Kevin_Kofler> Though you can never be sure.
14:20:23 <Southern_Gentlem> ltinkl,  what video card do you have
14:20:26 <jreznik> nouveau...
14:20:27 <rdieter> I'd propose someone currently experiencing it, upstream it, ltinkl ?
14:20:36 <Kevin_Kofler> It just doesn't behave like a graphics driver bug.
14:20:49 <jreznik> the question is - is it really upstream bug?
14:21:10 <ltinkl> hmm, now that I think of it, might be something to do with translations
14:21:15 <SMParrish> I am using the binary nvidia blob here on 1 machine and nouveau on another
14:21:41 <Kevin_Kofler> ltinkl: That could be it.
14:21:45 <ltinkl> jreznik: english/czech KDE?
14:21:53 <jreznik> ltinkl: english
14:22:01 <ltinkl> ok, that can't be it then :)
14:22:01 <Kevin_Kofler> What locales are the people using who have the issue? And those who don't?
14:22:30 <Kevin_Kofler> Yeah, if it can be reproduced with en_US.UTF-8 locale, it's likely not translations.
14:22:47 <thomasj> en_US  works
14:22:59 <svahl> de_DE too
14:23:08 <SMParrish> en_US here
14:23:10 <Kevin_Kofler> I think it's something else.
14:23:33 <Kevin_Kofler> Maybe it depends on whether you customized the shortcuts while using 4.2 or not?
14:23:36 <thomasj> Though keyboard is DE
14:23:43 <rdieter> Kevin_Kofler: you're prob right
14:23:49 <ltinkl> Kevin_Kofler: ye, maybe that as well, I had them customized before
14:23:58 <Kevin_Kofler> I think it must be that.
14:24:09 <thomasj> Not customized here
14:24:15 <rdieter> which makes it "fun" to try to test/reproduce
14:24:16 <jreznik> ltinkl: I didn't customize them
14:24:20 <Kevin_Kofler> The customization leaves some config file entries which apparently conflict with the new defaults.
14:24:31 <ltinkl> jreznik: sure?
14:24:32 <Kevin_Kofler> jreznik: Sure?
14:25:05 * jreznik is not 100% sure but usually don't customize global shortcuts
14:25:45 <jreznik> I'll check it home on laptops - I'll try to customize one and second I'll let without customizations
14:25:55 <ltinkl> jreznik: does  ~/.kde/share/config/kglobalshortcutsrc exist for you?
14:26:17 <ltinkl> if you hadn't customized it, it shouldn't exist
14:27:33 <jreznik> ltinkl: you're right it existed, I tried to remove it while checking to reproduce...
14:27:40 <Kevin_Kofler> jreznik: Keep the file!
14:27:43 <Kevin_Kofler> Rename it or something.
14:27:51 <Kevin_Kofler> But it's useful for testing.
14:27:55 <rdieter> ok, lt's clear we need more leads on this one, can discuss more after meeting, any other bugs to worry about?
14:28:00 <thomasj> ltinkl, i believe you, for sure, but i havent customized shortcuts at all and i have that file.
14:28:10 <jreznik> ops, too late... but I'll try to reproduce it on clean F11 image
14:28:22 <Kevin_Kofler> Ouch, I'm on 4.2.4 right now and my kglobalshortcutsrc is a trainwreck, I have lots of entries of the {01079d9c-a5c7-484e-900b-f554f526254b}=,none,PrintScreen form with varying UUIDs.
14:28:29 <Kevin_Kofler> No wonder PrintScreen doesn't work for some people.
14:28:38 <Kevin_Kofler> There are dozens of conflicting entries for it.
14:29:40 <Kevin_Kofler> I also have that fun entry there: {ee5b94c6-5735-4c35-8421-702dcb87fa2e}=,none,Perform DCOP call 'kdesktop KDesktopIface popupExecuteCommand()'
14:29:45 <Kevin_Kofler> Looks like a leftover from KDE 3.
14:30:22 <ltinkl> yup
14:30:37 <Kevin_Kofler> Anyway, I don't think that is really related to that issue.
14:30:48 <Kevin_Kofler> Those shortcuts got uuid-ized to get rid of them, it seems.
14:31:05 <Kevin_Kofler> There seems to be a different migration issue for 4.2 to 4.3.
14:31:12 <Kevin_Kofler> I guess the ID for "show dashboard" changed.
14:31:20 <Kevin_Kofler> There must be 2 types of "show dashboard" now.
14:31:23 <rdieter> ah
14:31:30 <Kevin_Kofler> And you end up with the same shortcut being assigned to both.
14:31:35 <ltinkl> yes
14:31:42 <Kevin_Kofler> I've read somewhere 4.3 has a different implementation of "show dashboard".
14:31:52 <Kevin_Kofler> There must be the old one too, still, and the same shortcut for both.
14:32:01 * ltinkl digging in the sources and config files
14:32:27 <Kevin_Kofler> The 4.2 one is:
14:32:28 <Kevin_Kofler> [plasma]
14:32:29 <Kevin_Kofler> Show Dashboard=Ctrl+F12,Ctrl+F12,Dashboard anzeigen
14:32:32 <jreznik> hmm, that could be our goal
14:32:51 <Kevin_Kofler> Figure out what the 4.3 one is.
14:32:59 <Kevin_Kofler> Then write a kconf_update scriptlet.
14:33:22 <Kevin_Kofler> That file has been touched by other kconf_update stuff, so it's possible to kconf_update it (we shouldn't get a mess like for plasma-appletsrc).
14:33:55 <ltinkl> Kevin_Kofler: there you go!
14:34:12 <ltinkl> it shouldn't contain the German string imo
14:34:32 <ltinkl> [plasma-desktop]
14:34:33 <ltinkl> Show Dashboard=Ctrl+F12,Ctrl+F12,Show Dashboard
14:34:44 <ltinkl> Kevin_Kofler: notice the different group name!
14:35:00 <ltinkl> [plasma] -> [plasma-desktop]
14:35:04 <Kevin_Kofler> Oh, that explains it.
14:35:05 <rdieter> looks like we have a winner
14:35:11 <Kevin_Kofler> Need a kconf_update script for that.
14:35:21 <ltinkl> yup, now I remember
14:35:39 <ltinkl> aseigo mentioning the plasma executable changed from plasma to plasma-desktop
14:35:45 <Kevin_Kofler> Who's going to write the script?
14:35:47 <jreznik> hmm, yes... I didn't realize that it's not plasma binary but plasma-desktop...
14:36:18 <ltinkl> can someone do this on 4.2: ps aux | grep plasma
14:36:27 <Kevin_Kofler> On 4.2 it's called just plasma.
14:36:32 <rdieter> let's get an upstream bug open first, document our findings, then a fix can be made (by us, upstream, whatever), sound like a plan?
14:36:35 <Kevin_Kofler> It got renamed in 4.3.
14:37:05 <Kevin_Kofler> That pointless rename is just causing problem. :-/
14:37:08 <Kevin_Kofler> *problems
14:37:31 <ltinkl> hmm, looks like Plasma has only one shortcut assigned, so we see this only with the Show Dashboard action, right?
14:37:39 <jreznik> it's plasma on 4.2, plasma-desktop on 4.3
14:37:45 <Kevin_Kofler> Wrong.
14:37:52 <Kevin_Kofler> I also have "Activate Application Launcher Widget=Alt+F1,Alt+F1,Application Launcher"
14:38:20 <thomasj> same here
14:38:37 <ltinkl> aha
14:38:40 <rdieter> the app launcher shortcut has never seems to work for me, is that the same class of problem?
14:38:44 <ltinkl> that one doesn't work either for me
14:38:51 <ltinkl> rdieter: definitely
14:39:31 <jreznik> alt+F1 shows Search
14:39:45 <rdieter> who wants the the task of working on this one?  (or should we assign it to than as penance for not being here? :) )
14:39:48 <ltinkl> jreznik: nothing here
14:40:14 <jreznik> ltinkl: KickOff search is focused
14:40:27 <jreznik> it's not application launcher but could be used that way
14:40:27 <ltinkl> jreznik: nothing happens here
14:40:28 <Kevin_Kofler> I'm using the classic menu on 4.2.4, Alt+F1 does nothing.
14:40:53 <thomasj> nothing here
14:41:43 <rdieter> well, let's move on then, we're running shortish on time.
14:41:54 <rdieter> #topic Extragear for KDE 4.3.0 (status)
14:42:06 <rdieter> who's working on the extragear stuff?
14:42:17 <svahl> than said that last week
14:42:29 <Kevin_Kofler> I don't think anybody is, right now.
14:42:38 <Kevin_Kofler> Than started importing some stuff, then nothing happened.
14:43:03 <rdieter> ok, another todo item them, any volunteers?
14:43:49 <rdieter> otherwise, we can punt on that until we have a chance to talk to than
14:44:16 <Kevin_Kofler> Well, I think we don't have to wait for him.
14:44:23 <SMParrish> I would but I'm booked solid until next week -(
14:44:30 <svahl> I'll do that (and also try to contact than)
14:44:49 <ltinkl> svahl: great, thanks
14:44:56 <svahl> but I don't have rights for all extragear packages
14:45:22 <Kevin_Kofler> svahl: Request the ones you're missing, we'll try to give you access to all.
14:45:42 <svahl> sure
14:46:26 <rdieter> boo, separate issue, just for team management, does pkgdb allow for assigning rights, acls for groups?  ie, have a kdesig group, so we don't have to go through this (like we did recenty for mathstuf, and now svahl)?
14:46:56 <Kevin_Kofler> No.
14:47:06 <rdieter> ok, was afraid of that.
14:47:12 <Kevin_Kofler> There's a hardcoded checkbox for provenpackager (which is supposed to be always checked these days) and that's all.
14:47:33 <rdieter> alright, next topic...
14:47:48 <rdieter> #topic KDE live images - status (see also http://lists.fedoraproject.org/pipermail/fedora-kde/2009-August/003555.html)
14:48:03 <rdieter> svahl ?
14:48:30 <svahl> I've uploaded i686 and x86_64 images yesterday: http://lists.fedoraproject.org/pipermail/fedora-kde/2009-August/003555.html
14:48:44 <svahl> but there were only one reporter yet
14:49:34 <rdieter> thanks!  I'll make sure to get it myself asap.
14:49:37 * jreznik is currently trying x86_64 image in VBox
14:49:39 <svahl> He also said that he needs to use prelink to get it working. Not sure what this means
14:50:03 <jreznik> first login - it crashed, on second attempt it was ok
14:50:35 <jreznik> with The process for the desktop protocol died...
14:50:39 <thomasj> svahl, it means one needs to run: prelink -f /usr/bin/kdeinit4   <== saw that in his rawhide installation already
14:50:42 <rdieter> there has been recent intermittent prelink-related breakage of kdeinit4 , prelink/glibc maintainer closed the bug once someone "fixed" things by manually re-running prelink (I don't consider that a fix)
14:50:57 <svahl> so is this something I have to use in the kickstart?
14:51:14 <rdieter> svahl: maybe we should omit prelink altogether for now.
14:51:17 <svahl> IMHO the packages should work after a clean installation
14:52:11 <jreznik> hmm, really kios are dead
14:52:36 <svahl> the images also contain newer PyKDE4 and kde-settings from koji to get printer configuration and wallpaper back. Are they already tagged for F12Alpha?
14:53:37 <rdieter> svahl: should be, I'll double-check after meeting
14:53:51 <svahl> thx
14:54:04 <rdieter> confirmed
14:54:05 <Kevin_Kofler> They are, they hit Rawhide last night.
14:54:24 <svahl> the biggest issue I've encountered is that firstboot isn't working after installation. bug #515419
14:54:25 <buggbot> Bug https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=515419 medium, low, ---, clumens, NEW, firstboot GUI crashes when launched
14:55:07 <svahl> clumens said that this probably won't be fixed before alpha: https://bugzilla.redhat.com/show_bug.cgi?id=515419#c6
14:55:09 <buggbot> Bug 515419: medium, low, ---, clumens, NEW, firstboot GUI crashes when launched
14:55:34 <rdieter> I guess the only related bug I can find is:  https://bugzilla.redhat.com/show_bug.cgi?id=514048
14:55:36 <buggbot> Bug 514048: medium, medium, ---, jakub, CLOSED RAWHIDE, kded crashes at kde startup
14:56:06 <rdieter> and we (Kevin) closed it. oh well.  time for another one I guess
14:56:14 <svahl> mhh. that one I've not seen in my tests
14:56:20 <mcepl> .
14:56:35 <Kevin_Kofler> I closed it because we thought the latest prelink in Rawhide fixed it.
14:56:47 <Kevin_Kofler> But it looks like it's still broken.
14:56:50 <Kevin_Kofler> So we should reopen it.
14:56:59 <svahl> but randomly the desktop protocol died unexpectactly (or so)
14:57:13 <svahl> right after login
14:57:29 <rdieter> desktop protocol is indeed broken on rawhide, that's been reported several times
14:57:41 <svahl> ah, ok
14:57:53 <svahl> for the package list: popular packages missing atm: amarok, kpresenter, konq-plugins, kdeartwork
14:58:39 <svahl> my images for x86_64 are 30-40 megs bigger than the i686 ones. Cannot explain, why. I think I need someone else for confirmation
14:58:55 <Kevin_Kofler> Any multilibs?
14:59:24 <rdieter> shouldn't be
14:59:25 <svahl> no
14:59:52 <rdieter> we're almost out of time, any last thoughts before we end the meeting?
14:59:58 <svahl> one time the install size of the kernel was 80 megs for i686 and 100 on x86_64
15:00:42 <rdieter> seems dracut is increasing kernel size a bunch, x86_64 even more, so maybe that explains it
15:01:15 <svahl> maybe. but my test run for the gnome spin was only 2 megs difference
15:01:32 <rdieter> oh, interesting, ok, more investigation is warranted then
15:01:44 <svahl> ok. I'll search for more information after the meeting
15:01:49 * rdieter too
15:01:50 <Kevin_Kofler> I think we can't say much more without more data, so let's look into it.
15:01:53 <svahl> I think that's all for live images
15:02:05 <Kevin_Kofler> Let's close the meeting, the bug zappers are knocking on our door.
15:02:09 <rdieter> ok, let's wrap things up, thanks everyone
15:02:11 <rdieter> #endmeeting