15:11:34 <Kevin_Kofler> #startmeeting KDE SIG Meeting
15:11:34 <zodbot> Meeting started Tue Mar  3 15:11:34 2015 UTC.  The chair is Kevin_Kofler. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:11:34 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:11:35 <helio|mtg> Kevin asked for someone take the lead
15:11:42 <Kevin_Kofler> Actually, Rex did.
15:11:48 <Kevin_Kofler> So I'm leading it.
15:11:50 <Kevin_Kofler> #meetingname kde-sig
15:11:50 <zodbot> The meeting name has been set to 'kde-sig'
15:11:52 <helio|mtg> Kevin_Kofler: Youthen
15:11:56 <Kevin_Kofler> #topic Roll call
15:12:00 * jreznik pretends to be here, hides
15:12:02 <rdieter> here
15:12:10 * helio|mtg is here
15:12:15 * jgrulich is here too
15:12:15 <tosky> me too
15:12:25 * satellit listening
15:12:38 <dvratil> hello again :)
15:13:00 <Kevin_Kofler> pino|work and than already reported present before we started.
15:13:12 * pino|work is still here
15:13:18 <Kevin_Kofler> #chair jreznik rdieter helio|mtg jgrulich tosky satellit dvratil pino|work than
15:13:18 <zodbot> Current chairs: Kevin_Kofler dvratil helio|mtg jgrulich jreznik pino|work rdieter satellit than tosky
15:13:41 <Kevin_Kofler> #info Kevin_Kofler, jreznik, rdieter, helio|mtg, jgrulich, tosky, satellit, dvratil, pino|work, than present.
15:13:46 <Kevin_Kofler> #topic Agenda
15:14:00 <Kevin_Kofler> So I think F22 Alpha debugging is the main topic.
15:14:02 <Kevin_Kofler> Anything else?
15:14:14 <helio|mtg> Status report on LXQt
15:15:19 <Kevin_Kofler> #topic F22 Alpha status/debugging
15:16:45 <Kevin_Kofler> So, the status right now is this: There were some kickstart and comps issues that prevented TC7 from working, those should be fixed now. There are however also issues with the actual binaries, which we think are GCC-5-related.
15:17:25 <dvratil> the rebuild did not help?
15:17:35 <Kevin_Kofler> The binaries currently in stable are affected by one or more known GCC bug(s) that has/have been fixed in the latest GCC build. In particular, plugin loading doesn't always work and KMix is crashing on startup.
15:17:44 <rdieter> adamw reports other issues, graphical glitches
15:17:47 <satellit> FYI live Plasma TC6 has KCMinit and KMIXctrl cause pop up bugs and defaults to openbox?
15:17:48 <Kevin_Kofler> The binaries in updates-testing fix that, but have a new regression.
15:18:07 <jreznik> rdieter: what kind of graphical glitches? you mean anaconda styling?
15:18:15 <satellit> tc8*
15:18:20 <rdieter> the scrollback in #fedora-kde from last night has a screenshot
15:18:35 * rdieter checks browser history
15:19:23 <Kevin_Kofler> https://www.happyassassin.net/temp/busted_kde.png
15:19:46 <rdieter> that's with the kf5 gcc5 rebuilds installed
15:20:09 <rdieter> what is currently queue'd for f22 updates-testing
15:20:22 * helio|mtg confirms that kmix has no issue with clanc compiled, so gcc5 is a good guess
15:20:34 <Kevin_Kofler> satellit: The Openbox thing is yet another issue that is not being considered a blocker.
15:20:48 <satellit> yes I know
15:20:49 <helio|mtg> openbox ?
15:20:53 <Kevin_Kofler> Basically, DNF drags in all sort of unwanted dependencies and openbox is part of those.
15:21:11 <Kevin_Kofler> And SDDM happily picks the first session in the alphabet as the default, openbox coming before plasma.
15:21:12 <satellit> alphabetical listing in chooser
15:21:25 <rdieter> .bug 1197135
15:21:26 <jreznik> rdieter: the black squares sometimes happens to me to with compositing off
15:21:26 <Kevin_Kofler> We need openbox the heck out of the spin.
15:21:28 <zodbot> rdieter: Bug 1197135 f22 TC6 live KDE x86_64 installs openbox, sddm defaults to openbox as a session on first boot - https://bugzilla.redhat.com/1197135
15:21:40 <rdieter> jreznik: so you think may be unrelated ?
15:22:01 <jreznik> rdieter: unrelated to gcc issues? I'd say so
15:22:09 <rdieter> jreznik: ok, good
15:22:15 <Kevin_Kofler> Hmmm, what happens if you "dnf remove openbox" on an install of the latest TC?
15:22:36 <jreznik> heh, seems like we have default KDE wp there, so without anything done it's not blocking :D
15:22:41 <Kevin_Kofler> Because if that doesn't remove anything important, we can just add "dnf -y --disablerepo=* -C remove openbox" to the kickstart %post.
15:22:57 <rdieter> full fix for 1197135 requires a fixed kwin build and patched initial-setup
15:23:24 <rdieter> kwin part is done, but not submitted as an update yet (part of plasma-5.2.1)
15:23:30 * satellit I used anaconda 22.20.1-1 passwords
15:24:02 <rdieter> .bug 1192182
15:24:04 <zodbot> rdieter: Bug 1192182 Tracker: issues caused by poor ambiguous provides resolution - https://bugzilla.redhat.com/1192182
15:24:04 <rdieter> the other part ^^
15:24:16 <rdieter> hrm, maybe not
15:24:22 <Kevin_Kofler> If we were using an installer that does the whole work (Calamares), we wouldn't need that crappy initial-setup hack to begin with.
15:24:59 <satellit> some lives require anaconda passwords before finishing
15:25:06 <rdieter> I guess I just need to change the component of the primary bug
15:25:28 <rdieter> Kevin_Kofler: offtopic
15:25:41 <rdieter> (ie, arguing to use calamares at this point is not constructive)
15:26:19 * satellit and it is not ready per my testing
15:27:23 <Kevin_Kofler> https://bugzilla.redhat.com/show_bug.cgi?id=1197135#c8
15:27:37 <Kevin_Kofler> IMHO, rejecting that as a freeze exception is very broken.
15:27:37 <satellit> change requirements to stop and require passwords in anaconda would bypass initial setup
15:28:14 * jreznik still hopes one day we will have ODM with preinstalled Fedora KDE and the, initial-setup is going to be needed :)
15:28:56 <Kevin_Kofler> ODM?
15:29:11 * danofsatx is here, busy setting up new system.
15:31:32 <rdieter> danofsatx: hi
15:31:56 <rdieter> so sounds like pulling in the kf5- rebuilds *may* not be so bad after all, but I'd feel better with a bit more testing
15:32:44 * danofsatx can start testing as soon as the new machine is loaded.
15:33:01 <danofsatx> I still have to grab the Windows license key before I wipe and load Fedora.
15:33:42 * rdieter tries booting tc8 on redone usb stick, seems to be working better this time
15:34:19 * satellit I used liveusb-creator in (dd) mode (direct write)
15:34:34 <satellit> makes nice USB for install
15:34:42 <rdieter> I may resort to that, used livecd-iso-to-disk
15:35:26 <rdieter> anyone else with f22-alpha testing feedback ?
15:36:59 <satellit> have not tried boot.iso of Plasma  but it di work 3 days ago
15:37:05 <satellit> did*
15:42:43 <rdieter> well, please test some more, I guess we're done with this topic then?
15:43:36 <jreznik> well, seems like TC8 is pretty solid...
15:43:50 <jreznik> I only see a lot of gcc crashes but mostly when closing apps
15:44:14 <jreznik> updating Qt
15:44:40 <helio|mtg> I get many kded5 crashed using LXQt, don't know the reason yet
15:45:06 <rdieter> helio|mtg: with rebuit qt-5.4.1 ?
15:45:11 <jreznik> so the only missing piece is theming now... but it's more likely FE now as it uses default backrounds from KDE, so it's not a blocker
15:45:12 <helio|mtg> rdieter: Yes, everything
15:46:18 <Kevin_Kofler> I think we have more important things to worry about than theming for the Alpha.
15:46:41 <satellit> fix missing favorites to logout menu backup..?
15:47:16 <satellit> have to alt f2 logout to fix it
15:47:19 <jreznik> rdieter: not sure if it's related but after updating qt, I can see that black squares too
15:47:49 * helio|mtg starts the f22 vm
15:47:52 <rdieter> gah
15:48:10 <jreznik> and more breakage than before
15:48:17 <Kevin_Kofler> jreznik: What are you running this under?
15:48:17 <jreznik> let me try clean reboot
15:48:27 <Kevin_Kofler> A VM?
15:48:32 <jreznik> Kevin_Kofler: TC8 + updated Qt in virt-manager
15:48:38 <Kevin_Kofler> If so, the bug is going to be very common even if it's a "hardware" bug.
15:48:48 <Kevin_Kofler> That's most likely what adamw used too.
15:48:59 <jreznik> well, still seems to be related to Qt update...
15:49:10 <Kevin_Kofler> I guess it could be worked around by using a different video emulation in virt-manager/qemu.
15:49:35 <Kevin_Kofler> QEMU does qxl, cirrus and "std" (standard VGA/VESA only).
15:49:52 <jreznik> let me check, if it's really caused by that rebuilded Qt
15:50:00 * rdieter gives up trying to get his trusty old dell d630 to boot f22-tc8, looks like lost cause (i965 kernel crash)
15:50:01 <satellit> need to test in boxes also
15:50:12 <Kevin_Kofler> But requiring the workaround would not be very nice to our users.
15:50:32 <Kevin_Kofler> satellit: Doesn't Boxes use the same libvirt setup as virt-manager under the hood?
15:50:42 <jreznik> Kevin_Kofler: yep, it does
15:51:06 <jreznik> Boxes are just crippled UI on top of the same hood
15:51:26 * satellit testing f21 boxes atm
15:54:22 <satellit> boot screen  with blue progress bar in boxes
15:55:02 <satellit> Plasma is up in boxes here  TC8 x86
15:55:12 * satellit live
15:56:11 <rdieter> we can continue hamering f22-alpha testing in #fedora-kde after meeting, our hour is almost up, and we had at least one more topic to hit, can we move on?
15:56:18 <Kevin_Kofler> satellit: Now try the Qt update from updates-testing and see what happens.
15:56:51 <Kevin_Kofler> rdieter: Yeah, though it doesn't really have all that much to do with KDE. ;-) But let's make Helio happy.
15:56:56 <Kevin_Kofler> #topic LXQt status report
15:56:58 * satellit nubie tester do I just use dnf update QT?
15:57:00 <Kevin_Kofler> helio|mtg: Your turn.
15:57:04 <jreznik> one more thing - no install icon on the desktop
15:57:06 <helio|mtg> Great
15:57:13 * helio|mtg holds
15:57:16 <Kevin_Kofler> satellit: qt5-qtbase, not QT
15:57:27 <satellit> k
15:57:38 <Kevin_Kofler> I guess you don't have to explicitly --enablerepo=updates-testing because it's on by default in Alpha.
15:57:47 <helio|mtg> Ok, can i go ?
15:57:56 <Kevin_Kofler> helio|mtg: Yes.
15:58:15 <helio|mtg> Ok, so finally allpackages, including pcman is in the same versioning
15:58:19 <helio|mtg> theme was added
15:58:33 <helio|mtg> And after new Qt and rebuild, everything works
15:58:49 <helio|mtg> There's minor inconveniences, like one layer menu
15:58:53 <helio|mtg> Becomes too big
15:59:06 <helio|mtg> And the xdg-open one, that is crashing kded5
15:59:11 <helio|mtg> Not known
15:59:27 <helio|mtg> We're able to do a live spin already
15:59:54 <Kevin_Kofler> Nice.
16:00:05 <helio|mtg> I appreciate if people can test it a little, because i'm unable to discover all things
16:00:15 <helio|mtg> I had one report that default is not all done, like terminal
16:00:21 <Kevin_Kofler> What are the hardware requirements for LXQt? Does it use QML2 and thus OpenGL 2? Or is it using QWidgets or something?
16:00:31 <helio|mtg> QWidgets, no opengl
16:00:36 <helio|mtg> No QML at all for now
16:00:47 <helio|mtg> Use some Frameworks libraries
16:00:54 <helio|mtg> Now KWindowSystem and soon Solid
16:01:07 * Kevin_Kofler might end up switching to LXQt on this machine if Plasma 5 performs too horribly on llvmpipe…
16:01:10 <helio|mtg> They thinking on kscreen, but is a little more complicated as dvratil talked already
16:01:35 <Kevin_Kofler> (but I guess I'll worry about that when F21 reaches EOL, no sooner :-) )
16:02:51 <helio|mtg> f21 is in the updates already
16:02:55 <helio|mtg> Testing soon
16:04:05 <helio|mtg> But that's it
16:04:34 <helio|mtg> I will endup doing the livecd myself, if ralphgro not want
16:04:47 * helio|mtg don't want meeting get longer
16:05:42 <rdieter> helio|mtg: thanks
16:05:44 <Kevin_Kofler> OK, I'm summing up for the summary.
16:06:23 <Kevin_Kofler> #info All packages including pcmanfm are updated to the current LXQt version, a live spin should be possible now.
16:06:48 <Kevin_Kofler> Thanks for the status update.
16:06:52 <Kevin_Kofler> #topic Open discussion
16:07:01 <Kevin_Kofler> Anything else or can I close the meeting?
16:07:20 <danofsatx> still some lingering errors with moving to plasma5 in F21, but I'll take that up in channel later
16:09:56 <Kevin_Kofler> OK, I guess that was all, thanks all for coming!
16:09:58 <Kevin_Kofler> #endmeeting