19:22:19 <ennael> #startmeeting
19:22:19 <Inigo_Montoya> Meeting started Mon Jun  8 19:22:19 2015 UTC.  The chair is ennael. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:22:19 <Inigo_Montoya> Useful Commands: #action #agreed #help #info #idea #link #topic.
19:22:24 <ennael> so let start meeting
19:22:30 <wilcal> ok
19:22:47 <Sebsebsebb> Hi
19:22:50 <ennael> last meeting about mga5 !
19:23:19 <marja> \o/
19:23:54 <ennael> so let see where we are :)
19:24:22 <marja> MrsB: meeting started ;-)
19:24:28 <Akien> Hi everyone
19:25:35 <wilcal> Hello
19:25:52 <ennael> so you must have mlive isos for final tests
19:27:33 <ennael> and I'm uploading classical isos at the moment
19:27:37 <marja> nice
19:27:38 <ennael> any ETA on tests ?
19:29:26 <MrsB> hi
19:30:26 <MrsB> tests have been ongoing on live isos, no major issues I'm aware of. There is the libx86 regression which has just been fixed so would be good to include that
19:31:13 <marja> MrsB: I saw a regression on the LiveCD (Gnome), a fail whale that wasn't there with the round before this one
19:31:17 <MrsB> My time has become quite limited unfortunately for the next few weeks :\
19:31:26 <marja> MrsB: https://bugs.mageia.org/show_bug.cgi?id=16088
19:31:28 <[mbot> [ Bug 16088 Round 6 Mageia-5-LiveCD-GNOME-en-i586-CD: Oh no! Something has gone wrong ]
19:31:48 <MrsB> Is that a new one or a new report of an old one?
19:32:19 <marja> MrsB: it's a new report, because I only saw it with the last round of isos
19:32:51 <MrsB> Ok, tmb any thoughts on that one?
19:33:23 <MrsB> /usr/lib/dri/i915_dri.so: undefined symbol: _glapi_tls_Dispatch
19:33:24 <marja> MrsB: I should have tested on other hw, didn't find time :-(
19:35:30 <Luigi12_work> hmm, I fixed the glapi_tls_Dispatch thing like a year ago
19:35:45 <marja> Luigi12_work: there is an error before that "systemd-logind: failed to get session: PID 1651 does not belong to any known session
19:35:49 <marja> "
19:35:57 <MrsB> two undefined symbols there
19:35:57 <marja> I don't know what that means
19:36:04 <Luigi12_work> I don't know about the logind thing
19:37:03 <marja> Luigi12_work: there are no problems with the GnomeDVD
19:37:37 <Akien> If there's someone who hasn't synced yet, it would be nice to compare the .idx from the previous round and from the new one
19:39:50 <tmb> I have them, and there is no missing package if thats what you mean
19:40:40 <marja> if no one else reports this, then let's just ignore it..... this was on old hw
19:41:46 <Luigi12_work> yeah that undefinied symbol shouldn't be happening, both x11-server and mesa still have glx-tls enabled
19:42:06 <Stormi> marja: oops, now I am
19:42:16 <marja> Stormi: :-)
19:42:18 <Luigi12_work> maybe you were using a wonky driver on that machine?
19:42:44 <tmb> btw, for those hitting "Oh no".... please try to close the "oh no" window with "alt-f4" ... it works for some...
19:43:08 <Akien> Good to know
19:43:11 <marja> Luigi12_work: I was just using the GnomeLiveCD
19:43:15 <marja> tmb: thx
19:43:24 <Luigi12_work> hmm, so alt-f4 = kool-aid man?  {oh yeahhhh!!!!!}
19:43:52 <Luigi12_work> marja: did you get the _glapi_tls_Dispatch problem?
19:44:55 <marja> Luigi12_work: there was an error before that, so not sure whether the first error wasn't the real problem
19:45:10 <tmb> there are timing issues that gets us, mostly because we have to restart gdm to change translations... but that reminds me... do I even need to do so on a english-only CD ?
19:45:24 <Luigi12_work> IINM, the lives use proprietary video drivers if applicable, maybe one of those is broken...not sure
19:45:42 <marja> tmb: no... you can switch language in the boot screen, but switching is useless
19:46:21 <marja> tmb: and maybe it could be set to only one kind of English, instead of giving the option to choose between 5 or so English varieties?
19:47:01 <tmb> Luigi12_work, yeah, the display-driver-helper it supposed to cope with that.... I wonder if I need to switch from ldconfig -X to the heavier full ldconfig to make sure we get all data updated
19:47:06 <marja> Luigi12_work: I doubt it, because the LiveKDE and the GnomeDVD don't have a problem
19:47:47 <tmb> marja, did you change any language settings during boot ?
19:48:06 <Luigi12_work> marja: are you sure?  Maybe they all have that error in the logs but it's only causing visible issues on the GNOMECD?
19:48:30 <MrsB> maybe we should just drop LiveCD's
19:48:54 <Luigi12_work> is another major distro still doing CD-sized Lives?
19:49:06 <tmb> I think we should atleast replace them with some of the lighter DEs
19:49:12 <spturtle> sounds like https://bugs.mageia.org/show_bug.cgi?id=14520
19:49:14 <[mbot> [ Bug 14520 Outdated ldconfig cache resulting in "Oh no! Something has gone wrong" in GNOME (liveDVD with xdriver=...) ]
19:49:26 <MrsB> makes sense tmb yeah
19:49:34 <marja> tmb: I did try (at least) 1. change lang to nl in boot screen, and later to English (the top choice) 2. change nothing in the boot screen, but again to the top English later on
19:50:02 <marja> tmb: I hope I tried not touching the language and keeping the defaults on my third try
19:50:13 <marja> tmb: but I'm not sure I did (only that I planned to)
19:51:47 <marja> Luigi12_work: ah, hadn't thought of that (same problem everywhere, but only visible in GnomeCD)
19:52:06 <marja> Luigi12_work: I should have compared the Xorg log :-/
19:53:25 <marja> MrsB: tmb: stormi suggested an XFCE Live CD for Mga6, and I agree
19:53:37 <Luigi12_work> ahh, we might have a real bug
19:53:41 <Akien> marja: Hm, could it be yet another "GNOME does not know how to handle locales properly as soon as it's a bit exotic" issue?
19:53:55 <Luigi12_work> upstream says glx-tls needs to be enabled in glamor too, and it doesn't appear to be in our package
19:53:59 <Stormi> marja: I said a lightweight DE, I haven't said XFCE specifically
19:54:06 <Akien> As in, maybe your system was setup with nl locale due to your choice, but locales-nl is not on the CD, so it should default to English
19:54:13 <Luigi12_work> I wonder if this is causing some of our GNOME issues
19:54:13 <marja> Akien: Is British English exotic
19:54:15 <marja> ?
19:54:18 <Luigi12_work> on ATI specifically
19:54:23 <marja> Stormi: sorry
19:54:25 <Akien> marja: I thought you said "nl"
19:54:28 <Luigi12_work> didn't we have some GNOME problems that were ATI-specific?
19:54:32 <Akien> Ah ok it was in boot screen, sorry
19:55:02 <marja> Akien: that was only the first try, but the 2nd and 3rd I certainly hadn't touched the language in the boot screen
19:55:27 <Akien> Luigi12_work: Two of them yeah, currently in errata
19:55:36 <Luigi12_work> glamor might be the cause
19:55:59 <Akien> https://bugs.mageia.org/show_bug.cgi?id=15653
19:56:01 <[mbot> [ Bug 15653 GNOME (mutter) calls an XRANDR function not supported on some AMD hardware when using proprietary driver (fglrx), causing a crash ]
19:56:07 <Akien> https://bugs.mageia.org/show_bug.cgi?id=16033
19:56:09 <[mbot> [ Bug 16033 Radeon fails to load hardware acceleration from DVD, works from USB ]
19:56:24 <Luigi12_work> I don't think fglrx would be affected
19:56:31 <Akien> Ah btw the issue looks similar
19:56:37 <Akien> s/issue/symptom/
19:56:45 <Akien> EE) AIGLX error: dlopen of
19:56:45 <Akien> > > /usr/lib64/dri/r600_dri.so failed (/usr/lib64/dri/r600_dri.so: undefined
19:56:45 <Akien> > > symbol: _glapi_tls_Dispatch)
19:56:59 <Luigi12_work> ahh there it is
19:57:27 <Luigi12_work> we should test if fixing glamor fixes that
19:57:45 <Akien> I think it's worth trying yes
19:58:21 <tmb> Luigi12_work, well, glamor is part of x11-server and that has tls enabled
19:58:48 <Luigi12_work> :v -d Mageia -r cauldron -s glamor
19:59:03 <Luigi12_work> oh no Sophie
19:59:24 <Luigi12_work> Sophie says we still have the glamor SRPM in Cualdron
20:00:02 <Luigi12_work> we do, libglamor0 and libglamor-devel
20:00:47 <Luigi12_work> although apparently it's supposed to autodetect it at build time in mesa, even though the cnfigure otuput says defualt disabled, according to fdo@66515
20:00:56 <Luigi12_work> so might not be the problem after all
20:01:26 <tmb> yeah I see we have only %rename glamour so it wont touch the libs
20:01:41 <Luigi12_work> ahh whoops
20:01:58 <Luigi12_work> we should fix that
20:02:32 <MrsB> nothing glamor on the ISO though
20:03:49 <MrsB> or glamour for that matter, which is clearly the proper spelling ;)
20:03:58 <marja> :-)
20:04:03 <Luigi12_work> that'd be a waste of letters
20:04:35 <tmb> so call it g+ :)
20:05:16 <Luigi12_work> I think Google might have trademarked that already :o
20:05:29 <Luigi12_work> how about g5r?  hehe
20:05:40 <Luigi12_work> oh g4r, I can't count
20:05:59 <ennael> (/me thinks she is in another chan in another secret and strange project)
20:06:17 <Luigi12_work> lol e4l
20:06:26 <marja> :=)
20:06:40 <Luigi12_work> so we're back to square one, this undefined symbol shouldn't happen
20:08:18 <marja> btw, it was only with the LiveKDE CD that I didn't hit the issue on the same system.... the GnomeLiveDVD was tested on a different one
20:11:04 <Akien> I guess we won't find a miracle right now to unbreak GNOME anyway...
20:11:23 <Akien> Hopefully GNOME will behave better in Mageia 6 once it's upgraded to a newer version (or it can be worse, too...)
20:12:07 <wilcal> Hopefully
20:13:35 <ennael> ok so classical isos uploaded
20:13:40 <Akien> Nice
20:13:53 <Akien> I plan to do some extensive testing on real hw tomorrow
20:14:07 * marja already started syncing
20:15:30 <tmb> I'll try some ideas tomorrow, like trying to call full ldconfig everywhere if I can find a system to reproduce on
20:17:05 <Luigi12_work> please nuke the old glamor packages too
20:19:31 <tmb> hm, I need to check deps for it
20:20:35 <MrsB> anything else about ISOs?
20:20:55 <wilcal> Except for the Gnome?firefox thing I found no
20:21:00 <Luigi12_work> nothing depends on it
20:21:07 <wilcal> https://bugs.mageia.org/show_bug.cgi?id=16092
20:21:09 <[mbot> [ Bug 16092 Gnome, All Live Media, Firefox, no sound ]
20:21:19 <MrsB> That's a codec missing wilcal, don't think it's a bug
20:21:28 <wilcal> k
20:21:39 <wilcal> Can it be fixed before release
20:21:40 <Stormi> why does it work in KDE Live and not GNOME live?
20:21:49 <wilcal> That I don't understand
20:21:55 <MrsB> phonon-vlc maybe?
20:21:56 <Akien> I guess KDE uses pulse, and GNOME gstreamer
20:22:03 <Akien> Ah yes phonon-vlc, not pulse, sorry
20:22:22 <Akien> I got the "p" right :p
20:22:29 <marja> :-)
20:22:46 <Luigi12_work> yeah nothing depends on libglamor0 and nothing BR's libglamor-devel or any of its provides
20:23:25 <Luigi12_work> hmm but it is a bug if it doesn't work and on a live it can't really be "fixed" later
20:24:24 * marja still doesn't understand what the first error means
20:24:33 <marja> "
20:24:35 <marja> Jun 06 17:14:07 localhost gdm-Xorg-:0[1651]: (EE) systemd-logind: failed to get session: PID 1651 does not belong to any known session"
20:24:49 <MrsB> try and identify which codec it's missing wilcal. Youtube gives some stats IINM which might tell you what it's missing
20:24:55 <tmb> well tehnically you can install rpms in live mode
20:24:56 <marja> can't that somehow be the real cause
20:25:08 <wilcal> "IINM"?
20:25:16 <MrsB> If I'm Not Mistaken
20:25:27 <MrsB> right click, stats for nerds
20:26:37 <wilcal> Plugins in Firefox?
20:27:18 <tmb> marja, that might be a fallout of us restarting gdm, and not all processes get the hint, so login manager looses track of it. but it shouldnt be the real cause for 'oh no'
20:27:49 <marja> tmb: yeah, I just found that coling said in another bug report that that's a red herring
20:29:15 <marja> something I also haven't tried is doing "
20:29:41 <marja> systemctl restart prefdm.service"
20:32:10 <tmb> yeah, the prefm stuff is something else we need to kill off for mga6...
20:34:09 <MrsB> So more experiments tomorrow and we'll know more after that I suppose
20:34:24 <MrsB> What else do we need to do for release?
20:34:55 <marja> mail mirrors ml (but only a sysadmin can do that)
20:36:01 <marja> wrong name, it's mirrors-announce@ml.mageia.org
20:36:03 <tmb> yep. I used to do it, but forgot last release
20:36:37 <marja> tmb: if you didn't get any complaints, then there isn't much urge :-)
20:37:03 <marja> tmb: but isn't it time to wipe Mga1 from the mirrors?
20:37:38 <tmb> yeah, but its nice to notify sysadmins
20:37:52 <marja> I think, with Mga5 repo + isos, a mirror will need more than 1TB for Mageia
20:38:02 <marja> tmb: it is :-)
20:38:20 <wilcal> need 2 keep the mirror under 1TB
20:39:01 <MrsB> ennael: did the classic isos pick up the libx86 fix?
20:39:10 <Luigi12_work> we probably have 1TB of game data alone now :o(  so yeah we should warn them
20:39:21 <Luigi12_work> the libx86 fix is only in updates_testing
20:39:22 <tmb> for mirrors there was a plan to move old releases to an archive section
20:39:29 <wilcal> Ya FlightGear :-O
20:39:44 <Luigi12_work> that one's been there
20:39:54 <Luigi12_work> Akien imported a whole bunch more huge games in mga5
20:39:56 <marja> Luigi12_work: no, I have a full mirror (1-cauldron + isos), yesterday it was 892,6 GB
20:40:15 <Luigi12_work> ok, so maybe 500GB of game data
20:40:24 <Stormi> Well if sometime games are too big we can still have a games repo
20:40:26 <Akien> Luigi12_work: No way
20:40:28 <marja> Luigi12_work: probably, lol
20:40:43 <Akien> Ah for 5 releases maybe
20:41:23 <MrsB> Is there anything else for tonight?
20:41:36 <marja> tmb: yeah, an archive section would be appreciated :-)
20:42:13 <marja> MrsB: not for me, except bedtime :-)
20:42:32 <ennael> MrsB: yep just synced repo before
20:42:46 <MrsB> it was in updates_testing apparently
20:43:02 <ennael> ah
20:43:09 <ennael> let me check
20:43:15 <MrsB> ok
20:43:33 <ennael> raaaa
20:43:36 <MrsB> :D
20:43:37 <ennael> why in testing
20:43:40 <ennael> ...
20:43:44 <MrsB> /o\
20:43:48 <tmb> and if you want it in installer you need to rebuild stage2 too
20:43:53 <Luigi12_work> because we can't push to core/release ?
20:44:07 <tmb> ennael, lockdown :)
20:44:13 <Luigi12_work> eatdirt actually pushed a broken build to core/updates_testing first but then I fixed it
20:44:50 <ennael> sorry I'm lost now
20:45:00 <ennael> is it still in testing?
20:45:12 <MrsB> yes, and you'll need to rebuild stage2 too
20:45:28 <ennael> my question is is it in testing
20:45:31 <Akien> Yes
20:45:35 <ennael> then we need to push it in release
20:45:36 <Stormi> like being tested
20:45:41 <Stormi> ?
20:45:46 <Akien> Not core/updates_testing
20:45:49 <Stormi> ok :)
20:45:52 <Akien> s/Not/no/
20:46:10 <Stormi> step 1 push to release
20:46:15 <Stormi> step 2 rebuild stage2
20:46:21 <Stormi> step 3 rebuild ISOs
20:46:21 <ennael> #chair MrsB Akien tmb
20:46:21 <Inigo_Montoya> Current chairs: Akien MrsB ennael tmb
20:46:23 <Stormi> am I good?
20:46:28 <MrsB> uhohh
20:48:08 <Luigi12_work> ok I bumped the release tag of libx86, it can be rebuilt in core/release
20:48:15 <marja> thx
20:48:17 <Luigi12_work> someone said it was a regression, it wasn't a regression
20:48:38 <Luigi12_work> it's been broken for several months.  I fixed it for i586 a few months ago, but it was never fixed for x86_64 until today.
20:48:47 <Akien> Ok, so apart from this last minute inconvenient, I guess we are somewhat done for tonight?
20:49:16 <tmb> yeah
20:49:19 <ennael> sorry for the "inconvenience"
20:49:23 <MrsB> Is that likely to have any impact on my 64bit screen corruption?
20:49:47 <Akien> I guess we got on ennael's nerves...
20:49:48 <Luigi12_work> maybe, get-montior-from-edid-using-vbe or whatever was broken without this fix
20:50:01 <Luigi12_work> hard to know
20:50:18 <Luigi12_work> guess you can test the new libx86_1 and see
20:50:23 <MrsB> yep
20:50:33 <MrsB> when it's done
20:50:49 <Luigi12_work> it's in updates_testing now, not sure if you need a new ISO to test
20:50:50 <Akien> So should classical + lives be rebuilt to get the new libx86?
20:50:50 <marja> well, so the i586 classical can be tested
20:51:00 <Sensebsebbb> Meeting still on or over?
20:51:14 <MrsB> a bit of both
20:51:56 <MrsB> Let's wrap this up then
20:52:03 <marja> and the dual on 32bit hw can be tested, too
20:52:28 <MrsB> Anybody have anything else to add before we end?
20:52:29 <wilcal> Off to testing ISO's
20:52:46 <Stormi> well, anybody not busy and having time, I accept help for Errata :)
20:52:46 <Akien> I'm good.
20:52:48 <wilcal> not from me
20:52:55 <Stormi> but it's not the main priority
20:53:01 <tmb> not from me
20:53:11 <marja> Stormi: thx for all your work on the bugs + errata
20:53:23 <Stormi> well thx for everyone's work
20:53:30 <Stormi> mine just sends more mails
20:53:32 <marja> +1
20:53:33 <Stormi> :)*
20:53:39 <MrsB> Let's do it then. Thanks all for being here. Sorry to Anne if we annoyed her.
20:53:46 <MrsB> #endmeeting