19:15:44 <ennael> #startmeeting
19:15:44 <Inigo_Montoya`> Meeting started Wed Oct  2 19:15:44 2013 UTC.  The chair is ennael. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:15:44 <Inigo_Montoya`> Useful Commands: #action #agreed #help #info #idea #link #topic.
19:15:47 <neoclust> ennael: yeah i know you :) i have some chouchenn for you :)
19:15:51 <sebsebseb> hi
19:15:51 <ennael> hi all
19:15:58 <ennael> thanks for attending this meeting
19:16:44 <ennael> #topic approaching freeze
19:17:03 <ennael> so first topic is about coming freeze planned for 15th of october
19:17:13 <ennael> this is version freeze
19:17:19 <ennael> is it clear for everyone ?
19:17:42 * coling nods like an obedient little packager
19:17:47 <marja> ennael: please say it again
19:17:53 <coling> (and runs off to package git versions of n+1)
19:18:05 <ennael> :)
19:18:17 <sebsebseb> n+1 whatever that is
19:18:28 <ennael> as a reminder, version freeze means that no new version will be allowed after 15th of october
19:18:38 <coling> The only new versions I know I want to push would be pulseaudio for proper bluez support
19:18:42 <ennael> meaning only new releases
19:19:15 <ennael> except if you have good reason to ask for a submit (bug fix, RC version in distro, ...)
19:19:46 <ennael> then you need to mail -dev to ask for it explaining: what package, what version, why we should let it in
19:21:13 <ennael> is it clear enough ?
19:21:40 <nmartin-mageia> yes
19:21:43 <malo> ennael: yes
19:22:02 <marja> ennael: thx
19:22:14 <coling> yup!
19:22:18 <ennael> no pb :)
19:22:48 <ennael> so we have just finalized alpha3 isos. It should be published tonight
19:23:02 <marja> \o/
19:24:28 <malo> congrats QA ennael tmb
19:25:27 <ennael> so next dates are version freeze for 15th of october
19:25:54 <ennael> full rebuild of the distro planned to start from 16th of october
19:26:14 <ennael> and beta 1 on 31rst of october
19:27:21 <coling> Cool. /me will try and fix his rpm debuginfo problem before mass rebuild...
19:27:30 <coling> Tho' reproducing is awkward.
19:27:41 <coling> Not a huge issues tho'.
19:28:52 <grenoya> coling: depend for who!
19:29:19 <Oro_Valley> hello, just got back from lunch
19:29:37 <ennael> what we will need also to organize is a meeting or at least a discussion between desktop maintainers
19:30:32 <tmb> coling: well, you can simply disable the "fix" before the rebuild as the BS uses clean chroots during build so it should not hit it...
19:30:39 <ennael> while testing alpha3 isos it seems we really need to rework desktops packaging and especially task packages
19:30:42 <coling> tmb, already done :)
19:30:49 <ennael> also menus, look'n feel...
19:30:51 <coling> tmb, but yeah true.
19:31:25 <malo> sorry guys, I have to leave for 1h, so please make it a very loooooooooong meeting
19:31:44 <AL13N> heh
19:31:50 <malo> :-)
19:32:07 <ennael> any question on this topic ?
19:32:07 * malo forgot the smiley :-P
19:32:13 <AL13N> ennael: is the desktopchooser ok now?
19:32:23 <AL13N> (in the installer?)
19:32:23 <coling> Does anyone know when Derek will be back to finish of that feature?
19:32:26 <coling> Ha!
19:32:32 <coling> Same Q as AL13N
19:32:38 <coling> :)
19:32:44 <AL13N> :)
19:33:14 <ennael> ok next topic
19:33:17 <AL13N> imho that first screen should really list only 3 options, the next custom screen, i have no objection
19:33:20 <ennael> #topic features review
19:34:01 <ennael> let start :)
19:34:03 <ennael> https://wiki.mageia.org/en/Feature:CheckboxesToChooseDE
19:34:09 <philippem> about this and future full rebuild , there are a lot of python modules that have build errors since we bytecomppile, usualy it's enough to fix the  file list
19:34:11 <ennael> is derek around ?
19:34:55 <philippem> sorry too soon
19:35:08 <ennael> it's back on stage2 but still not satisfying and usable
19:35:16 <ennael> we will mail -dev about this
19:35:30 <ennael> https://wiki.mageia.org/en/Feature:DiskDrakeRescan
19:35:32 <ennael> AL13N: ?
19:36:02 <AL13N> no progress for now, if i don't make it, we'll have to postpone :-(
19:36:21 <ennael> ok
19:36:26 <ennael> https://wiki.mageia.org/en/Feature:DoNotShipSysVInitScripts
19:36:30 <ennael> coling: ?
19:36:38 <coling> Yeah........
19:36:44 <coling> So I've done very little on that one..
19:36:49 <AL13N> it's sort of ongoing, so it's not that important, imho
19:36:58 <coling> I still maintain it's a general "aim" really anyway :p
19:37:17 <ennael> ok so maybe not completed for mageia4, but work in progress ? :)
19:37:23 <coling> Yeah.
19:37:27 <ennael> ok
19:37:28 <ennael> https://wiki.mageia.org/en/Feature:Kscreen
19:37:45 <ennael> checked with neoclust, this one is done now and available in alpha3
19:38:07 <ennael> https://wiki.mageia.org/en/Feature:Cinnamon
19:38:51 <ennael> https://wiki.mageia.org/en/Feature:Mate
19:39:25 <coling> I guess both are packages so that's OK (modulo installer selection)
19:39:28 <ennael> both of them should be part of coming discussion about destop integration
19:39:30 <coling> *packaged
19:39:31 <ennael> yep
19:39:45 <ennael> https://wiki.mageia.org/en/Feature:New_Mgarepo
19:39:49 <ennael> boklm: around ?
19:40:06 <leuhmanu> :g mate-panel
19:40:07 <Sophie> leuhmanu: Graphical desktop/MATE // core-release (Mga, cauldron, i586)
19:40:11 <leuhmanu> good
19:40:13 <coling> I don't think there is much progress here, but I'm not really sure it's tied specifically to release anyway.
19:40:19 <ennael> yep
19:40:35 <ennael> https://wiki.mageia.org/en/Feature:Rpmlint_enhancement
19:40:39 <ennael> same thing here
19:40:54 <coling> I will hopefully find some time to help with this one too with a view perhaps to deploying it shortly after mga4 release.
19:41:05 <ennael> great :) thanks
19:41:15 <ennael> https://wiki.mageia.org/en/Feature:MTP_(Media_Transport_Protocol)_Support
19:41:35 <coling> e.g. if we manage to do it on time, we can run at before unfreeze (will mean a rather prompt update of mgarepo but such is the way of things)
19:42:07 <coling> (tho' perhaps we should wait some weeks for security patches... anyway one for discussion/decision nearer the time)
19:43:59 <AL13N> for MTP, i didn't see much progress since last meeting we spoke about this
19:44:29 <AL13N> i had no contact with the responsible person
19:45:31 <ennael> ok will try to ping him
19:45:33 <ennael> https://wiki.mageia.org/en/Feature:MageiaWelcome
19:45:42 <ennael> this one has been added in alpha3 release
19:45:48 <ennael> we need feedbacks on it
19:46:10 <sebsebseb> nice :) default in alpha 3
19:47:04 <AL13N> so, far i've seen people react nicely to it
19:47:12 <ennael> https://wiki.mageia.org/en/Feature:Pillow
19:47:15 <ennael> philippem: ?
19:47:41 <philippem> yes, all is ok
19:47:52 <ennael> great :)
19:48:08 <ennael> https://wiki.mageia.org/en/Feature:Tool_for_writing_iso_on_USB_stick
19:48:29 <coling> Seems we now have four of those... :p
19:48:38 <ennael> yep I saw this today :)
19:49:15 <AL13N> heh
19:49:41 <AL13N> perhaps we could put a static build of this as a link on our downloads page or something
19:49:52 <AL13N> so that people can easily find it if they are on other distro's
19:49:54 <ennael> still we need proper packages and maybe choose one of them
19:51:04 <ennael> at least for communication for the final release of Mageia 4
19:51:21 <ennael> MrsB: would it be possible for QA to review these tools ?
19:51:37 <MrsB> yes, if somebody can package them
19:51:55 <MrsB> DavidWHodgins: is your one ready now?
19:52:16 <DavidWHodgins> I'll be concentrating on getting the rpm ready to submit. Should have it ready by tomorrow.
19:52:32 <ennael> ok nice
19:54:00 <ennael> https://wiki.mageia.org/en/Feature:PythonByteCompiling
19:54:04 <ennael> philippem: ?
19:54:55 <philippem> some package need files list change
19:55:34 <philippem> easy task, but time consumming so I asked help from padawans
19:55:34 <pterjan> ( http://pkgsubmit.mageia.org/autobuild/results.php?run=2013-09-29 )
19:55:35 <[mbot> [ 90.6% Success ]
19:55:58 <ennael> ok so still in progress then :)
19:56:28 <ennael> https://wiki.mageia.org/en/Feature:RpmDrakeShowsAdvisories
19:56:39 <philippem> yes, but better than I thought, thanks for all people helping
19:56:43 <ennael> pterjan: any news ?
19:56:48 <pterjan> no
19:57:44 <ennael> https://wiki.mageia.org/en/Feature:SystemdInRamdisk
19:57:54 <ennael> again your turn :)
19:57:55 <coling> So, I did some initial tests here
19:58:09 <coling> It went well, but more recent ones seemed to cause problems.
19:58:24 <coling> Will try again for next time and try and sync properly.
19:58:43 <coling> It's pretty trivial to turn on and revert back.
19:59:04 <coling> The dracut-based-stage1 discussions might impact this a little (I'd aim for systemd init)
19:59:42 <coling> But either way there is still upstream fixes in this area and Arch guys are actively working on this too, so it's a good time to be doing it I think.
20:00:09 <coling> it's somewhat delayed by me not using initrds on my main machine ;)
20:00:18 <AL13N> coling: at what kind of size increases are we looking at in stage1 if it becomes a dracut/udev/systemd stage1 ?
20:01:06 <coling> AL13N, not sure yet, but dracut is ~18Mb for a non-hostonly initrd with network support IIRC. I wouldn't expect it to be much bigger than that when our loader is included.
20:01:23 <coling> (but this is very hand-wavey guesses)
20:01:53 <coling> It will certainly be bigger for boot.iso, but stage2 and rescue should get slightly smaller in mitigation.
20:04:43 <ennael> ok
20:05:01 <ennael> https://wiki.mageia.org/en/Feature:SystemdTidyups
20:05:13 <coling> Progressing nicely enough.
20:05:55 <coling> It's a bit of a mixed bag, but polkit auth especially is getting there. Still need to add some symlinks for upgrade users for the moved paths, and there are a few drak* pkgs still to purge of consolehelper.
20:06:08 <coling> But over all it's in good shape and the policy files are all now translatable.
20:07:06 <ennael> and finally https://wiki.mageia.org/en/Feature:NetworkDeviceNameChange
20:07:27 <coling> So again, it's going OK, even if some people don't particularly like it :)
20:08:00 <AL13N> but the stage1 is a problem
20:08:04 <coling> Main "problem" is with the installer using different names and using that to write the device name to some files.
20:08:17 <AL13N> won't it take too long to get stage1 ok?
20:08:29 <coling> This is technically buggy in itself anyway for machines with multiple cards.
20:08:35 <coling> (under the old scheme)
20:09:04 <coling> because the generators are not run and the udev rules are not written.
20:09:19 <coling> So while it works for most people, some machines might still be a bit borked.
20:09:28 <coling> (those with dual net cards)
20:09:44 <coling> So the new scheme still solves bugs here too.
20:09:59 <coling> I think there is still plenty time to get stage1 working IMO
20:10:33 <coling> If we absolutely must revert if stage1 proves unreliable, then it can be done quite late IMO.
20:10:43 <coling> (without risking too much in terms of wasted QA)
20:11:22 <AL13N> coling: do you think you could do this for alpha4? so it can be tested?
20:11:38 <ennael> AL13N: there is no alpha4...
20:11:47 <AL13N> oops
20:12:15 <AL13N> before version freeze then, so it can be tested on cauldron?
20:12:30 <coling> I'll try for the beta certainly, might have to get some help from tmb but we can maybe convince neoclust to help a bit too as he's been interested in dracut recently :)
20:16:21 <AL13N> ennael: any other feature left?
20:16:51 <AL13N> (we should hurry, or malo will be back)
20:17:08 <ennael> no more feature
20:19:11 <ennael> about last topic... I'm still trying to find what malo wanted to speak about :)
20:21:32 <AL13N> i suppose it's about seeing how much rpm nobody has compared to last time? and maybe something about methods to get the teams to split up the packages between the packager as maintainer or something?
20:21:42 <AL13N> :nb_rpm nobody
20:21:43 <Sophie> AL13N: nobody (Mageia) maintains 2887 rpms
20:21:52 <AL13N> :nb_rpm jq
20:21:52 <Sophie> AL13N: jquelin (Mageia) maintains 2784 rpms
20:22:24 <AL13N> sort of related, pterjan posted his link on 90% rebuild
20:22:36 <AL13N> what are we going to do with stuff that we can't rebuild? should be drop that?
20:26:38 <marja> how does rebuilding work? I was surprised to see a lower release number of a package than was in last commit (from weeks ago)
20:27:47 <Oro_Valley> I noticed quite some packages not rebuilding anymore w/o making changes on the package. I got 2 new ones this morning?
20:27:48 <AL13N> yes, 700 newly broken
20:27:56 <AL13N> it might be new gcc or glibc that landed
20:28:14 <AL13N> http://pkgsubmit.mageia.org/autobuild/
20:28:27 <AL13N> you'll see that 200 were fixed, but 700 newly broken
20:28:40 <philippem> and python bytecompile too
20:28:52 <AL13N> philippem: would that actually break packages?
20:29:07 <AL13N> philippem: failed to build?
20:29:19 <pterjan> most breakage are java, some are python bytecompile
20:29:26 <pterjan> so work in progress
20:29:43 <AL13N> no gcc/glibc breakage?
20:29:47 <philippem> file list missing pyo and pyc mainly
20:29:48 <pterjan> see the yellow stars on the page
20:29:59 <pterjan> very few gcc/glibc
20:30:02 <AL13N> ic
20:30:15 <AL13N> i still have one that failed with newer autotools
20:30:19 <AL13N> beid-middleware
20:30:25 <AL13N> not easy to fix it
20:30:25 <philippem> some have bytecompile issue, but very few
20:31:13 <AL13N> i hope no big changes land, btw, i have a newer Xen ready, but i must wait until 10/10 to submit
20:32:41 <AL13N> in any case, what shall we do when we do mass rebuild after version freeze with the ones that don't build anymore? should be drop those on release freeze?
20:33:28 <marja> AL13N: at least check whether there are commits that weren't pushed yet
20:38:57 <sebsebseb> dead packagers eeting
20:39:09 <Oro_Valley> I just noticed a package (chiliproject) that is missing a dep (ruby). The surely will be more of these.
20:39:23 <ennael> ok it seems malo will not join tonight
20:39:40 <ennael> unless there are some questions, we can close the meeting
20:39:48 <coling> I guess ultimately we just make the call about the non-building ones on a case by case basis... some old, underused ones can likely be dropped, if they can be easily fixed, they should be, and if there are critical packages, well, I guess we'll have to decide.
20:39:49 <Oro_Valley> yep
20:41:12 <ennael> ok thanks for attending and see you in 2 weeks
20:41:15 <ennael> #endmeeting