16:09:45 <_TPG> #startmeeting
16:09:45 <chwido> Meeting started Tue Oct 28 16:09:45 2014 UTC.  The chair is _TPG. Information about MeetBot at https://wiki.openmandriva.org/om/en/MeetBot.
16:09:45 <chwido> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:09:48 <crisb> _TPG: kernel-firmware is the source but they get put into a sub-package called iwlwifi-agn-ucode
16:10:24 <_TPG> Ok
16:10:52 <_TPG> I've sent agenda to ml and project
16:11:02 <_TPG> First topic
16:11:19 <_TPG> #topic mass rebuill
16:11:31 <_TPG> #topic mass rebuild
16:11:44 <_TPG> Are we ready?
16:11:49 <crisb> i dont think so
16:12:22 <_TPG> Y?
16:12:41 <crisb> still too many broken packages, the july rebuild is still a good indicator of failed
16:12:44 <crisb> we should fix them first
16:15:17 <crisb> mirrors also need cleaning, there are a lot of duplicate packages as abf doesnt always remove the old version.
16:15:21 <_TPG> Notice that failed build list on Google doc is outdated
16:15:50 <_TPG> Not everybody were marking packages as fixed
16:16:56 <_TPG> Maybe mirrors are not syncing properly
16:17:35 <crisb> this is on the primary abf.rosalnux.ru
16:17:37 <bero> Yes, IMO the list is too outdated to still be useful
16:17:49 <bero> But it would probably be good to get clang 3.6 in before the mass build
16:18:08 <_TPG> Crisb any example?
16:18:11 <bero> Upstream maintainers advised to use current 3.6 branch rather than the 3.5 release because of threading performance when I talked to them at LPC
16:18:42 <crisb> _TPG: sure i have a list
16:19:03 <bero> IMO we should clean out the duplicate packages after the mass build, because the mass build will generate more of those
16:19:33 <_TPG> 3.6 got release
16:19:33 <_TPG> Crisb send me that list so ill clean repo
16:20:46 <crisb> repoclosure is in good state i've been going through and cleaning it
16:21:15 <bero> Is the qca Qt4 vs. Qt5 bit fixed?
16:21:44 <crisb> most on cooker/main repoclosure list are now either because of old dupes hanging around or the buildlists not being in sync with whats in the repo
16:22:49 <crisb> bero:i dont think so but tpgs suggestion was a good one no?
16:23:05 <bero> yes
16:23:22 <bero> It's just another thing we should fix before the mass build to make sure all applications are linked w/ the right one
16:23:35 <crisb> yeh
16:23:44 <crisb> lets delay by a week or 2 then?
16:24:22 <bero> Let's say 1, we need to get a clearer view of where we are
16:25:13 <crisb> ok
16:33:27 <crisb> shall we action that then?
16:34:49 <bero> IMO yes -- let's wait for _TPG to returb
16:35:48 <_TPG> ok i'm back
16:36:06 <_TPG> crisb: please do not look at FBA tests on cooker
16:36:20 <_TPG> especially on superseeded as these tests are broken
16:36:33 <crisb> _TPG: all repoclosure ones seemed fine
16:36:40 <crisb> ie they were valid breakages
16:36:42 <_TPG> i'm trying to clear the situation with Andrey Ponomarenko
16:37:43 <_TPG> ok what about qca qca2 situation ?
16:37:50 <_TPG> this is also a blocker for mass rebuild
16:37:58 <bero> Yes, let's fix that
16:38:09 <bero> and I'd prefer building the Qt4 and Qt5 versions from the same package
16:38:21 <bero> so we don't have to remember to update twice when there's a security bug or something
16:38:52 <bero> I'm in favor of creating a new package w/ both Qt4 and Qt5 versions based on Pulfer's qca2-qt5 package, that's cleaner than the older ones
16:41:15 <_TPG> ok so let's make some decision now
16:41:32 <itchka_> here
16:42:17 <itchka_> A mass rebuild would allow me to regenerate the list
16:43:16 <crisb> lets schedule one in a week then
16:43:36 <_TPG> crisb: +1
16:43:43 <bero> +1
16:43:56 <itchka_> +! though my contribution as been minimal
16:44:05 <_TPG> whos going to take qca issue ?
16:44:06 <_TPG> https://project.openmandriva.org/work_packages/140
16:44:25 <bero> I can take care of it
16:45:50 <_TPG> bero: ok thanks
16:46:24 <_TPG> #agree delay mass-rebuild for qca/qca2 fix and update llvm/clang to 3.6 - max one week more
16:48:09 <_TPG> #agreed delay mass-rebuild for qca/qca2 fix and update llvm/clang to 3.6 - max one week more
16:48:26 <_TPG> next topic
16:48:35 <_TPG> #topic New installer status - https://project.openmandriva.org/projects/installer/roadmap
16:49:05 <bero> Didn't make as much progress as I hoped to make, too much to do at the day job
16:49:16 <bero> Still optimistic that we can have it running by alpha time though
16:49:29 <bero> Things at work are calming down now that all the conferences are done
16:50:03 <_TPG> bero: do you have some sources ? maybe sharing them on abf would speed up things ?
16:51:10 <bero> Nothing useful yet, just a couple of fragments I've been testing, mostly just proper ways to unionfs mount various images, copy their contents without modifications from tmpfs and the likes
16:52:18 <itchka_> bero: We need to get together on EFI installs. I have it pretty much licked.
16:52:58 <bero> itchka_: Great. I figured out a lot of stuff there too, probably it'll work perfectly if we get both things together.
16:54:18 <itchka_> My personal repo has all the stuff you need to test
16:54:22 <_TPG> sounds good if it will be ready before alpha
16:56:27 <itchka_> bero if you are around latish evening we can discuss. I'm out till ten GMT.
16:56:44 <bero> itchka_: sure
16:56:50 <bero> I'll be around too
16:57:12 <itchka_> Ok
17:00:30 <_TPG> ok anything left for this topic ?
17:01:42 <itchka_> Not from me
17:02:32 <bero> Not from me either
17:02:36 <_TPG> #topic  2015.0 work packages status
17:03:08 <_TPG> ok we have some stuff from project which seems not updated
17:03:35 <_TPG> Make python 3.x main, start deprecating 2.x https://project.openmandriva.org/work_packages/55
17:03:55 <_TPG> i have feeling this is mostly done for main repository
17:04:21 <_TPG> there may be a lot of packages from other repos which will fail because python->python3
17:04:34 <bero> yes
17:05:03 <bero> also some special cases (e.g. something requires python 2.x and PyQt, but right now we only build PyQt for 3.x etc.)
17:06:32 <_TPG> what to do then ?
17:06:49 <_TPG> do we rebuild contrib also and try to fix all python issues there ?
17:06:59 <_TPG> or simply drop that packages ?
17:07:15 <bero> Preferrably: Fix it so it can use 3.x -- usually not THAT much work. Alternatively, we just have to build PyQt (and other libraries) for both 2.x and 3.x
17:07:25 <bero> We'll have to rebuild contrib too at some point
17:08:42 <Pulfer> For pyqt please note that it requires sip to be properly packaged
17:09:07 <itchka_> The sooner we mass build contrib the sooner people tyros like me can do the plodding work I guess.
17:09:12 <Pulfer> Both python2 and python3 versions of sip should co-exist
17:09:38 <_TPG> mass build for main and cooker together ?
17:09:48 <_TPG> s/cooker/contrib
17:09:49 <bero> IMO main first
17:09:50 <Pulfer> In Rosa I build sip like this: https://abf.rosalinux.ru/import/python-sip/tree/rosa2014.1
17:10:45 <Pulfer> python3 ships /usr/bin/sip3 and uses /usr/share/sip3 for data
17:13:03 <_TPG> aare you agree that all pyhon2->python3 issues needs to be fixed so this #55 may be closed as 100% done ?
17:13:32 <Pulfer> Then python3-qt4 builds without any patching. But python3-qt5 requires this patch: https://abf.rosalinux.ru/import/python3-qt5/blob/rosa2014.1/PyQt-gpl-5.3.2-sip3.patch
17:14:11 <Pulfer> There's a chance that Cooker should do the same thing but using sip2 for python 2 etc
17:14:16 <bero> Yes, agreed
17:14:25 <bero> sip should be for the default python
17:14:34 <bero> So we should have sip2 if anything
17:15:04 <_TPG> #agreed all pyhon2->python3 issues needs to be fixed so this #55 may be closed as 100% done ?
17:15:27 <Pulfer> Well, I agree
17:15:45 <Pulfer> Not that I package for Cooker much recently anyway...
17:15:52 <_TPG> next one is Integrate LXQt https://project.openmandriva.org/work_packages/58
17:16:00 <_TPG> Pulfer: :)
17:17:08 <bero> 58 is essentially done -- I'm running LXQt as my main desktop right now
17:17:39 <_TPG> bero: what about task-razorqt and task-lxde
17:17:44 <_TPG> and its deps ?
17:18:05 <_TPG> since lxde and razorqt have merged, would be good to clean our repo too ?
17:18:12 <bero> IMO yes
17:18:36 <bero> Not sure if anyone insists on keeping the gtk version of lxde for whatever reason, I certainly don't
17:19:00 <_TPG> +1 for droping lxde and razorqt
17:19:31 <bero> +1
17:20:27 <_TPG> i can do that if you like
17:21:13 <_TPG> bero: can you please update https://project.openmandriva.org/work_packages/58 status ?
17:21:19 <bero> sure
17:22:56 <itchka_> So does this mean we don't have a lightweight window manager that is not in contrib?
17:23:36 <crisb> itchka_: that would be razorqt
17:23:56 <_TPG> itchka_: we will have lxqt
17:24:24 <itchka_> Thanks just checking :)
17:24:49 <_TPG> so are we agree to drop lxde and razorqt from repo ?
17:24:49 <itchka_> so there will be nothing gtk-based
17:25:07 <_TPG> xfce is gtk based
17:25:16 <itchka_> That's in contrib
17:25:40 <_TPG> so ?
17:25:58 <crisb> sorry yes lxqt :)
17:26:18 <crisb> ok lets kill them
17:26:32 <crisb> right gtg now, bbl
17:27:15 <itchka_> Do we pull in packages from contrib when we build an iso. I suspect not this means if KDE breaks and you ain't got network there's nothing.
17:28:00 <_TPG> itchka_: no, and how this is related to current topic ?
17:29:56 <itchka_> Well presumably the two packages you are talking about are in main and get pulled in or do we not have a fallback strategy. I can raise this as a different item if you don't think it's relevent.
17:30:22 <itchka_> to this one
17:30:26 <_TPG> pulled in where ?
17:30:45 <itchka_> Onto the iso build
17:31:12 <_TPG> itchka_: look lxde and razorqt usptream project have merged by upstream, and currently these are obsoleted by lxqt
17:31:37 <_TPG> itchka_: it is about to clean repo from old unmaintained stuff which was replaced by something new
17:32:36 <itchka_> _TPG: I do understand that. I will bring the issue us as a different subject.
17:32:43 <itchka_> us=up
17:33:28 <_TPG> sure
17:33:38 <_TPG> but  other stuff are waiting
17:34:00 <itchka_> That's ok next week will do
17:34:54 <_TPG> next one https://project.openmandriva.org/work_packages/97
17:35:14 <_TPG> according to project this is done
17:35:32 <_TPG> but i haven't heard anythin from that guy for a long time
17:36:25 <bero> Me neither
17:36:25 <_TPG> according to this Cinnamon2 is ready, but seems not on the abf
17:36:55 <itchka_> _TPG: Do you want me to test it from his repo?
17:38:32 <itchka_> It appears to be complete for 2014
17:38:51 <_TPG> i've emailed him
17:39:03 <_TPG> itchka_: where is his repo ?
17:39:17 <itchka_> http://abf-downloads.rosalinux.ru/matthew_portner_personal/repository/cooker/x86_64/main/release/
17:39:44 <itchka_> Last build 10th May for cooker
17:40:09 <itchka_> Same for 2014
17:41:39 <_TPG> itchka_: seems like this is newer https://abf.io/mnportner/
17:43:13 <_TPG> seems like abandoned
17:43:53 <_TPG> next one
17:44:17 <_TPG> Drop Qt 3.x https://project.openmandriva.org/work_packages/95
17:44:43 <_TPG> there is a updated list https://docs.google.com/spreadsheets/d/1_IPP_Q3rNaGLeAQCNFi_hNNw7dWlcdw1E52W5dyeyR8/edit#gid=0
17:45:00 <itchka_> Last commits 6 months ago so that's amy as well, or near enough. I recall him having some horrible linking problem which he couldn't resolve. Come to think of it I had a go at it too and I couldn't fix it either. Not that that is anything significant:)
17:45:22 <itchka_> I way still have the build
17:45:47 <_TPG> according to that list 14 packages requires Qt3
17:45:53 <_TPG> what to do with them ?
17:46:48 <bero> I think we can drop the ones that remain
17:47:04 <bero> Chances are nobody has used them in years
17:47:20 <itchka_> Seems a waste to throw it aways. I guess since I'm the patron saint of lost causes I could have a look.
17:48:24 <_TPG> itchka_: i'll assing this to you then
17:48:43 <itchka_> Qt3 ? has ny alzheimers kicked in...:)
17:49:13 <itchka_> _TPG: I'll look and report back.
17:49:16 <_TPG> yes 14 packages left
17:49:36 <_TPG> if nothing will be done i'll drop all of them on next TC
17:49:49 <itchka_> Ok
17:50:11 <_TPG> #agreed itchka will take a look at these Qt3 packages - fix or drop before next TC
17:50:58 <itchka_> _TPG: Can you forward anything you get from M Portner
17:51:04 <_TPG> GNOME 3.14/3.16 https://project.openmandriva.org/work_packages/98
17:51:07 <_TPG> itchka_: sure
17:51:48 <_TPG> according to last entry on that #98 crisb updated all gnome packages to 3.14
17:52:10 <_TPG> does this mean we have a working gnome3 environment ?
17:52:26 <bero> No idea
17:52:39 <bero> broken grap ;)
17:54:10 <itchka_> Xu_R was working on Gnome too so there is a chance that it works
17:54:36 <_TPG> itchka_: yes but Xu_Rnever pushed his changes to cooker
17:54:45 <_TPG> and crisb now doubled the work
17:55:25 <itchka_> We shoould ask crisb then I guess.
17:59:38 <_TPG> few left for today
17:59:59 <_TPG> transition urpmi to use libzypp https://project.openmandriva.org/work_packages/25
18:00:08 <_TPG> what about this one ?
18:01:09 <itchka_> Save it for later we have enough on our plate without throwing something like that into the mix.
18:01:26 <bero> I don't think anyone has worked on it after I synced with upstream
18:03:08 <bero> which obviously should be done again since it's been 4 months ;)
18:06:18 <proyvind> regarding urpmi, I'll be pushing my branches for rpm, urpmi, perl-URPM sometime soon
18:06:36 <proyvind> same for drakx etc.
18:06:55 <_TPG> good :)
18:07:01 <proyvind> so these tools in general will receive  some major updates
18:07:21 <proyvind> urpmi in cooker is really, really fscked
18:07:38 <bero> brb, got to take Chwido outside
18:07:59 <_TPG> proyvind: speaking of rpm and stuff, have you seen that http://file-store.rosalinux.ru/api/v1/file_stores/3d2dd8b67a53df6aec28f31d437ea6f7c5baa680.log?show=true
18:09:08 <proyvind> _TPG: yes, I think I've pushed a fix for it, haven't I?
18:09:17 <proyvind> a while ago now
18:09:55 <proyvind> https://abf.io/openmandriva/glibc/commit/0d61b6a6e1f75070c70ef86d5a244d4373f8aba0
18:12:01 <_TPG> proyvind: publishing lxc container is running on Rosa2012.1 so probably Rosa did not apply your changes
18:12:21 <_TPG> proyvind: it's glibc commit :)
18:13:04 <proyvind> they're free to pick it up ;)
18:14:21 <_TPG> proyvind: which commit ?
18:15:59 <proyvind> 19:09:55 < proyvind> https://abf.io/openmandriva/glibc/commit/0d61b6a6e1f75070c70ef86d5a244d4373f8aba0
18:18:41 <_TPG> proyvind: how obove commit is related to db erros when generating sythesis files ?
18:27:22 <symbianflo> afternoon
18:27:50 <itchka_> Hi SF
18:33:12 <proyvind> _TPG: I'm not following.. is the glibc issue related to some issue with generating synthesis filse? (genhdlist2 in cooker is broken btw.)
18:37:44 <_TPG> proyvind: for me it is not related :)
18:39:39 <proyvind> I'm not really sure what you're asking about then..?
18:40:02 <_TPG> [19:07] <_TPG> proyvind: speaking of rpm and stuff, have you seen that http://file-store.rosalinux.ru/api/v1/file_stores/3d2dd8b67a53df6aec28f31d437ea6f7c5baa680.log?show=true
18:40:49 <_TPG> looks like when running ghenhdlist2 produces lot fo rpmdb errors
18:40:58 <proyvind> hm
18:41:12 <proyvind> I was only looking at the segfault during glibc build..
18:41:36 <_TPG> proyvind: please look at above link :)
18:41:56 <proyvind> oh
18:44:41 <proyvind> I don't have any immediate insight on this..
18:44:45 <proyvind> only happens on aarch64?
18:46:15 <proyvind> considering aarch64's rather recent appearamce, I wouldn't be surprised if it were som berkeley db bug pm aarch64..
18:47:47 <_TPG> proyvind: it appears also on i586 and x86_64 arches
18:47:56 <proyvind> hm
18:48:20 <_TPG> proyvind: notice that genhdlist2 is executed inside x86_64 rosa2012.1 lxc container :)
18:48:46 <_TPG> rpmdb: BDB1546 unable to join the environment
18:48:47 <_TPG> error: db_init:db3.c:1087: dbenv->open(11): Resource temporarily unavailable
18:48:49 <_TPG> error: cannot open Packages(0) index: Resource temporarily unavailable(11)
18:48:56 <_TPG> does not look good
18:49:49 <proyvind> well, I have no answers.. for all I know it could just as well be an issue related to it running within lxc..
18:50:15 <proyvind> unfortunately I have to admit that the database aspect of rpm is the one I've found the least interesting and have the least insight on.. :|
18:50:32 <_TPG> proyvind: this only occurs when regenerating whole repository
18:50:53 <proyvind> could be that it runs out of memory?
18:51:02 <_TPG> when running genhdlist2 with --clean option
18:51:03 <proyvind> that would most likely trigger similar errors
18:51:13 <_TPG> dunno
18:51:39 <_TPG> publishing workers are running on rosa's servers which i do not know what specifications they have
18:51:46 <_TPG> maybe they are low on RAM
18:52:38 <_TPG> sounds like a qood hint
18:53:20 <proyvind> without any access in order to personally investigate and debug, nor even any knowledge about system regarding memory etc., making wild guesses about possible causes that sounds plausible is as good as I get ;)
18:54:44 <_TPG> ok
18:54:52 <_TPG> anything left for a discussion ?
18:55:04 <_TPG> if no let's close TC meeting
18:56:35 <proyvind> yeah
18:56:39 <proyvind> take your top off!
18:58:15 <_TPG> [tpg@lazur ~]$ top
18:58:24 <_TPG> looks nasty
18:58:32 <proyvind> :>
19:03:25 <_TPG> proyvind: your glibc fails https://abf.io/build_lists/2324660 :>
19:06:59 <proyvind> let's try again then :)
19:21:26 <_TPG> proyvind: failed :)
19:21:30 <_TPG> #endmeeting