15:22:59 <itchka> #startmeeting 15:22:59 <chwido> Woof! Let's start the meeting. It's Wed Oct 14 15:22:59 2015 UTC. The chair is itchka. Information about me at https://wiki.openmandriva.org/en/Chwido. 15:22:59 <chwido> Useful Commands: #action #agreed #help #info #idea #link #topic. 15:22:59 <chwido> You should add extra chair(s) just in case, with #chair nick. 15:22:59 <chwido> Have a good meeting, don't bark too much! 15:23:29 <itchka> #chair klebedeff bero 15:23:29 <chwido> Current chairs: bero itchka klebedeff 15:24:08 <itchka> agenda here https://project.openmandriva.org/meetings/35/agenda 15:24:15 <itchka> #item 1 15:24:38 <itchka> #topic Project state report 15:24:52 <itchka> bero: This one I think is yours :) 15:28:08 <bero> So essentially just non-news ;) 15:28:31 <bero> Cooker is looking pretty good, works reliably on all of my boxes 15:28:56 <bero> theres still a few glitches we have to fix, like lock screen not working or a password prompt after login 15:29:03 <bero> but overall things are coming together 15:29:15 <bero> essentially that's all, unless people have questions 15:29:35 <itchka> Do we know whether the installer has been fixed people are only running live as far as I know 15:30:08 <bero> I don't know... I'm on an installed system, but it was installed ages ago and is just urpmi --auto-updated a couple of times a day 15:30:19 <bero> I know it worked a couple of weeks back 15:30:25 <bero> but may still need some tweaking 15:30:43 <itchka> Ok one more. Have you released your fixed up kdepim libs 15:31:00 <bero> no, not finished yet 15:31:28 <itchka> It blocks quite a lot when doing an update. 15:31:42 <ben79> calamares fails every time with 'Can not set user 'foo' full name'. 15:31:57 <ben79> https://issues.openmandriva.org/show_bug.cgi?id=1370 15:32:13 <itchka> ben79: Is that withh the very latest iso? 15:32:49 <ben79> Yes. 15:32:53 <rugyada> itchka: yes, it's 13132 iso 15:33:01 <itchka> Ok thanks 15:33:36 <crisb_mob> Is drakxtools staying in om3 or not? 15:33:44 <ben79> The last build I know of that will install is # 12442. 15:34:09 <itchka> I manage to install by mounting the install directory on and then doing a chroot. 15:36:14 <itchka> You need to delete all the password lockfiles in /etc and then you can assigned passwords and create user home dirs. Then you can run grub2-install and install the boot loader and get a running system. If you need help doing this talk to me after the meeting. 15:36:43 <itchka> cris_mob: That's the next agenda item. 15:37:11 <itchka> Any more questions for bero ? 15:37:34 <itchka> on this agenda item 15:39:07 <itchka> I'll give it till 40 minutes past the hour and then we will move on 15:41:02 <itchka> #item 3 15:41:29 <itchka> #topic Drop/replace MCC 15:41:59 <itchka> I'm bringing this item forward so that crisb_mob can be here to discuss 15:42:18 <itchka> bero: Will you kick this off please 15:43:01 <bero> So essentially we've decided that MCC used to be great when it was written, but it's showing its age and lack of maintenance (and pretty bad code). 15:43:59 <bero> So the idea is to start removing components that are already obsoleted by other tools 15:44:11 <bero> and start phasing out/replacing other bits 15:44:28 <crisb_mob> IMHO it's too late to do this for om 3 15:44:40 <bero> A particularly bad problem is that some tools that are part of MCC can actually create corrupted config files these days because they were meant for older versions of tools 15:44:45 <bero> we can't have that in a release 15:45:10 <crisb_mob> Better we have a full replacement 15:45:18 <bero> yes, no need to rush it 15:45:27 <bero> this is not something to be done immediately 15:45:58 <crisb_mob> On the other hand things are in a mess because we had pok updating stuff to his repo which is now closed 15:46:30 <crisb_mob> So we have some merged stuff with Mageia and some not (drakconf, rpmdrake) 15:49:10 <itchka> We must certainly sort out utilities that are creating corrupted config files. Can you be more specific as to which utils are doing this? 15:52:19 <itchka> bero: What needs fixing for this release 15:53:24 <bero> essentially, what I'm aware of: 15:53:29 <bero> - packages still failing the mass build 15:53:31 <bero> - lock screen 15:53:47 <bero> - password prompt after login 15:53:54 <bero> - install issues 15:54:36 <itchka> I was thinking more of drakx tools 15:56:28 <bero> I'm the wrong person to ask then... Haven't used them since the 1990s 15:57:37 <itchka> bero: Have there been any thoughts on what language the new tools would be written in? 15:57:46 <crisb_mob> Probably python 15:58:37 <bero> Up to whoever ends up writing them -- I don't think they'll all be the same language (but they should be the same framework so we can pull them together as needed). 15:58:44 <bero> Probably we're talking about some C++ and some python 15:59:14 <crisb_mob> I'm not sure how much of them is unique anyway 15:59:39 <bero> yes, for many things we'll probably just end up deciding to use a tool that exists already 16:00:00 <itchka> crisb_mob: I suppose that depends on how unique we want to make them :) 16:01:35 <itchka> bero: crisb_mob: There's a bit of an identity issue here though the tools have been to some extent part of the Mandrake/Mandriva identity 16:04:11 <itchka> I we replace tools with new ones we should consider putting our "stamp" on them perhaps. 16:04:47 <itchka> When I say new ones I mean existing tools that we would utilise. 16:08:53 <bero> yes, probably 16:10:02 <itchka> It looks to me as if this issue needs to be discussed at the next meeting so at least we will have some facts about which utilities need to have their configuration sorted out for the 2015 release. 16:12:28 <itchka> bero: crisb_mob: Shall I close the subject for now and if TPG makes it here he may know more and I'll re-open it. If not I'll do some checking up and we will discuss again at next weeks meeting. Please indicate if you are happy with this solution. 16:12:44 <bero> +1 16:12:53 <fedya> + 16:14:14 <itchka> crisb_mob: ping 16:14:43 <itchka> looks like crisb is in a tunnel 16:19:11 <crisb_mob> Yes I agree 16:19:23 <crisb_mob> Itchka I don't leave that early ;-) 16:19:41 <itchka> :) 16:19:58 <crisb_mob> I agree about identity there will be even less to make us unique if we just use generic tools 16:20:32 <crisb_mob> Its nice to have a consistent interface to the distro 16:21:14 <crisb_mob> Perhaps if everything was a kcm module you'd get that but then that is really anti any other desktop 16:22:08 <crisb_mob> And it has zero omv identity points 16:23:08 <crisb_mob> I will draw up something on the wiki so we know what we have and what options there are and what state our stuff is in compared to the competition 16:23:42 <itchka> I have had some ideas about this for time time. I believe that the community could design the interface most people could use QT designer. Once there is the graphical bit you clever guys can stitch it all together. Am I mad or could this work. 16:24:15 <fedya> btw my last conclusion 16:24:21 <itchka> cris_mob: That would be really helpful. Thanks :) 16:24:25 <fedya> oma 2014.0 not ready to use with raids 16:25:02 <itchka> fedya: We are just moving to that subject next :) 16:25:14 <itchka> #item 2 16:25:43 <itchka> #topic Server distribution 16:25:59 <itchka> Who wants to kick this off? 16:26:10 <fedya> i', 16:26:12 <fedya> i'm 16:26:23 <itchka> Off you go fedya :) 16:27:06 <fedya> wait a sec 16:28:55 <fedya> well, as we desided in Krakow, we want to move to own distro on soyoustart.eu servers 16:29:14 <fedya> i'm trying to do it 16:29:27 <fedya> well my conclusion is oma2014.0 100% not ready to use on server 16:29:46 <fedya> RAID1 not unworkable 16:30:07 <fedya> of course raid1 is working but other system is far far away from stable 16:32:24 <itchka> You mean our own abf yes? 16:32:28 <fedya> yes 16:33:07 <itchka> So we need to do some work on a server distribution. 16:33:27 <fedya> we need to do a lot of work 16:34:22 <itchka> It is not just raid1 that is no good? What else needs doing? 16:36:21 <fedya> itchka: network service, resolvconf service, dbus 16:36:45 <itchka> fedya: ouch!! 16:37:29 <fedya> https://bugzilla.redhat.com/show_bug.cgi?id=1010572 16:37:31 <fedya> this 16:37:56 <fedya> http://lists.opensuse.org/opensuse/2015-01/msg00774.html 16:38:24 <fedya> looks like cooker more stable than release 16:38:25 <fedya> haha 16:39:17 <crisb_mob> So what are the problems? 16:39:49 <fedya> ah 16:39:57 <fedya> look mail 16:40:01 <fedya> om-council 16:40:06 <fedya> i sent a letter 16:40:14 <fedya> it was first issues 16:40:32 <crisb_mob> I'm not in the cpunxil 16:40:36 <crisb_mob> Council 16:40:46 <itchka> fedya: I'll forward to crisb_mob 16:40:49 <fedya> Well after all i run oma14 on mdraid 16:41:17 <fedya> crisb_mob yesterday i sent for you too 16:41:18 <fedya> Crispin Boylan <cris@beebgames.com>, 16:41:50 <fedya> i run it, but distro14 not ready to be a server 16:42:00 <crisb_mob> I'm using mdraid on 2014 with raid 1 on two disks and raid 0 on 2 ssd 16:42:20 <crisb_mob> Ok not got the mail then 16:42:29 <itchka> crisb_mob: You have mail 16:42:47 <fedya> Well, yesterday i wrote "* I want to make an offer to Tomasz or Crispin, as good engineers, to reproduce this issue on server or in virtual machine to confirm (or not) my findings." 16:42:55 <fedya> crisb_mob ^^ 16:43:17 <itchka> _TPG: Hi 16:43:34 <fedya> crisb_mob for unknow reason md0 always detects as md127 16:43:45 <itchka> we are talking about oma 2014 server install that fedya has been working on. 16:43:45 <_TPG> hi 16:44:23 <crisb_mob> Yeah I confirm but I don't think it's an omv thing 16:45:01 <crisb_mob> I'm running lvm on top of mdraid 16:45:22 <fedya> raid1-127, raid0-126 16:45:58 <fedya> if crisb_mob or _TPG want to make server chroot, create raid1 on remote server without kvm, enjoy 16:46:00 <fedya> ;) 16:46:14 <fedya> i tried it 3-4 times 16:46:18 <fedya> without success 16:47:00 <_TPG> fedya: maybe better would be to copy stuff from iso into the chroot ? 16:47:05 <fedya> it would be easier to do with kvm on server, but price for kvm is horrible 16:47:10 <_TPG> rather thatn running urpmi --root 16:47:44 <fedya> _TPG how it help to run raid? 16:48:24 <_TPG> do you have any logs ? 16:48:35 <fedya> now raid working 16:48:38 <fedya> i did it 16:48:45 <_TPG> i do not have any raid here so i can't reproduce 16:50:40 <itchka> fedya: Lets get this clear :) you have succeeded in getting it working but it needs much manual work to do so? 16:50:50 <fedya> _TPG https://gist.github.com/fedya/a5c46dca871d13ccb6ef 16:51:02 <fedya> _TPG something wrong with dbus 16:51:50 <fedya> _TPG https://gist.github.com/fedya/711a4385ddd91d63d554 /lib64/security/pam_gnome_keyring.so in minimal installation 16:52:06 <fedya> bootloader-utils requires gtk+ and other graphical libs 16:52:12 <_TPG> fedya: systemctl enable dbus.socket 16:52:50 <fedya> [root@localhost ~]# systemctl enable dbus.socket 16:52:51 <fedya> The unit files have no [Install] section. They are not meant to be enabled 16:53:03 <_TPG> fedya: what status says ? 16:54:56 <fedya> _TPG https://gist.github.com/fedya/26d24b3620966d668808 16:58:01 <_TPG> fedya: this dbus is not fatal issue 16:58:14 <_TPG> it's related to userspace dbus 16:58:21 <_TPG> are you able to log in as root ? 16:59:17 <_TPG> according to status only problem is with resolvconf 16:59:25 <fedya> yes 16:59:26 <_TPG> Oct 14 16:53:50 localhost systemd-tmpfiles[439]: readlink(/run/resolvconf/resolv.conf) failed: Invalid argument 16:59:31 <_TPG> you have to recreate it 17:03:58 <itchka> We need a text mode installer for servers 17:04:12 <fedya> indeed 17:05:12 <itchka> Would the old drakx installer do for the time being? 17:05:18 <_TPG> no 17:05:30 <_TPG> drakx is not CLI 17:06:13 <itchka> wasn't there a test installer option where it reverted to a curses interface? 17:06:23 <_TPG> would be nice if someone started to work on this https://abf.io/openmandriva/omdv-cli-installer/blob/master/TODO#lc-24 17:06:32 <itchka> Perhaps I'm going too far back :) 17:06:39 <fedya> _TPG i suggest you to try install oma on the server 17:08:13 <_TPG> fedya: if i'll have such possibility i'll do that 17:08:51 <itchka> _TPG: I might take you up on that as it's bash :) 17:09:51 <_TPG> i did some research, and found couple of bash installers, and imho that one fits our needs 17:10:19 <fedya> _TPG i sent auth credentials to you 17:10:24 <_TPG> first it has to be ported to perform at least basic setup and installation 17:10:29 <_TPG> fedya: ok 17:13:58 <itchka> _TPG: I'll have a go 17:15:36 <fedya> ithcka i think we need to use standart installer, we just need an option 'Server install' 17:15:50 <fedya> http://howtoit.ru/linux/centos/item/36-ustanovka-centos-6-4-minimal.html look screenshots 17:18:12 <itchka> fedya: This will have to be done over ssh so we would need to make an X connection over ssh. We would have to make sure things were set up to allow that 17:26:46 <_TPG> fedya: see my emai 17:29:21 <itchka> To summarise Item 2 Server distribution 17:29:30 <itchka> We need to:_ 17:30:14 <_TPG> write/adapt to our needs CLI installer 17:30:59 <crisb_mob> Back what did I miss 17:31:50 <itchka> 1. Create a text based installer or alternatively create a spin which will include enough tools and libraries to allow the use of draklive-install suitably modified to provide the additional utilities required for server installation. 17:32:52 <_TPG> sed -i -e "s/draklive-install//g" 17:34:00 <_TPG> itchka: plz... you would have to put same effort in creating new CLI installer as adapting drakcrap to support server stuff 17:34:05 <itchka> crisb_mob: see pvt tab 17:34:56 <itchka> _TPG: :) Just reflecting the dialogue of the meeting..... 17:35:13 <_TPG> sure 17:35:46 <itchka> X-forwarding is very messy 17:36:33 <_TPG> and simply copying contents of squashfs to chroot in case of server does not work 17:36:57 <crisb_mob> In what way 17:38:59 <itchka> _TPG: Are you saying it's necessary to install from repos? 17:38:59 <crisb_mob> Itchka I've got no pvt tab? 17:39:07 <_TPG> itchka: no 17:40:10 <_TPG> itchka: draklive-install only copies squashfs, what we really need is a CLI installer that copies squashfs too, but is enough flexible to adapt it ti run additional commands, to setup working environment 17:41:14 <itchka> crisb_mob: Sorry I forgot you were on mobile. 17:42:05 <itchka> _TPG: Ok so we need to do post install stuff in a kind of pe environment ('scuse the Windoze term) 17:42:20 <_TPG> yes 17:42:43 <itchka> Ok I'm ok with that 17:43:31 <crisb_mob> Calamares doesn't/couldn't support text mode then? 17:47:13 <itchka> _TPG: So in summary we run in live mode and create the raid array then we copy the unpacked squashfs to the raid / we then chroot and do further configuration and finally we install the bootloader and reboot into the raid. We also need to be able to detect an existing OMA raid install and mount and chroot to it for rescue purposes. 17:48:35 <_TPG> crisb_mob: it could do that, but this needs some code refactoring to support CLI 17:48:37 <_TPG> itchka: yes 17:49:38 <crisb_mob> Sounds like it would be handy to avoid doing the same thing in two places if we adapted Calamares? 17:51:22 <_TPG> seems so, also calamares have some qt5 deps, conpared to cli installer 17:53:11 <crisb_mob> Brb running home now 17:53:49 <itchka> _TPG: won't calamares work without X. Somewhere in the vague past I remember somebody mentioning this. I accept there would still be some payload 17:54:32 <_TPG> it wasn't me who mentioned that 17:55:15 <itchka> No it wasn't I think it was in the calamares irc. 17:55:44 <itchka> I have the logs I'll look 17:57:16 <itchka> _TPG: Is the server fedya set up running properly now? 17:59:05 <_TPG> itchka: dunno 18:00:04 <_TPG> i can't connect with ssh even if my key i set up 18:03:41 <fedya> _TPG let me fix it 18:04:09 <fedya> _TPG try right now 18:06:11 <_TPG> fedya: nope 18:09:31 <itchka> Perhaps we should moveto the last item. We can always pick up the server install again if required. 18:09:55 <itchka> and you guys can work on it in the background 18:10:50 <itchka> We will give crisb a little while to get home though. 18:13:23 <_TPG> fedya: looks like it is cunning centos 18:16:10 <fedya> _TPG yes 18:16:33 <fedya> check /etc/sysconfig/network ifcfg-eth0 /etc/mdadm.conf /etc/default/grub 18:16:51 <crisb1> Home now 18:17:21 <itchka> crisb1: Wow that was quick !!! 18:17:53 <crisb1> Only two miles hehe 18:18:17 <itchka> both of them in four minutes what a runner!! :) 18:19:35 <itchka> I have to cook for the hoardes soon so let's deal with the final iem. 18:19:43 <itchka> #item 4 18:20:06 <itchka> #topic link KDE tool detecting/recording error/exceptions to QA mailing list or Errbit like web site 18:20:40 <itchka> Would anyone like to voluteer to tell me what this is about as I have absolutely no idea 18:21:47 <itchka> Is there a kde crash tool that can send emails? 18:22:03 <_TPG> itchka: yes there is a standard kde crash report tool 18:22:10 <_TPG> drkonqi 18:22:28 <itchka> Gosh I thought that died with kde3 18:22:42 <ashledombos_> me too :p 18:22:50 <_TPG> it turned into kcrash nowadays 18:23:02 <itchka> ah that would explain!! 18:23:36 <itchka> Something like this might seriously spam the list. 18:23:59 <itchka> wouldn't it be better to have a dedicated list for it. 18:24:00 <_TPG> idea was to modify it that way so each crash and each time user decides to send report that report will also hit QA mailinglist 18:25:49 <crisb1> Could be useful if someone triages them 18:26:47 <itchka> I guess if we could do it on a subscribed basis we could get users to participate in QA test programs. It certainly has the potential to catch the bugs that people tend to ignore because they are minor and they can't be bothered to file a bug report. 18:27:29 <_TPG> bbl 18:27:32 <crisb1> Well only crash bugs 18:27:38 <_TPG> gotta go 18:27:51 <itchka> _TPG: Ok see you later perhaps 18:28:05 <crisb1> Bye tpg 18:30:40 <itchka> I think it could provide useful information even if we can't triage everything it would give some clue as to popular applications that people like to use but which aren't working properly. QA Cannot possible test everything but this would help us target our effort. 18:33:20 <crisb1> Excuses excuses ;-) 18:33:47 <itchka> I know, I know :) 18:34:42 <itchka> I'll start a discussion on the QA ml and see what the QA'ers think as they seem to be backward in coming forward here :). 18:35:06 <fedya> _TPG any chanced that you try it today? 18:35:53 <itchka> fedya: _TPG is gone maybe 18:36:11 <itchka> crisb1: Could you take a look? 18:37:33 <fedya> well 18:37:37 <itchka> #action itchka to start discussion on QA ml regarding KDE crash tool. 18:37:42 <fedya> i'll go to play in world of tanks 18:37:43 <fedya> keke 18:38:01 <itchka> fedya: Have fun :) 18:40:06 <itchka> Ok folks I think the meeting is pretty much done so I'll end it at 45 minutes past the hour. Please raise any subjects you wish to discuss under any other business now. 18:40:17 <itchka> #topic AOB 18:40:49 <crisb> why are abf builds being cancelled? 18:41:38 <itchka> I don't know avokhmin any ideas? 18:41:55 <avokhmin> cancelled - timeout maybe? 18:41:58 <itchka> Have there been many cris? 18:44:01 <ashledombos_> is there a doc somewhere about where to download latest cooker isos, we could provide to public? 18:44:21 <crisb> today yes most of my chromium ones 18:44:28 <crisb> but they had only been going a couple of hours 18:47:40 <itchka> ashledombos_: I don't think we want to do that yet as they are uninstallable. 18:50:20 <ashledombos_> itchka: ah ok bit we alseady have some positive review in our forums though :) 18:50:20 <ashledombos_> already 18:51:34 <itchka> These will be people updating from 2014 I would think or people doing what I did and using a chroot to finish the install. 18:53:46 <itchka> crisb: Have you done some iso builds that work? 18:54:38 <crisb> for cooker? 18:54:42 <itchka> yes 18:54:45 <crisb> no 18:54:54 <crisb> i havent even tried booting a cooker iso 18:56:37 <itchka> Don't bother unless you want to mess with chroots; somethings broken in pam I think password.lock files get left behind so calamares can't set the passwords because they are locked. 18:57:10 <itchka> The lock files are not getting deleted 18:58:47 <itchka> Luca has filed some strace info but I've no idea how to fix the proble. 18:59:39 <itchka> I have to go and cook so I'm closing the meeting. I'll be back later. 18:59:46 <itchka> #endmeeting