14:00:19 <nils> #startmeeting modularity_wg
14:00:19 <zodbot> Meeting started Tue Aug 22 14:00:19 2017 UTC.  The chair is nils. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:19 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:00:19 <zodbot> The meeting name has been set to 'modularity_wg'
14:00:19 <nils> #meetingtopic Meeting of the Modularity Working Group (once every two weeks)
14:00:19 <nils> #chair dgilmore langdon mikedep333 tflink
14:00:19 <zodbot> Current chairs: dgilmore langdon mikedep333 nils tflink
14:00:31 <nils> #topic Roll Call
14:00:37 <nils> .hello nphilipp
14:00:38 <zodbot> nils: nphilipp 'Nils Philippsen' <nphilipp@redhat.com>
14:01:04 <jkurik> .hello2
14:01:05 <zodbot> jkurik: jkurik 'Jan Kurik' <jkurik@redhat.com>
14:01:31 <langdon> .hello2
14:01:32 <zodbot> langdon: langdon 'Langdon White' <langdon@redhat.com>
14:01:58 <karsten> .hello2
14:01:59 <zodbot> karsten: karsten 'Karsten Hopp' <karsten@redhat.com>
14:02:44 <tflink> .hello tflink
14:02:45 <zodbot> tflink: tflink 'Tim Flink' <tflink@redhat.com>
14:02:58 <nils> #topic Agenda
14:03:11 <nils> And... the agenda is empty!
14:03:17 <jkurik> :)
14:03:20 <nils> Does anyone want to add something?
14:03:49 <jkurik> I have a question, but not really an agenda topic
14:03:58 <asamalik> .hello asamalik
14:03:59 <zodbot> asamalik: asamalik 'Adam Samalik' <asamalik@redhat.com>
14:04:45 <nils> Well, crickets, so...
14:04:48 <nils> #topic Open Floor
14:04:51 <langdon> I was kinda expecting some people to show up with questions.. Maybe just open floor?
14:04:57 <langdon> Ha
14:05:08 * asamalik is getting distracted :/
14:05:11 <nils> jkurik, so what's up?
14:05:12 <asamalik> need few minutes
14:05:18 <jkurik> I was asked to push forward work on infrastructure for Fedora Modularity.
14:05:28 <nils> asamalik, I guess we'll be done by then :P
14:05:48 <jkurik> Unfortunatelly I do not have track of what is currently done and what needs to be done
14:06:32 <jkurik> is there any tracking for infrastructure ?
14:07:13 <jkurik> Are the tasks in Taiga also for infrastructure ?
14:07:16 <langdon> Almost entirely being managed by the factory folks.. Threebean would be the better consult
14:07:39 <jkurik> langdon: ok, that is the answer :-) thanks
14:08:16 <langdon> So.. We want it faster too but aren't the keepers
14:08:30 <nils> and, sidenote, we don't use Taiga anymore
14:08:39 <langdon> The one thing we need to follow up on that they aren't is copr support for mofules
14:08:41 <nils> perhaps we should delete our board there? langdon?
14:08:55 <langdon> Asamalik do you know anything ^^
14:09:00 <ttomecek> .hello ttomecek
14:09:01 <zodbot> ttomecek: ttomecek 'Tomas Tomecek' <ttomecek@redhat.com>
14:09:10 <langdon> Nils yeah.. We should or at least put a noticr
14:09:13 * dgilmore is kinda here, in another meeting also
14:10:14 <asamalik> langdon, yeah
14:10:32 <asamalik> langdon, I talked to Copr guys and we agreed they will wait for us to produce the compose
14:10:37 <asamalik> so they can consume it for dependencies
14:11:05 <asamalik> they don't plan to do anything before beta freeze, as:
14:11:19 <asamalik> 1) they need DNF we are going to produce
14:11:29 <asamalik> 2) they need compose we are going to produce
14:11:35 <asamalik> 3) it's a service on the side, so it's fine
14:12:00 <asamalik> and we is also DNF, Factory... the whole modularity project
14:12:22 <langdon> in the module creation process i recommend it as how to do "scratch builds" so .. its not "just" on the side
14:12:55 <asamalik> right, but they are basically blocked by the modular release, so.. not sure
14:12:58 <asamalik> do we need to revisit?
14:13:09 <langdon> but.. as for fedora infra, i assume you mean it isn't effected by infra freezes
14:13:28 <langdon> i figured they were blocked.. but.. is the code all done?
14:13:31 <asamalik> exactly that, yeah... just didn't know a better formulation
14:13:43 <asamalik> langdon, yeah
14:14:06 <asamalik> they can build modules now, but it gets build against normal fedora I guess... and the deps won't work...
14:14:40 <langdon> ok.. im ok with that answer.. i just want to be sure they have a plan to "drop everything" as soon as we can give them stuff and deploy it
14:14:51 <asamalik> but they use MBS, and they can submit a modulemd...
14:15:00 <langdon> but that isn't in production
14:15:44 <asamalik> ah, yeah, right... they also need the DNF
14:16:09 <asamalik> because they need to install it on the builders, on a traditional Fedora system
14:16:18 <asamalik> so they want it in fedora
14:16:25 <asamalik> but our current Boltron DNF is just in boltron
14:16:35 <asamalik> so when we get then DNF, they can go production
14:16:51 <asamalik> and when we give them composes, they can build on the right buildroot
14:17:13 <langdon> ok .. do we need any more of a plan with them? or just let it be until our side is available?
14:17:46 <asamalik> we agreed to just let it go until we are done..
14:17:59 <langdon> ok
14:18:04 <langdon> is geppetto herE?
14:18:07 <langdon> here even?
14:18:11 <geppetto> maybe
14:18:12 <langdon> ha
14:18:32 <langdon> i was wondering if you can get your hacked up f27-modular-preview in https://hub.docker.com/r/modularitycontainers/
14:18:45 <geppetto> get it in how?
14:18:55 <langdon> create a new repo, docker push
14:19:03 <geppetto> Ok, yeh
14:19:04 <langdon> maybe f27-modular-rawhide
14:19:17 <langdon> and then we can override it when we have real builds
14:19:26 <geppetto> It's still f26 data … just with the latest f27 dnf
14:19:42 <ttomecek> what are the changes in dnf?
14:19:45 <langdon> that's ok i think ?
14:19:59 <asamalik> langdon, when we get the installer done, we can do a real image.. and if DNF guys can push their stuff into dist git, even some different branch, contyk can use it for Platform
14:20:12 <asamalik> langdon, the problem is... the installer currently fails on s390x
14:20:16 <asamalik> :/
14:20:20 <geppetto> ttomecek: for all of them, you'd need to check upstream … biggest one in today's update is the new naming stuff.
14:20:33 <asamalik> karsten, any progress with the installer?
14:21:07 <ttomecek> geppetto, awesome, I'm gonna ask Martin about detail
14:21:09 <langdon> and, karsten have you talked to jwb?
14:21:44 <asamalik> langdon, and since we are going to start with builds against bootstrap, we can add new content very quickly
14:21:48 <langdon> geppetto: this "<langdon> that's ok i think ?" was to you
14:22:05 <langdon> asamalik: yep.. as soon as we have it.. ill believe it :)
14:22:28 <geppetto> langdon: Ok, just mention it due to the proposed name
14:22:35 <asamalik> langdon, the dependency report tooling can take care about runtime dependencies, so we don't have the problems we had with Boltron... https://github.com/fedora-modularity/dependency-report
14:22:39 <asamalik> langdon, ha, yeah... :)
14:23:30 <langdon> geppetto: yeah.. i hear ya.. but, we can just put a note on the docker hub description or something.. i mean it is rawhide.. so..
14:24:13 <langdon> any more questions for open floor? i think that was all of mine :)
14:24:43 <asamalik> just about the installer, but karsten doesn't seem to be here :(
14:25:16 <karsten> I'm here, I just was busy in another window
14:26:53 <karsten> anaconda ist still failing to build. glib-compile-resources aborts even when compiled with the latest packages.
14:27:19 <langdon> karsten: did you talk to jwb?
14:27:22 <langdon> for help?
14:27:30 <karsten> langdon:  you've pointed me at Samantha, not at jwb
14:27:42 <karsten> so that's who I've mailed
14:27:46 <langdon> karsten: both :) jwb is all about multi-arch
14:28:02 <langdon> you can find him on irc usally pretty quickly
14:28:05 <karsten> ok, I'll send another mail
14:28:13 <karsten> or that
14:28:45 <langdon> cool
14:30:01 <asamalik> cool!
14:33:44 <langdon> anybody got anymore?
14:34:29 <asamalik> I'm good
14:34:30 <nils> langdon, I'm not sure about next meeting with Flock, are there any followup meetings that will keep people away?
14:35:27 <langdon> nils: flock meeting would be office hours anyway..
14:35:40 <nils> langdon, I meant the week after Flock
14:35:42 <langdon> would you prefer to cancel?
14:35:49 <langdon> ohh.. not sure.. let me look
14:36:28 <langdon> i dont have any conflicts
14:36:44 <nils> ok
14:37:24 <langdon> although sept 4 (the day before the meeting) is a us holiday.. if that matters
14:37:55 <langdon> and canada and turkey to be more complete :)
14:37:55 <nils> langdon, not if someone doesn't tack some PTO on
14:38:39 <langdon> ok.. let's close this down?
14:38:58 <nils> yup
14:39:02 <nils> thanks everybody!
14:39:07 <nils> #endmeeting