15:00:53 <nils> #startmeeting modularity_wg
15:00:53 <zodbot> Meeting started Tue Nov 29 15:00:53 2016 UTC.  The chair is nils. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:53 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:53 <zodbot> The meeting name has been set to 'modularity_wg'
15:00:53 <nils> #meetingtopic Bi-weekly Meeting of the Modularity Working Group
15:00:53 <nils> #chair
15:00:53 <zodbot> Current chairs: nils
15:01:00 <contyk> hi
15:01:06 <langdon> .hello langdon
15:01:07 <zodbot> langdon: langdon 'Langdon White' <langdon@fishjump.com>
15:01:08 <jkurik> .hello jkurik
15:01:09 <zodbot> jkurik: jkurik 'Jan Kurik' <jkurik@redhat.com>
15:01:35 <nils> #chair dgilmore langdon tflink
15:01:35 <zodbot> Current chairs: dgilmore langdon nils tflink
15:01:42 <nils> #topic Roll Call
15:01:46 <contyk> .hello psabata
15:01:46 <nils> .hello nphilipp
15:01:47 <langdon> yay.. nils is back!
15:01:47 <zodbot> contyk: psabata 'Petr Šabata' <psabata@redhat.com>
15:01:50 <zodbot> nils: nphilipp 'Nils Philippsen' <nphilipp@redhat.com>
15:01:51 <sgallagh> .hello sgallagh
15:01:53 <langdon> to run the meeting!
15:01:54 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
15:01:57 <nils> and in one piece :D
15:02:21 <moto-timo> .hello ttorling
15:02:22 <zodbot> moto-timo: ttorling 'Tim Orling' <ticotimo@gmail.com>
15:02:22 <langdon> nils, you will have to tell me about why and how oman.. im kinda jealous
15:02:49 <contyk> put it on the agenda
15:02:58 <nils> langdon: sure, when do you have time ;)?
15:03:10 <nils> yeah, there's plenty of space there :D
15:03:10 <langdon> nils, weekends
15:03:27 <langdon> contyk, :P
15:03:34 <moto-timo> sleep is for the weak
15:03:36 <asamalik> .hello asamalik
15:03:37 <zodbot> asamalik: asamalik 'Adam Samalik' <asamalik@redhat.com>
15:03:43 <nils> moto-timo: sleep is for the wise
15:03:58 <langdon> asamalik wants to add "demos for sprint #18" to the agenda :)
15:03:59 <moto-timo> oh yeah. I always get that wrong. not enough sleep
15:04:09 <moto-timo> :)
15:04:09 <asamalik> do I? ok!
15:04:26 <nils> #topic Agenda
15:04:26 <nils> #topic Base Runtime prototype build status
15:04:32 <contyk> yay, that's me
15:04:35 <nils> #undo
15:04:35 <zodbot> Removing item from minutes: <MeetBot.items.Topic object at 0x4abb6290>
15:04:47 <asamalik> langdon: do you please have a link for the etherpad?
15:04:56 <nils> #info Base Runtime prototype build status
15:05:03 <nils> asamalik: http://piratepad.net/modularity-wg-agendas
15:05:10 <nils> but I'll just write it here
15:05:21 <nils> #info demos for sprint #18
15:05:44 <contyk> anything else?
15:06:01 <nils> doesn't look like it
15:06:12 <nils> #topic Base Runtime prototype build status
15:06:15 <contyk> ok
15:06:16 <nils> #chair contyk
15:06:16 <zodbot> Current chairs: contyk dgilmore langdon nils tflink
15:06:52 <langdon> nils, we should remember to reach out to threebean and ask for a status on f2
15:06:53 <contyk> so I just want to inform you all we recreated our prototype, using f25rc3 data, and then attempted to rebuild it to bootstrap our infrastructure (again)
15:07:22 <contyk> in other words, we're rebuilding a subset of f25rc3 using only the packages and versions in f25rc3
15:07:38 * threebean waves
15:07:47 <nils> hey threebean
15:07:53 <moto-timo> beetlejuice beetlejuice beetlejuice
15:08:01 <contyk> we encountered 152 build failures we're now investigating and will try to (with the primary maintainters' help) resolve over the upcoming few weeks
15:08:05 <langdon> Lol
15:08:17 <contyk> the logs are here -- https://psabata.fedorapeople.org/f25rc3-failures/
15:08:20 <nils> contyk: good luck :)
15:08:25 <sct> .hello sct
15:08:26 <zodbot> sct: sct 'Stephen Tweedie' <sct@redhat.com>
15:08:26 <ttomecek> contyk, any charts?
15:08:30 <contyk> I'll be mailing the devel@ list with details once we know some more
15:08:31 <nils> #chair sct
15:08:31 <zodbot> Current chairs: contyk dgilmore langdon nils sct tflink
15:08:45 <langdon> contyk no magic bullets like autoconf?
15:08:47 <contyk> and we'll file bug reports for stuff that's not trivial to fix
15:08:57 <contyk> langdon: not anymore, or at least it's not obvious
15:09:07 <langdon> :(
15:09:24 <contyk> some of those were caused by changes in the build dependency chains
15:09:36 <contyk> packages that were relying on something being in the buildroot is no longer there
15:09:56 <contyk> this affects many java packages but also a number of qt5 builds that are relying on perl, for example
15:10:13 * asamalik needs to be 3-4 mins afk :/
15:10:48 <contyk> other failures are more peculiar, such as sudden dependency incompatibilities
15:10:59 <langdon> So.. Stupid q.. Could we just add those things to the buildroot? At least now it would be explicit and we could kill them over time?
15:11:04 <contyk> where some packages were updated and other packages that require them can no longer be built against those versions
15:11:31 <contyk> langdon: we have them in the buildroot but it would bite us sooner or later
15:11:57 <contyk> langdon: also anything we build with this set is somehow tainted due its... let's say lack of quality
15:12:11 <langdon> threebean, assuming you didn't have scrollbars, I was asking for you to give an f2 status rep next mtg
15:12:42 <sgallagh> Going forward, it would be *really* nice if we could get a rolling-rebuild of every package in the Fedora repos.
15:12:43 <contyk> anyway, that's it from me; I'll include more details in the mail to devel@ later this or next week
15:13:08 <sgallagh> Basically a perpetually-running mass-rebuild that would alert maintainers if their packages suddenly go FTBFS due to dependency changes.
15:13:13 <nils> contyk: can you #info a summary?
15:13:20 <contyk> sure
15:13:26 <langdon> contyk like as in the rpms should be declaring these deps?
15:13:31 <contyk> sgallagh: that's not trivial
15:13:43 <sgallagh> contyk: Did I say it was? :)
15:13:48 <contyk> no :)
15:13:48 * asamalik is back
15:13:58 <sgallagh> But the scripts must already exist, since we do mass-rebuilds at least once a year.
15:13:59 <contyk> langdon: yes, but that's just a subset of the failures
15:14:26 <contyk> sgallagh: the thing is the udpated rpms can pull in additional dependencies that we wouldn't be included in the module
15:14:44 <sgallagh> contyk: I'm talking about traditional Fedora right now
15:14:50 <contyk> I see, okay
15:14:54 <contyk> alright, the info
15:14:58 <sgallagh> As a way to help us avoid these surprises
15:15:09 <sgallagh> And it would be generally-useful to Fedora as well
15:15:17 <contyk> #info Base Runtime team is focusing on analyzing F25RC3-based prototype build failures
15:15:41 <nils> contyk: thanks
15:15:47 <contyk> #info The team will cooperate with the respective primary maintainers to resolve these FTBFS errors over the new few weeks
15:16:06 <threebean> (sorry, I'm double booked in another meeting.  can't report today.)
15:16:07 <nils> contyk: typo, wait a sec
15:16:13 <langdon> contyk do we have a date for which we hope it will be working?
15:16:16 <nils> #undo
15:16:16 <zodbot> Removing item from minutes: INFO by contyk at 15:15:47 : The team will cooperate with the respective primary maintainers to resolve these FTBFS errors over the new few weeks
15:16:28 <nils> #info The team will cooperate with the respective primary maintainers to resolve these FTBFS errors over the next few weeks
15:16:34 <contyk> langdon: by the end of the year?
15:16:43 <nils> contyk: more #infos coming?
15:16:45 <langdon> threebean yeah.. Next time.. Two weeks
15:16:47 <contyk> nils: nope
15:16:54 <nils> 👍
15:17:05 <nils> #topic demos for sprint #18
15:17:09 <nils> #chair asamalik
15:17:09 <zodbot> Current chairs: asamalik contyk dgilmore langdon nils sct tflink
15:17:17 <asamalik> yes, we did demos!
15:17:24 * asamalik is looking for a youtube url
15:17:47 <asamalik> https://www.youtube.com/playlist?list=PLcwHJG45BmAMwio4RC3r95Px9by7dUazr
15:17:54 <asamalik> #link https://www.youtube.com/playlist?list=PLcwHJG45BmAMwio4RC3r95Px9by7dUazr Fedora Modularity Sprint 18 demos
15:18:09 <asamalik> this sprint we have been focused on producing container images
15:18:10 <langdon> Copr can build modules, right? And we have some containers..
15:18:41 <asamalik> but instead of creating the modules first, we are basing these images on fedora 25, we will figure out the configuration, etc
15:18:51 <asamalik> and we will package it as modules later
15:19:33 <asamalik> it might sound strange, but it's not important in which order exactly we are going to build them - and there will be a bit more mature infra later :)
15:19:40 <asamalik> so that's for container images
15:19:54 <asamalik> and yes, as langdon pointed out, the Copr team is working on building modules from modulemd
15:20:01 <asamalik> the demo is also included on that link
15:20:18 <asamalik> it is not deployed, yet - but it should be ready to use in few weeks
15:20:26 <langdon> But we have some modulemds for a couple of the containers.. They just don't actually produce containers yet
15:20:36 <asamalik> exactly
15:20:46 <asamalik> I will be working on a modulemd for my container
15:20:58 <asamalik> got the gen core API package set from contyk
15:21:10 <asamalik> so I can develop that and test with the mock backend of MBS
15:21:17 <langdon> And karsten did one
15:21:54 * asamalik got scared by how many messages he wrote
15:22:11 <nils> asamalik: do you want to #info some of it ;)
15:22:21 <asamalik> nils: good point!
15:23:11 <asamalik> #info in the sprint 18, the Modularity team have been focused on producing container images - configuration, etc. We will package them as modules later.
15:24:02 <asamalik> #info asamalik will be developing the modulemd for his FTP container, got an API package set for the gen core
15:24:21 <asamalik> #info the Copr team is working on building modules from modulemd - not ready yet - will be in a few weeks
15:24:43 * asamalik stops writing
15:24:54 <langdon> Totally thought it had been deployed
15:24:58 <nils> asamalik: :)
15:25:10 <nils> lemme check…
15:25:13 <asamalik> langdon: no, it's just a local vagrant box :)
15:25:20 <nils> nothing left!
15:25:24 <asamalik> cool!
15:25:26 <nils> #topic Open Floor
15:25:42 <nils> So, anybody has anything?
15:26:01 <contyk> you can talk about your vacation
15:26:33 <nils> Heh :)
15:26:48 <nils> I meant slightly more on topic than that ;)
15:27:14 <nils> I can do an AMA after we're done talking about modularity :)
15:27:33 <langdon> Hs
15:27:38 <langdon> Ha even
15:28:08 <langdon> It has been a pretty messy few weeks.. Hoping to get back to clean comms soon
15:28:20 <asamalik> yeah :)
15:28:50 <nils> in what way where comms dirty? anything I missed there?
15:29:23 <langdon> Just not as regular... I blame my illness..
15:30:00 <nils> langdon: you've been/are sick? speedy recovery then!
15:30:09 <langdon> i was .. mostly better
15:30:26 <langdon> lots of change with f2 moving full tilt.. and base runtime/gen core..
15:30:59 <langdon> thats all i got
15:32:29 <nils> okay, looks like we're done with on topic ;)
15:32:43 <langdon> +1
15:32:51 <nils> #endmeeting