14:00:20 <nils> #startmeeting modularity_wg 14:00:21 <zodbot> Meeting started Tue Oct 2 14:00:20 2018 UTC. 14:00:21 <zodbot> This meeting is logged and archived in a public location. 14:00:21 <zodbot> The chair is nils. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:21 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:00:21 <zodbot> The meeting name has been set to 'modularity_wg' 14:00:21 <nils> #meetingtopic Weekly Meeting of the Modularity Working Group 14:00:21 <nils> #chair dgilmore langdon 14:00:21 <zodbot> Current chairs: dgilmore langdon nils 14:00:24 <nils> \o 14:00:29 <nils> #topic Roll Call 14:00:33 <contyk> .hello psabata 14:00:35 <nils> .hello nphilipp 14:00:40 <zodbot> contyk: psabata 'Petr Šabata' <psabata@redhat.com> 14:00:40 <asamalik> .hello2 14:00:41 <langdon> o? 14:00:44 <zodbot> nils: nphilipp 'Nils Philippsen' <nphilipp@redhat.com> 14:00:47 <zodbot> asamalik: asamalik 'Adam Samalik' <asamalik@redhat.com> 14:01:59 <langdon> .hello2 14:01:59 <zodbot> langdon: langdon 'Langdon White' <langdon@redhat.com> 14:02:11 <nils> #topic Agenda 14:02:27 <contyk> I had an action regarding the Taiga board 14:02:34 <contyk> so I can comment on that 14:02:38 <contyk> do we have anything else? 14:02:50 <nils> asamalik, the "Managing module lifecycles" issue is still open 14:03:06 <nils> is there still something to be discussed here? 14:03:21 <nils> #info [contyk] Taiga board update 14:03:30 <contyk> the issue isn't really resolved, so probably 14:03:36 <langdon> any update on dnf for the record? 14:03:37 <contyk> okay, Taiga 14:03:43 <asamalik> nils: I don't think there was a lot of discussion in the past 2 weeks, but it's not resolved 14:03:52 <nils> ok I'll put it on the agenda 14:03:52 <contyk> langdon: +1, will also comment 14:03:57 <stickster> .hello pfrields 14:03:58 <zodbot> stickster: pfrields 'Paul W. Frields' <stickster@gmail.com> 14:03:59 <contyk> Taiga! 14:04:02 <asamalik> nils: I think we can remove the meeting label, but need to keep it before that 14:04:09 * stickster lurking -- sorry, in another meeting at the same time and missed the roll call earlier :-) 14:04:17 <nils> #info [asamalik] Managing module lifecycles 14:04:18 <contyk> so I've been creating epics in our instance for topics I know need some work 14:04:19 <asamalik> * before we have taiga 14:04:20 <asamalik> :) 14:04:34 <nils> shall I switch the order? 14:04:37 <contyk> it's still incomplete, there are a couple more I need to add for software management, third party story, and composes 14:04:50 <contyk> nils: should I wait? :) 14:05:02 <nils> I'm asking asamalik :) 14:05:23 <nils> ah yes, please wait until the agenda is done, thanks :) 14:05:36 <nils> anyway 14:05:37 * contyk misread the #info for #topic 14:05:47 * asamalik is cnfused 14:06:04 <nils> #info [langdon, contyk] dnf update 14:06:18 <nils> #topic Taiga board update 14:06:21 <nils> #chair contyk 14:06:21 <zodbot> Current chairs: contyk dgilmore langdon nils 14:06:25 <contyk> alright 14:06:31 <asamalik> nils: managing lifecycles — not done, nothing to report today 14:06:46 <contyk> so as I've just said, I've been adding epics for topics that need to be resolved in the near term (a release or two) 14:07:09 <contyk> there are still some more I need to define; this is just the first pass anyway -- once ready, I'd like to discuss it with everyone involved 14:07:31 <asamalik> contyk++ 14:07:33 <contyk> threebean also had some ideas but I need to clarify these with him first 14:07:52 <contyk> feel free to look at the current status 14:07:57 <contyk> #link https://tree.taiga.io/project/modularity-wg/epics 14:07:59 <asamalik> psabata++ 14:08:02 <contyk> most of these don't have any actions, yet 14:08:34 <contyk> I also talked with asamalik regarding how we do the tracking; we defined the columns (states) for epics, stories, tasks, and issues 14:09:01 <contyk> he already has a card on the board to document the current state 14:09:09 <contyk> of course that might change over time as we adopt it 14:09:18 <langdon> did igor add things like he wanted? 14:09:32 <contyk> I gathered input for him, for modulemd and ursa major 14:09:43 <contyk> we have an epic for ursa major, modulemd still needs to be defined 14:09:56 <contyk> that's mostly the buildorder redesign 14:09:57 <langdon> contyk: him? = asamalik? igor? 14:10:02 <contyk> langdon: Igor 14:10:14 <langdon> gotcha 14:10:23 <contyk> Bodhi changes/requests are also capured 14:10:24 <dgilmore> hey all 14:10:37 <contyk> hey 14:10:56 <contyk> I think that's all for now; I'll continue dumping the ideas there 14:11:08 <langdon> do you want others to add things? 14:11:16 <langdon> or more comment/review? 14:11:53 <contyk> comments and reviews would be great 14:12:10 <contyk> for epic ideas (ha), let's discuss those first 14:12:20 * stickster applauds definitions in the cards he's read so far 14:13:52 <contyk> if we could review the current list by the end the week, that'd be awesome 14:14:05 <contyk> we could start using this before our next formal meeting 14:14:23 <contyk> last time we also discussed jit.si meetings and there were some concerns about those 14:14:41 <contyk> when, how often, how to make them as inclusive as possible 14:14:47 <contyk> so that will require some more brainstorming 14:15:01 <contyk> oh and one more thing 14:15:32 <stickster> Can anyone summarize what the concerns were about jit.si? 14:15:45 <contyk> Taiga supports issues but to minimize the number of places people need to check and considering now FAS integration in taiga.io, I'd suggest we just keep using our pagure repo 14:16:16 <langdon> i think the issue is "video meetings" are tough to attend.. not jit.si per se 14:16:27 <contyk> stickster: people who might not attend meetings that are too frequent or can't join video calls might feel excluded 14:16:37 <contyk> what langdon says 14:18:00 <stickster> *nod... right, it's not unusual that volunteer contributors face challenges trying to be on a video call if they have other obligations, especially $DAYJOB 14:18:17 * contyk nods 14:18:21 <stickster> Are we considering doing this through IRC then? 14:18:39 <contyk> yes 14:18:48 <contyk> but I think there's still value in video calls 14:18:51 <stickster> And should we set some sort of expectation for how many people show up to those meetings, to help us calibrate whether we can switch to video? (Sorry if I'm retreading old topics.) 14:19:02 <langdon> and timezones.. 14:19:16 <stickster> IOW, if no one who shows up to do work has an issue, then we should use what works well for the groupl. 14:19:19 <stickster> *group, even. 14:19:23 <contyk> can't remember what meeting it was but I think bcotton was taking meetbot notes for a video call 14:19:41 <langdon> i think bcotton did an awesome job keeping a jit.si meetin "on irc" at the same time for the last council meeting.. so we may be able to make it work 14:19:43 <contyk> so it could be a little bit of both 14:19:51 <contyk> ah, the Council 14:19:52 <langdon> ha 14:20:07 <stickster> *nod. There are ways to do this that effectively put info out where others can see/participate. I'm not against video meetings. I'm also not against trying IRC first to see how it goes. 14:20:07 <langdon> i was busy typing and didn't see your remark :) 14:20:44 <stickster> just my $0.02 14:21:02 * stickster goes with the group flow here ;-) 14:21:19 <contyk> we can change it every day! 14:21:51 <langdon> how about every 1/4 in every meeting.. like sportball! 14:22:15 <contyk> we could just mail voice recordings to each other 14:22:43 * stickster sends carrier pigeon notes 14:22:45 <asamalik> on tapes? 14:23:04 <nils> In obscure formats. Like recorded on wire :) 14:23:09 <contyk> yes, as digital data on tapes 14:23:15 <nils> pfff digital 14:23:20 <contyk> good old tar 14:23:40 <contyk> anyway, needs some more thought still 14:23:54 <asamalik> glad we're exploring all the options 14:23:58 <nils> .oO(wave forms scratched into stone tablets) 14:24:09 <contyk> any more comments or should we move to the next topic? 14:24:29 <contyk> maybe some infos 14:25:11 <contyk> #info Still populating our Taiga board with epics and actionable cards; comments and reviews appreciated 14:25:38 <contyk> #info If we seem to agree it's in a usable state, we could start using it before the next WG meeting 14:26:22 <contyk> #info Still gathering ideas for new meetings, both in form and frequency 14:26:25 <contyk> EOF 14:27:49 * contyk hums 14:28:31 <contyk> nils? 14:29:29 <nils> oh sorry 14:29:46 <nils> my laptop is muted, didn't hear your humming ;) 14:30:05 <nils> #topic Managing module lifecycles 14:30:10 <nils> #info no update, postponed 14:30:17 <nils> #topic dnf update 14:30:22 <nils> langdon, contyk? 14:30:47 <langdon> mostly contyk 14:31:12 <contyk> this was discussed during the FESCo meeting yesterday 14:31:36 <contyk> we didn't really get to any conclusion 14:31:50 <langdon> is this the dnf database thing? or something else? 14:31:52 <contyk> so the decision was to give the modularity folks (us) and the dnf team another week to discuss things 14:32:04 <contyk> the goal is to identify an mvp for the blocker 14:32:10 <contyk> it's the dnf database thing, yes 14:32:36 <contyk> .fesco https://pagure.io/fesco/issue/1974 14:32:36 <zodbot> contyk: Error: 'https://pagure.io/fesco/issue/1974' is not a valid integer. 14:32:42 <contyk> bah 14:32:46 <contyk> .fesco 1974 14:32:47 <zodbot> contyk: Issue #1974: Problematic blocker for F29: dnf 'offline' module tracking - fesco - Pagure.io - https://pagure.io/fesco/issue/1974 14:32:56 <contyk> here's where we track it 14:33:28 <contyk> there's another major known dnf/modularity bug but it seems a fix already landed upstream 14:33:40 <langdon> so what can this group do to help with 1974? 14:33:52 <contyk> essentially you couldn't install profile packages unless those were part of your own module, making "meta-modules" fairly broken 14:34:39 <contyk> langdon: good question! we're meeting with the dnf team tomorrow to discuss the problem 14:34:48 <langdon> ok 14:34:49 <contyk> langdon: you included 14:34:57 <langdon> whoops 14:35:03 <langdon> :) 14:35:19 <contyk> but other than that... we'll see, I guess 14:35:47 <contyk> my proposal would be to keep the "database" requirement but acknowledge the issues it might introduce 14:36:21 <contyk> and not blocking the release on having mechanisms to mitigate these ready 14:36:27 <contyk> it's all corner cases 14:39:20 <langdon> so.. move on? 14:39:37 <langdon> contyk: actually one more thing 14:39:53 <langdon> i thought there was a workaround.. was that just mistaken? 14:40:12 <contyk> I'm not aware of that 14:40:22 * stickster neither 14:41:20 <langdon> contyk: maybe i misunderstood.. or something.. it came up in council meeting a few weeks ago.. maybe ill look at the logs and see if i can figure out where i got it from 14:41:24 <nils> anything to #info? 14:42:06 <contyk> #info Our WG will discuss FESCo issue 1974 with the DNF team to agree on an MVP later this week 14:42:37 <nils> so, anything for open floor? 14:43:23 <contyk> hmm 14:43:27 * stickster is prepping email to council-discuss on lifecycle objective 14:43:46 <contyk> I got some feedback from a confused user today when their ursine package was upgraded to a modular variant today 14:43:59 <stickster> hm 14:44:00 <contyk> I'm not sure whether it was f28 or f29, though 14:44:30 <contyk> this was because the content was modularized and we introduced defaults for the module replacing it 14:44:37 <nils> I noticed that an update enabled the libgit2 module for me, not sure if that's related, 14:44:49 <contyk> it was transparent and automatic but our "managing defaults" page says we should have a Change for these things 14:45:03 <contyk> #link https://docs.fedoraproject.org/en-US/modularity/making-modules/managing-defaults/ 14:45:18 <contyk> however, that page is a little unclear to me 14:45:51 <contyk> if I modularize my content, add defaults to and retire my ursine package in rawhide only, does that require a change? 14:46:10 <nils> #undo 14:46:10 <zodbot> Removing item from minutes: <MeetBot.items.Link object at 0x7f5650990b10> 14:46:17 <nils> #topic Open Floor 14:46:27 <nils> #link https://docs.fedoraproject.org/en-US/modularity/making-modules/managing-defaults/ 14:46:30 <nils> (oops) 14:46:37 <contyk> :) 14:46:43 <contyk> it probably does 14:46:55 <contyk> anyway, it might need some clarification (yet another process-type epic...) 14:47:13 <contyk> nils: and yeah, it was libgit2 in this case 14:47:48 <langdon> personally, on the point, I don't think that needs a change 14:47:50 <nils> some dependency that only is fulfilled when the module is enabled? 14:48:29 <contyk> langdon: on the other hand, what would need one? if anything 14:48:48 <langdon> if you wanted no default? 14:48:53 <contyk> that libgit2 transition happened in a stable release; that's not something you have changes for either :) 14:49:22 <contyk> langdon: hmm, yeah 14:50:00 <contyk> when users are somehow impacted 14:50:27 <contyk> content disappearing because it's only in non-default modules would be the case 14:50:44 <contyk> changes to default streams for previously modular content also 14:50:50 <contyk> but that might be it 14:52:05 <langdon> I think our initial guidelines thinking was all was to aggressive 14:52:17 <langdon> *way 14:53:14 <asamalik> yeah what I remember about the change requirement is that it was meant to give user a similar experience they had in the traditional release 14:53:30 <asamalik> so a new (default) version would be announced in the release notes 14:53:39 <asamalik> there won't be any new breaking version mid-release 14:53:40 <asamalik> etc 14:53:57 <asamalik> that page probably could use some love :) 14:56:34 * contyk will add an epic for it :) 14:57:05 <asamalik> contyk: awesome :) 14:57:57 <langdon> can it be called "make all the guidelines easier"? ;) 14:58:52 <nils> not sure if this can be made "time-bound" :D 14:59:50 <contyk> anything else in the last 12 seconds? 15:00:20 <nils> apparently not :) 15:00:24 <nils> thanks everybody! 15:00:27 <nils> #endmeeting