14:00:17 #startmeeting modularity_wg 14:00:17 Meeting started Tue Sep 4 14:00:17 2018 UTC. 14:00:17 This meeting is logged and archived in a public location. 14:00:17 The chair is nils. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:17 Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:00:17 The meeting name has been set to 'modularity_wg' 14:00:18 #meetingtopic Weekly Meeting of the Modularity Working Group 14:00:18 #chair dgilmore_ langdon mikedep333 14:00:18 Current chairs: dgilmore_ langdon mikedep333 nils 14:00:29 hey 14:00:31 #topic Roll Call 14:00:35 .hello psabata 14:00:36 contyk: psabata 'Petr Ĺ abata' 14:00:36 .hello nphilipp 14:00:38 .hello2 14:00:39 nils: nphilipp 'Nils Philippsen' 14:00:42 langdon: langdon 'Langdon White' 14:01:37 #topic Agenda 14:02:20 we have agenda in the issue list! 14:02:28 We actually do have two issues tagged for the meeting, but they're both by asamalik who's out today IIRC. 14:02:34 ha 14:02:34 :D 14:02:46 So, anything besides Adam's topics? 14:02:52 what are those? 14:03:01 - Replacing docs.pagure.org/modularity with a redirect to the Fedora Docs / Modularity 14:03:08 - Managing module lifecycles 14:03:08 we might be able to do the managing lifecycles one 14:03:17 ok I'll add it 14:03:26 #info Managing module lifecycles 14:03:56 anything else that doesn't fit AOB? 14:04:37 alright 14:04:40 #info AOB 14:04:49 #topic Managing module lifecycles 14:04:52 where AOB = any other business :) 14:04:54 langdon? 14:04:58 exactly :) 14:05:19 #link https://pagure.io/modularity/issue/107 meeting agenda item 14:05:29 I'm usually the last one to decipher acronyms, so didn't think explaining it was necessary ;) 14:05:35 #link https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/thread/K4FUOQHQSRAAI3PUUGXAC6CXEN27Y2JH/ mail thread 14:05:47 so 14:05:49 nils: AOB is pretty unusual in the US (in my experience) 14:05:54 ah ok 14:06:11 Adam started a thread about the lifecycles after the last meeting 14:06:22 he also said he would start another, more condensed one, easier to reply to 14:06:37 also regarding the other topic, I believe we ageed to do it two weeks ago 14:06:52 contyk: which other topic? 14:06:59 the redirect 14:07:07 oh 14:07:18 we can sitll put it on the agenda after this? 14:07:28 i would like to read that page.. but probably can do it in parallel 14:08:10 langdon, we can discuss it but the bot won't list it in the overview 14:08:15 AIUI 14:08:20 ahhh 14:08:44 (because the overview is just what we put as #infos in the Agenda topic) 14:09:19 i may just make some comments in the ticket 14:09:25 but.. module lifecycles? 14:09:33 do we have comments here? 14:10:17 option #2 sounds a lot like the policy we have for bare packages 14:10:31 just with the added recorded EOLs as trimming 14:12:11 the big problem with that is that it requires a module maintainer to continue to maintain it for at least all current releases 14:12:17 is that a high burden? 14:12:50 only if they forget to retire it from Rawhide in time 14:13:05 i would say, i think that, from a user perspective, it would be less confusing.. 14:13:26 we also don't have to assume that this policy is forever.. just the one we have for now 14:13:33 it's more implicit than option #1, but looks to be more what people are used to 14:14:09 One thing I don't like too much about option #1 is that it uses packages' EOLs to compute the one of the module, regardless of if they're API packages or not 14:14:19 ahh yes 14:14:27 i like the streams.yaml idea 14:14:32 contyk: opinions? 14:14:50 * contyk reads 14:15:09 But that may be doing the right thing as a side effect, i.e. non-API EOL influences module EOL, ... magic happens ..., maintainer knows it's something they need to look at. 14:16:07 nils: well.. it could very easily be that non-api-rpm is in the process of being replaced.. that shouldn't effect module eol at all IMO 14:16:41 I think what we currently have is more like #2 14:16:56 since you can only pick two dates each year 14:17:10 yeah, it's more a PR problem than a technical one -- if the maintainer wants to support the module until date X, then the module metadata should say that, regardless of some non-API component's shorter EOL 14:17:15 I also think people struggle with choosing this one thing; doing so for individual packages would be insane 14:17:37 I think you have to for stream branches of RPMs, right? 14:19:32 possibly 14:19:47 hmmm 14:20:02 I don't currently have an opinion on this, would need to come up with something on the spot :) 14:20:12 ha 14:20:26 well.. should we ask asamalik to send the new note.. and table this for now? 14:20:43 he's planning to, afaik 14:20:46 but sure 14:21:50 should this be #action'ed? 14:22:00 so ... #info needs more discussion, moving back to mailing list && #action asamalik to send out a new, more concise, note about module lifecycles 14:22:19 sounds good 14:22:37 #info needs more discussion, moving back to mailing list 14:22:45 #action asamalik to send out a new, more concise, note about module lifecycles 14:23:39 langdon, do you want to do the redirect thing now? 14:24:26 does anyone have comments? i am just going to make some comments in the ticket 14:25:02 for reference: https://pagure.io/modularity/issue/106 14:25:35 apparently no comments here 14:25:54 #topic Open Floor / Any Other Business 14:26:09 so, anything else? 14:26:31 i don't really have anything else 14:26:44 working on proposals for f30 & 31 14:27:11 me neither 14:27:46 okie dokie 14:27:55 thanks everybody! 14:27:57 #endmeeting