15:00:00 #startmeeting modularity 15:00:00 Meeting started Tue Feb 26 15:00:00 2019 UTC. 15:00:00 This meeting is logged and archived in a public location. 15:00:00 The chair is nils. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:00 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:00 The meeting name has been set to 'modularity' 15:00:00 #meetingtopic Weekly Meeting of the Modularity Team 15:00:00 #topic Roll Call 15:00:08 .hello nphilipp 15:00:08 nils: nphilipp 'Nils Philippsen' 15:00:13 .hello psabata 15:00:14 contyk: psabata 'Petr Šabata' 15:01:05 .hello2 15:01:06 langdon: langdon 'Langdon White' 15:01:39 .hello2 15:01:40 sgallagh: sgallagh 'Stephen Gallagher' 15:02:28 #topic Agenda 15:02:28 #info #112 Discussion: Module lifecycles 15:02:28 #info #115 Discussion: Stream branch ownership for packages & modules 15:02:38 anything else? 15:02:48 I think asamalik has some news from yesterday 15:03:01 yeah but I think it fits into one of these two, doesn't it? 15:03:08 yeah 15:03:11 .hello2 15:03:12 asamalik: asamalik 'Adam Samalik' 15:03:30 ok, let's get going 15:03:40 unfortunatelly I don't have any update to 115 15:03:44 but I do have an update to 112 15:03:46 #topic #112 Discussion: Module lifecycles 15:03:46 #link https://pagure.io/modularity/issue/112 15:03:46 .modularity 112 15:03:46 #link https://pagure.io/fesco/issue/2027 15:03:48 nils: Issue #112: Discussion: Module lifecycles - modularity - Pagure.io - https://pagure.io/modularity/issue/112 15:03:53 #chair asamalik 15:03:53 Current chairs: asamalik nils 15:04:05 here you go :) 15:04:49 So I have reworked the proposal completely, applying feedback I got so far, and had FESCo to discuss it yesterday. 15:05:58 Unfortunately, it's been rejected by FESCo, because there were people strongly against storing the EOL information in PDC, advocating for dist-git. 15:06:16 that sounds like an implementation detail 15:06:29 I mean, other aspects of the proposal would still be valid, right? 15:07:08 funny fact, the PDC (or FPDC) idea came from FESCo, and there were people in favour of it... but exactly what contyk says, it feels like an implementation detail? 15:07:45 maybe the proposal could consider both (or more) options to be more acceptable 15:08:04 did they say why they were against (F)PDC in favor of dist-git? 15:08:24 contyk: probably? 15:08:36 what would you say the next steps are then? 15:09:25 I think the proper step is to rephrase the proposal to eliminate implementation suggestions. 15:10:08 Instead of "should be stored in [F]PDC", it should read that "A database of this information must exist and a process for updating it must be written" 15:10:22 sgallagh: yeah probably, good point 15:10:38 yes, but it should still consider some implementations of the database, otherwise it's vague and feels incomplete 15:10:52 Well, it should set the requirements on the database, not the implementation 15:11:04 there were strong opinions about implementation details, but the overall concept seemed to be received ok 15:11:17 "database" is also an implementation detail 15:11:22 it could list the different approaches as examples: "..., for instance store it in (F)PDC, or in a file in the dist-git repository"? 15:11:27 langdon: "data store" then 15:11:28 you want an "api to the data" you don't care how it is stored 15:11:44 sgallagh: still impl ;) 15:11:46 the database is langdon; you have to ask him everytime you need to know 15:11:51 +1 15:11:58 Now I want his home phone number. 15:12:12 via comms channel defined by time and day in eastern 15:12:25 :D 15:12:32 and that info is stored in the database as well 15:12:39 asamalik: Do you want me to attempt to wordsmith it and send a PR? 15:12:40 dont forget daylight savings.. and a mistake on comms channel will delay response by, at a minimum, 2 weeks 15:12:49 sgallagh: that would be awesome! 15:12:56 OK, #action me then 15:14:00 #action sgallagh will send a PR to the lifecycles proposal removing implementation details 15:14:02 sgallagh++ 15:14:17 sgallagh: i might write it in terms of the data to be surfaced ... e.g. "GET /v1//eol" or whatnot 15:14:36 langdon: I'm not going to write the API spec, just a set of requirements it has to meet 15:15:46 anything else regarding that proposal? 15:15:56 * asamalik thinks we're good 15:16:05 sgallagh: re-reading.. i think the "storing the info" might be able to just be cut.. the requirement is just an externallly accessible api for the data.. 15:17:02 langdon: Well, there are storage implications to consider. 15:17:28 I'll see what it looks like as I write it and I'll link my PR around so others can send PRs to my PR :) 15:17:51 that's good PR 15:17:56 like patches to patches to dwm! 15:20:27 I guess we're done now, right? 15:20:39 nils: yeah 15:20:50 #topic #115 Discussion: Stream branch ownership for packages & modules 15:20:50 #link https://pagure.io/modularity/issue/115 15:20:50 .modularity 115 15:20:50 #link https://pagure.io/fesco/issue/2028 15:20:52 nils: Issue #115: Discussion: Stream branch ownership for packages & modules - modularity - Pagure.io - https://pagure.io/modularity/issue/115 15:21:06 asamalik, you said no updates, so I'll push this to next week 15:21:11 nils: thanks 15:21:20 #info postponed until next meeting 15:21:30 I might just remove the label and set it when it's ready? 15:21:40 👍 15:21:57 It sounds so obvious now you mention it. :D 15:22:12 #topic Open Floor 15:22:43 I have one: I've just submitted fedmod-0.5.0 as a Fedora update 15:23:08 #info fedmod 0.5.0 released and submitted as a Fedora update 15:23:11 nils: I think you should write a blog post or something about why it's helpful 15:23:27 * asamalik is not sure how many people know about the tool 15:23:35 yeah 15:24:10 we only mention it once in the docs, and that's for generating a modulemd 15:24:11 the issue with this is that it's really a collection of tools that do pretty different things 15:24:16 asamalik: +1 15:24:29 nils: so a long blog post :) 15:24:44 yeah, there's that and linting, and summarizing modules (from RPM repo metadata, meanwhile from build systems (MBS)) 15:24:47 haha 15:25:29 I would have to find where to blog. I guess I have this same question every time someone suggests writing a blog to me. :) 15:25:45 For completeness sake though... 15:25:47 communityblog.fp.o 15:26:04 langdon: +1 15:26:06 nils: just write the content and adam or i will get it published 15:26:20 #info summarize-module can query information from build systems (MBS) alternatively to RPM repositories now 15:26:52 the communityblog is technically just a wordpress you log into with your FAS and write 15:26:56 #link https://bodhi.fedoraproject.org/updates/FEDORA-2019-5f16d857b1 F28 update 15:27:07 and coordinate with the group (CommOps) who helps with publishing 15:27:09 #link https://bodhi.fedoraproject.org/updates/FEDORA-2019-773ef3b70b F29 update 15:27:15 asamalik: yeah... let's not scare him :) 15:27:16 asamalik, ahh good 15:27:28 I mean, WP is not good :) but I can manage I guess 15:27:33 just write it.. when we can guide you through the publishing fun 15:27:38 s/when/then 15:27:43 or email langdon and he'll do that for you (T's and C's apply, I'm sure) 15:27:56 and tips.. don't forget tips 15:28:24 #action nils write a blog post and get it published to communityblog.fedoraproject.org 15:28:29 yes, include at least 20 % of the email again as a tip :D 15:28:34 "Did you know...?" 15:29:22 "Did you know, you too can wait 20 minutes to get a summary of the latest module builds from MBS?" :D 15:30:02 So, anything else? 15:30:45 nothing from my side 15:31:44 good 15:31:51 Thanks everybody! 15:31:55 thanks 15:31:59 #endmeeting