15:00:53 <asamalik> #startmeeting modularity
15:00:53 <zodbot> Meeting started Tue Apr  9 15:00:53 2019 UTC.
15:00:53 <zodbot> This meeting is logged and archived in a public location.
15:00:53 <zodbot> The chair is asamalik. 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'
15:00:53 <asamalik> #meetingtopic Weekly Meeting of the Modularity Team
15:00:53 <asamalik> #topic Roll Call
15:00:53 <asamalik> #chair sgallagh langdon contyk ignatenkobrain
15:00:53 <zodbot> Current chairs: asamalik contyk ignatenkobrain langdon sgallagh
15:01:08 <langdon> .hello2
15:01:09 <zodbot> langdon: langdon 'Langdon White' <langdon@redhat.com>
15:01:11 <contyk> .hello psabata
15:01:12 <zodbot> contyk: psabata 'Petr Šabata' <psabata@redhat.com>
15:01:15 * sgallagh is here but split attention. Ping me directly if needed.
15:01:30 <asamalik> langdon: see? I'm chairing people now :P
15:01:41 <asamalik> sgallagh: ack!
15:01:47 <asamalik> .hello2
15:01:48 <zodbot> asamalik: asamalik 'Adam Samalik' <asamalik@redhat.com>
15:01:49 <langdon> asamalik++ \o/
15:01:52 <sgallagh> .hello2
15:01:53 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
15:02:40 <asamalik> #topic Agenda
15:02:40 <asamalik> #info 128 Discussion: naming common streams and profiles
15:02:44 <asamalik> anything else?
15:03:08 <langdon> asamalik++
15:03:15 <langdon> zodbot is ignoring me!
15:03:30 <asamalik> langdon: I think you can only do that once per release :(
15:03:35 * asamalik wants more cookies!
15:03:42 <asamalik> rations!
15:03:52 <langdon> asamalik: i thought zod told you if you had done it this release though
15:04:03 <asamalik> langdon: good point
15:04:34 <asamalik> #topic 128 Discussion: naming common streams and profiles
15:04:34 <asamalik> #link https://pagure.io/modularity/issue/128
15:05:10 <asamalik> So we discussed this on the mailing lists and I feel like we agreed there is a need for the two distinct "rolling" streams
15:05:23 <asamalik> *list
15:05:49 <asamalik> contyk: you had some doubts last time, what about now? :)
15:06:52 <contyk> well
15:07:00 <contyk> :)
15:07:23 <contyk> I'm fine with what you want to do
15:08:25 * asamalik waits for the but
15:08:37 <langdon> ha
15:08:37 <contyk> no buts
15:09:32 <asamalik> ok :)
15:10:17 <asamalik> there was also a good question from mattdm if we want to a) be closer to the upstreams with our branching or 2) consistent across the distro
15:10:24 <langdon> asamalik: do you have a pointer to the ml thread? i can't seem to find it atm
15:10:31 <asamalik> and my goal is the 2)
15:10:46 <langdon> just found it on the ticket
15:10:57 <asamalik> langdon: it's .. yes :)
15:12:11 <asamalik> ... and that's why I'm proposing those two streams, and giving packagers guidance about how to approach it
15:12:45 <asamalik> sgallagh: time to ping you! I'd appreciate you opinion on the last ~6 messages
15:13:40 <sgallagh> I thought the point here was to be consistent when upstreams are unclear.
15:14:07 <sgallagh> Like, if upstream has obvious names that would make sense to their user-base, that should always be preferred.
15:14:18 <sgallagh> When that's not the case, these are the fallback recommendations for consistency
15:14:25 <asamalik> sgallagh: right... and to have consistent naming in Fedora, not copying upstrams' names for these two particular cases
15:14:32 <sgallagh> ack
15:15:16 <asamalik> so if the upstream has a very clear "stable" branch, we'd call it "rolling" or whatever in fedora because that's how we help our users to identify what that means
15:16:49 <asamalik> any ideas on how to have a productive naming discussion? :)
15:17:37 <sgallagh> Hmm, I'm not actually sure I agree with that last.
15:17:41 <langdon> asamalik: go in to a different field?
15:18:05 <sgallagh> If upstream has a well-known set of releases called "stable"... we should probably let people stick with that in Fedora.
15:19:28 <asamalik> sgallagh: couldn't that result in the situation we're in now?
15:19:45 <sgallagh> yes
15:19:59 <sgallagh> 🤷
15:20:19 <sgallagh> I think we'll find that our packagers will just do that anyway
15:20:25 <sgallagh> So we might as well bless it.
15:20:32 <langdon> the problem with that is matthew's point.. "stable" in upstream means "blah" but fedora chose "stable" as a default name when there isn't a "stable" and those terms conflict
15:20:39 <sgallagh> And just cover those cases where no clear upstream preference exists
15:21:13 <sgallagh> It's impossible to avoid naming collisions altogether.
15:21:24 <sgallagh> We just need to pick the ones we think matter (and won't just be ignored)
15:21:34 <asamalik> would there be a point in standardizing anything? it would basically mean that "those standard names sometimes, and in some cases only, mean this"
15:21:37 <langdon> true.. stupid question, what is the rpm policy when there is no version upstream?
15:22:23 <sgallagh> langdon: SCM versioning
15:22:30 <langdon> i would also point out.. this is why there is a decription field.. so we don't HAVE to encode it in the name
15:22:41 <asamalik> langdon: yeah that's the thing...
15:22:43 <sgallagh> Fair point
15:23:09 <contyk> but the more you can deduce from NS the better
15:23:17 <asamalik> maybe I'm trying to shave this yak a bit more than necessary and we can just rely on the summary/description fields
15:23:42 <asamalik> contyk: defintiely agree.. but I'd say either enforce it for all modules or not enforce it at all
15:23:47 <langdon> i guess that is what i am getting at.. perhaps more focus on making "good descriptions" a policy..
15:26:10 <asamalik> ok... proposal:
15:27:45 <asamalik> there will be those two streams named "rolling" for the user-focused builds and "unstable" for the preview/master/pre-release things, and this will be just a suggestion to the packagers, but using whatever is established upstream is also a good option
15:28:22 <asamalik> and we'll not rename nor enforce renaming of anything, we'll leave it to packagers
15:28:50 <asamalik> and encourage a good use of the description/summary field for this purpose
15:28:54 <langdon> but we WILL enforce a good description of the stream?
15:30:08 <asamalik> langdon: or even that if we figure out criteria.. or we can just mention it and file bugs if we see something confusing
15:30:15 <sgallagh> langdon: How do we enforce that?
15:31:16 <langdon> 2 ways: 1) just state that it is extremely important 2) file bugs a la asamalik's remark.. so not really "enforce" but it should be a very strongly worded statement rather than "you are encouraged"
15:32:11 <sgallagh> Yeah, we can phrase it as "Stream description MUST include an indication of the stream's API/ABI compatibility policy"
15:33:01 <langdon> yeah.. exactly
15:33:02 * sgallagh suggests we make that a general guideline, not just for "rolling"/"unstable"
15:33:11 <langdon> sgallagh: +1
15:33:11 <asamalik> sgallagh: +1
15:35:02 <asamalik> contyk: what do you think about that?
15:35:12 <asamalik> (proposal ~10 messages above)
15:35:46 <contyk> not sure!
15:36:03 <contyk> so I abstain
15:36:49 <asamalik> yet another signal of yak shaving :D
15:37:00 <contyk> I don't really care
15:37:20 <contyk> as long as you provide some clear guidance on how my streams should be named going forward
15:37:28 <contyk> because it's just getting more and more confusing :)
15:38:03 <sgallagh> contyk: Call them "Alice", "Bob" and "Charlie"...
15:38:10 <asamalik> contyk: that's actually the point here :) if we want to keep it confusing and just have a fallback, or if we want to enforce consistency.. that's all :)
15:38:34 <contyk> I know
15:38:44 <asamalik> not really about the names that much
15:38:49 <sgallagh> asamalik: My point here is that we *can't* enforce consistency. Packagers will just ignore it when it disagrees with their upstream
15:38:58 <asamalik> sgallagh: and I agree with you
15:39:25 <asamalik> we could also just stop here and keep what we have now
15:41:28 <sgallagh> https://i.redd.it/5iwxlrym6tp01.gif
15:42:00 <asamalik> sgallagh++
15:43:05 <contyk> :))
15:44:19 <sgallagh> Ah, here's the one I really intended for that reply: https://i.pinimg.com/originals/36/47/c4/3647c47ebdb95ecadffc801bfc965618.jpg
15:45:13 <langdon> lol
15:45:18 <asamalik> so we just agree that we don't care anymore and just leave it as is? :)) </half-sarcastic>
15:45:54 <sgallagh> asamalik: I think I'm saying that we make a recommendation that they can use if they don't have a clear idea what it should be.
15:48:10 <asamalik> sgallagh: agree
15:48:13 <asamalik> #proposal There will be those two streams named "rolling" for the user-focused builds and "unstable" for the preview/master/pre-release things, and this will be just a suggestion to the packagers, but using whatever is established upstream is also a good option. We'll not rename nor enforce renaming of anything, we'll leave it to packagers and strongly encourage a good use of the description summary field for this
15:48:14 <asamalik> purpose.
15:48:15 <asamalik> +1
15:48:48 <asamalik> contyk sgallagh langdon ^
15:48:58 <sgallagh> s/strongly encourage/mandate/
15:49:21 <sgallagh> (even if "enforcement" is just "file a bug")
15:49:36 <asamalik> sgallagh: +!
15:49:38 <asamalik> +1
15:49:44 <langdon> +1
15:50:08 * contyk looks
15:50:31 <sgallagh> +1
15:50:34 <contyk> sounds reasonable
15:50:36 <contyk> +1
15:50:43 <asamalik> #agreed There will be those two streams named "rolling" for the user-focused builds and "unstable" for the preview/master/pre-release things, and this will be just a suggestion to the packagers, but using whatever is established upstream is also a good option. We'll not rename nor enforce renaming of anything, we'll leave it to packagers and mandate a good use of the description summary field for this purpose. +4 0
15:50:43 <asamalik> -0
15:50:53 <asamalik> ah no!
15:51:49 <langdon> asamalik: are you typing?>
15:51:49 <asamalik> #undo
15:51:49 <zodbot> Removing item from minutes: AGREED by asamalik at 15:50:43 : There will be those two streams named "rolling" for the user-focused builds and "unstable" for the preview/master/pre-release things, and this will be just a suggestion to the packagers, but using whatever is established upstream is also a good option. We'll not rename nor enforce renaming of anything, we'll leave it to packagers and mandate a good use of the description summary field for this purpose. +4 0
15:51:54 <asamalik> #agreed There will be two streams named "rolling" for the user-focused builds and "unstable" for the preview/master/pre-release things, and this will be just a suggestion to the packagers, but using whatever is established upstream is also a good option. We'll not rename nor enforce renaming of anything, we'll leave it to packagers and mandate a good use of the description summary field for this purpose. +4 0 -0
15:52:07 <asamalik> langdon: not anymore
15:52:13 <langdon> ha
15:52:41 <langdon> proposed edit: There will be two streams named "rolling" for the user-focused builds and "unstable" for the preview/master/pre-release things, and this will be just a suggestion to the packagers, but using whatever is established upstream is also a good option. We'll not rename nor enforce renaming of anything, we'll leave it to packagers and mandate the description summary field for this purpose.
15:52:58 <sgallagh> too late
15:52:59 <langdon> ooh.. well.. this isn't the final text right?
15:53:02 <langdon> just hte agreed?
15:53:05 <sgallagh> Correct
15:53:11 <asamalik> langdon: right
15:53:12 <langdon> so never mind
15:53:16 <asamalik> that won't go to the docs
15:53:42 <asamalik> #topic Next meeting's chair
15:54:22 <asamalik> any volunteers?
15:55:30 <sgallagh> I'm on PTO next week
15:56:04 <asamalik> ok I'll take it again, but there are at least two people on PTO so we might just skip it
15:56:16 <langdon> ha
15:57:06 <asamalik> #action asamalik to chair next meeting
15:57:09 <asamalik> ok that's it I guess
15:57:13 <asamalik> #topic Open floor
15:57:17 <asamalik> leaving this here for a minute
15:58:15 * sgallagh has nothing
15:58:32 <asamalik> #endmeeting