16:00:32 <geppetto> #startmeeting fpc
16:00:32 <zodbot> Meeting started Thu Sep  6 16:00:32 2018 UTC.
16:00:32 <zodbot> This meeting is logged and archived in a public location.
16:00:32 <zodbot> The chair is geppetto. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:32 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:00:32 <zodbot> The meeting name has been set to 'fpc'
16:00:32 <geppetto> #meetingname fpc
16:00:32 <zodbot> The meeting name has been set to 'fpc'
16:00:32 <geppetto> #topic Roll Call
16:00:41 <ignatenkobrain> .hello2
16:00:42 <zodbot> ignatenkobrain: ignatenkobrain 'Igor Gnatenko' <i.gnatenko.brain@gmail.com>
16:00:42 <tibbs> Hello.
16:00:46 <geppetto> #chair ignatenkobrain
16:00:46 <zodbot> Current chairs: geppetto ignatenkobrain
16:00:49 <geppetto> #chair tibbs
16:00:49 <zodbot> Current chairs: geppetto ignatenkobrain tibbs
16:00:52 <mhroncok> hi. I'm on train. the internets can go down any second
16:00:56 <geppetto> Hey
16:01:00 <geppetto> You made it!
16:01:04 <mhroncok> partially
16:01:04 <geppetto> #chair mhroncok
16:01:04 <zodbot> Current chairs: geppetto ignatenkobrain mhroncok tibbs
16:01:40 <decathorpe> hello
16:01:49 <geppetto> #chair decathorpe
16:01:49 <zodbot> Current chairs: decathorpe geppetto ignatenkobrain mhroncok tibbs
16:02:05 <geppetto> Hey, 5 ppeeps :)
16:02:24 <geppetto> Give it a couple more minutes for anyone else, then I'll start
16:05:06 <tibbs> I am still pretty snowed under.  Plus I've been up since 3:30AM so I'm kind of fried.
16:05:41 <geppetto> #topic Schedule
16:05:42 <geppetto> https://lists.fedoraproject.org/archives/list/packaging@lists.fedoraproject.org/message/6J2UC2OWNLAN7KIZIZPTX7Y2DFD5FYWA/
16:05:52 <geppetto> #topic  #727 convert guidelines to git and restructured text
16:05:55 <geppetto> .fpc 727
16:05:57 <zodbot> geppetto: Issue #727: convert guidelines to git and restructured text - packaging-committee - Pagure - https://pagure.io/packaging-committee/issue/727
16:06:13 <ignatenkobrain> I have to vote on this
16:06:16 <ignatenkobrain> let me write proposal
16:06:28 <tibbs> I don't think this is even a voting thing.
16:06:32 * limburgher here-ish
16:06:48 <tibbs> It just needs to actually happen.
16:06:59 <ignatenkobrain> PROPOSAL: convert latest versions of wiki packages to new format, publish it, make wiki read-only and improve that new format over time
16:07:08 <tibbs> And then we all figure out what other policies and procedures need to change to accommodate.
16:07:20 <ignatenkobrain> unfortunately if we try to make it perfect before, it is not going to happen ever
16:08:14 <tibbs> If we really want to vote about it, +1 even though it means I have to learn yet another markup language.
16:08:37 <decathorpe> I don't even care if it will use asciidoc or rst, I don't know either ;)
16:08:43 <mhroncok> ok, +1
16:08:51 <decathorpe> +1
16:08:54 <mhroncok> I don't think we need to vote, but whatever
16:09:07 <geppetto> sure +1 then
16:09:15 <tibbs> It's sad that "yet another markup language" is actually itself a markup language.  That's different from the one we'd be using.
16:09:17 <ignatenkobrain> it's mostly to get agreement that we are not going to touch wiki and just do the new format
16:09:28 <limburgher> +1
16:09:55 <tibbs> Well I need to know the last point at which I should be doing writeups.  Because there is at least one pending.
16:10:13 <ignatenkobrain> tibbs: how much time do you need to finish it?
16:10:17 <mhroncok> I'm also currently drafting a proposal in a wiki
16:10:32 <mhroncok> but I'll gladly convert it to a pull request later
16:10:37 <tibbs> Was hoping to do it today since I notice that it got enough votes.
16:10:46 * geppetto nods
16:10:58 <ignatenkobrain> tibbs: ok, I will be doing conversion then on weekend
16:11:15 <tibbs> And proposals in the wiki should still be OK, though we then lose the ability to do diffs.
16:11:35 <geppetto> #action  convert latest versions of wiki packages to new format, publish it, make wiki read-only (+1:6, 0:0, -1:0)
16:11:48 <tibbs> This will make diffing in general rather worse but I think it's worth it.
16:11:56 <geppetto> why?
16:12:06 <tibbs> Word wrapping, for one.
16:12:33 <tibbs> RIght now you can reflow wiki content and the diff doesn't really change.
16:13:02 <tibbs> But that's mainly because of how the wiki constructs the diff; with plain old git you'll just see blocks of changes.
16:13:35 <geppetto> some of the UIs on top of git can show better than line based diffs
16:13:36 <tibbs> Fancier diff viewers will at least show you sub-line changes, though you can't see that in pagure.
16:13:38 <geppetto> but, yeh
16:13:56 <ignatenkobrain> I think there was some diffs on antora docs
16:14:00 <ignatenkobrain> I mean the nice ones
16:14:15 <decathorpe> will you be using antora?
16:14:17 <ignatenkobrain> but let's move on
16:14:28 <tibbs> My biggest fear, really, is that this will keep changing due to the whims of whoever is running the docs site.
16:14:32 <geppetto> Do I need/want to know what antora is?
16:14:45 <tibbs> asciibinder, antora, next week's flavor.
16:14:57 <ignatenkobrain> geppetto: that's the engine for docs.fp.o ;)
16:15:04 <tibbs> Not that I know what either of those is, really.
16:15:07 <ignatenkobrain> all the docs are using antora which are published there I believe
16:15:16 <ignatenkobrain> and I hope it is the last iteration of changing engines
16:15:29 <decathorpe> using the same thing for all docs would be nice
16:16:09 <geppetto> #topic #719 Simplify packaging of forge-hosted projects
16:16:12 <geppetto> .fpc 719
16:16:14 <zodbot> geppetto: Issue #719: Simplify packaging of forge-hosted projects - packaging-committee - Pagure - https://pagure.io/packaging-committee/issue/719
16:16:34 <tibbs> I need to merge that PR.
16:16:40 <geppetto> Ok, I'm not entirely sure what's happening here … are we just waiting on mhroncok's pull request?
16:17:03 <tibbs> From FPC's perspective, this needs actual documentation somewhere.
16:17:13 <tibbs> Most likely in Packaging:SourceURL.
16:17:44 <geppetto> Do we have a victim to write that?
16:17:49 <mhroncok> mine?
16:18:12 <decathorpe> I think nim would work on it
16:18:32 <tibbs> I was going to get around to it one day.
16:18:40 <mhroncok> the internets are not great, what pull request do i have?
16:18:56 <tibbs> nim still keeps mentioning that the way he says it's supposed to work... doesn't work for me.
16:19:27 <tibbs> But I am actually using it in packages and it works.
16:19:39 <geppetto> mhroncok: 26 against redhat-rpm-config
16:19:57 <geppetto> There's also a 34 PR by nim
16:19:58 <tibbs> The only pending PR is the one against redhat-rpm-config which I'm hoping to deal with today.
16:20:50 <geppetto> #action tibbs to deal with PR soon, hopefully today
16:20:53 <tibbs> I don't know how PR26 is related.  "Fix kernel ABI related scripts".
16:21:06 <geppetto> #info Need some policy for this, probably in Packaging:SourceURL
16:21:26 <geppetto> bah 24
16:21:40 <tibbs> There are three open PRs related to these macros which I hope to deal with.
16:21:43 <decathorpe> 24 has been merged
16:21:57 <tibbs> But they can't be backported.
16:22:01 <geppetto> Ahh, cool
16:22:14 <tibbs> Anyway, there is as always still work to be done.
16:22:35 <tibbs> But from FPC's perspective, what remains is mentioning this stuff in the guidelines.
16:22:42 * geppetto nods … but it all doesn't have to happen in this ticket :)
16:23:16 <tibbs> True.
16:24:06 <decathorpe> I don't think nim wants to update his draft
16:24:45 <geppetto> #topic  #740 Add "%python_enable_dependency_generator" onto Packaging:Python
16:24:49 <geppetto> .fpc 740
16:24:50 <zodbot> geppetto: Issue #740: Add "%python_enable_dependency_generator" onto Packaging:Python - packaging-committee - Pagure - https://pagure.io/packaging-committee/issue/740
16:26:16 <geppetto> ignatenkobrain: mhroncok You both know what is going on here
16:26:41 <mhroncok> the fiff is at https://pagure.io/packaging-committee/issue/740#comment-527526
16:26:43 <mhroncok> diff
16:26:53 <mhroncok> it jut needs a vote I guess
16:27:37 <geppetto> So: https://fedoraproject.org/w/index.php?title=User%3AChurchyard%2FPackaging%3APython&type=revision&diff=523884&oldid=523876
16:28:42 <ignatenkobrain> obviously +1 from me
16:28:49 <decathorpe> looks good
16:28:56 <decathorpe> +1
16:29:05 <limburgher> +1
16:29:13 <mhroncok> onsider me +1
16:29:14 <mhroncok> c
16:29:28 <geppetto> Not sure we need to say "All requires MUST be resolvable within the targeted Fedora version."
16:29:31 <geppetto> but maybe we do
16:29:33 <geppetto> +1
16:29:55 <decathorpe> well, if they don't resolve, the package isn't installable, so ... MUST is a given, really
16:30:27 <geppetto> yeh
16:30:33 <mhroncok> I added it becasue it can easily bite
16:31:21 * geppetto nods … would just be much better as an auto CI/QA thing than as a text warning :)
16:31:21 <tibbs> +1
16:31:48 <ignatenkobrain> one thing I would add is "do not disable this if it generates incorrect data, fix the setup.py or whatsoever first"
16:32:31 <geppetto> #chair limburgher
16:32:31 <zodbot> Current chairs: decathorpe geppetto ignatenkobrain limburgher mhroncok tibbs
16:32:51 <geppetto> ignatenkobrain: You want to add that somewhere quickly and we can vote on that now?
16:33:32 <ignatenkobrain> I believe we can fix it up later, it should be very simple. So we probably do not need to vote on this thing separately
16:33:40 <geppetto> Assuming nobody else has any objections to that wording?
16:34:10 <mhroncok> ignatenkobrain: should not or must not?
16:35:00 <ignatenkobrain> mhroncok: must not I would say. because we want to automate dependency generation.. it's like we ask people to not link to libfoo.so rather than messing up with __requires_exclude
16:35:09 <tibbs> Well technically you can do what is done for Perl and just filter errant dependencies.
16:35:25 <mhroncok> i'm -1 to must not
16:35:29 <mhroncok> not ready yet
16:35:41 <mhroncok> I'd rather have this widely adopted fisrt
16:35:45 <mhroncok> and later we can forbid that
16:36:24 <ignatenkobrain> should not works for me too
16:36:25 <decathorpe> sounds good
16:36:37 <tibbs> I've forgotten; is this enabled by default on rawhide now?
16:36:48 <ignatenkobrain> not yet
16:36:49 <ignatenkobrain> but it will be for f30
16:37:04 <tibbs> The sooner the better assuming the feature has been approved.
16:37:06 <ignatenkobrain> I will submit proposal over the weekend
16:37:42 <mhroncok> +1 to should not
16:38:17 <decathorpe> +2
16:39:35 <geppetto> Ok
16:39:42 <geppetto> #topic Open Floor
16:39:47 <tibbs> I mean, I guess we can add something like that.
16:39:59 <tibbs> Seems dumb to say "don't turn this on if it doesn't work", though.
16:40:15 <mhroncok> what's the conclusion?
16:41:17 <ignatenkobrain> my point was mostly, if package has requires foo in setup.py (so the depgen will pick it up) and user doesn't want to see it... it will fail later if you omit the dependency
16:41:48 <mhroncok> unless it is bogus
16:41:52 <tibbs> Well, it will fail later if you don't filter the dependency.
16:42:00 <mhroncok> like upstreams when it adds specific stuuf for raspberyr pi if on arm
16:42:09 <tibbs> That's no different than all of the other automatic dependency generators we have currently.
16:42:14 <mhroncok> or when upstream repends on rpm-python-installer or how is it called
16:42:35 <mhroncok> i'd rather have an ability to explcitly opt out
16:42:50 <tibbs> RPM does give that to you.
16:43:41 <mhroncok> you mean filtering?
16:43:56 <tibbs> Both filtering and the ability to decide which dependency generators to run.
16:44:07 <tibbs> Or just disabling the automatic generation altogether if that's what you need.
16:44:34 <mhroncok> I just need %disable_python_generator
16:44:44 <tibbs> The point of these things is to save packagers time by not having to manually specify deps, and to help keep people from screwing it up.
16:44:53 <mhroncok> yes
16:44:58 <mhroncok> I agree
16:45:01 <mhroncok> nd I loe it
16:45:05 <mhroncok> and I love it
16:45:27 <tibbs> Whether there's a convenience macro for opting out... I don't really care about.
16:45:53 <mhroncok> and whether it is allowed to use it?
16:46:16 <tibbs> Would be nice if they existed and were consistent for all of the dep generators, but that's not really within FPC's scope.
16:50:10 <tibbs> I will try to get to all of the writeups today or tomorrow and get an announcement out so that the conversion to asciidoc can happen.
16:50:22 <geppetto> Ok
16:50:30 <geppetto> Anything else anyone wants to talk about?
16:50:34 <tibbs> I forget if this will preserve history or not.  I do hope so but I know it's a tough issue.
16:50:57 <mhroncok> it will
16:51:02 <geppetto> Cool
16:51:07 <mhroncok> there will be markup chnages on top
16:51:10 <limburgher> moin-blame
16:51:11 <mhroncok> so blame will not work
16:51:23 <mhroncok> but the history will be there
16:51:26 <tibbs> Ah, moin.
16:51:36 <limburgher> Right?
16:51:49 <tibbs> When we converted from that we lost the wiki history and that still causes problems occasionally.
16:51:52 <limburgher> The days of moin and plague and cvs. . .
16:52:22 <mhroncok> one more heads up
16:52:34 <mhroncok> as part of https://fedoraproject.org/wiki/Changes/Mass_Python_2_Package_Removal
16:52:44 <mhroncok> I'm drafitng https://fedoraproject.org/wiki/User:Churchyard/Python2Drop/Packaging:Python
16:52:51 <mhroncok> the appendix chnage is misisng now
16:53:27 <tibbs> FYI https://fedoraproject.org/w/index.php?title=User%3AChurchyard%2FPython2Drop%2FPackaging%3APython&type=revision&diff=524397&oldid=524390 is the diff
16:53:51 <tibbs> Seems reasonable at first glance.
16:54:24 <mhroncok> it's also just a first draft of the draft, I'll skim trough it once I'm by a real display and keyboard
16:56:28 <mhroncok> it seems the meeting is over...
16:57:24 * decathorpe shrugs
16:58:22 <geppetto> mhroncok: Not over quite yet. Did you want to discuss it?
16:58:29 <geppetto> mhroncok: Or wait until next week?
16:58:49 <geppetto> Should close in a couple of minutes though
16:58:51 <mhroncok> I just wanted to gives you aheads up that I'm working on this
16:58:57 <geppetto> Ok, no problem
16:59:11 <geppetto> I shall close then … see everyone next week.
16:59:15 <geppetto> #endmeeting