17:06:12 <Son_Goku> #startmeeting FESCO (2023-11-09)
17:06:12 <zodbot> Meeting started Thu Nov  9 17:06:12 2023 UTC.
17:06:12 <zodbot> This meeting is logged and archived in a public location.
17:06:12 <zodbot> The chair is Son_Goku. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
17:06:12 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:06:12 <zodbot> The meeting name has been set to 'fesco_(2023-11-09)'
17:06:19 <Son_Goku> #meetingname fesco
17:06:19 <zodbot> The meeting name has been set to 'fesco'
17:06:32 <Son_Goku> #chair nirik, decathorpe, zbyszek, sgallagh, mhroncok, dcantrell, mhayden, Conan_Kudo, Pharaoh_Atem, Son_Goku, King_InuYasha, Sir_Gallantmon, Eighth_Doctor, tstellar
17:06:32 <zodbot> Current chairs: Conan_Kudo Eighth_Doctor King_InuYasha Pharaoh_Atem Sir_Gallantmon Son_Goku dcantrell decathorpe mhayden mhroncok nirik sgallagh tstellar zbyszek
17:06:35 <decathorpe> oof has the meeting shifted to one hour earlier or is the US still on DST?
17:06:44 <Son_Goku> no, we're not on DST anymore
17:06:47 <Son_Goku> that's why it shifted
17:06:57 <Son_Goku> this meeting is pegged to UTC instead of a US timezone, so it floats for Americans
17:06:57 <decathorpe> so ... one hour earlier? got it ...
17:07:11 <Son_Goku> #topic init process
17:07:23 <decathorpe> now it's concurrent with FPC meeting, yay
17:07:27 <nirik> yeah, which isn't great for me, but oh well, we should perhaps check for a new time for everyone.
17:07:29 <Son_Goku> ugh joy
17:07:36 <Son_Goku> well we need to move this to Tuesday anyway
17:07:44 * Son_Goku needs to make that ticket
17:08:21 <Son_Goku> .hello ngompa
17:08:22 <decathorpe> btw, I'm connected via IPoAC so don't worry if I drop off
17:08:22 <zodbot> Son_Goku: ngompa 'Neal Gompa' <ngompa13@gmail.com>
17:09:27 <zbyszek> .hello2
17:09:28 <zodbot> zbyszek: zbyszek 'Zbigniew Jędrzejewski-Szmek' <zbyszek@in.waw.pl>
17:09:37 <zbyszek> Sorry for being late.
17:10:41 <zbyszek> Son_Goku: I floats for most people, as long as DST is still a thing.
17:11:15 <nirik> decathorpe: IP over air conditioning?
17:11:33 <Son_Goku> at least we're not Australia, who decided to split timezones on both longitude and latitude
17:12:10 <Son_Goku> and they've got funky half hour offsets and hour offsets
17:12:24 <zbyszek> Or Canada, based on percentage of French speakers
17:12:33 <Son_Goku> lol
17:12:46 <Son_Goku> is it just us three? then we may not have enough for quorum
17:12:49 <zbyszek> https://en.wikipedia.org/wiki/Time_in_Canada#/media/File:Canada_time_zone_map_-_en.svg
17:13:24 <decathorpe> nirik: https://datatracker.ietf.org/doc/html/rfc1149
17:14:05 <nirik> ah right. ;)
17:16:35 <Son_Goku> I just pinged fesco folks in devel on Matrix to remind people that it is in fact now
17:17:01 <tstellar> .hello tstellar
17:17:02 <zodbot> tstellar: tstellar 'Tom Stellard' <tstellar@redhat.com>
17:17:12 <Son_Goku> we only have one topic that involves carlwgeorge and sergiomb
17:17:25 <tstellar> I was mixed up by the time change.
17:17:31 <Son_Goku> so was I :)
17:18:09 <carlwgeorge> .hi
17:18:10 <zodbot> carlwgeorge: carlwgeorge 'Carl George' <carl@redhat.com>
17:18:21 <Son_Goku> I think with 5 (decathorpe, zbyszek, nirik, tstellar, me) we have quorum
17:19:12 <nirik> timechange--
17:20:10 <Son_Goku> #topic #3089 retiring redhat-lsb in Fedora
17:20:14 <Son_Goku> .fesco 3089
17:20:15 <zodbot> Son_Goku: Issue #3089: retiring redhat-lsb in Fedora - fesco - Pagure.io - https://pagure.io/fesco/issue/3089
17:21:33 <Son_Goku> #info We have (+5, 0, -2) for in-ticket voting to retire redhat-lsb
17:22:37 <nirik> I support retiring it, but I wish we could convince the maintainer that that is the right thing to do.
17:22:40 <zbyszek> To be clear, I don't disagree with the retiring, I just don't think that FESCo should tell a maintainer to retire a package like that.
17:22:43 <zbyszek> Yeah.
17:23:27 <Son_Goku> My point of view on this is that a package that's about the distribution's conformance to a standard (thus making a statement for multiple packages or a collection of packages) is a special corner case where we *do* have to say something one way or another
17:23:34 <sgallagh> Sorry folks. Time zone issues. I’m here
17:23:46 <Son_Goku> because it has implications for other packagers' work too
17:24:28 <sgallagh> I’m with Son_Goku here. For most packages, I’d leave it to the maintainer
17:24:54 <Son_Goku> and to be fully clear, I don't like this either
17:25:09 <zbyszek> Sure, but if we say that we conform to LSB-5.0, but we actually don't have support for tape archives or carrier pidgeons, does anybody care?
17:25:13 <Son_Goku> but Fedora _cannot_ be LSB compliant as the standard exists today
17:26:22 <carlwgeorge> LSB 5.0 requirements aren't as antiquated as tape archives.  It's things like Python modules (which have been removed from the standard library and don't exist elsewhere).
17:26:58 <tstellar> Son_Goku: Can the tool report that Fedora isn't compliant or does the presence of the tool make people think Fedora is compliant.
17:27:35 <zbyszek> tstellar: yes, it can report noncompliance. The other lsb_release does that.
17:27:58 <Son_Goku> as zbyszek said, the version I packaged explicitly does
17:28:16 <zbyszek> OK, so would there be support to say that lsb_archive (all versions) MUST report lack of compliance? (i.e. just print 'n/a' afaik)
17:28:54 <Son_Goku> lsb_release, but sure
17:29:08 <zbyszek> Yeah, this needs to be phrased better.
17:29:25 <carlwgeorge> I believe that would be better than nothing, but we will still run into instances (and bug reports) from people that think the mere installation of the package means LSB compliance.
17:30:00 <Son_Goku> and I believe the continued existence of the LSB package in Fedora is a bit of an X/Y problem
17:30:27 <Son_Goku> people assume the only way to resolve all the bugs around lsb_release(1) missing is to have the full LSB package, because that's how it was always done in Fedora (though nobody else did it that way)
17:30:58 <nirik> so I guess the maintainer wasn't able to make it today. ;(
17:31:39 <Son_Goku> I was tempted to go look for the bug reports asking for lsb-release to be split out of the main LSB package, but I didn't feel like searching through bugzilla for that
17:32:07 <Son_Goku> I know they exist because when I worked on some projects a few years ago to replace lsb_release(1) with os-release(5), I found comments about it
17:36:03 <Son_Goku> nirik: I pinged the maintainer via the fesco ticket (as well as carlwgeorge) and Bcc'd them to the meeting announcement email
17:37:21 <Son_Goku> as it currently stands, with no new information from the maintainer or anyone else, I'm going to let the voting in the ticket stand
17:38:27 <zbyszek> Quick question: is it possible to conform to some parts of LSB, or our current lack of things prevents any type of compliance?
17:39:30 <nirik> I'm not sure, but I don't think there's a 'partial compliance' path... it's either compliant or not.
17:39:58 <nirik> (but I could be wrong, it's been a long time since I last looked)
17:40:06 <Son_Goku> there is no partial compliance path last I checked
17:40:16 <zbyszek> Proposal: lsb_release (all implementations) must not report compliance with LSB (because various components are missing from Fedora, so compliance is not possible).
17:40:28 <carlwgeorge> There are different "modules" of compliance, but many of the flaws I found were in the core module, which I assume the other modules sort of require.
17:40:34 <Son_Goku> carlwgeorge: they do
17:40:41 <sgallagh> zbyszek: +1
17:40:56 <nirik> sure, +1
17:40:58 <Son_Goku> if you strip all the modules out of the redhat-lsb package, you're left with nothing
17:41:30 <Son_Goku> but sure, I guess
17:41:31 <carlwgeorge> Except `/usr/bin/lsb_release`, which is what the lsb_release package is :)
17:41:35 <Son_Goku> +1
17:41:37 <Son_Goku> right
17:41:49 <decathorpe> +1
17:42:13 <Son_Goku> lsb_release is a fork of the lsb_release script from the dead LSB upstream that uses os-release and explicitly doesn't report LSB compliance unless some kind of LSB configuration declares it
17:42:56 <tstellar> +1
17:43:41 <zbyszek> Jesus, ISO distributes the standards as a zip file. Should I print it out to read?
17:43:49 <Son_Goku> lol
17:44:15 <carlwgeorge> https://refspecs.linuxfoundation.org/LSB_5.0.0/LSB-Core-generic/LSB-Core-generic/book1.html
17:44:17 <Son_Goku> zbyszek: we should probably explicitly state that LSB modules that cannot be satisfied must not be shipped
17:44:20 <nirik> zbyszek: only if you have a dot matrix printer. ;)
17:45:01 <Son_Goku> while our guidelines actually kind of cover this already with the "it should work" statement, the crux of the problem around this package is that the maintainer seems to want to ignore it
17:45:36 <Son_Goku> this whole thing kicked off with https://bodhi.fedoraproject.org/updates/FEDORA-EPEL-2023-336dbb57e0#comment-3237125
17:45:43 <carlwgeorge> There are also other "books" that are architecture dependent, and I don't know if they specify additional requirements or not.
17:45:53 <carlwgeorge> And of course multiple that by all the other modules.
17:46:11 <carlwgeorge> https://refspecs.linuxfoundation.org/LSB_5.0.0/index.html
17:47:08 <zbyszek> We're at +6.
17:48:03 <Son_Goku> I think we need an explicit proposal that LSB modules that we can't comply with must be removed
17:48:11 <carlwgeorge> +6 for all implementations of `/usr/bin/lsb_release` not claiming LSB compliance, correct?
17:48:35 <Son_Goku> yes
17:49:08 <Son_Goku> #agreed lsb_release (all implementations) must not report compliance with LSB, because various components are missing from Fedora, so compliance is not possible. (+6, 0, 0)
17:49:13 <Son_Goku> I also have a proposal
17:49:15 <carlwgeorge> That's a fine first step, but doesn't cover the incorrect and misleading summaries/descriptions in the redhat-lsb package, or the misleading and arguably incorrect package name.
17:50:00 <zbyszek> carlwgeorge: Yeah, the package %description should be updated too.
17:50:18 <Son_Goku> Proposal: No implementation of an LSB package should expressly state or offer compliance for any LSB module that Fedora does not or cannot comply with.
17:50:43 <carlwgeorge> Also if the idea is that having a best effort lsb package is fine, it should comply as close as possible, and so far the maintainer has been unwilling to correct the correctable flaws I've identified.
17:51:20 <sgallagh> Straw man proposal: "Fedora explicitly declines to support the LSB 5.0 or earlier. Packagers will remove any information that implies otherwise."
17:51:45 <Son_Goku> sgallagh: we can add that to mine? I think it works for this
17:52:00 <sgallagh> WFM
17:52:24 <Son_Goku> Proposal: Fedora explicitly declines to support the LSB 5.0 or earlier. Packagers will remove any information that implies otherwise. No implementation of an LSB package should expressly state or offer compliance for any LSB module that Fedora does not or cannot comply with.
17:52:43 <sgallagh> +1
17:52:47 <Son_Goku> +1
17:53:23 <tstellar> +1
17:53:30 <nirik> +1
17:53:34 <zbyszek> +1
17:53:41 <mhroncok_web> hello folks, I am sorry for not making it to the meeting in time
17:53:53 <Son_Goku> no worries
17:54:01 <Son_Goku> we're voting on a proposal, I'll repaste for you
17:54:05 <Son_Goku> Proposal: Fedora explicitly declines to support the LSB 5.0 or earlier. Packagers will remove any information that implies otherwise. No implementation of an LSB package should expressly state or offer compliance for any LSB module that Fedora does not or cannot comply with.
17:54:12 <sgallagh> mhroncok_web: I'd berate you, if not for my own tardiness :-)
17:54:15 <decathorpe> +1
17:55:37 <mhroncok_web> Forgive my ignorance, but what does "no <package> should <do something>"?
17:55:53 <mhroncok_web> is it forbidden or frowned upon?
17:56:07 <sgallagh> mhroncok_web: Currently, the LSB package in Fedora falsely includes descriptions that imply that we comply with things we cannot
17:56:07 <Son_Goku> forbidden
17:56:22 <sgallagh> s/should/may/
17:56:34 <Son_Goku> s/should/must/
17:56:44 <Son_Goku> err
17:56:45 <sgallagh> must doesn't work in that sentence construction
17:56:45 <Son_Goku> may
17:56:49 <Son_Goku> you're right
17:57:02 <Son_Goku> I'll swap should for may after voting
17:57:23 <mhroncok_web> in rality, this means the redhat-lsb-thing can stay but needs to be patched to say nay for a lot of things?
17:57:42 <mhroncok_web> *reality
17:58:05 <mhroncok_web> I guess +1 on principle, but I wonder who writes that patch :D
17:58:59 <sgallagh> If no one writes the patch, we fall back to the original plan of forcibly retiring it, I think?
17:59:13 <Son_Goku> yep
17:59:18 <Son_Goku> #agreed Fedora explicitly declines to support the LSB 5.0 or earlier. Packagers will remove any information that implies otherwise. No implementation of an LSB package may expressly state or offer compliance for any LSB module that Fedora does not or cannot comply with. (+6, 0, 0)
17:59:28 <Son_Goku> I hope that's not too long for IRC logging
18:00:45 <mhroncok_web> what are the practical implications of this? we wait and if nothing changes, another fesco ticket is opened?
18:00:56 <Son_Goku> if nothing changes, we will ask releng to retire the package
18:01:12 <sgallagh> I think we just leave this ticket open for a bit, since it originally called for retirement
18:01:15 <sgallagh> We can reus eit
18:01:18 <Son_Goku> yup
18:01:22 <zbyszek> OK.
18:01:50 <Son_Goku> so I think that's it for this?
18:01:50 <carlwgeorge> Proposal: redhat-lsb be renamed to lsb and go through a package review.
18:01:54 <Son_Goku> ah
18:01:55 <Son_Goku> well
18:02:02 <Son_Goku> I don't have a problem with this, actually
18:02:22 <zbyszek> -1
18:02:23 <Son_Goku> IIRC, most distributions called the package lsb and did subpackaging for each component of conformance or tooling
18:02:42 <carlwgeorge> This is dual purpose.  The name is misleading, and there are correctable flaws to get closer to LSB compliance, which is worthwhile if the package is worth keeping.
18:02:45 <zbyszek> If the package should be renamed, this can done at the binary package level.
18:03:01 <nirik> only-parts-of-lsb-we-happen-to-still-support
18:03:19 <Son_Goku> lol
18:03:30 <Son_Goku> lsb-swiss-cheese
18:03:49 <mhroncok_web> I don't care what it is called, but I guess Red Hat has the power to say it cannot be called redhat
18:04:02 <Son_Goku> that's true
18:04:15 <tstellar> Has the maintainer objected to renaming the package?
18:04:22 <Son_Goku> it has never been brought up
18:04:29 <zbyszek> Tell that to redhat-rpm-config ;)
18:04:36 <Son_Goku> there is a flip side, too
18:04:36 <carlwgeorge> To be clear in case there is any doubt or suspicion, I work for Red Hat, but I'm not asking for this on behalf of Red Hat or because anyone in Red Hat asked me to.
18:04:50 <tstellar> So maybe there should be a discussion with the maintainer first, before FESCO considers this.
18:04:50 <Son_Goku> redhat-lsb, like redhat-rpm-config, is API now
18:05:18 <Son_Goku> admittedly, I'm fine breaking this API since we already aren't conformant to whatever this package claims today
18:05:49 <Son_Goku> and IIRC, the redhat-lsb package has virtual names for lsb (without the redhat- prefix)
18:06:12 <Son_Goku> flipping it the other way makes it... less of a statement I guess
18:06:29 <mhroncok_web> I don't like brainstorming new ideas and voting for them on the same meeting, without giving time for the involved parties to speak up
18:06:46 <Son_Goku> I'm fine with punting this to devel@
18:06:51 <Son_Goku> or the maintainer
18:07:08 <sgallagh> Yeah, I think with the guidance we provided on the agreed-upon proposal, we can take that back to the list and see where it leads
18:07:19 <nirik> +1
18:07:31 <smooge> i agree with mhroncok_web .. i think at this point it is time to move on
18:07:35 <Son_Goku> carlwgeorge: would you be willing to make a devel thread about it?
18:08:01 <Son_Goku> I'd like to say we're done with the things we needed to deal with on this ticket
18:08:08 <carlwgeorge> I'd prefer discourse if that's ok
18:08:17 * Son_Goku mumbles deprecations
18:08:22 <Son_Goku> sure I guess
18:08:58 <Son_Goku> with that, I'm calling this topic done and moving on
18:09:08 <Son_Goku> #topic Next week's chair
18:09:14 <Son_Goku> who wants to do it?
18:09:48 * Son_Goku hears crickets
18:09:49 <zbyszek> I can do it.
18:09:54 <Son_Goku> cool
18:10:05 <Son_Goku> #action zbyszek will chair next meeting
18:10:10 <Son_Goku> #topic Open Floor
18:10:28 <tstellar> What do we need to do for the elections?  The nomination period has started already.
18:10:31 <Son_Goku> I'll remind everyone that tomorrow is the Fedora Linux 39 Release Party! https://hopin.com/events/fedora-linux-39-release-party/registration
18:10:42 <Son_Goku> oh, do we have elections?
18:11:13 <mhroncok_web> The nomination period starts 2023-11-08
18:11:19 <mhroncok_web> https://fedorapeople.org/groups/schedule/f-39/f-39-elections-tasks.html
18:12:06 <Son_Goku> so yesterday
18:12:11 <sgallagh> We probably need to get Aoife working on that ASAP
18:12:11 <mhroncok_web> except it did not happen
18:12:13 <nirik> I think we need to ask if this is something that amoloney will process, of if we need to come up with others to do it.
18:12:26 <nirik> man, I can't type today.
18:12:51 <Son_Goku> yeah, I'm pretty sure this is under Aoife's remite
18:12:55 <Son_Goku> *remit
18:13:14 <Son_Goku> anyone want to poke her about it?
18:13:28 <sgallagh> I'll reach out
18:13:33 <Son_Goku> sounds good
18:13:57 <zbyszek> https://docs.fedoraproject.org/en-US/fesco/#fesco-members says 5 seats are up for election.
18:14:10 <Son_Goku> #action sgallagh will reach out to amoloney to get the elections going
18:15:26 * nirik nods
18:16:53 <Son_Goku> unrelated/related, release party schedule: https://fedoraproject.org/wiki/Fedora_Linux_39_Release_Party_Schedule
18:17:20 <Son_Goku> it's not in hopin, so if you wanted to know what's going on, that's the list
18:17:41 <Son_Goku> anything else anyone wants to bring up before I end it all?
18:18:03 <smooge> well that took a turn
18:18:05 <sgallagh> Son_Goku: Don't do it, you have so much to live for?
18:19:05 <Son_Goku> going once...
18:19:08 <zbyszek> Son_Goku: please end it ;)
18:19:08 <Son_Goku> going twice...
18:19:11 <Son_Goku> rip!
18:19:17 <nirik> If anyone knows anyone making massive queries against koji, could you tell them to stop doing that? ;)
18:19:24 <Son_Goku> 🪦
18:19:29 <Son_Goku> #endmeeting