19:00:16 <zbyszek> #startmeeting FESCO (2021-08-16)
19:00:16 <zodbot> Meeting started Mon Aug 16 19:00:16 2021 UTC.
19:00:16 <zodbot> This meeting is logged and archived in a public location.
19:00:16 <zodbot> The chair is zbyszek. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:00:16 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
19:00:16 <zodbot> The meeting name has been set to 'fesco_(2021-08-16)'
19:00:16 <zbyszek> #meetingname fesco
19:00:16 <zodbot> The meeting name has been set to 'fesco'
19:00:16 <zbyszek> #chair nirik, decathorpe, zbyszek, sgallagh, mhroncok, dcantrell, defolos, mboddu, Conan_Kudo, Pharaoh_Atem, Son_Goku, King_InuYasha, Sir_Gallantmon, Eighth_Doctor
19:00:16 <zbyszek> #topic init process
19:00:16 <zodbot> Current chairs: Conan_Kudo Eighth_Doctor King_InuYasha Pharaoh_Atem Sir_Gallantmon Son_Goku dcantrell decathorpe defolos mboddu mhroncok nirik sgallagh zbyszek
19:00:19 <zbyszek> .hello2
19:00:20 <zodbot> zbyszek: zbyszek 'Zbigniew Jędrzejewski-Szmek' <zbyszek@in.waw.pl>
19:00:28 <sgallagh> .hello sgallagh
19:00:29 <Eighth_Doctor> .hello ngompa
19:00:29 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
19:00:32 <zodbot> Eighth_Doctor: ngompa 'Neal Gompa' <ngompa13@gmail.com>
19:00:33 <bcotton> .hello2
19:00:35 <zodbot> bcotton: bcotton 'Ben Cotton' <bcotton@redhat.com>
19:00:39 * Eighth_Doctor is getting a sandwich
19:00:41 <mhroncok> .hello churchyard
19:00:43 <zodbot> mhroncok: churchyard 'Miro Hrončok' <mhroncok@redhat.com>
19:00:43 <nirik> morning
19:00:46 <nirik> .hello kevin
19:00:47 <zodbot> nirik: kevin 'Kevin Fenzi' <kevin@scrye.com>
19:01:38 <zbyszek> We have quorum. Do we expect anyone else?
19:01:47 <dcantrell> .hello2
19:01:48 <zodbot> dcantrell: dcantrell 'David Cantrell' <dcantrell@redhat.com>
19:01:53 <asosedkin> what's this ritual?
19:02:12 <zbyszek> asosedkin: thanks for joining. I was looking up your nick to ping you ;)
19:02:26 <zbyszek> dwmw2: we'll be discussing 2659 next.
19:02:37 <mhroncok> asosedkin: it summons the good spirits of IRC to be with us
19:02:41 <decathorpe> .hello2
19:02:44 <zodbot> decathorpe: decathorpe 'Fabio Valentini' <decathorpe@gmail.com>
19:02:58 <zbyszek> #topic #2659 Arbitration request: Crypto policy prevents VPN connections
19:02:58 <zbyszek> .fesco 2659
19:03:04 <zodbot> zbyszek: Issue #2659: Arbitration request: Crypto policy prevents VPN connections - fesco - Pagure.io - https://pagure.io/fesco/issue/2659
19:03:05 <mhroncok> asosedkin: it's for the chair to know if we have quorum and for the logs to be clear who's who
19:03:30 <mhroncok> as usually, I don't think fesco tickets are good discussion forum
19:03:38 <mhroncok> this should be discussed on devel mailing list
19:03:47 * nirik nods in agreement
19:04:20 * asosedkin hopes there ain't much left to discuss on 2659
19:04:24 <mhroncok> as for emergency stop gap solution I think we should revert and then discuss of devel of what the nexts steps should be
19:04:28 <mboddu> Morning
19:04:45 <dwmw2> I've just put a final response in the ticket; sorry for being late
19:04:49 <mhroncok> it's much less heated when things are not broken
19:05:50 <asosedkin> sure. I think I've understood the need for revert well enough now
19:05:57 * zbyszek is reading the latest comments on the ticket
19:06:25 <dwmw2> Allowing DTLS0.9 *only* in GnuTLS should be fine. Let's make sure ocserv has it disabled by default too, and there are no other applications that would even try to use it.
19:06:45 <asosedkin> I was previously under the impression that openconnect's gonna opt out with an envvar trick anyway and we'll need to win their hearts back when we have a better solution =)
19:06:48 <dwmw2> Don't care about OpenSSL; although you *can* build OpenConnect against OpenSSL that isn't the preferred configuration and isn't what Fedora does
19:07:15 <dwmw2> I did that for upstream. As noted, it's too incoherent to do it in *Fedora*; I didn't think that was right.
19:07:17 <asosedkin> dwmw2: can we tighten that to "not hard-disabling DTLS0.9 should be fine"?
19:07:21 <zbyszek> Yeah, I think we should revert as a stop gap solution, both in F35 and rawhide, until we have a better option.
19:07:26 <dwmw2> asosedkin: yes.
19:07:47 <nirik> zbyszek: +1
19:07:48 <dwmw2> Users of other protocols will care about DTLS1.0 in similar ways but it's Cisco AnyConnect which is the major user base
19:08:03 <dwmw2> So they can wait for the prio-string based solution
19:08:11 <defolos> .hello2
19:08:12 <zodbot> defolos: defolos 'Dan Čermák' <dan.cermak@cgc-instruments.com>
19:08:13 <mhroncok> ok, is reverting a general agreement or do we need to put our subber stamp on it?
19:08:13 <asosedkin> dwmw2: cool, this way it's not enabled by default unless apps enable it with a prio-string
19:08:18 <defolos> sorry for being late
19:08:21 <mboddu> mhroncok: +1
19:08:28 <mhroncok> *rubber
19:08:30 <dwmw2> asosedkin: Indeed. It never was anyway, was it?
19:08:35 <asosedkin> mhroncok: +1 for reverting to not-hard-disabling in both f35 and rawhide
19:08:48 <asosedkin> dwmw2: I'm new to this, I can't answer
19:08:51 <mhroncok> asosedkin: you are the one who can do that?
19:09:08 <asosedkin> mhroncok: yes
19:09:13 <dwmw2> This protocol never existed for anything but Cisco. I added it to OpenSSL and worked with nmav to put it in GnuTLS too.
19:09:14 <mhroncok> asosedkin++
19:09:14 <zodbot> mhroncok: Karma for asosedkin changed to 1 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
19:09:17 <mhroncok> thanks asosedkin
19:09:22 <dwmw2> Thanks
19:09:24 <mhroncok> I don't think fesco needs to vote
19:09:31 <zbyszek> mhroncok: agreed
19:09:58 <sgallagh> Always best when such things sort themselves out, indeed
19:10:17 <dwmw2> asosedkin: I would *love* to be able to handle older protocols like we handle bad certs. At *negotiation* time, we realise we're in that situation and we ask the user to confirm acceptance.
19:10:44 <zbyszek> OK, any last comments, or can we close the fesco ticket, and asosedkin and dwmw2 will take it from here?
19:10:46 <dwmw2> In some cases when we realise we only get DTLS1.0 when we wanted DTLS1.2 we can't know any earlier. We want to do it just like a cert verify.
19:11:18 <asosedkin> dwmw2: that'd be quite a revamp to all gnutls algorithm selection, basically a third or fourth override to the existing mechanisms, but feel free to file a ticket upstream
19:11:53 <dwmw2> Yeah. We should take the implementation discussion elsewhere, I suppose. zbyszek no further comments in FESCo context; thanks.
19:12:16 <zbyszek> OK, thank you both.
19:12:19 <zbyszek> dwmw2++
19:12:19 <zodbot> zbyszek: Karma for dwmw2 changed to 1 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
19:12:22 <zbyszek> asosedkin++
19:12:25 <zodbot> zbyszek: Karma for asosedkin changed to 2 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
19:12:28 <defolos> thanks guys
19:12:48 <zbyszek> #2660 F35 incomplete changes: Code complete (testable) deadline
19:12:49 <zbyszek> .fesco 2660
19:12:50 <zodbot> zbyszek: Issue #2660: F35 incomplete changes: Code complete (testable) deadline - fesco - Pagure.io - https://pagure.io/fesco/issue/2660
19:13:07 <asosedkin> sorry, very quick question since you're probably the right folks
19:13:16 <mhroncok> zbyszek: how do you want to dicuss this? one by one?
19:13:21 <zbyszek> #undo
19:13:21 <zodbot> Removing item from minutes: <MeetBot.items.Topic object at 0x7f4737d81fd0>
19:13:28 <asosedkin> what's the deadline to fixing things the right way, if it requires a F36 system-wide change?
19:13:44 <asosedkin> like, change filing deadline and change implementation deadlines?
19:13:48 <mhroncok> asosedkin: https://fedorapeople.org/groups/schedule/f-36/f-36-key-tasks.html
19:13:49 <sgallagh> I thought this was about F35?
19:13:52 <defolos> if it's only about one package, then it's self contained
19:14:08 <mhroncok> asosedkin: Change Checkpoint: Proposal submission deadline (Self Contained Changes)         Tue 2022-01-18
19:14:24 <sgallagh> Or do you mean doing the deprecation in 36?
19:14:32 <zbyszek> asosedkin: I guess for F35 this would be handled like a bug, i.e. no Change page. For F36, there's plenty of time as others said.
19:14:34 <Eighth_Doctor> I think that's what asosedkin is talking about
19:14:39 * mhroncok needs to reconnect, brb
19:14:52 <asosedkin> I want a better fix in f36 as well, switching everything gnutls to soft-disabling from hard-disabling
19:15:45 <zbyszek> asosedkin: I guess this could be considered a system-wide change, the deadline for that is Tue 2021-12-28.
19:16:04 <asosedkin> zbyszek: filing or implementing?
19:16:29 <zbyszek> That was filing.
19:16:34 <zbyszek> Implementing is Change Checkpoint: Completion deadline (testable) 	Tue 2022-02-08.
19:16:50 <asosedkin> zbyszek: awesome, and sorry for derailing the meeting a bit
19:17:03 <zbyszek> But the important part is getting the necessary changes to the code done. The bureacracy in Fedora is adaptable.
19:17:42 <zbyszek> OK, let's move to the other Changes.
19:17:49 <zbyszek> #2660 F35 incomplete changes: Code complete (testable) deadline
19:17:49 <zbyszek> .fesco 2660
19:17:50 <zodbot> zbyszek: Issue #2660: F35 incomplete changes: Code complete (testable) deadline - fesco - Pagure.io - https://pagure.io/fesco/issue/2660
19:17:53 <zbyszek> Let's go one by one.
19:18:40 <zbyszek> #topic #2660 F35 incomplete changes: Code complete (testable) deadline
19:18:41 <zbyszek> .fesco 2660
19:18:42 <zodbot> zbyszek: Issue #2660: F35 incomplete changes: Code complete (testable) deadline - fesco - Pagure.io - https://pagure.io/fesco/issue/2660
19:19:02 <zbyszek> Hmm, I think messed up the minutes. Sorry.
19:19:20 <sgallagh> Not a big deal.
19:19:23 <zbyszek> #topic GNU Toolchain update (gcc 11, glibc 2.34, binutils 2.37, gdb 10.2)
19:19:29 <zbyszek> It's all good, no?
19:20:06 <nirik> I think it's done yeah...
19:20:11 <sgallagh> binutils has some issues, but otherwise I think so
19:20:11 <mhroncok1> no idea if there are more minor versions to update to
19:20:15 <Eighth_Doctor> iirc, it's all implemented except the glibc ABI stability hasn't been done
19:20:24 <defolos> looks done, I think
19:20:33 <mhroncok1> unrelistic to revert anyway
19:20:40 <mhroncok1> *unrealistic
19:20:45 <Eighth_Doctor> oh that landed last week
19:20:46 <Eighth_Doctor> looks good then
19:20:50 <zbyszek> OK, let's move on.
19:21:00 <zbyszek> Should I go over the ones that are "most done" too?
19:21:13 <zbyszek> (or just the "mostly not done ones?)
19:21:24 <mhroncok> no preference
19:21:45 <zbyszek> OK, I'll go over all, maybe we'll have some insights.
19:21:48 <zbyszek> #topic CompilerPolicy Change
19:22:17 <zbyszek> mhroncok: you are on the FPC, right? Can we get this done more quickly?
19:22:49 <mhroncok> zbyszek: yes I am
19:23:06 <mhroncok> zbyszek: until your comment on the ticket I was unaware this is blocked on fpc
19:23:33 <mhroncok> fpc unfortunatelly does not deal with tickets the way fesco does
19:23:41 <mhroncok> they tend to... wait
19:24:02 <mhroncok> I can certainly try to get it done, do you have a link?
19:24:08 <zbyszek> I think it's two steps: https://src.fedoraproject.org/rpms/redhat-rpm-config/pull-request/138
19:24:14 <decathorpe> https://pagure.io/packaging-committee/pull-request/1066
19:24:20 <mhroncok> redhat-rpm-config is not maintained by FPC
19:24:23 <zbyszek> … and what decathorpe linked.
19:24:28 <mhroncok> lately, it feels like it is not maintained
19:24:37 <tstellar> Yeah the redhat-rpm-config change should probably be merged first.
19:24:39 <Eighth_Doctor> well, that's because it's not
19:24:43 <decathorpe> the packaging guidelines PR looks fine
19:24:46 <mhroncok> it's "get a provenpackager to merge stuff" maintained
19:24:55 <Eighth_Doctor> redhat-rpm-config hasn't been properly maintained for years
19:25:05 <zbyszek> A volunteer to proven-packager this?
19:25:21 <Eighth_Doctor> I just merged it
19:25:26 * defolos isn't one
19:25:27 <tstellar> I have proven packager fwiw, but I never know what the process is for getting redhat-rpm-config chanages approved.
19:25:29 <zbyszek> Eighth_Doctor: thanks!
19:25:47 <Eighth_Doctor> I can't trigger builds right now (probably later), but it's merged now
19:25:51 <zbyszek> tstellar: I think it's better style if somebody else presses the merge button ;)
19:26:03 <mhroncok> tstellar: there is no process
19:26:08 <mhroncok> it's no man's land
19:26:10 <zbyszek> OK, so now we only need to do the FPC ticket.
19:26:21 <Eighth_Doctor> I'm doing builds now
19:26:46 * bcotton will be mobile for a few minutes to get the kids from the school bus
19:26:46 <Eighth_Doctor> building for f35 and rawhide now
19:26:47 <tstellar> mhroncok, zbyszek: This is a discussion for another day, but I think we need to find official mantainers for this package.  It's pretty important for the distro.
19:26:55 * nirik considers if we should assign/get some folks to actually maintain that package
19:27:09 <zbyszek> #info The redhat-rpm-config pull request has been merged. We still need FPC#1066 to be merged.
19:27:19 <zbyszek> tstellar: definitely
19:27:37 <Eighth_Doctor> the wording is fine
19:27:38 <Eighth_Doctor> mhroncok or I can merge the FPC PR later
19:28:00 <zbyszek> #action mhroncok or Eighth_Doctor to handle the FPC ticket 1066
19:28:00 <Eighth_Doctor> it's tagged with meeting, so I won't do it today
19:28:03 <mhroncok> Eighth_Doctor: I guess we should bring it up on Thursday's meeting
19:28:07 <Eighth_Doctor> yep
19:28:18 <mhroncok> zbyszek: ack
19:28:27 <mhroncok> sorry for being so slow at fpc
19:28:48 <Eighth_Doctor> tstellar: note that you'll need to sync it to rhel yourself
19:28:49 <mhroncok> (I've considered resigning, but for a lack of other volunteers, i haven't)
19:28:55 <Eighth_Doctor> and like in fedora, redhat-rpm-config is essentially unmaintained in rhel
19:29:02 <zbyszek> #action zbyszek to open a ticket to discuss redhat-rpm-config maintanance
19:29:13 <mhroncok> zbyszek: no tickets please
19:29:13 <Eighth_Doctor> (as in, it has maintainers, but they don't do anything)
19:29:17 <mhroncok> zbyszek: threads
19:29:33 <zbyszek> #undo
19:29:33 <zodbot> Removing item from minutes: ACTION by zbyszek at 19:29:02 : zbyszek to open a ticket to discuss redhat-rpm-config maintanance
19:29:43 <nirik> it's been in the past the rpm maintainer(s), but they don't actually do much maintaining on it.
19:29:44 <zbyszek> #action zbyszek to open a thread to discuss redhat-rpm-config maintanance
19:29:55 <mhroncok> zbyszek++
19:29:55 <zodbot> mhroncok: Karma for zbyszek changed to 1 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
19:30:04 <zbyszek> #topic Autoconf-2.71
19:30:15 <zbyszek> Nothing to see here, I think bcotton will reassign it.
19:30:21 <zbyszek> (to F36).
19:30:25 <mhroncok> ack
19:30:26 <defolos> hasn't that been deferred?
19:30:26 <sgallagh> This was deferred, right?
19:30:46 <mhroncok> it is being deferred
19:30:48 <zbyszek> Yeah, by the change owners.
19:30:58 <zbyszek> I only mentioned it for completeness.
19:31:15 <zbyszek> #topic Golang 1.17
19:31:56 <zbyszek> Copr builds are happening. I think we can expect this to be done in 1–2 weeks.
19:32:01 <Eighth_Doctor> yep
19:32:03 <nirik> this is... sort of done
19:32:11 <Eighth_Doctor> we just talked about it an hour ago in the Go SIG meeting
19:32:14 <mhroncok> do we want this in f35?
19:32:29 <defolos> well, it's getting pretty late
19:32:38 <mhroncok> beta freeze is in 1 week
19:32:44 <zbyszek> mhroncok: I think the general answer is yes, if the change seems stable.
19:32:51 <mhroncok> done in 1-2 weeks would land after beta
19:32:55 <defolos> Conan Kudo: any big/maportant packages that could be impacted by this?
19:32:57 <nirik> beta freeze is likely the deadline
19:33:06 <defolos> s/maportant/important
19:33:09 * Eighth_Doctor shrugs
19:33:21 <Eighth_Doctor> the container circus packages are the main ones anyone cares about
19:33:35 <decathorpe> I'd be more likely to say "push to f35 before beta freeze" if this was already in rawhide without causing trouble
19:34:11 <Eighth_Doctor> well, they're doing the build in copr right now, so it's really not a huge deal
19:34:15 <zbyszek> Proposal: the change can be done in F35 if it can be finished before the beta freeze.
19:34:20 <mhroncok> proposal: if this lands by beta freeze, it will be in f35. if not, don't do it
19:34:28 <mhroncok> zbyszek: +1, same as mine
19:34:30 <Eighth_Doctor> zbyszek: +1
19:34:35 <dcantrell> is there anything we absolutely want from golang 1.17?  if not, I say it missed the deadline and just postpone it
19:34:44 <decathorpe> well ... I have experience with golang stuff failing on ppc64le and s390x, which are both not going to get caught in COPR
19:35:11 <defolos> so build it in a side tag and merge it before the beta freeze?
19:35:13 <nirik> copr does have ppc64le now
19:35:13 <zbyszek> dcantrell: I'd leave that to the maintainers…
19:35:15 <sgallagh> zbyszek: +1
19:35:50 <mhroncok> defolos: I'd only do that side tag after rawhide was succesfull
19:35:52 <defolos> zbyszek: +1
19:35:52 <dcantrell> ok, then what I'm reading in the change proposal doesn't look like something that can't be postponed.
19:36:19 <nirik> I think the plan was to update golang and a few things, but not mass rebuild everything go...
19:36:42 <mhroncok> dcantrell: one aspect is if we need the distro packages to be rebuilt with new go
19:36:52 <decathorpe> most packages only ship sources so that should be fine
19:36:55 <mhroncok> dcantrell: I say we don't necessarily
19:37:17 <mhroncok> dcantrell: the other aspect is that we want f35 developers to have new go compiler
19:37:26 <dcantrell> mhroncok: right, which was a concern I have and it seems like we've run out of time for that.  all I'm saying
19:37:46 * Eighth_Doctor shrugs
19:37:58 <mhroncok> good thing is that packages that won't build will still function
19:38:05 <zbyszek> dcantrell: I think our an important part of the Fedora appeal is that you get the recent version of various development packages. I think we should strive to have latest stable versions, when reasonable.
19:38:32 <dcantrell> I agree with that, but if we're cutting out our time window to validate that then it's a gamble.
19:38:40 <mhroncok> if only there was a way to offer new go to users without impacting the distro...
19:38:42 * mhroncok hides
19:38:45 <dcantrell> and if this isn't something that is urgent then I say why worry and just postpone it
19:38:59 <sgallagh> If only we had a technology in Fedora that would allow us to ship multiple versions of Go...
19:39:01 <dcantrell> but if no one seems to care, I don't really care
19:39:09 <mhroncok> sgallagh: exactly
19:39:12 <Eighth_Doctor> if only our infrastructure didn't suck for this kind of thing...
19:39:24 <sgallagh> mhroncok: I see we are on the same page here :-D
19:39:42 <zbyszek> sgallagh, mhroncok: maybe two sides of the same page?
19:39:43 <nirik> anyhow, I'm fine giving folks until freeze to get it in...
19:39:44 * zbyszek ducks
19:39:47 <mhroncok> yeah, see Python :)
19:39:58 <mhroncok> count the votes?
19:40:01 * Eighth_Doctor is not bitter at all...
19:40:05 <zbyszek> dcantrell, decathorpe: vote?
19:40:19 <dcantrell> give it until the freeze  +1
19:40:25 <mboddu> +1 give it until the freeze
19:40:25 <decathorpe> +1 to let them get it done before beta freeze
19:40:41 * mboddu 's browser died, fixing it...
19:41:01 <Eighth_Doctor> +1 give until freeze
19:41:32 <zbyszek> #agreed The change can be done in F35 if it can be finished before the beta freeze.
19:41:34 <zbyszek> #undo
19:41:34 <zodbot> Removing item from minutes: AGREED by zbyszek at 19:41:32 : The change can be done in F35 if it can be finished before the beta freeze.
19:41:42 <zbyszek> #agreed The change can be done in F35 if it can be finished before the beta freeze (+7, 0, 0)
19:42:09 <zbyszek> #topic LLVM 13
19:42:19 <zbyszek> #info https://fedoraproject.org/wiki/Changes/LLVM-13
19:42:35 <zbyszek> tstellar: are you still around?
19:42:39 <tstellar> zbyszek: Yeah.
19:43:11 <zbyszek> Oh, I missed this in the ticket:
19:43:12 <zbyszek> I have 13 of the 16 packages built in f35-build-side-43964, and I expect to finish the rest this week.
19:43:24 <zbyszek> So… it's all good?
19:43:29 <nirik> cool. +1 to wait on that
19:43:42 <sgallagh> Yeah, this is in good hands
19:43:45 <mboddu> +1 to wait
19:43:54 <defolos> +1 to wait for the rest to finish
19:43:56 <tstellar> zbyszek: I ran into some issues with annobin, that set me back a little, but I have that resolved now and have resumed doing builds today.
19:43:56 <mhroncok> oviously, until the beta freeze
19:43:56 <zbyszek> ack, I don't think we need an explicit vote.
19:44:11 <zbyszek> mhroncok: f35 beta freeze ;)
19:44:19 <decathorpe> explicit +1 to let tstellar do his job anyway :)
19:44:29 <zbyszek> Ok, ok
19:44:38 <mhroncok> zbyszek: yeah, that one
19:44:55 <zbyszek> #agreed  let tstellar do his job (+6, 0, 0)
19:45:02 * Eighth_Doctor sighs about annobin
19:45:11 <Eighth_Doctor> it continues to haunt us
19:45:32 <zbyszek> Eighth_Doctor: well, a discussion for another time.
19:45:39 <zbyszek> #topic Adding Selected Flathub Applications
19:46:01 <dcantrell> Eighth_Doctor: heh
19:46:31 <Eighth_Doctor> that's all done, we're waiting for GNOME 41 to land
19:46:34 <zbyszek> #info https://fedoraproject.org/wiki/Changes/Filtered_Flathub_Applications
19:46:54 <Eighth_Doctor> Fedora specific work is complete, upstream work is complete, it'll be part of GNOME 41 alpha
19:46:58 <zbyszek> I see that bcotton wrote "Adding Selected Flathub Applications — set to MODIFIED"
19:47:28 <zbyszek> Eighth_Doctor: ack, so I think we don't need to worry about this.
19:47:49 <zbyszek> #topic Introduce Module obsoletes and EOL
19:47:52 <Eighth_Doctor> yep
19:47:59 <zbyszek> dmach is not here…
19:48:17 <mhroncok> my understanding here is that dnf supports this, but our infra does not
19:48:29 <zbyszek> #info https://fedoraproject.org/wiki/Changes/Module_Obsoletes_and_EOL
19:48:29 <mhroncok> and there has been no work in infra to support that
19:48:53 <sgallagh> I'm not sure DNF support for this landed either.
19:49:02 <mhroncok> or maybe even that
19:49:05 <Eighth_Doctor> 😩
19:49:07 <sgallagh> And there's no packaging guidelines for how to use it.
19:49:07 <defolos> it looks overall pretty stalled
19:49:17 <defolos> iirc this was discussed at some point a year ago?
19:49:20 <mhroncok> also, this has been driven by some rhel needs, and at the end I think they say it will not be done in rhel that way
19:49:21 <nirik> well, it's just a document, hosting that should not be a problem if it existed.
19:49:24 <sgallagh> So I'd have to assume this is not happening in F35
19:49:47 <zbyszek> .bugzilla 1834844
19:49:51 <sgallagh> nirik: What's just a document?
19:50:04 <zbyszek> Hmm, what's the magic phrase to link to a bug?
19:50:04 <mhroncok> nah, it's not happening. and I'd propose to reject instead of deferring, resumbitting if wanted
19:50:09 <mhroncok> .bz
19:50:15 <zbyszek> .bz 1834844
19:50:18 <defolos> sgallagh: the module obsoletes
19:50:19 <decathorpe> .bug , I think.
19:50:19 <zodbot> decathorpe: Error: That URL raised <Invalid URL: URL can't contain control characters. '/, I think.' (found at least ' ')>
19:50:36 <decathorpe> .bug 1834844
19:50:38 <zodbot> decathorpe: 1834844 – Introduce module Obsoletes and EOL - https://bugzilla.redhat.com/1834844
19:50:44 <decathorpe> yeah that did it.
19:50:49 <mhroncok> mea culpa
19:50:51 <mboddu> mhroncok: What infra work is needed? I thought it will be all part of module metadata file?
19:51:14 <mhroncok> mboddu: I don't know about specifics
19:51:25 <nirik> sgallagh: the document that describes the eol ? the thing dnf could download/parse?
19:51:49 <sgallagh> OK, I was just making sure I knew what you meant.
19:51:58 <mhroncok> and a way for maintainers to sumbit chnages to that thing
19:52:03 <nirik> "Proposed new modulemd document: "
19:52:20 <nirik> I was assuming it would be in a git repo with the other modular defaults
19:52:30 <Eighth_Doctor> modular obsoletes are implemented in libdnf already, from what I can tell
19:52:31 <sgallagh> I haven't followed this effort, so I don't know whether it landed (other than that the module document format is handled by libmodulemd now)
19:52:37 <sgallagh> No idea if it's actually consumed anywhere
19:52:42 <sgallagh> OK, maybe it is
19:52:53 <Eighth_Doctor> it was already implemented in January
19:52:54 <defolos> it's been deferred once already and there doesn't appear to be any real movement, so maybe reject and let them resubmit it?
19:52:56 <mhroncok> nirik: that would require a group of people in Fedora that would maintain it
19:53:06 <mhroncok> defolos: +1
19:53:14 <sgallagh> mhroncok: I think the idea was that it wouldn't be centralized
19:53:18 <Eighth_Doctor> at least based on this: https://github.com/rpm-software-management/libdnf/commits/dnf-4-master/libdnf/module
19:53:23 <nirik> well, the change says when eoling or obsoleting a module, maintainers of that module would submit this
19:53:25 <zbyszek> defolos: -0, there *has* been movement. It seems most of the work is done.
19:53:32 <sgallagh> But would be part of the stream
19:53:34 <sgallagh> But I'm not sure.
19:54:03 <nirik> how about we delegate someone to ask change owners and revisit next week?
19:54:12 <sgallagh> Proposal: Defer decision one week. Email mcurlej and find out its status
19:54:17 <zbyszek> I think we want this at some point. But it seems rather late to introduce changes to dnf/repos/module structure right before beta.
19:54:21 <Eighth_Doctor> sgallagh: +1
19:54:40 <mhroncok> dmach is the chnage owner
19:54:47 <mhroncok> not mcurlej
19:54:52 <Eighth_Doctor> the bigger problem is nobody maintains MBS right now
19:55:03 <Eighth_Doctor> with no maintainer of MBS, the server side components can't get implemented
19:55:18 <mhroncok> my understanding was that this is exactly the blocker
19:55:23 <zbyszek> sgallagh: +1, but I expect to postpone this to F36 unless it's all done are ready now.
19:55:40 <mboddu> sgallagh: +1, but I think it will be moved to F36
19:55:52 <Eighth_Doctor> iirc, all the DNF work was done, but nobody is there to do MBS stuff
19:55:54 <zbyszek> mhroncok: mcurlej seems to be doing the implementation, he's assigned on the ticket.
19:56:13 <nirik> mbs is maintained by the koji team now
19:56:26 <Eighth_Doctor> oh dear
19:56:44 <nirik> They got our mbs upgraded eariler this year...
19:57:15 <zbyszek> mhroncok, defolos, decathorpe, dcantrell: vote?
19:57:23 <nirik> I have no idea how active they are on implementing new features, etc
19:57:25 <mhroncok> 0
19:57:32 <zbyszek> nirik: vote?
19:57:39 <dcantrell> 0
19:57:41 <nirik> sgallagh: +1
19:57:56 <decathorpe> -1, reject, resubmit, if needed
19:58:10 <zbyszek> (Sorry folks for missing one or another person. It's close to 22:00 here, and only a constant stream of very dark tee is keeping me awake.)
19:58:31 <mhroncok> decathorpe: I'd +1 that
19:58:40 <sgallagh> I'm changing my vote to -1
19:58:58 <dcantrell> yeah, save room for me on that bandwagon  -1
19:59:04 <sgallagh> So I'll follow Fabio Valentini (decathorpe) 's lead
19:59:11 <defolos> i'm for rejecting
19:59:15 * Eighth_Doctor shrugs
19:59:23 <Eighth_Doctor> accepting or rejecting basically has no impact
19:59:32 <Eighth_Doctor> at this point, it's an out-of-band change
19:59:43 <Eighth_Doctor> we get it whenever MBS gets updated, whether anyone likes it or not
19:59:45 <mhroncok> rejecting menas they would need to resumbit to show there is some interest
20:00:15 * zbyszek counts
20:00:17 <mhroncok> Eighth_Doctor: is it that rady?
20:00:20 <mhroncok> *ready
20:00:30 <Eighth_Doctor> yep
20:00:39 <Eighth_Doctor> libmodulemd and dnf already know what to do, more or less
20:00:44 <nirik> what was the mbs update needed?
20:01:09 <Eighth_Doctor> well, MBS generates the modules.yaml documents for repodata, and it needs to handle the new fields and validate them
20:01:23 <zbyszek> So, we're at +4 to wait one week, and -4 to reject now.
20:01:28 <nirik> huh, oddly, it's not listed even once on the change page
20:02:13 <Eighth_Doctor> oh hold up
20:02:19 <Eighth_Doctor> it's actually done in MBS too
20:02:21 <Eighth_Doctor> https://pagure.io/fm-orchestrator/pull-request/1667
20:02:28 <Eighth_Doctor> modulemd v3 is supported
20:03:09 <defolos> uh, so this thing is done now?
20:03:24 <Eighth_Doctor> apparently?
20:03:33 <mhroncok> I don't think it is "done"
20:03:41 <mhroncok> can module maintainers use this?
20:03:51 <Eighth_Doctor> MBS 3.6.1 was released 4 months ago
20:03:57 <Eighth_Doctor> if that's in prod, then I guess yeah they can
20:04:18 <Eighth_Doctor> mmd v3 was added in 3.6.0
20:04:26 <sgallagh> This would be nice to know... or have some documentation about it...
20:04:26 <nirik> well, there still needs to be files somewhere and docs on how to use it, and such
20:04:42 <mhroncok> and policies :/
20:04:45 <nirik> we are running 3.6.1 since eariler this year.
20:04:55 <zbyszek> sgallagh: I think your original proposal was good. We need some input from dmach and mcurlej to make an informed decision.
20:04:55 <Eighth_Doctor> yeah, so I'd be inclined to say this needs poking for a comprehensive update
20:05:02 <Eighth_Doctor> because it looks like it's code-complete
20:05:15 <sgallagh> yeah, that seems fair
20:05:31 <mhroncok> there seem to be no consensus, so explcitly asking is the safer thing to do
20:05:31 <defolos> yeah, let's poke them first and get some info about the current state
20:05:51 <defolos> given that we'd need docs & policies, I'd suggest that we defer this to f36 though
20:06:02 <mhroncok> let's ask the change owners about the status
20:06:20 <zbyszek> OK, are we all agreed on this?
20:06:32 <Eighth_Doctor> yup, let's ask them
20:06:43 <defolos> yes, +1 to asking them
20:06:47 <decathorpe> fine with me
20:06:48 <mhroncok> yes
20:06:54 <mhroncok> any volunteers?
20:07:14 <mhroncok> I can ask on the meeting we get each week, but that would be next week
20:07:25 <zbyszek> #agreed Defer decision one week. Email mcurlej and find out current status.
20:07:27 <mhroncok> *each other week
20:07:52 <zbyszek> #action I'll ask in the ticket and maybe also send an email if there's no response.
20:07:58 <zbyszek> #undo
20:07:58 <zodbot> Removing item from minutes: ACTION by zbyszek at 20:07:52 : I'll ask in the ticket and maybe also send an email if there's no response.
20:08:06 <zbyszek> #action zbyszek to ask in the ticket and maybe also send an email if there's no response.
20:08:26 <zbyszek> #topic Patches in Forge macros - Auto macros - Detached rpm changelogs
20:08:35 <mhroncok> proposal: reject
20:08:52 <mhroncok> this is outdated and there has been no response in months
20:08:54 <zbyszek> Maybe "reject and resubmit if still wanted" ?
20:08:57 <defolos> isn't nim MIA anyway?
20:09:07 <mhroncok> zbyszek: that goes without saying
20:09:11 <decathorpe> yeah
20:09:13 <sgallagh> +1 reject
20:09:13 <decathorpe> +1 to reject and resubmit if needed
20:09:17 <defolos> plus we got rpmautospec now
20:09:26 <nirik> +1
20:09:28 <zbyszek> mhroncok: but it's still nicer to be explicit about this.
20:09:35 <dcantrell> +1 reject
20:09:35 <mhroncok> sorry for being brief, I am not powered by tea, but alcohol
20:09:40 <defolos> +1
20:09:55 * decathorpe thinks Miro has the right idea there
20:10:14 <defolos> mhroncok: why not both 😜
20:10:14 <Eighth_Doctor> +1 reject
20:10:38 <Eighth_Doctor> defolos: I think alcoholic tea might be too much for most :D
20:10:43 <zbyszek> +1 to reject and resubmit
20:10:54 <Eighth_Doctor> there won't be a resubmission
20:10:59 <Eighth_Doctor> nim disappeared off the face of the earth
20:11:01 <zbyszek> mboddu?
20:11:09 <defolos> Conan Kudo: I'm sure mhroncok is a tough fella 😉
20:11:10 <Eighth_Doctor> in fact, this wasn't even supposed to carry over to f35
20:11:27 <defolos> anyone knows what happened to nim?
20:11:30 <nirik> I hope he's well and is able to come back someday.
20:11:30 * Eighth_Doctor shrugs
20:11:32 <zbyszek> #agreed Reject and resubmit if still wanted (+8, 0, 0)
20:11:35 <mboddu> zbyszek: +1 reject
20:11:39 <zbyszek> #undo
20:11:39 <zodbot> Removing item from minutes: AGREED by zbyszek at 20:11:32 : Reject and resubmit if still wanted (+8, 0, 0)
20:11:43 <zbyszek> #agreed Reject and resubmit if still wanted (+9, 0, 0)
20:12:03 <zbyszek> #topic Enhanced Inscript as default Indic IM
20:12:37 <mhroncok> no idea here
20:13:23 * Eighth_Doctor shrugs
20:13:29 <mhroncok> the IM changes wwere always beyond me
20:14:04 <nirik> https://fedoraproject.org/wiki/Changes/Enhanced_Inscript_as_default_Indic_IM
20:14:34 <zbyszek> .bug 1982546
20:14:37 <zodbot> zbyszek: 1982546 – Make inscript2 engines rank higher than inscript engines - https://bugzilla.redhat.com/1982546
20:14:37 <zbyszek> This one is done…
20:14:50 <mhroncok> "m17n-db and ibus-m17n packages will be updated"
20:15:09 <defolos> that's kinda unspecific
20:15:35 <zbyszek> https://gitlab.gnome.org/GNOME/gnome-desktop/-/commits/master/libgnome-desktop/default-input-sources.h
20:15:54 <nirik> that seems not done
20:15:58 <zbyszek> This one hasn't seen any changes in the last 8 months…
20:16:25 <nirik> how about asking and if it's not done by next week defer or reject?
20:16:31 <zbyszek> nirik: +1
20:16:34 <mhroncok> nirik: +1
20:16:39 <sgallagh> +1
20:16:42 <dcantrell> +1
20:16:49 <Eighth_Doctor> +1
20:16:54 <defolos> +1
20:16:56 <decathorpe> +1
20:17:12 <zbyszek> #agreed Ask for input from change owners and defer decision until next week (+8, 0, 0)
20:17:36 <zbyszek> OK, that's all in my list.
20:17:46 <mhroncok> \o/
20:17:58 <defolos> 🎉
20:18:03 <zbyszek> #topic Next week's chair
20:18:03 <zbyszek> I'm likely to miss the next meeting.
20:18:26 <zbyszek> Vlntrs?
20:18:44 <defolos> I can try if someone would be willing to help me out
20:18:45 <sgallagh> I can do it next week, I think.
20:18:53 <sgallagh> Sure, I'll back up defolos
20:19:05 <defolos> thanks sgallagh
20:19:15 <zbyszek> defolos: https://fedoraproject.org/wiki/FESCo_meeting_process has very good instructions.
20:19:33 <zbyszek> #action defolos is it. sgallagh will provide backup.
20:19:38 * defolos bookmarks that
20:19:45 <zbyszek> #topic Open Floor
20:19:55 <bcotton> shameless plug for the text meeting training video I recently made, too https://www.youtube.com/watch?v=3hG4Y09reZg
20:20:39 <mhroncok> bcotton: nice
20:20:45 <zbyszek> Actually, I have a thing.
20:20:50 <zbyszek> .bug 1803302
20:20:52 <zodbot> zbyszek: 1803302 – Review Request: github-cli - The GitHub CLI - https://bugzilla.redhat.com/1803302
20:21:19 <zbyszek> I think it's important for developers who use github, which is pretty much everyone nowadays, to have this…
20:21:58 <zbyszek> Joe Doss wrote:  I am unsure on the dependencies for the current version. I have not had the free time to push this over the finish line because the dependencies of the upstream github-cli keeps changing and packaging golang is a deathmarch. If someone has the time and means to pick up my work here and get it out the door, I am all for handing over the maintainership of this package and all dependencies. :(
20:22:20 <zbyszek> Dunno, could we get some more folks to look into this?
20:22:36 <sgallagh> Also, its predecessor `hub` is umaintained in Fedora
20:22:39 <mhroncok> if it was Python, I'd go for it :D
20:22:53 <defolos> it's go, right?
20:22:58 <mhroncok> indeed
20:23:05 <mhroncok> sgallagh: retired 9 months ago
20:23:05 <sgallagh> Yeah, it's in Go
20:23:36 <sgallagh> mhroncok: Not surprising. It broke, it was getting replaced by github-cli and I was just not able to keep it going
20:23:44 <decathorpe> I was in a similar situation with syncthing (changing dependencies with every version) which is why it's the only one of my packages that bundles sources ... I would suggest doing that with github-cli too
20:23:54 <mhroncok> sgallagh: yeah, i recall the announcement
20:24:11 <sgallagh> Is there any chance we could get away with including this in the third-party repos?
20:24:21 <sgallagh> Github provides a yum repo, IIRC
20:24:35 <mhroncok> sgallagh: I'd prefer bundling over pre-built binaries by github
20:24:58 <zbyszek> Yeah, maybe bundling would be an option.
20:25:22 <decathorpe> theres a few examples of Go packages that do the bundling thing "right" (caddy, syncthing, are two examples)
20:25:35 <sgallagh> I don't know Go bundling well enough to know if that's feasible.
20:26:15 <decathorpe> it's pretty easy to set up, but checking all necessary licenses and adding the required "Provides: bundled (foo)" is a chore
20:26:24 <decathorpe> (which is why I automated the latter)
20:26:48 <defolos> maybe we should provide some automation for this?
20:26:56 <defolos> I mean rpm has automated provides generators
20:27:01 <zbyszek> OK, I don't think we're going to solve this here. But before we move on, decathorpe, do you have a link for the syncthing stuff?
20:27:25 <defolos> I'm not a huge fan of bundling, but the go and rust and node ecosystems make this increasingly a must
20:27:41 <Eighth_Doctor> it's less of a problem for rust because of how it works compared to node and go
20:27:57 <Eighth_Doctor> node has an insane dep graph
20:28:00 <decathorpe> https://src.fedoraproject.org/rpms/syncthing/blob/rawhide/f/vendor2provides.py generates bundled provides from golang's standard vendor manifest
20:28:16 <Eighth_Doctor> and go is anti-dep charting
20:28:19 <zbyszek> decathorpe: thanks.
20:28:31 <zbyszek> Does anyone else have anything for open floor?
20:28:33 <sgallagh> We actually have a reliable script for Node deps
20:28:33 <decathorpe> the .spec file has some more comments.
20:28:38 <sgallagh> *bundled deps
20:29:22 <zbyszek> If not, I'll close at 22:30 ;)
20:29:44 <defolos> nothing from my side
20:29:54 <sgallagh> I'm good
20:29:55 <defolos> besides a chili cooking overcooking on the stove 😉
20:30:01 <mhroncok> zbyszek: thanks for chairing
20:30:08 <zbyszek> Thank you all for attending.
20:30:09 <zbyszek> #endmeeting