17:00:07 <sgallagh> #startmeeting ELN (2022-12-16)
17:00:07 <zodbot> Meeting started Fri Dec 16 17:00:07 2022 UTC.
17:00:07 <zodbot> This meeting is logged and archived in a public location.
17:00:07 <zodbot> The chair is sgallagh. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
17:00:07 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:00:07 <zodbot> The meeting name has been set to 'eln_(2022-12-16)'
17:00:07 <sgallagh> #meetingname eln
17:00:07 <zodbot> The meeting name has been set to 'eln'
17:00:07 <sgallagh> #topic init process
17:00:07 <sgallagh> .hi
17:00:08 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
17:00:50 <yselkowitz[m]> .hello yselkowitz
17:00:51 <zodbot> yselkowitz[m]: yselkowitz 'Yaakov Selkowitz' <yselkowi@redhat.com>
17:03:31 <sgallagh> Great turnout today :eyeroll:
17:03:55 <yselkowitz[m]> it is that time of year...
17:04:19 <sgallagh> Fair enough.
17:04:22 <sgallagh> #topic The Final Countdown
17:04:37 <sgallagh> yselkowitz has been doing a fantastic job closing out the remaining build failures.
17:05:04 <sgallagh> At last count, we were down to seven remaining that we don't know what to do with yet.
17:05:49 <sgallagh> yselkowitz: Do you mind if I copy in that YAML doc for the meeting record?
17:06:05 <yselkowitz[m]> sure
17:06:23 * sgallagh sent a code block: https://libera.ems.host/_matrix/media/v3/download/libera.chat/a83f61e047b340a9ec215b8551430380b26c403c
17:06:54 <sgallagh> I'm looking at rust-rav1e at the moment.
17:07:44 <yselkowitz[m]> that's the one we're most likely to need, provided libavif and/or the stripped ffmpeg are really going to be in RHEL10
17:08:16 <sgallagh> Yeah, I think I have a hunch what's going on there, so I'll see if I can fix that up.
17:08:42 <yselkowitz[m]> that would be great, because I'm not that familiar with rust ecosystem yet
17:09:38 <sgallagh> yselkowitz: webkit needs it, so I have to assume it's going to come in
17:10:17 <yselkowitz[m]> it's probably configurable, since we have webkit in RHEL9 w/o libavif
17:10:32 <yselkowitz[m]> but it seems like a likely candidate
17:10:49 <sgallagh> I'm not so sure that it's optional in WebKit 6.0
17:11:31 <yselkowitz[m]> idk I'd have to look but I think it's fair to assume we'll need libavif
17:12:01 <yselkowitz[m]> the others on this list, not so much
17:12:16 <sgallagh> Let's just go through them quickly.
17:12:51 <sgallagh> #info We probably need rust-rav1e for RHEL 10, the others unlikely.
17:13:22 <sgallagh> coffee-script: I think we can extract this from sinatra and pcs will still work.
17:13:24 <sgallagh> We need to discuss it with the maintainer
17:14:30 <sgallagh> golang-github-stretchr-objx: I'm pretty sure this will end up bundled in toolbox
17:14:39 <yselkowitz[m]> 100%
17:14:57 <sgallagh> What was that "100%" referring to?
17:15:03 <yselkowitz[m]> I'm working on a patch to use vendored sources %if 0%{?rhel}
17:15:12 <yselkowitz[m]> 100% certain it will be vendored
17:15:31 <sgallagh> Excellent
17:15:41 <yselkowitz[m]> and using vendored sources should drop a bunch of golang packages from ELN
17:16:23 <yselkowitz[m]> the tradeoff being someone will need to maintain the vendoring going forward, but the upstream doesn't seem to be moving very quickly nowadays
17:16:47 <yselkowitz[m]> I hope that's not a bad sign, because I really like toolbox
17:17:18 <sgallagh> llvm11: It will NOT be a package in RHEL 10. That needs to get fixed in the Content Resolver input
17:17:31 <sgallagh> Yeah, I'm using toolbox all the time too
17:17:39 <sgallagh> (I switched my main system to Silverblue)
17:17:52 <yselkowitz[m]> and dropping llvm11 will cause libffi3.1 to drop
17:18:18 <sgallagh> Right
17:18:34 <sgallagh> maven-scm: I think this should drop out once the OpenJDK changes for F38 land.
17:19:01 <yselkowitz[m]> what changes specifically?
17:19:48 <sgallagh> See https://pagure.io/fesco/issue/2907
17:19:55 <tdawson> So, llvm11 is part of the "LLVM Compat Packages"  That's getting dropped from RHEL 10?
17:20:49 <yselkowitz[m]> what does "llvm compat" even mean?  afaik we only support the latest version in rhel
17:20:50 <sgallagh> tdawson: I'm reasonably confident, yes. I'll double-check with the Toolchain team
17:21:12 <tdawson> That would be nice.  The sooner the better for that.
17:21:42 <yselkowitz[m]> why does that openjdk change mean that byteman will be dropped?
17:21:45 <tdawson> The compat packages are usually because we've promised that we'd keep something for so many releases.
17:22:23 <tdawson> But once that obligation is done, it can be dropped.
17:22:25 <sgallagh> Yes, but it's not in the ACG
17:22:57 <tdawson> Ohh ... if it's not in the ACG, then all the more likely it will be dropped.
17:23:03 <sgallagh> Or rather, it's only ACG level 3
17:23:04 <sgallagh> And it's in AppStream, which is even less strict
17:23:25 <sgallagh> yselkowitz: I *think* it should fall out with the bundling, but I might be mistaken
17:24:11 <yselkowitz[m]> I don't think it's a bundling, byteman is part of the openjdk workload
17:24:28 <yselkowitz[m]> and it grew a new dependency since RHEL9 branched, which eventually leads to maven-scm
17:24:43 <sgallagh> Maybe we should speak to jvanek about this
17:25:23 <yselkowitz[m]> I imagine it would be buildroot-only at most, but still might be needed
17:26:03 <sgallagh> ack
17:26:18 <sgallagh> rubygem-sassc-rails: Probably will end up bundled
17:26:46 <yselkowitz[m]> agreed
17:27:29 <yselkowitz[m]> so how do we handle the failures for packages that we're pretty sure will be bundled in rhel?  just ignore them?  block the packages entirely from eln?
17:28:03 <sgallagh> Ideally, we work with the maintainer to do the bundling in ELN
17:28:05 <yselkowitz[m]> or institute and maintain the bundling in rawhide?
17:28:30 <yselkowitz[m]> what you just said
17:28:30 <sgallagh> Which can be either in Rawhide or if needed they can have a separate eln branch
17:29:01 <sgallagh> (crypto-policies has its own branch because it's too complicated to conditionalize, for example)
17:30:19 <sgallagh> OK, so that's the whole list.
17:30:48 <sgallagh> #info We're fairly confident that six of the remaining seven packages will not be in RHEL 10, so we'll engage with the maintainers to get that cleaned up
17:31:05 <yselkowitz[m]> I've already started working on the toolbox vendoring for rawhide
17:31:05 <sgallagh> yselkowitz: Thanks for tracking down the dependency chains, by the way.
17:32:16 <sgallagh> So we are in really good shape
17:32:31 <sgallagh> In fact, we're exactly where I wanted to be by the end of January, so... applause all around!
17:32:56 <tdawson> Ya!!
17:33:17 <sgallagh> #topic Open Floor
17:33:28 <sgallagh> I just have one thing to say here:
17:33:58 <sgallagh> Thank you to everyone who got involved with ELN this year. It has been a tremendous success and every one of you were a part of that!
17:34:18 <sgallagh> #info Thank you to everyone who got involved with ELN this year. It has been a tremendous success and every one of you were a part of that!
17:34:43 <tdawson> You are very welcome.  I enjoyed it.
17:34:50 <sgallagh> Any other topics for today?
17:34:51 <tdawson> And I think it was a good year.
17:35:03 <yselkowitz[m]> should we link to the outstanding PRs?
17:35:31 <sgallagh> Our next meeting will be held on January 13, due to the winter holidays.
17:35:46 <sgallagh> yselkowitz: Be my guest. Let me just chair you...
17:35:53 <sgallagh> #chair yselkowitz
17:35:53 <zodbot> Current chairs: sgallagh yselkowitz
17:36:08 <sgallagh> Just preface the links with #link
17:36:21 <yselkowitz[m]> #info outstanding ELN PRs:
17:36:23 <yselkowitz[m]> #link https://src.fedoraproject.org/rpms/rsyslog/pull-request/35
17:36:38 <yselkowitz[m]> #link https://src.fedoraproject.org/rpms/texlive-base/pull-request/10
17:36:40 <yselkowitz[m]> #link https://src.fedoraproject.org/rpms/unbound/pull-request/9
17:36:48 <yselkowitz[m]> #link https://src.fedoraproject.org/rpms/xorg-x11-drv-vmware/pull-request/1
17:36:58 <yselkowitz[m]> #link https://src.fedoraproject.org/rpms/ruby/pull-request/143
17:37:19 <yselkowitz[m]> #link https://src.fedoraproject.org/rpms/python-cryptography/pull-request/18
17:37:34 <yselkowitz[m]> #link https://pagure.io/releng/issue/11186
17:37:48 <yselkowitz[m]> I think that's it
17:37:48 <sgallagh> I'll go through those and merge any trivial ones today
17:38:28 <sgallagh> The rest, I'll merge after the new year if the maintainers have not responded
17:38:40 <yselkowitz[m]> I think there is unresolved discussion on most of those, and one is a version bump
17:39:12 <sgallagh> #info The above tickets may be merged and built by a provenpackager if the maintainer has not responded.
17:39:34 <yselkowitz[m]> but please take a look and let's see what we can move forward
17:39:34 <sgallagh> The python-cryptography one at least should be fine to merge
17:39:40 <sgallagh> Even if it is just a stopgap solution
17:40:19 <yselkowitz[m]> that one should be ?rhel instead of ?eln
17:40:23 <yselkowitz[m]> no, sorry, we said to make that unconditional
17:41:14 <yselkowitz[m]> so do we wait for the PR author to change that, or just make a new PR ourselves?
17:41:22 <sgallagh> I'll take care of creating a new one
17:41:36 <sgallagh> I'll give them until next week to review it, then I'll merge it if no response
17:42:36 <sgallagh> OK, anything else for today?
17:43:50 <tdawson> Nothing from me
17:44:53 <yselkowitz[m]> I'll be around but I expect not much will happen until January
17:45:35 <yselkowitz[m]> let's call it
17:46:18 <sgallagh> Alright, thank you everyone
17:46:19 <sgallagh> See you all in 2023!
17:46:23 <sgallagh> #endmeeting