13:00:30 <FranciscoD> #startmeeting NeuroFedora - 2023-09-11
13:00:30 <zodbot> Meeting started Mon Sep 11 13:00:30 2023 UTC.
13:00:30 <zodbot> This meeting is logged and archived in a public location.
13:00:30 <zodbot> The chair is FranciscoD. Information about MeetBot at
13:00:30 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
13:00:30 <zodbot> The meeting name has been set to 'neurofedora_-_2023-09-11'
13:00:36 <FranciscoD> #meetingname neurofedora
13:00:36 <zodbot> The meeting name has been set to 'neurofedora'
13:00:45 <FranciscoD> #chair davdunc Penguinpee principis
13:00:45 <zodbot> Current chairs: FranciscoD Penguinpee davdunc principis
13:01:06 <FranciscoD> #topic Agenda
13:01:11 <FranciscoD> #info Agenda:
13:01:21 <FranciscoD> #info meetbot/zodbot guide:
13:01:57 <Penguinpee> .hello gui1ty
13:01:58 <zodbot> Penguinpee: gui1ty 'Sandro .' <>
13:02:31 <FranciscoD> #topic Intros and roll call
13:02:34 <FranciscoD> Penguinpee: o/
13:02:42 <Penguinpee> \o
13:02:57 <FranciscoD> let's wait a few minutes for others to join, I've given the link etc. in the matrix channel too
13:05:28 <FranciscoD> ok, let's proceed!
13:05:41 <FranciscoD> #topic Tasks from last meeting
13:06:17 <FranciscoD> #info Last meeting was on July 17:
13:07:08 <Penguinpee> That's almost two month ago. Tempus fugit.
13:07:22 <FranciscoD> I was to update this one, but I can't remember what I was to update it about.. ->
13:07:27 * FranciscoD looks at logs
13:08:15 <FranciscoD> ah, I think we're just going to close the ticket
13:08:16 <Penguinpee> With the conclusions of the brief discussion, maybe?
13:08:43 <Penguinpee> It's probably in the logs of a previous meeting.
13:08:51 <FranciscoD>
13:09:04 <FranciscoD> yeh, you said "Looks like this can be closed" :D
13:09:42 <FranciscoD> I'll update docs with a section on EPEL and open a PR
13:10:45 <Penguinpee> Yeah. I believe we concluded that we have no joint agreement on what to do with EPEL. Some packagers like supporting it, others only do it upon request if feasible.
13:12:10 <FranciscoD> yeh, I'm just going to add your suggested text to our docs
13:13:32 <FranciscoD> one sec, just opening the pr now
13:14:49 <FranciscoD> #info -> add snippet about EPEL support
13:15:01 <Penguinpee> Sounds good.
13:15:34 <FranciscoD> #info AniketPradhan[m] to update #552 -> DONE
13:16:03 <FranciscoD> #info adding packit is easy: copy over conf file from one of the other repos, like snakemake, update the github url, open a PR, then use packit commands.
13:16:46 <FranciscoD> #info FranciscoD_ to update RHBZ#2213973 -> DONE
13:17:07 <FranciscoD> those were all the tasks, all done
13:17:16 <FranciscoD> #topic Open pagure tickets
13:17:24 <FranciscoD> #info
13:17:38 <Penguinpee> I saw some activity and your recent commit on python-mizani. I haven't looked into it yet, but if it's as easy as adding the config, I need to make a habit of doing so whenever I submit PRs.
13:18:43 <FranciscoD> yeh, here's an example:
13:19:39 <FranciscoD> copied .packit.yaml from snakemake, tweaked the first line, checked if tags use a `v` prefix, in which case this is required
13:19:41 <FranciscoD> and then commit
13:19:56 <FranciscoD> I do it as a PR, because one has to have a PR to then use the packit commands
13:20:14 <Penguinpee> Thanks. Do we have a preference or is there some limitation on which packages should be Packit enabled?
13:20:39 <FranciscoD> it only works for packages from github/gitlab/bitbucket at the mo
13:20:41 <FranciscoD> not for pypi
13:20:57 <FranciscoD> and of course, if there are patches etc., then we need to make further tweaks manually
13:21:15 <FranciscoD> but one can check out the packit branch, make the tweaks and open a new PR etc.---as we'd normally do
13:21:32 <Penguinpee> All right. I'll start with simple GitHub packages then.
13:21:35 <FranciscoD> this is how I did this one IIRC:
13:22:15 <FranciscoD> took the packit PR, then added another one to update license
13:22:46 <FranciscoD> #info packit dashboard:
13:23:13 <FranciscoD> #info note, dashboard does not show koji builds that packit does for the "pull from upstreams" jobs
13:24:12 <FranciscoD> #info vxl update: -> WIP
13:24:30 <FranciscoD> chipping away on this one, but it's more work than lots of simpler FTBFS fixes and updates, so I keep doing those too
13:24:50 <FranciscoD> no other tix, we've already looked at the EPEL support one.
13:24:58 <Penguinpee> If upstream does not publish releases on GitHub, will `trigger: release` still work?
13:25:12 <FranciscoD> probably not
13:25:37 <FranciscoD> but again, I'm not sure if it relies on tags or releases..
13:25:46 * Penguinpee hopes there's a `trigger: tag` as well then
13:25:53 <FranciscoD> I think it looks at tags
13:26:05 <Penguinpee> Anyway, I've got some reading to do.
13:26:09 <FranciscoD> which is why one has to provide the tag prefix if upstream uses `v1.2.3` instead of just `1.2.3`
13:26:20 <FranciscoD> Penguinpee:  here's the page:
13:26:48 <Penguinpee> Yeah, the tag thingy is a constant cat and mouse game with forge macros as well.
13:27:11 <FranciscoD> #topic Package health check
13:27:12 <Penguinpee> The forge macros actually expect the 'v' prefix.
13:27:19 <FranciscoD> ah, really?
13:27:25 <FranciscoD> but I assume one can tweak that?
13:27:39 <FranciscoD> #info neuro-sig packager dashboard is here:
13:27:41 <Penguinpee> Yup. If not present, one needs to define 'tag' manually.
13:27:45 <FranciscoD> cool
13:28:03 <FranciscoD> I'm glad the forge macros are being maintained, they are very helpful
13:28:18 <Penguinpee> They are being revamped too...
13:29:04 <FranciscoD> #info Order of urgency/priority: CVE/FTI bugs in stable releases, then in rawhide along with FTBFS bugs, then other bugs + new version updates
13:29:36 <FranciscoD> #info Note that the beta freeze is in place, so all updates will remain in testing unless they're proposed and accepted as freeze exceptions
13:29:39 <Penguinpee> I think in my mind I don't even look into f37 issues anymore.
13:30:23 <Penguinpee> Most of the Python3.12 FTBFS/FTI bugs in neuro-sig are now dealt with.
13:30:41 <FranciscoD> yeh, f37 is now old, folks should be upgrading to f38
13:30:43 <FranciscoD> at least
13:31:12 <Penguinpee> I'm curious regarding the f38 bugs, though.
13:31:20 <FranciscoD> #info To propose a freeze exception, use the blocker bugs app:  ; bugs are proposed as freeze exceptions or blockers, so link your update to a bug and then use the bug
13:32:03 <FranciscoD> #info FTI bugs that have updates fixing them are generally accepted as freeze exceptions
13:32:24 <FranciscoD> #info SOP:
13:32:43 <FranciscoD> #info and:
13:33:20 <FranciscoD> #info Most of the py3.12 FTI/FTBFS bugs have been sorted; some others are waiting for upstream interventions
13:33:40 <FranciscoD> uh, lots of CVEs for tiff in neuron again ..
13:33:46 * FranciscoD has half a mind just to close them all..
13:34:28 <Penguinpee> Yeah. They keep coming, but our bundled TIFF is kinda a mummy already...
13:34:41 <FranciscoD> yeh, with neuron, I'm really not that worried
13:34:59 <FranciscoD> such a niche tool, and very unlikely to ever have anything privacy/security sensitive being analysed in it
13:35:22 <FranciscoD> biophysical models of neurons do not contain any private data
13:36:06 <Penguinpee> Depends on who's neurons these are. 🙃
13:36:10 <FranciscoD> #info a few packages have new versions available: please take these on and open PRs where possible
13:36:34 <Penguinpee> I guess I can lean into updating packages now...
13:36:48 <FranciscoD> Penguinpee:  usually they're no ones neurons---they're models with params fitted from experimental data, so they don't really contain any of the data themselves
13:37:31 <FranciscoD> #info just FYI: here's an example of a human cortical circuit model in NEURON (and LFPy):
13:38:06 <FranciscoD> #info accompanying paper:
13:38:44 <FranciscoD> #info Packages are in good health
13:38:47 <Penguinpee> I was just being silly. I totally agree that impact would be minimal.
13:39:01 <FranciscoD> :D
13:39:10 <FranciscoD> #topic Open package reviews check
13:39:25 <FranciscoD> #info Neuro-sig package reviews should block the tracker bug:
13:39:33 <FranciscoD> no new reviews at the moment
13:39:49 <FranciscoD> #info no new reviews at the moment
13:39:56 <Penguinpee> How difficult a package / language is nimble?
13:40:24 <FranciscoD> I really don't know. I think the tricky bit is the first package, when one has to bootstrap it so that it can build itself
13:40:29 <FranciscoD> I'vve just never done it before
13:40:47 <FranciscoD> I also don't know how useful a language it is tbh. Seems to have been around for a while
13:40:56 <FranciscoD> I wish taskopen upstream hadn't switched to it..
13:42:05 <Penguinpee> Yeah. It feels like introducing a new language is beyond my skill set. Yet, I'm intrigued...
13:42:42 <FranciscoD> it doesn't seem to have lots of parts, and is self-contained, so should be fine
13:42:51 <FranciscoD> just gotta figure out the initial bootstrapping bit
13:42:57 <FranciscoD> folks on -devel would know
13:43:23 <FranciscoD> #topic CompNeuro lab image status check
13:43:29 <FranciscoD> #info
13:43:37 <FranciscoD> #info All green ticks: building nicely
13:43:48 <FranciscoD> #topic Neuroscience query of the week
13:43:54 <FranciscoD> not a lot new recently
13:43:57 <Penguinpee> If I can find the time, I may poke around a bit.
13:44:01 <FranciscoD> +1
13:44:31 * FranciscoD takes a quick peep at planet neuroscientists
13:44:33 <FranciscoD>
13:47:38 <FranciscoD> hrm, nothing really noteworthy there
13:47:47 <FranciscoD> #topic Next meeting day, and chair
13:47:57 <FranciscoD> same time, in two weeks, hopefully over matrix if the bot is ready?
13:48:18 <Penguinpee> Works for me.
13:49:01 <FranciscoD> #info Next meeting in two weeks at same time, hopefully over matrix (otherwise, fall back again to irc)
13:49:05 <FranciscoD> #topic Open floor
13:49:09 <Penguinpee> I'm in the bot dev channel on Matrix. If there's any movement towards a release, I'll let you know.
13:49:14 <FranciscoD> Anything that wasn't on the agenda can go there
13:49:15 <FranciscoD> awesome!
13:49:25 <FranciscoD> I see it's coming along, think it's being tested in staging?
13:50:00 <Penguinpee> Yeah. It's working rather well. But it uses STG backend still,
13:50:55 <FranciscoD> cool, looking forward to having it working
13:51:31 <FranciscoD> OK, if there's nothing else, let's close the meeting and use the remaining ~10 minutes for a break :D
13:51:36 <FranciscoD> closing in 3
13:51:43 <FranciscoD> 2
13:52:22 <FranciscoD> 1
13:52:23 <Penguinpee> 💣 <-- bad detonator
13:52:27 <FranciscoD> LOL
13:52:30 <FranciscoD> #endmeeting