#meeting:fedoraproject.org log

<@ankursinha:fedora.im>

13:02:17

!startmeeting NeuroFedora - 2024-02-26

<@meetbot:fedora.im>

13:02:17

Meeting started at 2024-02-26 13:02:17 UTC

<@meetbot:fedora.im>

13:02:17

The Meeting name is 'NeuroFedora - 2024-02-26'

<@ankursinha:fedora.im>

13:02:41

!topic neurofedora

<@ankursinha:fedora.im>

13:02:50

!undo

<@ankursinha:fedora.im>

13:02:55

!meetingname neurofedora

<@ankursinha:fedora.im>

13:03:08

!topic Agenda

<@ankursinha:fedora.im>

13:03:12

!link https://neuroblog.fedoraproject.org/2024/02/26/next-open-neurofedora-meeting-26-february-1300-utc.html

<@gui1ty:fedora.im>

13:03:48

!hi

<@zodbot:fedora.im>

13:03:49

Sorry, could not get info from FASJSON (code 500)

<@ankursinha:fedora.im>

13:04:18

!topic New introductions and roll call

<@ankursinha:fedora.im>

13:04:22

!hi

<@zodbot:fedora.im>

13:04:23

Sorry, could not get info from FASJSON (code 500)

<@ankursinha:fedora.im>

13:04:27

oh no..

<@music:fedora.im>

13:04:28

!hi

<@zodbot:fedora.im>

13:04:28

Sorry, could not get info from FASJSON (code 500)

<@nerdsville:fedora.im>

13:04:31

!hi

<@zodbot:fedora.im>

13:04:32

Sorry, could not get info from FASJSON (code 500)

<@music:fedora.im>

13:04:40

There is a Kerberos problem right now.

<@music:fedora.im>

13:05:08

`fkinit -u music` gives me `kinit: Generic error (see e-text) while getting initial credentials`, and I just had a CI job fail the same way

<@gui1ty:fedora.im>

13:05:11

Mondays be like that ...

<@ankursinha:fedora.im>

13:05:35

Ah, do infra know, would you know music ?

<@ankursinha:fedora.im>

13:05:43

status.fp.o doesn't say anything

<@music:fedora.im>

13:06:00

I commented in the infrastructure chat about a minute ago

<@ankursinha:fedora.im>

13:06:08

I think this ticket covers it too: https://pagure.io/fedora-infrastructure/issue/11793

<@ankursinha:fedora.im>

13:06:43

Anyway, let's move on with the agenda then

<@ankursinha:fedora.im>

13:06:53

!topic Tasks from last meeting

<@ankursinha:fedora.im>

13:06:59

!link https://meetbot.fedoraproject.org/meeting_matrix_fedoraproject-org/2024-02-12/neurofedora.2024-02-12-13.01.html

<@ankursinha:fedora.im>

13:07:09

Penguinpee speak to design/mindshare etc. about unique logos for the meeting rooms on matrix to make it easier for folks to get to them

<@ankursinha:fedora.im>

13:07:19

any update here Penguinpee ?

<@gui1ty:fedora.im>

13:07:33

I opened a ticket, but no response yet. Will chase ...

<@ankursinha:fedora.im>

13:07:38

cool

<@ankursinha:fedora.im>

13:07:58

!info Penguinpee speak to design/mindshare etc. about unique logos for the meeting rooms on matrix to make it easier for folks to get to them -> ticket filed -> DONE

<@ankursinha:fedora.im>

13:08:13

reassigning the steps issue

<@ankursinha:fedora.im>

13:08:17

!action everyone please look at https://pagure.io/neuro-sig/NeuroFedora/issue/566 and see if you can help out with the compilation issue

<@ankursinha:fedora.im>

13:08:34

Ankur: file ticket with fesco about nodejs etc. -> not done yet, reassigning

<@ankursinha:fedora.im>

13:09:04

!action Ankur: file ticket with fesco about nodejs etc. (related to https://pagure.io/neuro-sig/NeuroFedora/issue/563)

<@ankursinha:fedora.im>

13:09:15

Penguinpee look into the possibility of automatically disabling anitya scratch builds when packit is enabled (instead of us having to manually disable scratch builds for all repos)

<@gui1ty:fedora.im>

13:09:22

Discussion is ongoing, Implementation is blocked on missing API: https://pagure.io/fedora-infrastructure/issue/11791

<@gui1ty:fedora.im>

13:09:53

Oh wait, that was in relation to PRs. nm

<@ankursinha:fedora.im>

13:10:14

!info Penguinpee look into the possibility of automatically disabling anitya scratch builds when packit is enabled (instead of us having to manually disable scratch builds for all repos) -> DONE (discussion ongoing)

<@gui1ty:fedora.im>

13:10:19

Anyway, tickets filed. Developers are discussing it.

<@ankursinha:fedora.im>

13:10:35

Ankur file issue against dashboard about FTBFS issues---python-lsp-server as example -> also PENDING :(

<@ankursinha:fedora.im>

13:10:38

!action Ankur file issue against dashboard about FTBFS issues---python-lsp-server as example

<@ankursinha:fedora.im>

13:10:50

!info Ankur prioritise fixing and updating python-pynn FTBFS (since it breaks our comp-neuro iso image compose task) -> DONE

<@ankursinha:fedora.im>

13:11:03

but 0.12.2 is out now, but will hopefully be an easier update

<@ankursinha:fedora.im>

13:11:16

!topic Open Pagure tickets

<@ankursinha:fedora.im>

13:11:39

!link https://pagure.io/neuro-sig/NeuroFedora/issues?status=Open&tags=S%3A+Next+meeting

<@ankursinha:fedora.im>

13:11:50

I don't see any new ones that need commenting here---they're all WIP

<@ankursinha:fedora.im>

13:12:14

vxl is at 90% compilation, but needs some patching from the looks of it

<@ankursinha:fedora.im>

13:12:33

!topic Packages health check

<@ankursinha:fedora.im>

13:12:37

!link https://packager-dashboard.fedoraproject.org/dashboard?groups=neuro-sig

<@music:fedora.im>

13:13:04

I looked at https://pagure.io/neuro-sig/NeuroFedora/issue/566, followed the steps, and got ``` Failed to resolve the transaction: No match for argument: sundials2-devel No match for argument: sundials2-mpich-devel No match for argument: sundials2-openmpi-devel ``` What *should* I be seeing?

<@ankursinha:fedora.im>

13:14:30

sundials2 isn't in F40+ from the looks of it: https://packages.fedoraproject.org/pkgs/sundials2/sundials2/

<@ankursinha:fedora.im>

13:14:42

maybe that needs to updated to only `sundials`?

<@ankursinha:fedora.im>

13:15:24

https://github.com/CNS-OIST/STEPS/issues/23

<@gui1ty:fedora.im>

13:15:29

Let me revisit the ticket and update it with current information. It's been a while. I may have cleaned up Copr since.

<@ankursinha:fedora.im>

13:15:42

it doesn't build with new sundials, though, so upstream suggests using the bundled one (2.6.x)

<@music:fedora.im>

13:17:01

I’m trying again `--without system_sundials`.

<@music:fedora.im>

13:18:27

``` In file included from /builddir/build/BUILD/STEPS-4.1.1/test/unit/cpp/test_mol_state.cpp:1: /builddir/build/BUILD/STEPS-4.1.1/src/steps/mpi/dist/tetopsplit/mol_state.hpp:9:10: fatal error: mpi/dist/tetopsplit/definition/fwd.hpp: No such file or directory 9 | #include "mpi/dist/tetopsplit/definition/fwd.hpp" | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ```

<@music:fedora.im>

13:18:38

That’s not what was blocking the update before, is it?

<@ankursinha:fedora.im>

13:18:52

the netpyne update worked for me locally, but fails zuul etc. I'll look into that: https://src.fedoraproject.org/rpms/python-netpyne/pull-request/6

<@ankursinha:fedora.im>

13:19:57

not sure, haven't checked this one recently

<@music:fedora.im>

13:20:31

Penguinpee: Pull me in after you double-check the status, and I’ll see if there’s anything I can help with?

<@music:fedora.im>

13:21:05

Looks like you might need `xvfb-run` for even the smoke test: ``` + /usr/bin/python3 -sP /usr/lib/rpm/redhat/import_all_modules.py -f /builddir/build/BUILD/python-netpyne-1.0.6-2.fc41.s390x-pyproject-modules -e '*optuna*' -e '*sbi*' Check import: netpyne Warning: no DISPLAY environment variable. --No graphics will be displayed. Traceback (most recent call last): ```

<@gui1ty:fedora.im>

13:21:28

Will do. I'll probably have to start over. It's been a while since I last worked on it.

<@ankursinha:fedora.im>

13:21:54

possibly, but upstream doesn't use it, and the "no graphics will be displayed" bit means that matplotlib's UI bits won't be used at all

<@ankursinha:fedora.im>

13:22:08

I'll dig into it more, just weird that it works with a local mockbuild

<@ankursinha:fedora.im>

13:22:31

``` Traceback (most recent call last): File "/usr/lib/rpm/redhat/import_all_modules.py", line 171, in <module> main() File "/usr/lib/rpm/redhat/import_all_modules.py", line 167, in main import_modules(modules) File "/usr/lib/rpm/redhat/import_all_modules.py", line 100, in import_modules importlib.import_module(module) File "/usr/lib64/python3.12/importlib/__init__.py", line 90, in import_module return _bootstrap._gcd_import(name[level:], package, level) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "<frozen importlib._bootstrap>", line 1387, in _gcd_import File "<frozen importlib._bootstrap>", line 1360, in _find_and_load File "<frozen importlib._bootstrap>", line 1331, in _find_and_load_unlocked File "<frozen importlib._bootstrap>", line 935, in _load_unlocked File "<frozen importlib._bootstrap_external>", line 995, in exec_module File "<frozen importlib._bootstrap>", line 488, in _call_with_frames_removed File "/builddir/build/BUILDROOT/python-netpyne-1.0.6-2.fc41.x86_64/usr/lib/python3.12/site-packages/netpyne/__init__.py", line 24, in <module> from netpyne import batch File "/builddir/build/BUILDROOT/python-netpyne-1.0.6-2.fc41.x86_64/usr/lib/python3.12/site-packages/netpyne/batch/__init__.py", line 13, in <module> from .batch import Batch File "/builddir/build/BUILDROOT/python-netpyne-1.0.6-2.fc41.x86_64/usr/lib/python3.12/site-packages/netpyne/batch/batch.py", line 33, in <module> from .grid import gridSearch, getParamCombinations File "/builddir/build/BUILDROOT/python-netpyne-1.0.6-2.fc41.x86_64/usr/lib/python3.12/site-packages/netpyne/batch/grid.py", line 27, in <module> import imp ModuleNotFoundError: No module named 'imp' ```

<@ankursinha:fedora.im>

13:22:36

so could be related to this too ^

<@ankursinha:fedora.im>

13:22:43

again, should work if works locally in mock..

<@music:fedora.im>

13:23:16

I think that traceback is just what an import failure looks like. But it’s weird that it’s OK in local mock.

<@ankursinha:fedora.im>

13:23:57

Alessio: you around?Any chance you'll be able to look at alizams? https://bugzilla.redhat.com/show_bug.cgi?id=2260976

<@ankursinha:fedora.im>

13:25:05

a few other FTBFS bugs that we should prioritise---the python ones are probably py3.13 related

<@ankursinha:fedora.im>

13:25:29

a few other FTBFS bugs that we should prioritise---some of the python ones are probably py3.13 related

<@gui1ty:fedora.im>

13:25:50

Build fails with: ``` CMake Error at /usr/share/cmake/Modules/FindPackageHandleStandardArgs.cmake:230 (message): Could NOT find Boost: Found unsuitable version "1.83.0", but required is exact version "1.81.0" (found /usr/include, ) ```

<@ankursinha:fedora.im>

13:25:51

rather, since F40 isn't going to see py3.13: https://fedoraproject.org/wiki/Changes/Python3.13

<@ankursinha:fedora.im>

13:26:30

boost is now at 1.83 in F40+: https://packages.fedoraproject.org/pkgs/boost/boost/

<@music:fedora.im>

13:26:32

aha, `fkinit` worked!

<@music:fedora.im>

13:26:34

!hi

<@zodbot:fedora.im>

13:26:36

Benjamin Beasley (music) - he / him / his

<@music:fedora.im>

13:26:43

welcome back, zodbot

<@ankursinha:fedora.im>

13:26:44

maybe loosening the bound would be worth trying to begin with

<@music:fedora.im>

13:27:15

definitely agree with that. boost compatibility is usually pretty good

<@music:fedora.im>

13:27:41

(they don’t have enough people to change all that spaghetti code at once 😆)

<@gui1ty:fedora.im>

13:28:10

I'll take a look at `alizams`. Shouldn't be too hard to fix.

<@ankursinha:fedora.im>

13:28:48

weirdly enough, searching the GH repo for Boost doesn't return a lot: https://github.com/search?q=repo%3AAlizaMedicalImaging%2FAlizaMS%20Boost&type=code

<@ankursinha:fedora.im>

13:28:57

I must be doing something wrong there..

<@music:fedora.im>

13:32:47

I think you’re right about `netpyne`. The `DISPLAY` thing is just a warning and is a red herring.

<@music:fedora.im>

13:33:15

Also, I scrolled down in the output you pasted, and `ModuleNotFoundError: No module named 'imp'` is certainly real.

<@ankursinha:fedora.im>

13:33:55

yeh, I've contributed some code to netpyne (we look after the neuroml bits) and my understanding is that they provide an option to run without DISPLAY so that it can be used on clusters

<@ankursinha:fedora.im>

13:34:15

otherwise with MPI, matplotlib tries to create a fontcache etc. on each process and crashes

<@ankursinha:fedora.im>

13:34:44

ah, here is my PR: https://github.com/suny-downstate-medical-center/netpyne/pull/788

<@music:fedora.im>

13:36:05

It is certainly true that `netpyne/batch/grid.py` imports `imp`, and that’s not OK because it was removed in Python 3.12.

<@music:fedora.im>

13:36:32

If it’s succeeding anywhere, it’s because one of the dependencies still brings in `python3-zombie-imp` indirectly.

<@ankursinha:fedora.im>

13:37:01

I'd even updated their CI to test for 3.12, but they don't run import tests so maybe that's why it hasn't been caught: https://github.com/suny-downstate-medical-center/netpyne/pull/789

<@ankursinha:fedora.im>

13:37:26

yeh, should be a simple enough patch to fix that one error I'd hope

<@ankursinha:fedora.im>

13:37:52

I can take a look at it---I've been working on the netpyne NeuroML bits recently already

<@music:fedora.im>

13:38:37

Trivial uses of `imp` are trivial to replace; the long tail of obscure uses are sometimes *really* hard.

<@music:fedora.im>

13:38:50

`fedpkg --release rawhide mockbuild --enablerepo=local -- --dnf` fails the same way as CI/koji does

<@music:fedora.im>

13:39:06

so some dependency dropped `python3-zombie-imp` and it hasn’t hit a compose yet

<@ankursinha:fedora.im>

13:39:50

hrm, I'll scrub my mock cache also just to be sure it's clean and up to date

<@ankursinha:fedora.im>

13:40:03

Any bugs/packages that we need to prioritise?

<@ankursinha:fedora.im>

13:40:13

otherwise we'll keep plodding along as we usually do :)

<@music:fedora.im>

13:40:23

LMK if you run into any trouble patching it.

<@ankursinha:fedora.im>

13:40:32

cool, will do

<@ankursinha:fedora.im>

13:40:44

!action Ankur patch out imp import in netpyne to fix FTBFS

<@ankursinha:fedora.im>

13:41:32

Note: python-limits was orphaned, but nothing depends on it, so we'll probably let it die

<@ankursinha:fedora.im>

13:42:13

!info A few package updates are blocked by needing new deps

<@gui1ty:fedora.im>

13:43:12

Also: `raptor` was orphaned. It's a dependency for `COPASI`.

<@ankursinha:fedora.im>

13:43:53

it does have another admin from the looks of it

<@music:fedora.im>

13:43:58

Ankur Sinha (FranciscoD@fedora) (he/him/his): Thanks for reviewing and merging the `python-pyspike` PR in advance of Cython 0.29 retirement.

<@ankursinha:fedora.im>

13:44:17

well, someone with commit access

<@gui1ty:fedora.im>

13:44:28

Yeah, we "only" co-maintain. I'd rather the main admin picks that up.

<@music:fedora.im>

13:44:43

Totally selfishly and subjectively, I really don’t enjoy working on that package…

<@ankursinha:fedora.im>

13:45:18

let's let it die then, it isn't even really a science related package, was just a dep for something from the looks of it

<@gui1ty:fedora.im>

13:45:31

I gave it an update just before letting go of it ...

<@ankursinha:fedora.im>

13:45:32

I can ping rex to see what they think

<@ankursinha:fedora.im>

13:45:54

seems to be a fairly low maintenence package too, so I could pick it up to keep it alive

<@ankursinha:fedora.im>

13:46:07

!action Ankur ping raptor co-maintainer to see if they'd take it over

<@gui1ty:fedora.im>

13:46:20

It should be ported to `raptor2` to be honest.

<@ankursinha:fedora.im>

13:46:27

!info python-limits orphaned: not needed by anything, not a science package either, so we let it go

<@ankursinha:fedora.im>

13:47:14

COPASI? we can ask sagitter about that too---I'm sure they'll look at it when they get a chance, they're very active

<@ankursinha:fedora.im>

13:47:32

!topic Open package reviews check

<@ankursinha:fedora.im>

13:47:55

!link https://bugzilla.redhat.com/show_bug.cgi?id=fedora-neuro

<@gui1ty:fedora.im>

13:48:03

Yea, `COPASI` should port to `raptor2` - it's the reason `raptor` was orphaned. Too old, no longer maintained.

<@ankursinha:fedora.im>

13:48:31

!info 4 packages in review, 4 more looking for reviwers

<@ankursinha:fedora.im>

13:48:45

any of the new ones urgent or blocking updates?

<@ankursinha:fedora.im>

13:48:56

i think there's another snakemake plugin that's required, right?

<@gui1ty:fedora.im>

13:49:15

`sandia-omega-h` is an update dependency.

<@gui1ty:fedora.im>

13:49:36

for `python-steps`

<@ankursinha:fedora.im>

13:49:55

!info sandia-omega-h is required for steps: we'll prioritise it

<@gui1ty:fedora.im>

13:50:01

I've reviewed that already.

<@music:fedora.im>

13:50:32

Thanks to everyone for reviewing a ton of snakemake packages so we could make snakemake 8 happen…

<@gui1ty:fedora.im>

13:51:13

`pyuca` is now waiting on a response from legal regarding my PR proposal for license change.

<@ankursinha:fedora.im>

13:51:39

I'll take up sandia-omega-h now and try to get it done this week

<@ankursinha:fedora.im>

13:51:44

!action Ankur review sandia-omega-h

<@ankursinha:fedora.im>

13:52:35

!topic CompNeuro lab compose status check for Fedora 40/rawhide

<@ankursinha:fedora.im>

13:52:44

!link https://koji.fedoraproject.org/koji/packageinfo?packageID=30691

<@ankursinha:fedora.im>

13:52:53

!info Back to green tick again

<@ankursinha:fedora.im>

13:52:58

(now that pynn was fixed)

<@ankursinha:fedora.im>

13:53:05

!topic Neuroscience query of the week

<@ankursinha:fedora.im>

13:53:17

I didn't see anything too interesting this week

<@gui1ty:fedora.im>

13:53:25

You've pynned it down. 🤣

<@ankursinha:fedora.im>

13:53:36

https://copyleaks.com/about-us/media/copyleaks-research-finds-nearly-60-of-gpt-3-5-outputs-contained-some-form-of-plagiarized-content

<@ankursinha:fedora.im>

13:53:41

https://www.schneier.com/blog/archives/2024/02/ais-hacking-websites.html

<@ankursinha:fedora.im>

13:53:47

more LLM/generative AI bits ^

<@ankursinha:fedora.im>

13:54:15

A role for the cerebellum in motor-triggered alleviation of anxiety: https://www.cell.com/neuron/fulltext/S0896-6273(24)00007-2?rss=yes

<@ankursinha:fedora.im>

13:55:19

^this one is intersting: "Physical exercise is known to reduce anxiety, but the underlying brain mechanisms remain unclear. Here, we explore a hypothalamo-cerebello-amygdalar circuit that may mediate motor-dependent alleviation of anxiety. This three-neuron loop, in which the cerebellar dentate nucleus takes center stage, bridges the motor system with the emotional system. Subjecting animals to a constant rotarod engages glutamatergic cerebellar dentate neurons that drive PKCδ+ amygdalar neurons to elicit an anxiolytic effect. Moreover, challenging animals on an accelerated rather than a constant rotarod engages hypothalamic neurons that provide a superimposed anxiolytic effect via an orexinergic projection to the dentate neurons that activate the amygdala. Our findings reveal a cerebello-limbic pathway that may contribute to motor-triggered alleviation of anxiety and that may be optimally exploited during challenging physical exercise."

<@ankursinha:fedora.im>

13:56:11

TLDR: when anxious, move and get some exercise

<@ankursinha:fedora.im>

13:56:54

this one was good too, related to navigation: https://www.nature.com/articles/d41586-024-00230-5 "A neural circuit for navigation keeps flies on target"

<@ankursinha:fedora.im>

13:57:44

for folks interested in neuromorphic, spinnaker v2 for ML: https://arxiv.org/abs/2401.04491: SpiNNaker2: A Large-Scale Neuromorphic System for Event-Based and Asynchronous Machine Learning

<@ankursinha:fedora.im>

13:59:00

An improved Hopfield network: https://arxiv.org/abs/2008.02217

<@ankursinha:fedora.im>

13:59:11

so, turns out there were a few interesting things..

<@ankursinha:fedora.im>

13:59:46

!topic Next meeting day, and chair

<@ankursinha:fedora.im>

13:59:49

same time, 2 weeks?

<@gui1ty:fedora.im>

14:00:12

Works for me.

<@nerdsville:fedora.im>

14:00:41

👍️

<@ankursinha:fedora.im>

14:00:54

!agreed next meeting in 2 weeks at 1300 UTC

<@ankursinha:fedora.im>

14:00:59

!topic Open floor

<@ankursinha:fedora.im>

14:01:20

a couple of minutes here for anything we missed, otherwise we can continue in the channel after I close the meeting

<@gui1ty:fedora.im>

14:01:21

Have fun tonight! 😉

<@music:fedora.im>

14:01:49

I am not sure if anyone else had packages that were waiting on it, or had the dependency patched, out, but `python-pyside6` just passed package review: https://bugzilla.redhat.com/show_bug.cgi?id=2265554

<@music:fedora.im>

14:02:23

It took a long time for someone to get it packaged. I took a look at it once and ran away screaming.

<@ankursinha:fedora.im>

14:03:11

!endmeeting