Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • Aug 19 17:21
    isuruf closed #1813
  • Aug 19 17:21
    isuruf commented #1813
  • Aug 19 15:42
    ssurbhi560 synchronize #1812
  • Aug 19 12:58
    markounikau labeled #1813
  • Aug 19 12:58
    markounikau opened #1813
  • Aug 17 08:30
    ssurbhi560 review_requested #1812
  • Aug 17 08:30
    ssurbhi560 opened #1812
  • Aug 17 04:09
    isuruf closed #1800
  • Aug 17 04:04

    github-actions[bot] on gh-pages

    deploy: f477db8b950d94a9b4d6a02… (compare)

  • Aug 17 04:00

    isuruf on main

    add announcement for dropping P… Merge branch 'conda-forge:main'… remove mention of dropping inst… and 4 more (compare)

  • Aug 17 04:00
    isuruf closed #1811
  • Aug 17 03:47
    h-vetinari synchronize #1811
  • Aug 17 03:35
    h-vetinari synchronize #1811
  • Aug 17 03:32
    h-vetinari commented #1811
  • Aug 17 03:32
    h-vetinari synchronize #1811
  • Aug 17 03:31
    h-vetinari synchronize #1811
  • Aug 17 03:28
    h-vetinari synchronize #1811
  • Aug 17 02:47
    isuruf commented #1811
  • Aug 13 00:11
    ngam commented #1784
  • Aug 12 20:12
    ocefpaf commented #1052
hcwinsemius
@hcwinsemius
Sebastian Ehlert
@awvwgk:matrix.org
[m]
Can the outputs section in a recipe contain noarch: python and compiled packages (without Python requirement in host)? The conda-build documentation doesn't mention anything on this, but it doesn't include much details on outputs anyway.
3 replies
Christian Brueffer
@cbrueffer
A PR I merged today ended up in an empty feedstock (https://github.com/conda-forge/perl-file-chdir-feedstock), what's the best way to fix this?
Matt Chan
@thewchan
Is it possible to use dev version of poetry in a recipe? Seems like that's what's required for the pacakge. I see that it's available for installing generally, but I'm not sure how to specify for it in a recipe (the instruction I found is here: conda-forge/poetry-feedstock#64)
9 replies
Nathan Roach
@NatPRoach
Hi all, I'm running into a small issue and I'm not sure if it's something I've done wrong or if it's just a bug in CI. A PR I made was merged recently and this feedstock was setup https://github.com/conda-forge/conda-env-builder-feedstock. However it seems like the actual package never ended up building. Is there something I can do to get the CI to actually kick off the first time build? I looked through the documentation and found this https://conda-forge.org/docs/maintainer/infrastructure.html#ci-build-services but didn't see anything that I could figure out how to get it working.
2 replies
Curtis Mayberry
@curtisma
Good evening, could someone from staged-recipes please help me merge softworks recipe PR?
1 reply
Sebastian Ehlert
@awvwgk:matrix.org
[m]

Sorry for bothering again. I have a couple of Python recipes ready for review, some have been setting in staged recipes for >1 month now:

Reviews, comments and merges are much appreciated. Thanks in advance.

Curtis Mayberry
@curtisma
@beckermr or another core member:
Would you mind helping me apply to join staged-recipes?
2 replies
Emily Jane McTavish
@snacktavish

Hello! I am trying to locally test my recipe before submitting a PR to staged recipes ( as described in https://conda-forge.org/docs/maintainer/adding_pkgs.html#staging-test-locally). I have forked the repo, and updated my recipe in /recipes repo: https://github.com/snacktavish/staged-recipes

However, when I try to run the tests using
python build-locally.py from the top level dir, a bunch of stuff runs (I could share output), but at the end I get "There are no tempfile(s) to remove.
There are no logfile(s) to remove.
Found no recipes to build".
Any tips?

2 replies
Thomas A Caswell
@tacaswell
The CI skipping does not seem to be working: conda-forge/mongoquery-feedstock#15
Ben Mares
@maresb
Heads up, setuptools recently deleted distutils.msvccompiler and this has been wreaking havoc with many different packages. pypa/setuptools#3532
Should have mentioned: a quick workaround is to pin setuptools <65.
alisterburt
@alisterburt

Hello, sorry for what I'm sure is a naive question, I'm working on my first recipe...

The package has requires CUDA so in the build requirements I have specified - {{ compiler('cuda') }} # [linux64], this is effective for the CUDA builds but dependencies fail to resolve for linux64 with

-   - nothing provides requested nvcc_linux-64 None.*

conda-forge/staged-recipes#20084, there are other real problems here but this would be a nice step forwards! Thanks :-)

Thomas A Caswell
@tacaswell
We are having issues with getting the latest Matplotlib patch release to build on linux CPython. I think I have isolated it to a change in conda-build, but if that is the case I would have expected other projects to also be failing conda-forge/matplotlib-feedstock#325
1 reply
Isuru Fernando
@isuruf:matrix.org
[m]
Isuru Fernando
@isuruf:matrix.org
[m]
It's a pkg-config bug. We really should move to pkgconf as our pkg-config implementation instead of the reference implementation from freedesktop
See conda-forge/freetype-feedstock#49 for a workaround in freetype2
Thomas A Caswell
@tacaswell
Thanks @isuruf:matrix.org ! It would have taken me a very long time to get there...
Hameer Abbasi
@hameer-spire
Hello! I'm trying to build a package with maturin, but I'm getting the following error:
- package maturin-0.13.0-py37hbd0741f_0 requires python_abi 3.7.* *_cp37m, but none of the providers can be installed
1 reply
Harvey South
@HarveySouth
Hello, I am wondering if there is any way to view download statistics across all packages, just relating to information kind similar to how PyPi has a "trending" section to pip package searches
5 replies
Giancarlo Castro
@giancastro
Hey! Has anyone seen this feedstock check error? CircleCI Pipeline — Could not find a usable config.yml because you deleted the CircleCI OAuth app.
1 reply
Jack Sundberg
@jacksund
I have the same issue as @giancastro (message right above this one), but rerendering didn't fix it. Is there another way to fix this? conda-forge/simmate-feedstock#30
1 reply
Akira Koseki
@akoseki2

Hi, I would like to contact conda-forge/core. The site github.com/conda-forge/dpm360-lightsaber-feedstock has no active maintainers now. So dear staff, could you merge the following PR to add myself to the maintainers?

conda-forge/dpm360-lightsaber-feedstock#3

1 reply
Hameer Abbasi
@hameer-spire

Hello, I'm running into an issue: /home/ubuntu/mambaforge/envs/.../conda-bld/pygnssr2_1660827664678/_h_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_pl/bin/python: No such file or directory

Happens with both boa and conda-build locally.

7 replies
Mike Sarahan
@msarahan

I seem to have messed up a couple of my feedstocks. PRs pass, but when the feedstock tries to upload the packages, it fails.

Passing PR: conda-forge/dart-sdk-feedstock#5
Failing build on merge: https://dev.azure.com/conda-forge/feedstock-builds/_build/results?buildId=553498&view=results

What did I mess up here? I'm seeing it across a few of my feedstocks, but not in the general queue of all builds, thus I suspect something I have done.

Chris Burr
@chrisburr
I think it's an anaconda.org issue https://github.com/conda/infra/discussions/592
1 reply
Isuru Fernando
@isuruf
Did someone restart jaxlib-feedstock 35 times (1000+jobs)?
18 minutes ago
Leo Fang
@leofang
btw the Azure job retrigger seems to be broken (it used to work fine), we've noticed in the past few days "Rerun failed checks" would rerun everything (with a new parent job) , not just the failed ones
it might be the cause?
Uwe L. Korn
@xhochy
Didn't we have had a flag in the past to limit the parallelism in Azure for certain packages? I think getting that back into smithy would come in handy for the arrow-cpp-feedstock :(
Isuru Fernando
@isuruf
@xhochy, it's still there
but at 50 jobs
maybe we should reduce that for arrow-cpp
this limit is per build though
If you have 4 builds each with 70 jobs, it will still do 4*50 jobs in parallel
Keith Kraus
@kkraus14
I apologize for the arrow-cpp-feedstock churn :(. The recent abseil changes caused a lot of triaging and troubleshooting that involved lots of CI jobs unfortunately.
3 replies
Tom de Geus
@tdegeus
On macOS (and only on macOS) I intermittently get : Could not find suitable distribution for Requirement.parse('FrictionQPotFEM==0.20.0'). Rerunning solves it...?! See e.g. conda-forge/frictionqpotfem-feedstock#76 . Note that the package has a pybind11 extension, but it seems that it is bugging even on the pure Python part. Does anyone have an idea?
lorenz
@looooo
what does this mean?
Failed to upload due to Command '['anaconda', '--quiet', '-t', '/tmp/tmpvgo5w872/binstar.token', 'upload', '/home/conda/feedstock_root/build_artifacts/linux-64/freecad-0.21.pre-py310h606b19c_25.tar.bz2', '--user=freecad', '--channel=dev']' returned non-zero exit status 1.. Trying again in 87.96388244628906 seconds
Using BINSTAR_TOKEN for anaconda.org uploads to freecad.
INFO:binstar:Using Anaconda API: https://api.anaconda.org
[ERROR] ('Method Not Allowed: Specified method is invalid for this resource. ([POST] https://api.anaconda.org/commit/freecad/freecad/0.21.pre/linux-64/freecad-0.21.pre-py310h606b19c_25.tar.bz2 -> 405)', 405)
Chris Burr
@chrisburr
It means anaconda.org is broken :) conda-forge/status#132
When did it happen? If it was more than 12 hours ago I'd say retry the job
(as it seemed to be working again when I tried last night, but maybe the issue is back)
lorenz
@looooo
ah, thanks, yes it happened yesterday
smautner
@smautner:matrix.org
[m]
Hello all, Thanks for helping to get my recipe accepted. I have a feedstock now: https://github.com/conda-forge/biofilm-feedstock, the package however was not added to the confa-forge channel and the build failed, i assume this is because i addded the dependencies at the same time. Can somebody give me a hint on how to make the system retry? (pushing to a branch might help but that is discouraged in the documentation)
3 replies
Tom de Geus
@tdegeus
This message was deleted
2 replies
Tom de Geus
@tdegeus

I get a conda segmentation fault on macOS arm:

/Users/runner/miniforge3/etc/profile.d/conda.sh: line 9:  3572 Segmentation fault: 11  "$CONDA_EXE" $_CE_M $_CE_CONDA "$@"
/Users/runner/miniforge3/lib/python3.9/multiprocessing/resource_tracker.py:216: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown
  warnings.warn('resource_tracker: There appear to be %d '
##[error]Bash exited with code '139'.

https://github.com/conda-forge/frictionqpotfem-feedstock

How should I proceed. (I don't really know what the error comes from, neither where to report it, or to work around it)

Isuru Fernando
@isuruf
@tdegeus, try restarting that job
Seems like a lief bug with the new version
Jan Janssen
@jan-janssen
Can somebody from core merge conda-forge/pyneqsys-feedstock#1 ? It is a recipe which was created two years ago but never successfully build, so it seems to me the maintainers are inactive.
Jan Janssen
@jan-janssen
And there is one more conda-forge/chempy-feedstock#2 ?