{data,lib}[foss/2023b] cfchecker v4.1.0, cfunits v3.3.7, cftime v1.6.4 - #25995
Conversation
Updated software
|
|
@boegelbot please test @ jsc-zen3 |
|
@sassy-crick: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de PR test command '
Test results coming soon (I hope)... Details- notification for comment with ID 4708740212 processed Message to humans: this is just bookkeeping information for me, |
|
Test report by @boegelbot |
Crivella
left a comment
There was a problem hiding this comment.
In general we are trying to move to using Bundles also for single packages.
I added an example to convert one of the 2 ECs to one, could you convert also the other?
Co-authored-by: Davide Grassano <34096612+Crivella@users.noreply.github.com>
|
Test report by @Crivella |
Crivella
left a comment
There was a problem hiding this comment.
There is already an EC for cfunits with the same version in 2024a.
https://github.com/easybuilders/easybuild-easyconfigs/blob/develop/easybuild/easyconfigs/c/cfunits/cfunits-3.3.7-gfbf-2024a.eb
There cftime is added as an extension instead of through ntecdf4-python which allows to keep the TC to a lower level (gfbf instead of foss).
Is there a reason here it was done in a such a manner?
It not i would follow the same approach as the existing EC as avoids introducing a lot of not needed dependencies
@Crivella Ah I see, my thought was just to use an already installed package rather than installing it again as an extension. It does make sense though to keep it down to In that case, though, would it make more sense to revert the dependencies? We install What is confusing to me is having the same library repeated multiple times as an extension across different modules, feels a bit unorganised and to me it doesn't really fit the idea of a "single" version per toolchain, as we can end up with different versions also across these multiple modules. |
|
In general if a package is used multiple times as a dependency i agree it would be best to add it as a separate dependency for improved re-usability, and ensuring we do not ship multiple versions in a TC. There are currently talks on how we can improve this, either through CI checks or other methods. I think in general it is about finding a balance on how to granularize bundles (i think the original idea was to avoide having a huge amount of entries in the PATH-like environment variables) In this case whether we want to split |
Okay, I think for the sake of this PR I'll separate @Crivella Regarding |
|
I would not do it in this PR and keep it for a separate one |
Updated software
|
Perfect, I'll open it when I have some time;) |
There was a problem hiding this comment.
Judging from cftime docs and the diff with the previous cfunits EC we will probably need Cython to make sure the build works atleast for cftime
EDIT:
== 2026-06-22 17:46:45,931 run.py:650 INFO Output of '/home/crivella/.local/easybuild/software/Python/3.11.5-GCCcore-13.2.0/bin/python ...' shell command (stdout + stderr):
Using pip 23.2.1 from /home/crivella/.local/easybuild/software/Python/3.11.5-GCCcore-13.2.0/lib/python3.11/site-packages/pip (python 3.11)
Processing /home/crivella/.local/easybuild/build/cftime/1.6.4/gfbf-2023b/cftime/cftime-1.6.4
Preparing metadata (pyproject.toml): started
Running command Preparing metadata (pyproject.toml)
Compiling /home/crivella/.local/easybuild/build/cftime/1.6.4/gfbf-2023b/cftime/cftime-1.6.4/src/cftime/_cftime.pyx because it changed.
[1/1] Cythonizing /home/crivella/.local/easybuild/build/cftime/1.6.4/gfbf-2023b/cftime/cftime-1.6.4/src/cftime/_cftime.pyx
running dist_info
I guess on my system it picked it up from somewhere else but we should definitely add it as a builddep
Emmmmmmm true, you're right:) Just added it now That's quite confusing though, since early ECs for But it has to be there, checking the |
…asyconfigs into 20260514164119_new_pr_cfchecker410
|
Test report by @Crivella |
|
@boegelbot please test @ jsc-zen3 |
|
@Crivella: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de PR test command '
Test results coming soon (I hope)... Details- notification for comment with ID 4778329137 processed Message to humans: this is just bookkeeping information for me, |
|
Test report by @boegelbot |
|
Going in, thanks @stamenminkov! |
No AI usage