🎉 Conda 23.3.1 & conda-build 3.24.0 Releases

News
, ,
1

March releases are here! Both conda 23.3.1 and conda-build 3.24.0 are now available on both main and conda-forge:

To update conda run:

conda install -n base conda=23.3.1

To update conda-build run:

conda install -n base conda-build=3.24.0

Both conda and conda-build have dropped Python 3.7 support and efforts are underway to add Python 3.11 support for the May release. Conda and conda-build continue to deprecate and remove old, unused, and outdated code (e.g. conda-build drops conda <4.13 logic).

Conda adds improved error messages for environment creation via conda env create, introduces new configuration parameters, adds 64-bit RISC-V architecture compatibility, enhances code coverage, and adds experimental support for faster repodata.json parsing via incremental fetching.

Conda-build addresses a Git LFS failure, reduces false-positives when detecting Perl modules in conda skeleton cpan, and starts to address a subdir issue with testing downstream packages.

Checkout conda’s changelogs for 23.3.0 and 23.3.1 and conda-build’s changelog for 3.24.0 for a complete list of improvements.

Snake away!

4 Likes
2

Here’s a less than two minute screen capture demonstrating the speed difference between conda search -c conda-forge "conda>23" in conda 23.1 and the new 23.3.1 release on an empty cache. On my machine it takes 16.9 seconds on conda 23.1 but only 5.9 seconds on conda 23.3.1.

2 Likes
3

That’s good news! Thank you all.
Please confirm or correct: these updates are also compatible with a miniconda distribution.

4

Yes, you should be able to upgrade conda inside the base environment of any miniconda or anaconda distribution. Some users have found that it is necessary to pass --repodata-fn=repodata.json to solve the upgrade based on all packages instead of those in the smaller current_repodata.json.

1 Like