nextstrain / conda-base

Conda package build for nextstrain-base
https://anaconda.org/Nextstrain/nextstrain-base
1 stars 1 forks source link

Add nextclade2 #48

Closed victorlin closed 6 months ago

victorlin commented 7 months ago

Description of proposed changes

The newest version of nextclade contains a symlink to nextclade2.¹ This will eventually replaced with a symlink to nextclade3.

To continue availability of nextclade2, add a package that provides the nextclade2 binary.²

¹ https://github.com/bioconda/bioconda-recipes/commit/c20aa93b2f346df36d26f9c00bd059ae0c1a5dda ² https://github.com/bioconda/bioconda-recipes/commit/b08d5c611de349a55396ecf0735ed1f3f26e8fff

Related issue(s)

Checklist

corneliusroemer commented 7 months ago

I'm not sure what will happen when we have two times the same binary. Once as a symlink to Nextclade and once as a binary itself.

Will conda notice and complain? It probably should.

Nextclade==2.* and nextclade2 should be mutually exclusive, there might be a way to encode this into conda.

victorlin commented 7 months ago

The nextstrain-base package builds fine, but there is a warning when installing it and linking dependencies:

warning  libmamba [nextclade2-2.14.0-h9ee0642_0] The following files were already present in the environment:
    - bin/nextclade2

I don't see why they should be made mutually exclusive. While both of them contain nextclade2, those are functionally identical so it doesn't matter that one overwrites the other. Once nextclade no longer has nextclade2 but instead has nextclade3, this allows what is currently supported in the Docker image - access to both nextclade2 and nextclade3.

corneliusroemer commented 7 months ago

Yep this should work - though the warning is a little annoying. Up to you whether we wait or not with merging.

corneliusroemer commented 6 months ago

We should merge this PR soon (tomorrow) so that nextclade2 remains available when we release nextclade(3) on bioconda tomorrow