Maxime U Garcia

OK, I have seen it this is BAD. Let's try to figure out what is in the container and reproduce as much as we can in the conda env

I found a hack for testing topic in subworkflows, adding this emission at the end of the emit block will not disrupt anything and collect the topic versions for the...

and I found another hack for the modules test: ``` ['versions', process.out[1], process.out[2], process.out[3]] ``` with the first output ([0]) being the index, and the last 3, 3 different topic...

I don't think it should be that complex (famous last words)

Currently we have stubs implemented properly at the module and subworkflow level, but pipeline level stubs are incomplete. Main issue for pipeline level stubs is that rnaseq is checking the...

Issue is most likely due to read group names coming from the test data, could be probably be solved with some smart args cf https://gatk.broadinstitute.org/hc/en-us/community/posts/360074797471-Read-Names-of-Public-Fastq-Files-incompatible-with-Picard-s-MarkDuplicates

This is currently fixed by a patch that keep the same version of picard, but issue is most likely still there and will need fixing in future update, so let's...

Having this issue now on `25.10.0`, with `NXF_SYNTAX_PARSER=v2`, but I assume that was expected

probably need to update the snapshots

Do we really need all that for tests?