Hello, I tested the build on two different bullseye machines, just had to first manually compile and install libsis-jhdf5-java and fastqc. https://tracker.debian.org/pkg/fastqc explains that it was removed because of a problem with libsis-jhdf5-java that is already in new.
My hunch is that this is also the reason why multiqc fails in p/cowbuilder, but this needs some extra investigation. I have at least one local project for which I want to use this package and likely will find/fix the one or other problem with it on they way. Today, 8 packages that are dependencies for the pigx-scRNAseq single-cell package were accepted. This is great. Almost there. Sadly, I yet failed to overcome an error during testing for scRNAseq. Additional eyeballs would be welcome. On the data front I have not yet made up my mind. If we have a few reference installations, say one for structural bioinformatics and another one for transcriptomics, then I started to like the idea to use datalad to distribute such readily deployable bits. How each reference installation then maintains itself is a different problem. And where do we put tutorials? Jupyter somewhere? Best, Steffen
