Hello, What do you guys think? If it is not completely over the top then I would like to have the effort as part of the continuous integration testing. The result should then go into a git repository from where the world can retrieve it. I know, bandwidth, but it should not be not too bad, I tend to think, as in "below 1TB/month".
Why git? I think to spot differences when rerunning bits locally, a presumed-efficient transfer of large files, and to somehow arrange for multiple genomes in separate branches that all diverge from a "code-only" branch, so we always have the complete info on the provenance of the files. And I am not aware of any git repository for genomic data, yet, so I would like know how this would be accepted. What do you think? Would there be an audience for such a repository? It would be completely distribution-agnostic from how I imagine any such effort. And we should identify all the gaps we possibly can to have it adopted by the typical NGS service groups. Best, Steffen
