Wolfgang Maier

Ah, yes, sorry for missing the workflow group meeting (had a different meeting at the same time). Tool-based is certainly an alternative though the computation of the metadata flag I'm...

> because now the behavior depends on when a tabular dataset has been produced (which is almost impossible to tell for users Yes, that's the big caveat I thought of,...

see https://bioconda.github.io/recipes/snipit/README.html#package-snipit

https://docs.galaxyproject.org/en/latest/dev/schema.html#tool-tests-test-output decompress | When this attribute is true and compare is set to diff, try to decompress files if needed. This flag is useful for testing compressed outputs that are...

> but that's about tool outputs, this is about inputs, or did I miss something again ? that's what I thought, too, but from the usage pattern in tools-iuc it...

> > but that's about tool outputs, this is about inputs, or did I miss something again ? > > that's what I thought, too, but from the usage pattern...

> https://github.com/galaxyproject/tools-iuc/search?p=3&q=decompress%3D%22true%22 seems to only be contained in `` tags, are there other cases ? see above: quite some tests ask for output decompression but then compare it to a...

This is still an issue in 24.2. We had a complaint about this just yesterday. passwords seem to get sanitized at the tool wrapper level so when users have passwords...

Ok, so with some experimenting I found that the wrapper can access the unsanitized password with sth like: ``` #set $username = $__user__.extra_preferences.get('account|username', "") #set $password = $__user__.extra_preferences.get('account|password', "") #if...

More input/output and analysis description is required in the README and as part of input annotations for me to review this properly.