Michael Hanke

Yes, that is correct, and also the desired behavior of `download()`. There will likely be *no* feature to make `download` add a file to an annex, or register a URL...

FTR: I have spent the last few days looking at the problem and what it would take to support them. I believe it boils down to being able to mangle...

https://github.com/datalad/datalad-next/pull/589 now brings an implementation that is a basis for an alternative `status` implementation that properly supports such typechange scenarios.

Given that the typechange detection update never made it into the mainline, this got stalled. However, https://github.com/datalad/datalad-next/pull/589 brings an alternative basis for a diff command that does typechange detection and...

https://github.com/datalad/datalad-next/pull/589 has an implementation that behaves as proposed here.

Also this is no longer an active issue, I want to put on record another undesired complexity that this decorator causes: platform path variability. For example, I can pass a...

Thanks for the analysis! FTR: #3456 is concerned with a related topic.

This issue is fixed when installing datalad-next 1.0.0b3 and enabling its datalad-runtime-patching via ``git config --global --add datalad.extensions.load next``.

I was using a simple method of a Python class instance, not a static method. For now I have made it work by wrapping ```py ` wsh = _WebDavSiblingHandler( url=url,...

The following patch would restore this behavior: ```diff diff --git a/datalad/cli/main.py b/datalad/cli/main.py index bb68e46c8..1106b99db 100644 --- a/datalad/cli/main.py +++ b/datalad/cli/main.py @@ -208,6 +208,14 @@ def _run_with_exception_handler(cmdlineargs): # in general we do...