I should find a discussion of this before committing, but I think leaving superclass arguments out of signatures for derived classes (and leaving only **kwargs) can leave available arguments opaque, particularly when just looking at the function definitions.
I should find a discussion of this before committing, but I think leaving superclass arguments out of signatures for derived classes (and leaving only **kwargs) can leave available arguments opaque, particularly when just looking at the function definitions.
Part of an overall push to better document things