Stefan Appelhoff

that's really neat, thanks for the example! but it seems that the state of adoption in big projects is still very early. issue for support in numpydoc: https://github.com/numpy/numpydoc/issues/196

Yes, I don't particularly like the inheritance principle either. There have been several calls to formalize it "better" (or formalize in the first place), see e.g., https://github.com/bids-standard/bids-specification/issues/102 For the sake...

> I'd say @sappelhoff is in charge :sweat_smile: :sweat_smile: Let's just continue our mixture of consensus-based decision making and do-ocracy, so far it seems to work. > if they make...

Just noticed that perhaps #617 is related

+1 to both your suggestions (especially the first one, which should be relatively straight forward)

Hi @adam2392, as you already point out, ... > there are quite a number of "optional" fields and on top of that, users can also define arbitrary fields in e.g.,...

> However, there is no stimulus channel type in I guess `TRIG` is kind of a stimulus channel  to me this error could be due to either the dataset,...

> After seeing the issues @sappelhoff ran into with annotations and orig_time and first_samp and all these weird interactions again (some recent MNE-Python PR), I don't think we should allow...

> I don't really see the point of annotations.tsv anyway? I also think that it'd be better to make full use of events.tsv than to start on a "new file"...

> or as anonymize can be a dict we can add a valid key to say if age should be written that sounds good to me, re-using our existing API,...