<inserted> scribenick: kaz
Kaz: Zoltan is OK with the above
minutes
... any problems/comments form the others?
Daniel: np
Mizushima: np
Kaz: I'm also OK, so let's approve the minutes
(previous minutes approved)
(Zoltan joins)
<inserted> scribenick: zkis
Zoltan: We have some ReSpec issues. Maybe we should update the source, following the best practices in https://github.com/w3c/web-nfc
Daniel: we should work on defining InteractionOptions
Zoltan: Currently at https://w3c.github.io/wot-scripting-api/#idl-def-interactionoptions
... We need to remove the markdown line from the doc header in the spec. Then it renders fine.
<dape> https://www.w3.org/TR/wot-scripting-api/#idl-def-interactionoptions
Daniel: the simplest is to extend with another optional field that provides an index to the preferred binding
Zoltan: should it be a single value or an Array?
... what is the policy, just a hint or if that index doesn't work then should it fail?
Daniel: the easiest would be to fail, since the script could choose another one, or remove the index hint
Zoltan: formIndex as name?
Daniel: sounds fine
Zoltan: so we agreed how to extend, in the name and what does it mean
Daniel: what about it being optional?
Zoltan: InteractionOptions is a dictionary all are optional at the moment
Daniel: ok, it should not be required
<zkis> https://github.com/w3c/wot-scripting-api/issues/200
Daniel: will prepare a proposal with Ege in the issue
Zoltan: about writeProperty return values we have already agreed. Zoltan takes the task to create PR.
Daniel: this is a breaking change in the API
... problem with scripts already running
Zoltan: what I planned to do at the moment was to return "any" from reads, but explain in more detail how to deal with Forms, content types and DataSchema
Daniel: in node-wot the "any" gets interpreted according to the media type.
Zoltan: the thing is that there are already solutions in the Web Platform and would be nice to use them here as well
... maybe we can create a new issue about it
Zoltan: having issue doing smart things instead of the scripts
Daniel: it should be the same as reading the property
Zoltan: when do we provide the value?
Daniel: we raised that in the TD spec
Zoltan: indeed we could handle it saying it's the same as a read done with the same bindings
... but there are issues
... not all devices will behave in the same way
[adjourned]