W3C

– DRAFT –
WoT Scripting API

22 January 2025

Attendees

Present
Daniel_Peintner, Jan_Romann, Kaz_Ashimura
Regrets
-
Chair
Cristiano
Scribe
dape

Meeting minutes

Previous minutes

Jan 8

Cristiano: Minutes look fine
… minutes are approved

PRs

Extended return type of invokeAction() PR 561

<kaz> PR 561 - Extended return type of invokeAction()

Cristiano: PR that created new issues in TD repo
… TD is underspecified
… a *real* fix will happen in TD2.0
… as part of manageable actions
… for now we can not expect anything
… for the Scripting API we cannot assume anything
… the application layer should be in charge
… we can have 2 methods
… queryaction and cancelaction
… application can just read byte stream

Daniel: fear the calls might (should) change in the future .. and that might conflict

Cristiano: deprecate functions for TD 1.1
… since it is a major version we can loose backward compatibility

Kaz: We cannot fix the core problem quickly
… in any case, there might be incompatible functions in TD 2.0 (possibly not only the return type changes but also some more possible features)

Daniel: unsure what it gives us to have very open calls now

Cristiano: We are getting closer to feature complete ..
… anyhow, there are others like read/write multiple

Cristiano: can add experimental features

Daniel: Need to change the return type of invokeAction

Cristiano: Disagree
… can have simple calls next to invokeAction

Daniel: What about parallel actions?

Cristiano: Just one action at the time
… a problem of the TD in the first place

Daniel: need to think about :-)

Jan: Couldn't we use a similar pattern like observing
… having a dedicated action invokeAsyncAction ?

Cristiano: Assume URI variable to be used to identify which action has been started
… invoke action returns ID

Cristiano: The runtime cannot know that it needs to pass the ID to cancel/query

Daniel: not much difference in having the same optionality on external method or method related to the ActionObject

Cristiano: deprecation in future would cause less issues

Kaz: The discussion so far is nice, but we should clarify which we're talking about, (1) how to fix the bug for 1.1 or (2) how to improve the situation for 2.0.

Cristiano: idea is to fix the bug in 1.1
… staying conservative

Jan: I agree with Cristiano
… should be focusing on fixing the bug
… wondering whether there is another use-case for data mapping

Cristiano: Yes, good point
… for the future

Daniel: I think we can describe here where we are heading too
… but create another PR

Jan: Is this going to be the last "fix" for TD1.1 ?

Cristiano: I don't have an answer right now
… some TD1.1 features we will not be implemented.. I think.. like read/write multiple
… we can have 2 version .. one in a branch etc

Jan: Sounds good to me

Cristiano: Yes, having branch "1.x" that tackles TD1

Kaz: Should report back in main call later today

<Cristiano describing the outcome in w3c/wot-scripting-api#561>

Kaz: Additional features to the note should not cause any problem
… no additional implementation report is needed since Scripting API is a note

Cristiano: Good discussion
… see w3c/wot-scripting-api#561 (comment)

[Adjourned]

Minutes manually created (not a transcript), formatted by scribe.perl version 242 (Fri Dec 20 18:32:17 2024 UTC).