W3C

WoT Scripting API

02 August 2021

Attendees

Present
Cristiano_Aguzzi, Daniel_Peintner, Kaz_Ashimura, Tomoaki_Mizushima
Regrets
Zoltan
Chair
Daniel
Scribe
cris, kaz

Meeting minutes

minutes

Daniel: minutes are ok

<kaz> July-26

Daniel: ok minutes approved

Quick Updates

vacation plans

Daniel: no updates
… it might be diffcult to join on 16th

Cristiano: yeah also from my side

Daniel: zoltan will be able to join on the 9th

PRs

<dape> https://github.com/w3c/wot-scripting-api/pull/329

Daniel: we have still a pending PR from Cristiano: 329 PR 329 - feat: move thing-description types to a dedicated package
… we are waiting for feedback from Zoltan

Cristiano: no pressure from my side

issues

Daniel: we are small group and there are not so many updates
… not sure what to do

Cristiano: we could just do small clean up

Issue 310

<kaz> Issue 310 - Handling pagination

Daniel: issue 310 is about discovery right?

Cristiano: not only it talks about pagination in discovery but it can be used as a built in functionality

TypeScript - Type DataSchemaValue circularly references itself #301

<dape> Issue 301 - TypeScript - Type DataSchemaValue circularly references itself

Cristiano: I think it was solved

Daniel: I would close it

Issue 107

<kaz> Issue 107 - Very high frequency updates

Daniel: what happens when a thing fires many events in a short time?

Cristiano: is it more protocol specific?

Daniel: we could state the maximum updates in the TD and follow the regulation

Daniel: I think we should add the propose closing label here, and just wait for feedback

Kaz: we should have clear detailed use case, especially for complicated features like this
… if we don't get any we could defer it

issue 117

<kaz> Issue 117 - Extract normative assertions for testing

Daniel: this issue is about assertions in the document
… since scripting in not on the rec check
… do we need this normative assertions?
… or even a test plan?

Kaz: we don't have to work on testing, since Scripting API is not a group Note at the moment.
… maybe we could define the test plan if we have extra time

Cristiano: might be really difficult to define a testing suite

Daniel: I agree we could do behavioral testing (i.e. unit test suites)

Cristiano: it might be a future improvement

Daniel: there was I company who could develop those tests
… but it never took off
… dape close in favor 190

Issue 193

<kaz> Issue 193 - Should writeProperty() return a value?

Daniel: my take is to allow both use cases
… if the td allows it we should allow to return the data
… it is more flexible

Daniel: it should not be required

Cristiano: the api will be more complex, my opinion is that we should well document this behavior

Daniel: yes understand, but as you said there are already concrete implementations that does so

Cristiano: we have to keep in mind that this feature might only work for a specific protocol
… so when checking the return value developers should somehow know the protocol used

Cristiano: do we have a concrete TD of a Philips HUE device

Kaz: we can think about this specific device implementation, but the discussion should be moved to profile calls or work with them

Cristiano: should we move this issue somewhere else then?

Kaz: yeah or at least link

Daniel: we have an open issue about it in the thing description repository

Kaz: we could ask this in the main call then

Daniel: ok!

Issue 200

<kaz> Issue 200 - Error Handling

Daniel: (adds comments)

<kaz> Daniel's comments

issue 212

<kaz> Issue 212 - Partial TD validation for input of produce() method

Cristiano: I propose to close it and open small issues for left open topics

Daniel: I agree
… we are out of time
… let's close the meeting now

<kaz> [adjourned]

Minutes manually created (not a transcript), formatted by scribe.perl version 136 (Thu May 27 13:50:24 2021 UTC).