W3C

– DRAFT –
WoT Scripting API

28 November 2022

Attendees

Present
Cristiano_Aguzzi, Daniel_Peintner, Jan_Romann, Kaz_Ashimura, Tomoaki_Mizushima, Zoltan_Kis
Regrets
-
Chair
Daniel
Scribe
dape

Meeting minutes

Minutes

-> November-21
… No objections, minutes are approved

Quick updates

Publication Schedule within Jan 2023

Daniel: Plan to have updated draft within January

Holidays plans

Daniel: available Dec 5, Dec 12 and Dec 19

<cris_> +1

Daniel: restart Jan 9

<Mizushima> +1

Zoltan: Makes sense for everyone

Kaz/Mizushima: Jan 9 holiday in Japan

Kaz: most Japanese people will not be available on Jan 9
… I could join call on the 9th

Zoltan: We can have a *smaller* call

Daniel: Updates over Github?
… resume on 16th

Daniel: Try to be active week of Jan 9

PRs

Daniel: No updates yet

Zoltan: draft PR
https://github.com/zolkis/wot-scripting-api/commits/discovery-experiment1

Zoltan: need to add more algorithmns
… some more work
… need to refactor as well
… try to complete by next week

Daniel: Great

Zoltan: Please look at PR
… single ThingFilter only
… url used in case of needed
… from Discovery spec, do we need URL parameter ?

Cristiano: searching in directory... Discovery spec does not have this use
… filter based on body of TD

Zoltan: I see

Cristiano: url can be used in SPARQL

Zoltan: single ThingFilter enough

Cristiano: Keep it simple is good
… not much interest of standardizing different discovery mechanisms across bindings
… we have steps for HTTP and CoAP
… we miss other protocols
… could be moved to binding templates

Zoltan: for scripting. Shall we use local discovery ?

Cristiano: Not sure
… we might have such use-cases

Zoltan: Okay, lets omit url
… we don't use it nor do we know how to use it
… suggest to remove url

Cristiano: Need to ask ourselves whether there is a use-case

Zoltan: WoT runtime needs internal process
… local (Bluetooth or NFC) might be a use-case
… at this time we should keep it generic
… hence I suggest to remove

Cristiano: Agree

Zoltan: Bluetooth bridge in Discovery spec?

Cristiano: out of spec...

Zoltan: separate spec ?

Cristiano: Discovery spec just lists them
… I wonder whether we can encapsulate it in a generic fashion

Zoltan: e.g, runtime groups bluetooth devices in own directory

Daniel: will bluetooth/NFC part of next spec?

Zoltan: don't think we need that
… I wonder whether local discovery is allowed
… private directory?

Daniel: Out of scope of Discovery

Cristiano: local host does not make sense for BLE

Zoltan: As a bridge
… how to tell discovery that i would like to have local things

Cristiano: not possible to describe "locality"
… obtain URL via Multicast

Zoltan: discover local things via Discovery API

Cristiano: Yes, possible
… privacy concerns

Zoltan: dedicated method or "local" filter

Cristiano: Hard to decide
… tend to filter

Cristiano: need to define what local means

Zoltan: correct. Might depend on implementation also

Cristiano: for node-wot things in runtime are local
… WebThings as multicast plugin

Daniel: How does discovery spec for local things?

Cristiano: intro phase vs exploration
… one method in intro phase allows MDNS etc
… it generates an URL
… there are some constraints
… BLE is completely open

Jan: whether local method would do 2-step process in one?

Cristiano: Acts almost the same
… follow all the phases
… URL discovery and fetch
… others do just exploration

Zoltan: I think we can define local method later
… combines intro+exploration
… not just a filter
… I will not do it know
… should track it in an issue
… future work

Jan: postponing sounds good

<cris__> +1

Zoltan: Will summarize in issue

Daniel: May I ask you do create draft PR for your commits

Zoltan: Will do

[adjourned]

Minutes manually created (not a transcript), formatted by scribe.perl version 196 (Thu Oct 27 17:06:44 2022 UTC).