W3C

WoT Scripting API

15 Feb 2021

Agenda

Attendees

Present
Kaz_Ashimura, Cristiano_Aguzzi, Zoltan_Kis, Philipp-Alexander_Blum, Tomoaki_Mizushima, Daniel_Peintner
Regrets
Chair
Daniel
Scribe
zkis

Contents


<scribe> scribe: zkis

Agenda

<dape> https://www.w3.org/WoT/IG/wiki/WG_WoT_Scripting_API_WebConf#Agenda

approval of previous minutes

<dape> --> https://www.w3.org/2021/02/08-wot-script-minutes.html

Daniel: there are discussions about partial TD in the TD spec as well

Minutes approved.

Introduction of Philipp Blum (Invited Expert)

Philipp: writing a RIOT implementation for WoT

Daniel: any running implementation in RIOT?

Philipp: yes, on the last plugfest already provided a TD to work with the impl
... in the next plugfest more extensive examples will be provided

<inserted> Philipp's contribution to the PlugFest during TPAC

Cristiano: welcome Philipp, also interested in the RIOT implementation

call time change

Daniel: CA proposed changing the time

Cristiano: yes, we need more time to discuss the issues
... the agenda keeps growing, so we need more time

Daniel: there is an issue: security call after Scripting
... also, before the Scripting meeting I am not available

Zoltan: other time on Monday would be difficult as well

Daniel: we could check the available slots

<Zakim> kaz, you wanted to suggest we ask Philipp to introduce himself briefly after this prev minutes topic :)

Kaz: schedule is pretty busy, it's kind of overkill to have a call every week

Zoltan: we could be more efficient by starting earlier and solve fast the past minutes approvals

Daniel: difficult for me to come earlier

Kaz: we could also use github in a better way

Cristiano: yes, we could be more productive

Daniel: so we keep the time for now

Virtual F2F

<dape> --> https://www.w3.org/WoT/IG/wiki/F2F_meeting,_March_2021

Daniel: we need to craft an agenda for Scripting

<cris> +1 for the skeleton

Zoltan: in general, we will have a status update, implementations update, testing update and current issues that need attention from the whole group

Daniel: yes, we can sketch that and elaborate later

PR 289

<inserted> PR 289 - Add definitions for partialTD

Cristiano: there were some minor edits

Zoltan: asks about the tab selectors

Cristiano: the examples are changed on click
... security issue still open
... the algorithm was improved in small ways
... there are portions that are not well defined yet anywhere

Zoltan: change "applying schema to init" to " return the validation of init with schema" and refer to outside steps

Cristiano: we have 2 levels of validations, one as a Thing init, one as Thing model

Daniel: JSON Schema is not a spec, so we can't really use it in algorithms
... maybe we want to call "schema" as "ExposedThing" schema

Cristiano: yes, we can improve that

Daniel: anyone else is also welcome to comment on the PR

issue#299

Daniel: Chose a particular security schema for an ExposedThing, https://github.com/w3c/wot-scripting-api/issues/299

Cristiano: this comes from the Security TF discussion

Zoltan: the notions of Scripting API and management API are mixed up, we should clarify the scopes
... we could have all management APIs as service exposed via a Thing

Cristiano: should we standardize management API?

Daniel: in node-wot we do have lifecycle methods etc
... there is no solution to fit all

Zoltan: right, there are multiple levels of standardization opportunities. Some APIs could be exposed as service (Thing), some others could be standardized.

Cristiano: the need to tackle with security from app level is quite well defined

Daniel: could we handle it with security handlers?

Zoltan: that would be a man in the middle for the API

Cristiano: node-wot issue: everything is clear-text

Daniel: we have the conformance classes, a 4th could be a management API?

Zoltan: no, because we'd need e.g. 2 separate security sections, as they are not at the same level
... the issue can be left open, as the Scripting TF has to solve it, but it's out of scope for the Scripting API

<dape> Should title be mandatory in ExposedThingInit?, https://github.com/w3c/wot-scripting-api/issues/300

issue#300

Cristiano: MMc pointed out that titles should not be generated, but given by developers, so we should require it

Daniel: but if not provided, there are use cases that don't need a title, should it be empty?

Zoltan: so could we use an empty title as default?

Cristiano: doesn't solve the problem

Zoltan: true enough

Daniel: will take a look how node-wot will handle it

Adjourned.

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
$Date: 2021/03/01 06:17:24 $