[wot-discovery] minutes - 24 February 2020

available at:
  https://www.w3.org/2020/02/24-wot-discovery-minutes.html

also as text below.

Thanks a lot for taking the minutes, Christian!

Kazuyuki

---
   [1]W3C

      [1] http://www.w3.org/

                               - DRAFT -

                             WoT-Discovery

24 Feb 2020

   [2]Agenda

      [2] https://www.w3.org/WoT/IG/wiki/WG_WoT_Discovery_WebConf#24_Feb_2020

Attendees

   Present
          Kaz_Ashimura, Michael_McCool, Andrea_Cimmino,
          Christopher_Ambler, Christian_Glomb, Daniel_Peintner,
          Tomoaki_Mizushima, Zoltan_Kis, Kunihiko_Toumura

   Regrets

   Chair
          McCool

   Scribe
          glomb

Contents

     * [3]Topics
         1. [4]Invited Guest
         2. [5]Previous minutes
         3. [6]Agenda for today
         4. [7]Repo Reorg
         5. [8]Issues and PRs
         6. [9]DID review
         7. [10]Discovery in Scripting API
         8. [11]AOB
     * [12]Summary of Action Items
     * [13]Summary of Resolutions
     __________________________________________________________

   <kaz> scribenick: glomb

Invited Guest

   McCool: Chris Ambler is invited

Previous minutes

   <inserted> [14]Feb-17 minutes

     [14] https://www.w3.org/2020/02/17-wot-discovery-minutes.html

   minutes feb 17 review: note on publication schedule done

   previous minutes approved

Agenda for today

   McCool: would include michael kosters slides

   Zoltan: different things: discovery in scripting api vs
   semantic discovery

Repo Reorg

   repo reorg: wot discovery re-structured

   background needs update

Issues and PRs

   <inserted> [15]Issue 3

     [15] https://github.com/w3c/wot-discovery/issues/3

   issue #3

   e.g. mdns

   clarify core resource directory

   proposal: two-phase-directory-based discovery

   PR wanted for alternative ideas

   clean up directory.md

   readme for use-cases

   use issues (problem) and PR (suggestion)

   or add to already existing PRs

   Zoltan: proposal as a whole (PR)

   then split up

   issues and PRs

   <inserted> [16]Issue 2

     [16] https://github.com/w3c/wot-discovery/issues/2

   issue #2: relevant for security TF

   <inserted> [17]Issue 3

     [17] https://github.com/w3c/wot-discovery/issues/3

   issue #3: McCool will elaborate

DID review

   review of DID:

   <McCool>
   [18]https://github.com/w3c/wot-architecture/blob/master/proposa
   ls/2020-02-WoT-DID.pdf

     [18] https://github.com/w3c/wot-architecture/blob/master/proposals/2020-02-WoT-DID.pdf

   w3c wg

   e.g. with the help of blockchain

   url scheme includes method + method specific id

   resolves JSON document

   kind of directory service (decentralized)

   how to use in WoT: thing ids

   dump TD info into DID document...

   or point to TD directory

   Andrea: point to discovery service (deal better with security)

   McCool: or point to thing directly (.well-known/core)

   authenticate before

   similar approaches

   lacking: more general concept

   device may not host own TDs -> directory service

   typed links

   some other non-directory based way to retrieve TDs?

   Andrea: blockchain could do
   ... upcoming paper - how to retrieve RDF from blockchain

   DID will have multiple methods how to retrieve documents, no
   specification how it works behind the scene

   not just Core RD in our proposal

   to be elaborated

Discovery in Scripting API

   <zkis> [19]https://w3c.github.io/wot-scripting-api/

     [19] https://w3c.github.io/wot-scripting-api/

   discovery in scripting api

   discover method defined in Scripting API

   thing filter: search everything / directory / local / multicast

   local: bluetooth, nfc, beacon, local hardware, ...

   McCool: propose to make it more abstract

   directories might be represented as a thing, then multicast
   might find it

   discovery - two phase approach might meanigful

   "multicast" might be re-phrased

   DOMstring e.g. TD fragment

   how to authorize - to be done before

   application should figure out whether it is directory or thing

   or: are just things returned?

   location important - currently not included, maybe query
   property

   infer location may be a problem - see issue #2

   location of things as well as browser

   definition in scripting api to be alligned with discovery TF
   work

   in sensitive context disallow multicast

   Zoltan: thing discovery object - enum

   better way?

   return TD and not URLs pointing to TDs

AOB

   skip topic semantic discovery and move it to the next call on
   March 2

   [adjourned]

Summary of Action Items

Summary of Resolutions

   [End of minutes]
     __________________________________________________________


    Minutes manually created (not a transcript), formatted by
    David Booth's [20]scribe.perl version 1.154 ([21]CVS log)
    $Date: 2020/03/09 03:32:09 $

     [20] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm
     [21] http://dev.w3.org/cvsweb/2002/scribe/

Received on Monday, 9 March 2020 03:36:51 UTC