- From: Kazuyuki Ashimura <ashimura@w3.org>
- Date: Wed, 01 Apr 2020 18:45:06 +0900
- To: public-wot-wg@w3.org
available at:
https://www.w3.org/2020/03/23-wot-discovery-minutes.html
also as text below.
Thanks,
Kazuyuki
---
[1]W3C
[1] http://www.w3.org/
- DRAFT -
WoT Discovery
23 Mar 2020
[2]Agenda
[2] https://www.w3.org/WoT/IG/wiki/WG_WoT_Discovery_WebConf#Agenda
Attendees
Present
Christian_Glomb, Kaz_Ashimura, Kevin_Olotu,
Kunihiko_Toumura, Michael_McCool, Takahisa_Suzuki,
Tomoaki_Mizushima, Zoltan_Kis
Regrets
Chair
McCool
Scribe
kaz
Contents
* [3]Topics
1. [4]Agenda
2. [5]Previous minutes
3. [6]EdgeX
4. [7]Two-phase proposal
* [8]Summary of Action Items
* [9]Summary of Resolutions
__________________________________________________________
<scribe> scribenick: kaz
Agenda
[10]Agenda
[10] https://www.w3.org/WoT/IG/wiki/WG_WoT_Discovery_WebConf#Agenda
McCool: (goes through the agenda above)
Previous minutes
[11]March 2 minutes
[11] https://www.w3.org/2020/03/02-wot-discovery-minutes.html
[12]Virtual F2F minutes - Discovery session
[12] https://www.w3.org/2020/03/16-wot-minutes.html#item08
Kaz: March 2 minutes had been already approved
McCool: let's review the minutes from the virtual f2f then
... (goes through the virtual f2f minutes)
... need terms for Vorto
... one issue for discovery on Vorto to be created
... (creates an issue)
... align with Eclipse Vorto
... Kevin, do you use directory, etc?
Kevin: for Vorto, we use a virtual repository
McCool: Vorto repository stores and manages models
... (adds comments on Vorto based on Kevin's clarification)
Kevin: TD template is conceptually similar to the Vorto model
McCool: partial TD may describe classes
... we need some URL template to specify template
... we still need to discuss the detail
... also we need to see the relationship between TDs and Vorto
models
... my proposal is using links
... URLs point to models
... if we do that, Vorto repo and TD directory are different
... Vorto repo stores and manages models
... TD directory stores and manages TDs describing device
instances
... need to look into existing IANA media-types that might work
here
... Vorto model is a JSON file?
Kevin: domain-specific language
McCool: content-type already registered?
Kevin: no
McCool: probably need registered IANA MIME type for Vorto
models
... this discussion should occur during the WoT Architecture
calls
Kevin: note that we don't use TD Template inside Vorto model
McCool: yeah
... what is the exact mapping between Vorto concepts and
built-in vocabulary in TDs?
... we need to do a mapping with the information model and
... (goes through the information model from the TD spec)
[13]TD spec - 5. TD Information Model
[13] https://www.w3.org/TR/2020/PR-wot-thing-description-20200130/#sec-vocabulary-definition
McCool: how to align Vorto with TD?
... similar image to the TD Information model Figure 1
[14]Figure 1
[14] https://www.w3.org/TR/2020/PR-wot-thing-description-20200130/visualization/td.png
<McCool> [15]https://github.com/w3c/wot-discovery/issues/7
[15] https://github.com/w3c/wot-discovery/issues/7
McCool: would suggest you create a Pullrequest to put contents
Kevin: ok
[16]fyi, pullrequests for wot-discovery repo
[16] https://github.com/w3c/wot-discovery/pulls
EdgeX
McCool: (quickly goes through the slides)
[17]EdgeX slides
[17] https://github.com/w3c/wot-discovery/blob/master/presentations/2020-03-18-WoT-EdgeX-Conexxus-PoC.pdf
McCool: feedback from users, e.g., during TPAC 2020
... would like to add TD for devices supported by EdgeX
... EdgeX has its own application stack
... (core services section of the EdgeX platform archtiecture
include microservices intercommunicating with each other via
APIs)
... Proposed WoT/EdgeX Integration
... can do JSON matching or something
... have directory service and APIs
... would registere myself in the directory service
... to demonstrate 2-phase concept
... also node-wot inside EdgeX
... for WoT orchestration
... upload scripts and invoke them
... item 4 is optional and nodeRED orchestration is higher
priority
... still in planning phase
Christian: where node-wot would fit within the architecture?
McCool: all EdgeX is service-based docker container
... orchestration inside EdgeX
... 2 issues
... how to build the container
... and instantiating it
... might have some examples
... retail examples could be resources for scripts
... but in general need to see how to build our own scripts
Christian: there are already some endpoints existing
... upload script, start and stop it
... would be a good timing to extend the interface
McCool: any resources about that?
Christian: let me take an action about doc container
McCool: it is critical to integrate EdgeX
... they currently use open APIs
... part of this effort, we define APIs for a directory service
... my intention is making it public
... will work on it temporally at my private repo due to NDA
things, though
... (creates a new issue on the wot-discovery repo)
... define a concrete network API for a TD directory service
[18]Issue 8
[18] https://github.com/w3c/wot-discovery/issues/8
Christian: thing-web repo is a bit outdated
McCool: yes
... we should see the current status existing implementation
for TD directory
... (adds a question about that for Victor)
Kaz: we should add some description about our plan itself on
the wot-discovery/README.md
... how to do our PoC including EdgeX, node-wot, wot discovery,
etc.
McCool: yeah, maybe we could have a specific directory for the
PoC purposes
... maybe need to clean up the directory structure
... e.g., move prior.md to the prior-work directory
... and update the README.md
... let's create an issue
... create structure to organize PoCs
[19]Issue 9
[19] https://github.com/w3c/wot-discovery/issues/9
McCool: also reorganize prior work
[20]Issue 9
[20] https://github.com/w3c/wot-discovery/issues/9
toumura: question about the main purpose of this PoC
... just for discovery?
... or maybe all the related building blocks?
... maybe better to put the PoC work under wot repo rather than
wot-discovery?
McCool: this PoC is specifically for discovery
... but maybe would be better to go under wot repo
... let's discuss in the main call
... can be sensitive due to NDA things, etc.
... anyway we still need this issue 9 itself
Two-phase proposal
McCool: questions?
Christian: rather a comment
... we already have TD
... the proposal doesn't really include how to generate TD
... we should include some description about that
... also the mechanisms how existing protocols support
discovery
McCool: likewise "discovery" doesn't mean "broadcast"
... the actual meaning of the current text is a bit different
[21]Two-phase proposal
[21] https://github.com/w3c/wot-discovery/blob/master/proposals/directory.md
McCool: let's create an issue to improve the text
... (create a new issue)
[22]Issue 11
[22] https://github.com/w3c/wot-discovery/issues/11
McCool: let's put concrete descriptions to the issue during the
next call
[adjourned]
Summary of Action Items
Summary of Resolutions
[End of minutes]
__________________________________________________________
Minutes manually created (not a transcript), formatted by
David Booth's [23]scribe.perl version 1.154 ([24]CVS log)
$Date: 2020/03/30 14:59:40 $
[23] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm
[24] http://dev.w3.org/cvsweb/2002/scribe/
Received on Wednesday, 1 April 2020 09:45:11 UTC