W3C

- DRAFT -

WoT IG Meeting

07 Dec 2016

Agenda

See also: IRC log

Attendees

Present
Kaz_Ashimura, Michael_McCool, Uday_Davluru, Victor_Charpenay, Dave_Raggett, Feng_Zhang, Kirby_Shabaga, Matthias_Kovatsch, Takuki_Kamiya, Johannes_Hund, Sebastian_Kaebisch, Michael_Koster, Katsuyoshi_Naka, Masato_Ohura, Frank_Reusch, Takeshi_Yamada, Yingying_Chen
Regrets
Chair
Matthias_Kovatsch
Scribe
victor

Contents


<kaz> scribenick: victor

AC Review

MMC: new PR to add more collaborations

assume this can be closed next week

One review was "don't remove the APA reference"

Kaz: APA = Accessible Platform Arch.

<mkovatsc> ?q

Link to the most recent version of the charter to be sent

Kaz: Wendy has a meeting today with the W3C Managment, results will be sent over soon

F2F Logistics

MMC: still waiting for a response on the budget. The person responsible has just come back from vacation. Hope to have a response this week

<mkovatsc> https://www.w3.org/WoT/IG/wiki/F2F_meeting,_February_2017,_USA,_Santa_Clara

MMC: 13k budget in total, MK, could you investigate in parallel?

MK: OK

Matthias: finding a hotel in this area is hard. Reserving early is good for attendees

MMC: maybe do a reservation with an option to cancel it later?

Kaz: clarify the dates before? What happens on Sunday?

MMC: Let's be conservative, let's say 15k (to reserve half a day extra)

Matthias: I've just updated the wiki to add info about housing

oneM2M liaison

MMC: apparently there is a meeting this week, which is why Yongjing is not here this week. Shift this point to next week

OCF liaison

MMC: we'll create a liaison committee. Discussed with OCF lawyers about that. This will be chartered.

Members: Yongjing, Michael (Koster), Michael (McCool), Jonathan Jeon.

Translating TD into RDF

Dave: I've been doing some work on NodeJS (shows screen)

link sent to the mailing-list

<kaz> Dave's demo

includes examples of things like water tank or ECG

it also shows compound (nested) properties

example of real-world running web service: airport status (provided by FAA)

URI is a template

Sebastian: can you show us the context file?

what do Web developers have to do?

Dave: nothing particular, depending on what they want to do with the RDF data

Matthias: the structure of the JSON seems to be changing across examples

Dave: it is syntactic shortening mostly

and the rest, e.g. "source": "some URI" in the ECG+A example

depends on the semantics of the meta-data. Here, source is expected to be a URI (by clients aware of it)

Matthias: a summary of the overall model used in this proposal?

Dave: part of the answer is in the rest of my demo (follows after this question round)

(Dave shows the source code of the application )

MMC: The JSON-LD was generated automatically or written by hand?

Dave: generated

<mjkoster> dropping of now, thanks!

MMC: a generic demo where RDF data is generated for any TD (given as input) would be good

so that we understand the rules that apply to turn TDs into RDF

Sebastian: remark: TD should be platform/language-agnostic (shows a slide presented this morning in the TD telco)

<kaz> Sebastian's slides

Dave: here it is more about choosing JSON or JSON-LD.

<dsr> No, I believe that Web developers will prefer a simple JSON based represention, but I agree that there could be multiple serializations

Matthias: maybe, another way to look at the question is to say that Dave is trying to provide a "debug" view of a TD, similar to what CBOR does

what we maybe need is a more abstract view, a diagram with boxes and arrows to understand what the model you are using is.

victor: Suggested improvements: comply to N-Triples syntax (add brackets and quotes) and use the exact context so that deserializing the JSON-LD version would lead to the exact same triples as the N-Triples version

The JSON version and the JSON-LD version look almost the same. It would be a pity if they required two separate recommendations for RDF deserialization. If Dave is about to provide more explanations on his model, it would be good that he indicates the common parts with JSON-LD processing.

scribenick: kaz

<kaz> WoT Servient architecture

kaz: not sure what Dave wants to do here and would like to know which module in the WoT Servient architecture should be updated

or do you want to restart TD discussion based on RDF?

scribenick: victor

Matthias: at last TPAC, the conclusion was that we should have this discussion at a high-level, on diagrams instead of on JSON examples

scribenick: kaz

dsr: tool for drawing diagrams?

scribenick: victor

Matthias: a proposal was to use PlantUML

Turtle was also proposed, we could use then tools to generate a graphical representation of the model

Matthias: next topic was TD update. Shift to next week.

Kaz: update on security? ok to invite people outside of the WoT IG, e.g., Automotive?

MMC: OK. still looking for people. On-going

<kaz> [ adjourned ]

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.148 (CVS log)
$Date: 2016/12/07 20:59:11 $