AW: [TF-TD] proposal for JSON format, its application to OCF, and mapping to RDF

Hi Dave,

I am totally OK if "some" Web developers use internally another format that suits their needs better. Other platforms might do the same.

The TD should be an abstract format that is meant to provide a stable and well-defined thing description for many uses cases, platforms, and programming languages.

-- Daniel

________________________________
Von: Dave Raggett [dsr@w3.org]
Gesendet: Montag, 5. Dezember 2016 17:38
An: Peintner, Daniel (ext) (CT RDA NEC EMB-DE)
Cc: Public Web of Things IG
Betreff: Re: [TF-TD] proposal for JSON format, its application to OCF, and mapping to RDF


On 5 Dec 2016, at 14:59, Peintner, Daniel <daniel.peintner.ext@siemens.com<mailto:daniel.peintner.ext@siemens.com>> wrote:

Hi Dave, all,
While I understand the desire to convince Web developers and  I acknowledge your work and your background I do not fully agree with your conclusions.
Yes, the format should be easy to understand. Yes, it would be good to have it as concise as possible. Yes, JSON might be a good choice also.
Having said that, we should also take into account the tools that have been built around a given technology and the actual standing/dissemination it has.
JSON-LD seems to be a good solution (tradeoff?)
* it uses JSON format
* it supports RDF (for those who really like/need it)
* there are many tools around it
* ...
Inventing a new JSON format might be more concise BUT does not have any tool ready. Nor is it proved like a standard that has been worked  on for years.
Moreover, I doubt that people gonna write thing description by hand. There will be tools built on top of JSON-LD. For a new format there is no such support.
Hence, conciseness for plain-text thing descriptions might not be an issue (efficiency for restricted devices is another argument but is better solved differently).
Do these arguments make sense?

Why don’t you use JSON-LD to model the suite of use cases and OCF resources in my two demos and we can then ask web developers and SMEs for their comments on the side by side examples?  This would ensure that we cover a broad range of requirements as provided by the large set of use cases.

In respect to tooling, it took very little time to write the JSON to RDF translator, and making it easy to the exploit the mature tooling for RDF for reasoning over rich models of things, their relationships and context. Are you interested in helping with an exploration of semantic reasoning for the Web of things?

[1] https://www.w3.org/WoT/demos/td2ttl/
[2] https://www.w3.org/WoT/demos/td2ttl/oic.html

Best regards,
—
   Dave Raggett <dsr@w3.org<mailto:dsr@w3.org>>

Received on Wednesday, 7 December 2016 08:36:22 UTC