Re: JSON-LD graph should be an RDF graph

On 2/26/13 5:25 PM, David Booth wrote:
> A comment on
> http://www.w3.org/TR/2012/WD-json-ld-syntax-20120712/
>
> The definition of "Linked Data" does not say that a linked data graph 
> is an RDF graph. 

I think its better to qualify RDF based Linked Data as being distinct 
from the very loose "Linked Data" phrase.

> I think it is important to state explicitly that a linked data graph 
> is an RDF graph. 

I think its important to state that JSON-LD is about RDF based Linked 
Data graphs.

RDF qualification is inherently useful. As you know, RDF graphs do 
possess unique qualities:

1. IRI based denotation of entities
2. Explicit (rather than implicit) entity relationship semantics
3. Loose coupling of Syntax and Notations.

> Otherwise JSON-LD would be a competitor to RDF (since both are used to 
> represent directed graphs) but without being grounded on RDF semantics.

Not if we qualify matter now, bearing in mind all the negative results 
of conflation from the past.

> Appendix B suggests that JSON-LD is intended to be an RDF format, but 
> this should be stated explicitly in the definition.

How about it being a notation for expressing RDF model data while also 
being an across-the-wire data serialization format?

>
> Thanks,
> David
>
>
>


-- 

Regards,

Kingsley Idehen	
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen