W3C home > Mailing lists > Public > public-rdf-wg@w3.org > March 2011

Re: [JSON] The best of both worlds?

From: Ivan Herman <ivan@w3.org>
Date: Mon, 21 Mar 2011 09:58:28 +0100
Cc: RDF Working Group <public-rdf-wg@w3.org>
Message-Id: <0F6A107A-BA67-4107-AE12-3EA1F5CBC2F3@w3.org>
To: Manu Sporny <msporny@digitalbazaar.com>
Manu,

forgive me, but I think this also confusing. Mixing two completely different world-views into one language is really a recipe for lots of disasters in my view. Also, your approach, as Nathan also said, looks very much like reification, which is also a source of confusion. Furthermore... the "[" ... "]" already has the usage in JSON-LD for "regroup JSON-LD objects or resources", and you are adding a pretty different meaning to the same syntactic construction. This is yet another source of confusion...

If the WG decides that there is room and need for both approaches, then both approaches should be defined separately (with different media types, etc) in my view.

Ivan

P.S. As an aside: this made me realize that the "[" ... "]" syntax _cannot_ be used for rdf lists. Is there any shorthand for those in JSON-LD?





On Mar 21, 2011, at 03:25 , Manu Sporny wrote:

> Thinking a bit more about how this object-based vs. triple-based
> approach has turned into a divisive discussion (partly my fault, I'm
> sure). There is another option - we support both approaches in a single
> format by providing something in the default context to switch between
> the two mechanisms, so object based would be this:
> 
> {
>   "#": { "name": "http://xmlns.com/foaf/0.1/name" },
>   "@": "http://example.com/people#david",
>   "name": "David Wood@en"
> }
> 
> and triple-based would be something like this:
> 
> {
>   "#": "RDF-IN-JSON-RAW-TRIPLES",
>   "@":
>   [
>      {
>         "subject": "http://example.org/people#david"
>         "predicate": "http://xmlns.com/foaf/0.1/name"
>         "object":
>         {
>            "value": "David Wood"
>            "language": "en"
>         }
>      },
>      ...
>   ]
> }
> 
> The framing of these objects above is a bit ugly, but it's something
> that I know would work in JSON-LD (just as a proof of concept).
> 
> Basically, the RDF in JSON processor would have two modes of operation,
> the first is the object-based mechanism. The second is the triple-based
> mechanism. The interesting thing with this approach is that you could
> switch between the two fairly easily. The other way that we could do
> this is by switching off of the top-level JSON object:
> 
> { ... } -> object-based processing mode
> [ ... ] -> triple-based processing mode
> 
> So, we may not have to pick - this approach could effectively combine
> RDF/JSON / Flat Triples / JTriples with JSON-LD / JSN3 / RDFj.
> 
> We could call it: FlatJTriplesinJSONwithRDFandJSN3jGRDDLforLinkedData :P
> 
> -- manu
> 
> -- 
> Manu Sporny (skype: msporny, twitter: manusporny)
> President/CEO - Digital Bazaar, Inc.
> blog: Payment Standards and Competition
> http://digitalbazaar.com/2011/02/28/payment-standards/
> 


----
Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
PGP Key: http://www.ivan-herman.net/pgpkey.html
FOAF: http://www.ivan-herman.net/foaf.rdf






Received on Monday, 21 March 2011 08:56:20 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:04:04 UTC