Re: representing deserialized RDF - N3, JSON-LD, RDFa, XML/RDF from Ian Dickinson on 2013-12-11 (public-rdfjs@w3.org from December 2013)

From: Ian Dickinson <i.j.dickinson@gmail.com>
Date: Wed, 11 Dec 2013 11:53:29 +0000
To: Ruben Verborgh <ruben.verborgh@ugent.be>
Cc: public-rdfjs@w3.org
Message-ID: <CAHvAVxY7kkKu7OUDwkvozS3ci1riq1xUNkTNHomj_FZ-oyB4wg@mail.gmail.com>

On Wed, Dec 11, 2013 at 10:21 AM, Ruben Verborgh
<ruben.verborgh@ugent.be> wrote:
> If we go for a statement-based representation, we are abandoning the JSON-LD route.
> Which is okay if that’s what we want.
> If the goal is to exchange data between systems that use triples internally, it probably is what we want.
>
> Here is the representation format I use:
>
> { subject: 'http://dbpedia.org/resource/Daft_Punk’,
>   predicate: 'http://www.w3.org/2000/01/rdf-schema#label',
>   object: '"Daft Punk"@en' }
That looks lossy from an RDF perspective. How would you differentiate
the following two triples?

ex:s ex:p <http://foo/bar>.
ex:s ex:p "http://foo/bar".

Regards,
Ian.

Received on Wednesday, 11 December 2013 11:53:57 UTC