RE: Profiles in Linked Data

Hi Lars

> On May 8, 2015 at 5:44 PM "Svensson, Lars" <L.Svensson@dnb.de> wrote:
>
>
> John, Kingsley,
>
> I wrote:
> >>> OK, I can understand that. Does that mean that if I have under the same
> >>> URI
> >>> serve different representations (e. g. rdf/xml, turtle and xhtml+RDFa) all
> >>> those
> >>> representations must return exactly the same triples, or would it be
> >>> allowed to
> >>> use schema.org in the RDFa, W3C Organisation Ontology for rdf/xml and foaf
> >>> when returning turtle? After all it's different descriptions of the same
> >>> resource.
>
> John wrote:
>
> >> My take on this is each representation (with negotiation only on format via
> >> HTTP Accept header) *should* contain the same set of RDF statements
> >> (triples).
> >> Also one could define a different URL for each representation which can be
> >> linked to with Content-Location in the HTTP headers.
> >>
> >> We’re you to introduce an additional (orthogonal) way to negotiate a
> >> certain
> >> profile, this would be orthogonal to the format. Following on from above,
> >> one
> >> could then have a separate URL for each format-profile combination.
>
> Kingsley wrote:
>
> > Yes.
> >
> > For the sake of additional clarity, how about speaking about documents and
> > content-types rather than "representation" which does inevitably conflate
> > key
> > subtleties, in regards to RDF (Language, Notations, and Serialization
> > Formats)?
>
> The terminology is fine with me, as long as we don't forget the entities we
> describe.
>
> So to repeat my question in another mail: I have an entity described by a
> (generic) URI. Then I have three groups of documents describing that entity,
> the first uses schema.org, the second group uses org ontology and the third
> uses foaf. All documents are available as RDF/XML, Turtle and xhtml+RDFa. How
> does a client that knows only the generic URI for the resource tell the server
> that it prefers foaf in turtle and what does the server answer?
 
I believe that the two options are in HTTP headers or in the query string part
of the URI. In the latter case I guess you would say that is no longer the
generic URI.
 
I note in the JSON-LD spec it is stated "A profile does not change the semantics
of the resource representation when processed without profile knowledge, so that
clients both with and without knowledge of a profiled resource can safely use
the same representation", which would no longer hold true if the profile
parameter were used to negotiate which vocabulary/shape is used.
 
John

>
> Best,
>
> Lars