Re: comments on current version of SHACL document

If we cannot even assume that the shapes are represented in RDF, how 
would the spec look like? We could not even talk about the property 
sh:minCount or the class sh:Shape then. If you think this needs to be 
changed, please propose an alternative. NB: It may require a rewrite of 
the whole spec.

Holger


On 9/18/2015 3:48, Arnaud Le Hors wrote:
> Holger Knublauch <holger@topquadrant.com> wrote on 09/16/2015 07:40:59 PM:
>
> > ...
> > This is not my understanding of how SHACL works. I believe the SHACL
> > spec always assumes that the shapes are represented in RDF, and in a
> > dedicated shapes graph, using exactly the specified vocabulary. If
> > someone wants to use another (compact) syntax then these syntaxes need
> > to be translated into RDF triples prior to execution.
> > ...
>
> Why? As long as the result is the same I don't see why an 
> implementation would have to go through this step. This seems similar 
> to being able to implement SHACL without using SPARQL. It's the result 
> that counts. How implementers get that result is up to them.
> --
> Arnaud  Le Hors - Senior Technical Staff Member, Open Web Technologies 
> - IBM Software Group
>
>
>