Re: RDF Data Shapes WG agenda for 15 October 2015

On 10/14/2015 11:11 PM, Dimitris Kontokostas wrote:
> 
> 
> On Wed, Oct 14, 2015 at 9:55 PM, Peter F. Patel-Schneider
> <pfpschneider@gmail.com <mailto:pfpschneider@gmail.com>> wrote:
> 
> 
> 
>     ISSUE-86 (mostly a repeat from last week)
> 
>     I vote -1 on any proposal that requires or advocates putting shape and data
>     information or shape and ontology information together.  SHACL is not a
>     modelling language.  SHACL can function with shape information fully separated
>     from
>     data and ontology information and this separation should be the suggested way
>     to use SHACL.
> 
>     This means that I vote -1 on Dimitris's proposal.
> 
> 
> Personally I would prefer only option c (linking with sh:schemaShapes) but I
> think Holger would object to that and that's why I added options a & b.
> The truth is that we cannot force people to write ontology axioms and shapes
> in separate graphs even if that can be considered a better practice.
> In addition, the current spec already mentions the mixing of shapes and
> ontology information with sh:ShapeClass which is a syntactic sortcut for
>  rdfs:Class, sh:Shape and sh:scopeClass (issue-23 / unresolved yet)
> 
> Dimitris

Users and tool builders will end up doing whatever they feel best.  SHACL is
about how to validate a data graph against shapes plus control information.
This can be done without SHACL specifying how to make connections between
data, ontologies, and shapes.

peter

Received on Friday, 16 October 2015 16:12:42 UTC