Re: Update and opportunities with SHACL

On 8/11/2015 10:51, Tom Johnson wrote:
> > This is not the goal. SPARQL is not an extension language but a 
> genuine part of SHACL.
>
> I'm tempted to read this as "SPARQL is a subset of SHACL" (and I think 
> it's strictly true, on a current reading), but I suspect you might me 
> something different... under the current description,JavaScript is 
> also a subset, yes?

Maybe this Axis helps:

http://www.slideshare.net/HolgerKnublauch/shacl-specification-draft/22

SHACL with SPARQL is more expressive than SHACL Core, and JavaScript 
would allow more use cases to be represented than SPARQL. However, with 
JavaScript you'd be API-specific and imperative, while SPARQL is 
standardized and declarative. Likewise someone could define a Java 
binding, or Python, or Ruby or whatever. All these languages would 
overlap in the use cases that they can cover. In contrast to just 
embedding the constraints in executable code, SHACL would provide a 
declarative framework to exchange constraint definitions, with the Core 
vocabulary as the common glue to communicate the structure of the data 
(classes, properties, cardinalities etc)

Holger

Received on Tuesday, 11 August 2015 03:40:38 UTC