multimericRDF people: why does everyone still use RDFS and not SHACL for writing schemas, when all the actual validation tools are for SHACL?
Are there even any RDFS validators? I can't find any.
Jeen@multimeric I think they serve different purposes. #rdfs is not really a schema language, at least not in the "here is a prescribed structure you can validate against" sense. It was more intended as a descriptional thing, something to define basic vocabulary ("my model is about kinds of vehicles, like cars, and bikes"). A lightweight ontology language if you will. SHACL is about setting structure constraints, and data validation. They're complementary IMO.
Thanks, that helps me understand the mindset.
I guess it also makes sense that the simpler language has more adoption. I suspect that SHACL can be simple if you want it to, though.
Still, automated validation is massively important to me!
emidiostani@multimeric it depends on your use cases, in general you split the model from the constraints.
For the concepts you defines (classes and properties), you would create persistent uri, therefore providing HTML and RDF représentations.
Then some users might demand to have a collection of your concepts, you could create an ontology.
Some of your classes might be defined as subclass of other vocabulary, so then you use rdfs (shacl leverages rdfs).For model constraints you could release just shacl.
For the concepts you defines (classes and properties), you would create persistent uri, therefore providing HTML and RDF représentations.
Then some users might demand to have a collection of your concepts, you could create an ontology.
Some of your classes might be defined as subclass of other vocabulary, so then you use rdfs (shacl leverages rdfs).For model constraints you could release just shacl.