Re: [poe] Use of cardinality restrictions in the ontology?

> In any case, a user of ODRL doesn't have to be familiar with OWL or SHACL at all anyway. That's why we have our information model.

@simonstey you may be right by an academic view but from my experience with teams having to implement a standard they need a) something which shows in a strict way structures, properties and their cardinality as guideline for their serializations and b) something which checks if their instance of a serialisation complies with the specification.

(Aside: why was something like the JSON Schema created? It is far from perfect but there are development tools supporting feature a) and b) above - making developers almost happy.)

Currently ODRL provides to potential adopters:

* the Information Model document: it specifies classes and properties and should include their cardinality, but (currently) details are disputed. And it shows examples as JSON serialisation which is already a step out of an abstract definition.
* the Vocabulary document: is a human readable rendition of an also existing OWL ontology, it defines classes and properties and could define cardinality, but the ontology cannot be used for validation.
* an emerging SHACL document: it should be able to validate a graph against the Information Model.
* an XML Schema: it could be used to specify the XML serialisation of a policy instance and could be used to validate it against the schema. But the current schema has bugs and a design different from the JSON in the IM examples - #197 

Honestly, this makes it hard to encourage developers to use ODRL.



-- 
GitHub Notification of comment by nitmws
Please view or discuss this issue at https://github.com/w3c/poe/issues/198#issuecomment-309732305 using your GitHub account

Received on Tuesday, 20 June 2017 12:04:13 UTC