[poe] Issue: On requiring mutually disjoint sets of policy/rule types marked as CR review

simonstey has just labeled an issue for https://github.com/w3c/poe as "CR review":

== On requiring mutually disjoint sets of policy/rule types ==
Although never explicitly mentioned in the IM, the Vocab (i.e. ontology) defines all types of [rules](https://w3c.github.io/poe/vocab/#ruleConcepts) and all types of [policies](https://w3c.github.io/poe/vocab/#policySubClasses) as being mutually disjoint (which was also briefly touched upon in #247) .

and while that arguably makes sense for e.g. some types of rules (as pointed out by @vroddon  https://github.com/w3c/poe/issues/247#issuecomment-327355711):
> Another example we have always had in the spec is the disjointness between Permission, Prohibition and Duty. I don't see how can a Permission be a Prohibition a the same time, for example.

it doesn't for certain policy types or ones added by profiles, e.g.:

> For the examples in this document, the ODRL Policy subclasses are mapped to the JSON-LD @type tokens. **The above example could have also used `Policy` type instead of `Set` type (as they are equivalent).**

-> `odrl:Policy owl:equivalentClass odrl:Set`

> Parent class:  Policy

-> `odrl:Agreement rdfs:subClassOf odrl:Policy`

> Disjoint classes: Assertion, Offer, Privacy, Request, Set, Ticket

-> `odrl:Agreement owl:disjointWith odrl:Set, ... `

`odrl:Agreement owl:disjointWith odrl:Set, odrl:Policy, ...`
`odrl:Agreement rdfs:subClassOf odrl:Set, odrl:Policy .`

See https://github.com/w3c/poe/issues/280

Received on Wednesday, 25 October 2017 06:14:50 UTC