search

w3c / dx-prof

The Profiles Vocabulary
https://w3c.github.io/dx-prof/prof/
Other
5 stars 2 forks source link

Conceptual model: Profile: class vs. property #11

Open heidivanparys opened 5 years ago

heidivanparys commented 5 years ago

I have a comment regarding this part of the conceptual model:

conceptual_model

Why is "Profile" present as a class?

I would be inclined to model this part as follows (UML-style model, as that's what I'm used to...):

specification

So

This model also implies not using dct:Standard ("A basis for comparison; a reference point against which other things can be evaluated.") at all.

The information that is currently modelled with ResourceDescriptor and its relations, couldn't that be generalized to specifications in general, and be useful?

Note: I used "specification" as that term is used in several places in the set of documents of profiles. Other place mention "standard".

smrgeoinfo commented 5 years ago

This approach makes good sense to me; a profile is just a particular kind of specification (has some relationship to one or more base specifications). The ResourceDescriptor class could be modeled as a qualifiedAssociation to related resources, and any specification could/should have an array of related resources for guidance, validation, examples etc.
The important thing for the content negotiation is the ability to identify a specification that the target representation conforms to, such that machine agents can identify and retrieve representations they can work with.

rob-metalinkage commented 5 years ago

Yes, in some ways this model would work - however dct:Standard is already defined and is not incorrect, so we should reference it.

We could introduce a new role for a resource being a profiled specification.

the domain of all properties could be dct:Standard or open. Asking if something is a Profile is easier than asking if it has no profileOf relationships, and its possible to declare something is a Profile even if the canonical URIs of all the base specifications are not known.

rob-metalinkage commented 5 years ago

I think this could be left as an open issue in the next PWD, if there is a consensus that dct:Standard is all with need with some predicates this could be changed.