Re: [dxwg] Create generic properties such as "hasExpression" or "documentedIn" instead of hasResource/(hasRole | hasArtifact) (#747)

@makxdekkers I don't understand "there is no semantic difference between defining a small set of generic properties and defining a small set of generic roles, especially if there is no maintenance agency for the role vocabulary". The fact that there is or not a maintenance agency is not about semantics. There is a semantic difference between SKOS concepts and RDFS properties, quite a big one actually. Or do you mean something else?

And I'm not sure that we have the same perspective on flexibility: I'm interested in the flexibility of the representation, and SKOS semantics makes it more flexible than RDFS. This said, SKOS doesn't solve all the data management issues of handling extensions, as we've noted in #536. But again the argument is a bit moot as a solution based on extending a small set of core RDFS properties (as proposed [here](https://github.com/w3c/dxwg/issues/747#issuecomment-463341120)) would have the same issues. Can we agree on this?

Maybe it will be easier if we can also agree on the two dimensions of the problem we're considering:
1. pattern for representing roles: (a) SKOS concepts vs. (b) RDFS properties
2. degree of inclusion of roles in core spec: (a) having a core list of roles in PROF and relying on future extensions for specialized needs vs. (b) entirely relying on extension(s).

To me these dimensions are to a great extent orthogonal: a choice on 1 doesn't dictate a choice on 2 and a choice on 2 does not dictate a choice for 1.
Well, I think SKOS makes it easier to handle extensions, but it's not an absolute pre-requisite, and it would benefit for 2a and 2b anyway.

-- 
GitHub Notification of comment by aisaac
Please view or discuss this issue at https://github.com/w3c/dxwg/issues/747#issuecomment-463625702 using your GitHub account

Received on Thursday, 14 February 2019 13:25:47 UTC