W3C home > Mailing lists > Public > public-dxwg-wg@w3.org > September 2019

Re: [dxwg] Definition of prof:isProfileOf (#1068)

From: Rob Atkinson via GitHub <sysbot+gh@w3.org>
Date: Thu, 12 Sep 2019 01:53:58 +0000
To: public-dxwg-wg@w3.org
Message-ID: <issue_comment.created-530631825-1568253237-sysbot+gh@w3.org>
The questions are already addressed:

prof:isProfileOf rdfs:range dct:Standard

doesnt need any further decisions, and answers the questions posed.

The axiom makes the relationship clear within the formalism of OWL semantics, and its been stated in text already. This was relaxed from rdfs:range prof:Profile even though that would be correct, because it seems too hard to understand that any dct:Standard may be a profile of something else in an "Open World" interpretation, and of itself under principles of reflexivity.  Possibly we should explicitly declare profileOf as a reflexive property (w3.org/TR/owl2-syntax/#Reflexive_Object_Properties).

(To understand Profiles are analogous to rdfs:Classes - they describe a set of individual instances that conform to them - where the instances are datasets, not sets of RDF statements about a subject - and hence Profiles naturally exhibit transitivity and reflexivity. We avoided doing this just in the most formal and technical way because it should be easier for _users_ of the spec to deal with concepts of Profiles and data without being versed in computer science - but likewise where we need to be specific about behaviour then the _authors_ need to understand the computer science to design the specification so it says just what it needs to say and doesnt introduce logical errors)

The proposal for this issue is that including the axiom is necessary because text is too easily misinterpreted.  There is simply no intention to try to create a model for all possible relationships between specifications where no assertions about conformance can be made. 

The misunderstanding here is about what a statement about conformance actually implies ;

"expand on" is just adding constraints about the semantics and cardinality of additional terms - it doesnt change any requirements about conformance to the profiled specification.  Just because languages dont use the generic term "constraints" doenst mean it doesnt apply - OWL semantics are constraints, sko:Concepts constrain interpretation of terms etc.  

GitHub Notification of comment by rob-metalinkage
Please view or discuss this issue at https://github.com/w3c/dxwg/issues/1068#issuecomment-530631825 using your GitHub account
Received on Thursday, 12 September 2019 01:53:59 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 30 October 2019 00:15:57 UTC