- From: Charpenay, Victor (ext) <victor.charpenay@siemens.com>
- Date: Fri, 28 Aug 2015 15:27:47 +0000
- To: "ruben.verborgh@ugent.be" <ruben.verborgh@ugent.be>
- CC: "public-hydra@w3.org" <public-hydra@w3.org>
Thank you very much. > These are good places to start: > - https://lists.w3.org/Archives/Public/public-hydra/2015Jan/0022.html > - https://github.com/HydraCG/Specifications/issues/90 After I read this, I share your point of view. However, it has not really been a debate yet: nobody argued to keep the vocabulary as it is... > There's probably no place in such a system where you'd have to use hydra:Resource and hydra:Class explicitly. Sorry, I wrote this too fast. It is less related to this hydra:Class/hydra:Resource story than to the issue tracker demo. In the latter, the classes vocab:User and vocab:Issue are defined as instances of hydra:Class, while they have no rdfs:subClassOf relation. It sounds rather strange to me. Let assume I'd like to semantically describe some automation devices -sensors or actuators- and both types support the property "ex:deviceID" (or whatever). I'd naturally wish to define an abstract resource "ex:Device", so that "hydra:Class" > "ex:Device" > "ex:Sensor", "ex:Actuator" (hydra:Class is still there, say). I guess I'm free to do it. But why then isn't it this way in the issue tracker demo? Is there a specific reason? Cheers, Victor Charpenay -----Ursprüngliche Nachricht----- Von: Ruben Verborgh [mailto:ruben.verborgh@ugent.be] Gesendet: Montag, 24. August 2015 10:48 An: Charpenay, Victor (ext) Cc: public-hydra@w3.org Betreff: Re: Hi // Comments on the Hydra spec Hi Victor, > If so, may you give me a hint on where to search in the mail archive (e.g. which mail subject)? These are good places to start: - https://lists.w3.org/Archives/Public/public-hydra/2015Jan/0022.html - https://github.com/HydraCG/Specifications/issues/90 > I'd like to develop a method to automatically turn ontological elements into REST-accessible resources (i.e. automatically generate Hydra specs out of an OWL ontology). Having such a design where classes are practically seen as individuals/class instances does not facilitate the thing, it would be of great help if I understood why. There's probably no place in such a system where you'd have to use hydra:Resource and hydra:Class explicitly. Also, in RDF, classes are instances by definition, i.e., all classes are instances of rdfs:Class. Best, Ruben
Received on Friday, 28 August 2015 15:28:21 UTC