- From: Holger Knublauch <holger@topquadrant.com>
- Date: Fri, 09 Jan 2015 09:43:20 +1000
- To: public-hydra@w3.org
On 1/9/2015 9:40, Markus Lanthaler wrote: > On 8 Jan 2015 at 23:40, Holger Knublauch wrote: >> On 1/9/15, 7:48 AM, Markus Lanthaler wrote: >>> Ignoring the dereferenceability stuff for a moment: do you think it >>> makes sense to keep something like hydra:Class given that they are >>> quite different from RDFS classes due to hydra:supportedProperty? Of >>> course assuming we keep hydra:supportedProperty in more or less this >>> form :-) >> Why could hydra:supportedProperty not be used for any rdfs:Class? > Of course it can be used on any rdfs:Class, hydra:Class is, after all, a > subclass of rdfs:Class. It is just that the model is quite different from > RDFS and much closer to the one of most OO programming languages. As such, > it might be worth to give it a name. I am all in favor of being as close as possible to OO languages. However, then I believe you have a contradiction here: in an OO system, you would need to attach hydra:supportedProperty to rdfs:Class, not its subclass hydra:Class, so that any rdfs:Class can declare supported properties. Holger
Received on Thursday, 8 January 2015 23:46:37 UTC