- From: Kjetil Kjernsmo <kjetil@kjernsmo.net>
- Date: Thu, 11 Jun 2015 11:05:12 +0200
- To: public-hydra@w3.org
- Cc: greg@evilfunhouse.com
On Thursday 11. June 2015 09.24.28 Miel Vander Sande wrote: > On 11 Jun 2015, at 09:15, Ruben Verborgh <ruben.verborgh@ugent.be> wrote: > > Another option is to have a very specific property like "has TPF > > interface", though that might be overly specific then. > > Since a resource of the TPF interface needs to be dereferenced in any > case in order to interact with it, should an rdfs:seeAlso not cut it? Hmmm, I don't think that's specific enough. :-) I think the semantics should be clear on that there is a query interface that can be used to query the same data that has been queried through the current interface (e.g. a SPARQL endpoint), but that it can be any Hydra-expressed interface. I think there may be two cases here: In one case, you're linking directly to a IriTemplate, in which case, hydra:search would be appropriate. In the other case, you're linking to a VoID description, a LDF result, or something else that contains control information, but where the control information typically is a subgraph of the response. I suspect the latter will be the most common case, and that's were we might need another predicate. Best, Kjetil
Received on Thursday, 11 June 2015 09:05:54 UTC