Open damooo opened 2 years ago
It may be resolved by saying all concept resource ids MUST end with #
, while information resource ids do not. And id of information resource describing a concept resource MUST be same fragment-less part.
This way, there is distinction between a concept resource, and information resource describing it, all the while allowing deererencing affordance of atomic data architecture, and having no semantic clashes with rdf.
(Atomic Data is) limited only to self describing information resources / documents.
I think this is the case, currently. I agree that it should be possible to share information about a thing that is hosted somewhere else. However, re-using that information becomes a challenge, as the ones who know only of the original resource will not find the newly added information.
Anyways, I think this problem could probably be solved by defining a Property that does what you want it to do. Maybe it could be called describes
, and it's definition could be: The URI of the thing that is described by this resource. This can be a link that does not resolve to any Atomic Data Document.
Currently all subject urls are mandated to be dereferable. That is the issue for concept resources.
I think I'm not completely understanding the problem then. What are concept resources
? I wasn't able to find this on the internet, let alone the RDF spec.
Could you describe a usecase example?
Following documents can give proper context. And in general helpful for understanding uri.
Thanks for sharing the links, I think I'm familiar with most of them, but I still don't understand what the issue is. I think you're referring to Tim Beners-Lee's ConceptualWork
, right? If you want to make a distinction between resources that represent something and resources that describe something, then I think my suggested describes
property would cover the usecases I have in mind.
Can you provide a usecase for a concept resource
and describe your wanted outcome? The more concrete the example the better!
Current spec as it exists, doesn't distinguish between information resources and concept resources.
I.e, as every subject's identifier has to deref to an information resource representation(like json-ad doc), the architecture is
From docs, it seems second case. This may be a serious ontological issue for serious usage. And also makes the data semantically not a subset of rdf as claimed.