- From: Pierre-Antoine CHAMPIN <champin@cpe.fr>
- Date: Thu, 09 Dec 1999 10:02:00 +0100
- To: Jonas Liljegren <jonas@paranormal.o.se>
- CC: RDF Intrest Group <www-rdf-interest@w3.org>
Jonas Liljegren wrote: > > Pierre-Antoine CHAMPIN wrote: > > > > Jonas Liljegren wrote: > > > With schemas, you could specify that an URI is an instance of some > > > type. Literals is said to be an instance of the Literal class. You > > > can subclass the literal class and set the type of the URI to be a > > > specific type of value. > > > > such types already exist on the web : mime-types. > > I already suggested that Litterals should not be different from resources in RDF, > > but should be anonymous resources with text/plain mime type. > > How would you represent that in RDF? I guess the mime-type is an attribute of the resource. If it can be retrieved (typically, by HTTP for 'http:*' URIs) then the mime-type will come with it. If it can't be retrieved, well you won't really care about the mime-type... Which doesn't exclude explicit representation of mime-types in RDF, like <rdf:Description about="..."> <rdf:type resource="http://somewhere.org/mimetypes/text/plain" /> </rdf:Description> > There could be many things that would be plain text but that you would like to > specify closer, like a date. To say that it is text/plain is not enough. We have XML-Schema working on the defining of datatypes. I guess we could have new mime-types like "x-xmls/string" for built-in datatypes ; for user-defined datatypes, I still don't see a better way than explicit typing (like above). I really think considering resource-types in general is more important than considering resources on one hand, and literals on the other hand. RDF authors admit the dichotomy Resource/Literal is not satisfying, I'm sure their would be a great profit in unifying them, and I guess considering (mime?) types of resource content is the key. Pierre-Antoine
Received on Thursday, 9 December 1999 04:01:28 UTC