Re: does RDF require understanding all 82 URI schemes?

Graham Klyne wrote:
> 
> At 02:03 PM 2/7/01 -0500, Peter F. Patel-Schneider wrote:
> >In an e-mail discussion Dan Connolly <connolly@w3.org> mentioned the
> >existance of a data: URI Scheme.  I was wondering if RDF parsers are
> >supposed to understand all 82 URI schemes and what they are supposed to do
> >with them.
> 
> I don't think an RDF parser should "understand" any URI scheme -- just
> handle all URIs according to the syntactic rules of RFC2396.

I agree totally.
And I don't think Dan's proposition of using 'data:' URI scheme contracicts with that :
currently, RDF handle two kinds of things (URIs and Literals) and handle them in different ways
(Literal can not be subject nor predicate of a statement, which happens to bother much people,
 -- the subject issue, at least)

Using 'data:' URIs instead of Literals has teh following advantages :

 - RDF has to manage one unique sort of things : URI
 - hence, there is no limitation on literals (since they are not distinct from resources)
 - 'data:' URIs have a mime/type, which literals do not !
 - surely some others I don't see...

That does not mean we have to change the spec :
 - the syntax remains unchanged
 - old-style parsers parse embeded literals as plain literals
 - new-style parsers convert embeded literals as 'data:' URI
 - hence, old-style literals are sill allowed in the model, but discouraged

Note that old-style parsers will have no problem with RDF-pieces containing 'data:' URIs,
since they are valid URIs...

  Pierre-Antoine

-- 
Sometimes I think the surest sign that intelligent life exists elsewhere in the
universe is that none of it has tried to contact us.
(Bill Watterson -- Calvin & Hobbes)