rdf:PlainLiteral (was Re: RDF 2 Wishlist)

> It still remains a bad idea in my opinion because

Both your reasons are based on a misunderstanding.  The key bit you're
missing:

       applications that employ this datatype MUST use plain literals
       (instead of rdf:PlainLiteral typed literals) whenever a syntax
       for plain literals is provided, such as in existing syntaxes for
       RDF graphs and SPARQL results.

In other words, rdf:PlainLiteral literals MUST NOT appear in RDF/XML,
N-Triples, Turtle, RDFa, etc.  On the other hand, they MAY appear in
documents using the OWL XML syntax and RIF XML syntax, which do not
have any other way to state plain literals.  No old code needs to
change, and new code (for some new syntaxes) can be simpler (having only
datatyped literals).

> 1) It attempts to update an existing REC without updating the document by
> adding a new term to the rdf: namespace
> 
> It has this interaction with the rdf/xml recommendation (2004):
> http://www.w3.org/TR/2004/REC-rdf-syntax-grammar-20040210/#section-Namespace
> "Any other names are not defined and SHOULD generate a warning when
> encountered, but should otherwise behave normally."
> 
> That'll be showing up in logs and validators probably.

Since it's an error to use rdf:PlainLiteral in RDF/XML, giving a warning
remains a good idea.

> 2) Adds two ways to do one thing
> 
> The RDF(2004) REC (concepts, syntax) way, which includes all derived formats
> and implementations that are based on rdf(2004) concepts - syntaxes, APIs
> and the OWL2(2009) / rdf:text way.

As you can see now, there remains only one way to serialize an RDF Plain
Literal in a given syntax.

This design is tricky, and kind of odd, I know, but I don't think it
causes the problems you were thinking it did.

     -- Sandro

Received on Tuesday, 3 November 2009 07:28:24 UTC