RE: N3 and N-Triples (was: RDF in HTML: Approaches)

> I disagree XML in itself is the sticking point. The thing to do is
> beat on the XML syntax until it's as simple as RDF is supposed to
> be.

I'm just curious; what exactly is wrong with n-triples?  I can understand if people want to serialize the triples as XML, but that is trivial.  It seems that n-triples is exactly "as simple as RDF is supposed to be".

(Or, look at it another way: n-triples is preferred for achieving repeatable and comparable tests across processors -- which sounds a lot like the definition of "canonical" to me.)