On Tue, 28 Oct 2003, pat hayes wrote: > Well, OK, but guys, he has a point. Why not just tweak Ntriples to > allow all of ASCII? That wouldn't break any of our stuff and it would > be generally good-citizen-ish. I find the argument underwhelming. The current situation is that the possible identifiers in Ntriple are a proper subset of those in RDF/XML. The proposed new situation is that the possible identifiers in Ntriple are a proper subset of those in RDF/XML. If you are generating identifiers and wish them to be interoperable between RDF/XML and NTriple it can be done in either case (generate Ntriple ids) If you are reading in RDF/XML and wisht to use the identifiers in Ntriple it can't be done in either case. (You need to do some work). We should only change things that are broken - I am not even convinced this is an improvement. JeremyReceived on Wednesday, 29 October 2003 03:58:52 EST
This archive was generated by hypermail pre-2.1.9 : Wednesday, 29 October 2003 03:58:55 EST