IRIs and NFC: RDF issue williams-02

This message is sent on behalf of the RDF Core WG to the I18N WG.

We would like a one bit confirmation of a conversation I had with Martin  
Dürst.

We are considering the last call comment:
http://www.w3.org/2001/sw/RDFCore/20030123-issues/#williams-02
s/URIref/IRI/

raised by Stuart Williams and Tex Texin.

In a private conversation with Martin Dürst, he informed me that IRIs do *not* 
have to be in Unicode NFC.
However, parts of the current design of the RDF abstract syntax:

http://www.w3.org/TR/2003/WD-rdf-concepts-20030123/#section-Graph-URIref

are based on advice given to RDF Core by the I18N WG at the Cannes plenary 
2002, which included the constraint that the internationalized URIs should be 
in NFC.

Please send an e-mail labelled as on behalf of the I18N-WG to www-rdf-comments 
saying something like:
[[
We [affirm|withdraw] our advice at the Cannes plenary that IRIs should be in 
NFC.
]]

We note that without this constraint the intent of that section is the same as 
the intent of the definition of IRI reference in XML Namespaces 1.1.

Jeremy Carroll on behalf of RDF Core

Further Personal comment:
We have not yet decided our disposition on this issue and have, surprisingly, 
not yet had an official comment from I18N concerning this. Options range from 
doing nothing through to deferring to the XML Namespaces definition and doing 
a global substitution of "IRI" for "URI". Please however, don't delay in 
providing the single bit [affirm|withdraw] in order to discuss the wider 
issue.

Received on Friday, 28 March 2003 13:39:30 UTC