- From: Bijan Parsia <bparsia@cs.man.ac.uk>
- Date: Wed, 21 Jan 2009 20:24:20 +0000
- To: W3C OWL Working Group <public-owl-wg@w3.org>
See: http://www.w3.org/2007/OWL/wiki/LC_Responses/MD1
Copied below.
Ivan, please check that I got your point right.
Cheers,
Bijan.
The basic reason is that RDF/XML and XML are (by and large) tied to
older versions of Unicode, and we're following them.
We can:
* Stick with 3.0.
* Update to 5.0.
* Let it be extensible, adapting to the "latest version".
Where it matters:
* Identifiers, i.e., what counts as a legal spelling for IRIs
* String/plain literals (for obvious reasons).
Possible impact:
* Semantically -- new characters mean that counts of strings
might change, which mean that some unsatisfiabilities would change.
* Syntactically -- the structural model might be able to contain
IRIs and strings which it cannot serialize into some syntaxes
(because, for example, they were defined against Unicode 3, or 5).
This seems to be the case for RDF/XML and XML
* http://www.w3.org/TR/rdf-syntax-grammar/#section-Terminology
* http://www.w3.org/TR/REC-xml/#Unicode
Note that "release 5" of XML changed the reference from 3 to 5, but
with a lot of controversy. RDF seems to normatively reference 3.
* http://www.w3.org/TR/rdf-concepts/#section-Normative-References
* http://www.w3.org/TR/2004/REC-rdf-syntax-grammar-20040210/
#section-References
Since it's not just the RDF/XML syntax, but the RDF model, it may be
the case that we would have ontologies that are not representable in
the RDF Model.
Ivan's thought was that we could be "open ended" with respect to
Unicode (i.e., not reference a version) but delegate to other specs
(i.e., RDF or XML). That way, we would be, at least, in synch with
them if they should update (as XML is already trying to).
(Jeremy would be a good person to ask about it.)
BijanParsia 20:16, 21 January 2009 (UTC)
Received on Wednesday, 21 January 2009 20:20:58 UTC