- From: Sandro Hawke <sandro@w3.org>
- Date: Fri, 16 Feb 2001 12:54:01 -0500
- To: pat hayes <phayes@ai.uwf.edu>
- cc: www-rdf-logic@w3.org, ht@w3.org
Pat Hayes writes:
>.... But since the same number might be
>written in binary, decimal, hexadecimal or octal; all of them, for
>various purposes, with a claim to be the canonical representation,
>why not use a terminology which indicates the encoding, say for
>example: rdf:decnum rdf:binnum rdf:octnum and so on? Notice
>that 'num' here means numeral, not number.
That makes sense, but XMLSchema happens to define only one "lexical
representation" mapping and one "canonical lexical representation"
mapping for its "integer" datatype, and they happen to use base 10.
[1]
So I think you're heading somewhere a little different than the
simplest practical merging of XMLSchema and DAML+OIL (which I think is
what is being proposed [1]).
Would you like to propose something else, like actually defining
datatypes axiomatically? I personally like that idea, but people say
that writing the axioms and writing the systems to work with them
efficiently are both difficult or maybe impossible. I haven't managed
to prove them wrong yet (and since I haven't really tried, I'm still
optimistic).
-- sandro
[1] http://www.w3.org/TR/xmlschema-2/#integer
[2] http://www.cs.man.ac.uk/~horrocks/daml+oil/Datatypes/datatypes.html
Received on Friday, 16 February 2001 12:54:07 UTC