W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > August 2003

Re: xml literal lex space question

From: pat hayes <phayes@ihmc.us>
Date: Fri, 1 Aug 2003 11:53:48 -0500
Message-Id: <p06001a69bb504a043be1@[10.0.100.23]>
To: Graham Klyne <gk@ninebynine.org>
Cc: w3c-rdfcore-wg@w3.org

Regarding XML literals, I have one request to make to the WG, in its 
deliberations on this topic. Speaking as the semantics editor, I 
really do not give a damn what XML literal values are defined to 
actually BE. What I do care about, a lot, is this:  whatever they are 
defined to be, that the lexical-to-value mapping from the legal 
lexical forms is precisely and unambiguously 1:1. We have already 
canonicalized the lexical forms about as strongly as the combined 
forces of XML and RDF are able to canonicalize them; the operational 
point of doing this is so that an RDF engine can compare equality of 
literal values by comparing identity of lexical form, and we don't 
have to have special inference rules which substitute different but 
co-referring XML forms one for another in all possible ways.  If we 
lose this 1:1 nature of the mapping, then we might as well not have 
bothered to have a canonicalized lexical form. SO please don't do 
anything that runs a risk of losing that property. I do not know if 
having lexical-to-value be UTF-8 encoding does so or not. I would 
rather not have to know, to be honest.

Pat
-- 
---------------------------------------------------------------------
IHMC	(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.	(850)202 4416   office
Pensacola			(850)202 4440   fax
FL 32501			(850)291 0667    cell
phayes@ihmc.us       http://www.ihmc.us/users/phayes
Received on Friday, 1 August 2003 12:54:01 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:59:32 EDT