W3C home > Mailing lists > Public > www-webont-wg@w3.org > May 2003

RE: Guide Comments: Suggested response w/ question

From: Smith, Michael K <michael.smith@eds.com>
Date: Wed, 7 May 2003 10:22:21 -0500
Message-ID: <B8E84F4D9F65D411803500508BE3221415CF2FB9@USPLM207>
To: Jeremy Carroll <jjc@hpl.hp.com>, www-webont-wg@w3.org

Jeremy,

> the sentence quoted was concerning the abstract syntax, which is 
> well removed from surface representation issues such as XML Base. 

I agree with that answer, but I think it is a slightly weasly way out of 
the readers question.  Maybe a reminder at that point in the S&AS to the 
effect that this URI represents a translation from any one of the 
multiple surface syntax possibilities?  Since we don't anywhere really 
tell directly how the translation from rdf/xml to abstract syntax occurs.

Your example of the utility of xml:base is good.  Will incorporate in 
my response.

- Mike

-----Original Message-----
From: Jeremy Carroll [mailto:jjc@hpl.hp.com] 
Sent: Friday, May 02, 2003 2:03 PM
To: www-webont-wg@w3.org
Subject: Re: Guide Comments: Suggested response w/ question




sorry, forgot, S&AS does not conflict ...

the sentence quoted was concerning the abstract syntax, which is well
removed 
from surface representation issues such as XML Base. 

The usage of xml:base that already happens, for example with our owl:imports

test cases, is that the xml:base and rdf:about idiom specifies a logical 
location, and the document might actually have been retrieved from somewhere

else, such as in a zip file, or a local cache.

The S&AS sentence does suggest that it would be a mistake to use a URL for
an 
ontology that was different from the one that can be retrieved from that
URL; 
and would discourage the use of a non-retrievable URI with some private 
mechanism to relate URIs with ontologies. (The Web get action is the
ontology 
retrieval action). The quotes from Guide do not contradict this.

A further reason why the xml:base mechanism is good is that many different 
URLs retrieve the same physical bits. By including an xml:base within the 
bit-stream then one of those equivalent URLs is given as preferenced, by the

document author. This minimizes the need for the receiver to make good.

e.g.

http://www.w3.org/2002/07/owl

can be retrieved with:

http://www.w3.org/2002/07/owl.rdf
HTTP://www.w3.org/2002/07/owl
http://www.w3.org:80/2002/07/owl HTTP://18.7.14.127/2002/07/owl.rdf

However, because of the xml:base in it, all of these correspond to identical

RDF graphs.

See
http://www.w3.org/RDF/Validator/ARPServlet?PARSE=Parse%20URI:%20&URI=HTTP://
18.7.14.127/2002/07/owl.rdf

Jeremy
Received on Wednesday, 7 May 2003 11:22:35 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:00 GMT