W3C home > Mailing lists > Public > www-webont-wg@w3.org > October 2002

RE: importing and entialment

From: Smith, Michael K <michael.smith@eds.com>
Date: Wed, 23 Oct 2002 13:39:44 -0500
Message-ID: <B8E84F4D9F65D411803500508BE32214115E6C2C@USPLM207>
To: Jeremy Carroll <jjc@hpl.hp.com>, www-webont-wg@w3.org

An alternative to an error is an assertion.  

Consider treating the imports statement as a macro.  Successful macro
expansion pulls in the appropriate imported triples/XML.  Plus it could add
an assertion

 thisOntology owl:importSuccess http://www.example.com/ontology2.owl

indicating that that ontology2 was pulled in.  In order to make it simple to
detect an incomplete importation chain we could provide

 thisOntology owl:importFailure http://www.example.com/ontology2.owl

Just a thought.  Of course this is really just a translation of the 404
error into something accessible to ontological reasoning.  Fundamentally,
the question is where in the processing layers this information should best
be made  available.  

An example of the use of such a feature would occur where you are importing
many instances from all over the web based on entries in a database.  You
don't want processing to stop if one document is missing.  But it might be
handy to be able to count successes and failures and report those numbers.

- Mike

Michael K. Smith, Ph.D., P.E.
EDS - Austin Innovation Centre
98 San Jacinto, #500
Austin, TX  78701

* phone: +01-512-404-6683
* mailto:michael.smith@eds.com


-----Original Message-----
From: Jeremy Carroll [mailto:jjc@hpl.hp.com]
Sent: Tuesday, October 22, 2002 1:22 PM
To: www-webont-wg@w3.org
Subject: Re: importing and entialment



> However, I believe an error condition is still better than just
> saying "well let's assume the imports statement refers to an empty
> document."

Agreed.

certainly a 404, or a 500 is an error - I was pedantically pointing out that

being different they have different characteristics. 

I would always want to report such errors with some appropriate mechanism
that 
an end user might eventually see.

Jeremy
Received on Wednesday, 23 October 2002 14:39:52 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:57:53 GMT