W3C home > Mailing lists > Public > public-rdf-dawg-comments@w3.org > April 2005

Re: replies on named graphs? containers use case

From: Bob MacGregor <bmacgregor@siderean.com>
Date: Thu, 14 Apr 2005 09:11:56 -0700
Message-ID: <425E964C.7010504@siderean.com>
To: Giovanni Tummarello <giovanni@wup.it>
CC: public-rdf-dawg-comments@w3.org
RDF's containers are a sorry blot on the language.  We would be much 
better off
if they had never been invented.  For declaring a set of objects, one 
can choose their
favorite 'hasMember' predicate, and assert the members of the set, without
needing to extend the RDF language.  However, its frequently the case 
that there
is a stronger semantics associated with a collection, in which case the 
chosen to map from a collection to its members can use that stronger 
The strong-semantics option is not available with collections.

The numbered predicates used for sequences are a bastardization of the
predicates that is particularly sorry.  'first' and 'rest' for lists are 
verbose, but they capture the semantics accurately, and should be used
in preference to sequences.

- Bob

Giovanni Tummarello wrote:

> A few days ago, Kampman made an important remark about named graph 
> semantics but i havent seen any reply, did i miss some?
> Also, i also would like to argue for the support for containers.
> For a very interesting use case (that i cant possibly solve without 
> query language support) see .
> http://giovanni.ea.unian.it/temp/rdftef.pdf
> Textual encoding using RDF (requiring chains of words/symbols)
> the fact that there is no knowledge on how to implement this 
> efficiently shouldnt probably be an issue?.. i guess they'll be 
> inefficient at first, but at least people can use them if needed.
> Giovanni


Bob MacGregor
Chief Scientist

	Siderean Software Inc
390 North Sepulveda Blvd., Suite 2070
El Segundo, CA 90245
bmacgregor@siderean.com <mailto:bmacgregor@siderean.com> 	
tel: 	+1-310 647-4266
fax: 	+1-310-647-3470




Received on Thursday, 14 April 2005 16:12:14 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:52:06 UTC