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

Re: Issue timbl-03 "collection clutter" proposal to close

From: Dan Brickley <danbri@w3.org>
Date: Thu, 24 Apr 2003 06:19:41 -0400
To: Dave Beckett <dave.beckett@bristol.ac.uk>
Cc: w3c-rdfcore-wg@w3.org
Message-ID: <20030424101940.GB2012@tux.w3.org>

* Dave Beckett <dave.beckett@bristol.ac.uk> [2003-04-24 11:01+0100]
> 
> 
> Summary: reject
> 
> The comment raised in
> http://lists.w3.org/Archives/Public/www-rdf-comments/2003JanMar/0586.html
> 
> [[
>   I believe that in 7.2.19 Production parseTypeCollectionPropertyElt
>     the wording
> 
>   """For each event nin s, the following statement is added to the graph:
> 
>   n.string-value <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> 
>   <http://www.w3.org/1999/02/22-rdf-syntax-ns#List> .
> 
> 
>   """
> 
>   adds redundant triples to the graph.
>   I believe that waste of time and space at this level in the 
>   architecture is unnecessary, and that that wording should be removed 
>   (and any other reference to the adding type statements for Lists where 
>   a rdf:first is there).
> 
>   It is trivial to restore the triples for anyone who wants them fro a 
>   graph without them,
>   using
> { ?x rdf:first ?y } =>  { ?x  a rdf:List }.
> ]]
> 
> 
> Here is a draft response:
> 
> [[
> The RDF Core WG has considered your last call comment captured in
> 
>    http://www.w3.org/2001/sw/RDFCore/20030123-issues/#timbl-03
> 
> (raised in
>  http://lists.w3.org/Archives/Public/www-rdf-comments/2003JanMar/0586.html
> )
> 
> and decided
> 
>    URL-HERE
> 
> to reject it on the grounds that it is being used by OWL and could
> not be added at a higher level since it is closedly tied to an
> RDf/XML syntax abbreviation.
> 
> This triple is part of the closed collection form added to the RDF
> model and RDF/XML syntax for use by OWL based on the DAML+OIL
> daml:Collection syntax extension to RDF/XML. 
> 
> The reason this could not be added at the OWL level is that it is
> generated by the rdf:parseType="Collection" syntax which is in the
> RDF/XML specification.  There is no "hook" to allow optional adding
> of <x> rdf:type rdf:List for the generated notes.
> 

I thought TimBL's point was that the 'this is a List' statements could
be inferred through knowledge of the rdfs:domain of rdf:first and 
rdf:rest. Anything familiar with the meaning of these rdf properties 
will know that they can be truly applied only to lists. In that light,
forcing parsers to emit this data explicitly is indeed rather redundant.

The OWL use case should be fine without the explicit extra triples, 
since OWL is all about inferring extras.

http://www.w3.org/TR/rdf-schema/#ch_first notes this quite explicitly,
[[
5.2.2 rdf:first

rdf:first is an instance of rdf:Property that is used to indicate the first item of a list.

A triple of the form:

    L rdf:first O

states that L is an instance of rdf:List and that O is the first item of the list.

The rdfs:domain of rdf:first is rdf:List. The rdfs:range of rdf:first is rdfs:Resource.
5.2.3 rdf:rest

rdf:rest is an instance of rdf:Property that is used to indicate the 
sublist that contains the items of a list other than the first.

A triple of the form:

    L rdf:rest O

states that L is an instance of rdf:List, that O is an instance of 
rdf:List and O is the list L without L's first item.

The rdfs:domain of rdf:rest is rdf:List. The rdfs:range of rdf:rest is rdf:List.
]]

Dan
Received on Thursday, 24 April 2003 06:19:45 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:56:58 EDT