- From: <Patrick.Stickler@nokia.com>
- Date: Tue, 13 Nov 2001 22:36:01 +0200
- To: w3c-rdfcore-wg@w3.org
It occurred to me that both the X and P++ proposals have a similar advantage in that qualifications of values do not displace the value itself in the graph representation as is now the case with qualified anonymous nodes. Thus, because literal nodes may act as the subjects of statements, one need not deal with variants such as _:X _:someProperty "foo" . _:Y _:someProperty _:Z . _:Z rdf:value "foo" . _:Z _:someType . but rather, in either case, the value is in the same place in the graph, and statements are simply made about it. I.e P++ _:X _:someProperty _:1:"foo" . _:Y _:someProperty _:2:"foo" . _:2:"foo" rdf:type _:someType . X [1,S,2,3,4] [2,U,{X}] [3,U,{someProperty}] [4,L,"foo"] [5,S,6,7,8] [6,U,{Y}] [7,U,{someProperty}] [8,L,"foo"] [9,S,5,10,11] [10,U,rdf:type] [11,U,{someType}] Note that in both cases, the representations for the statements about _:X and _:Y have identitical structure. This is not the case with the anonymous node form based on the present graph model. (someone please tell me if I've munged up the P++ or NTriples notations) Patrick -- Patrick Stickler Phone: +358 50 483 9453 Senior Research Scientist Fax: +358 7180 35409 Nokia Research Center Email: patrick.stickler@nokia.com
Received on Tuesday, 13 November 2001 15:36:22 UTC