- From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
- Date: Tue, 5 Nov 2002 11:25:47 +0100
- To: <w3c-rdfcore-wg@w3.org>
Going on-list .... Dave: > I'd stick it in 2.3.2 since it is there where you first mention > "Blank". Identification of distinct but blank nodes deserves to go > here. Introducing a term for syntax-specific identifiers for such > blank nodes could be in another doc, but I expect we all want to talk > about them since otherwise we say things like: > nodes are [URI refs] or [Literals] or <nothing to point at here> > > I can live with defining the term blank node identifiers per-syntax; > it'll just have to be done twice for rdf/xml and n-triples, and again > in other docs? I am looking at this para and the latest editors version of concepts: http://sealpc09.cnuce.cnr.it/jeremy/RDF-concepts/2002-11-05/rdf-concepts.htm l (I currently have the lock). Graham defines "local node identifier" (I prefer "blank node identifier", I will chat with him as to why the change). I carefully avoid a constructive definition of blank node, a definition just gets in the way. [[ The blank nodes of an RDF graph are those nodes that are not RDF literals or RDF URI References. ]] is what I say; so that as far as the abstract syntax is concerned my mum can be a subject or an object of a triple in a graph. Then she is a node in the graph, and by virtue of not being a literal or a URIref she is a blank node. All that means is that her internal structure is not relevant to RDF. Jeremy
Received on Tuesday, 5 November 2002 05:25:53 UTC