- From: Richard Cyganiak <richard@cyganiak.de>
- Date: Fri, 27 Mar 2015 11:49:49 +0000
- To: Holger Knublauch <holger@topquadrant.com>
- Cc: "public-data-shapes-wg@w3.org" <public-data-shapes-wg@w3.org>
> On 26 Mar 2015, at 23:31, Holger Knublauch <holger@topquadrant.com> wrote: > > I noticed that we should rename sh:nodeType so that it doesn't conflict with the term "type" in the possible variations of sh:valueType, sh:datatype etc. Looking at the RDF spec [1], what we call the node types are introduced as: > > "There can be three kinds of nodes in an RDF graph: IRIs, literals, and blank nodes." > > Question: Does anyone object to renaming sh:nodeType to sh:nodeKind? Any better proposals? How about sh:termKind? Background: “IRIs, literals and blank nodes are collectively known as RDF terms” [1]. Why do we need bot words, “node” and “RDF term”? Because RDF terms can exist outside of an RDF graph (for example, in a SPARQL SELECT result, or as arguments and results of SPARQL functions), and it wouldn’t be appropriate to talk of “nodes” in these contexts. So, the nodes in an RDF graph must be RDF terms, but RDF terms are only nodes in the context of a particular RDF graph that happens to contain them. Another option would be to just go with sh:kind. Best, Richard [1] http://www.w3.org/TR/rdf11-concepts/#dfn-rdf-term
Received on Friday, 27 March 2015 11:50:14 UTC