- From: Antoine Isaac <aisaac@few.vu.nl>
- Date: Tue, 26 Jul 2011 11:28:59 +0000
- To: Richard Cyganiak <richard@cyganiak.de>
- Cc: "public-rdf-wg@w3.org, SKOS , Ivan Herman , Coralie Mercier , Guus Schreiber" <public-esw-thes@w3.org>
Hi Richard, > On 18 Jul 2011, at 21:21, Antoine Isaac wrote: >>> Alan Ruttenberg wrote: >>> without a language tag it is a string, with a language tag it is a pair of >>> strings. The set of plain literals without language tags is *not* the >>> set of pairs (string , ""). >> >> (which I think matches what is written in the RDF Concepts and Abstract Syntax) > > Alan is right. "foo" does *not* have a language tag. And RDF Concepts is the normative reference here. > > Other specifications and implementations may explicitly treat an empty string as absence of a language tag, but that doesn't change anything. > > Note that in the editor's draft for the upcoming RDF 1.1 Concepts [1] (Section 6.5), this is a lot clearer because all literals now are either "typed" or "language-tagged". > > Best, > Richard > > > [1] http://dvcs.w3.org/hg/rdf/raw-file/default/rdf-concepts/index.html Thanks for the clarification. That's something solid to work on I'm still slightly uncomfortable with your: > Other specifications and implementations may explicitly treat an empty string as absence of a language tag, but that doesn't change anything. It changes something: it can make a reader puzzled, and may result in loosing time in discussions that are merely caused by what seems to be (but in fact is not) lack of synch between specs. It's an editorial issue, granted. But still it's an issue, no? And perhaps the solution could be straightforward: just emphasizing that the reference is really the "RDF Concepts" doc, and that some syntax-specific handling of language tags in RDF/XML should not be read in a way that questions this reference. Cheers, Antoine
Received on Tuesday, 26 July 2011 11:48:57 UTC