- From: Alex Hall <alexhall@revelytix.com>
- Date: Thu, 17 Nov 2011 12:10:02 -0500
- To: Richard Cyganiak <richard@cyganiak.de>
- Cc: Sandro Hawke <sandro@w3.org>, Ivan Herman <ivan@w3.org>, RDF Working Group WG <public-rdf-wg@w3.org>
- Message-ID: <CAFq2biyv_HN-5ifV-E-Yy1=dx472XbAGaiGxjzynVk+yYcmG8Q@mail.gmail.com>
On Thu, Nov 17, 2011 at 10:55 AM, Richard Cyganiak <richard@cyganiak.de>wrote: > On 17 Nov 2011, at 14:58, Sandro Hawke wrote: > > Are we really breaking backward compatibility that much? > > The design and terminology of the rdf:PlainLiteral spec make no sense at > all when read with RDF 1.1 goggles. This is only partly true. I think there were two motivations for the rdf:PlainLiteral datatype: 1. Give a name to the class of untyped literals in RDF, for use in domain/range statements and in data models that require a datatype IRI for all literals. 2. Allow values from the RDF data model, which has 3 slots (lexical, language, datatype) to be expressed in OWL and RIF data models which only have 2 slots (lexical, datatype) by mashing together the lexical and language parts of the RDF model. The first issue might no longer be a concern with the addition of rdf:LanguageString, but the second issue is still relevant. Short of modifying the OWL and RIF data models to align with RDF, they'll still require something along the lines of rdf:PlainLiteral, even if it's no longer called rdf:PlainLiteral. That said, I agree that rdf:PlainLiteral will need a rewrite to accurately reflect changes in RDF 1.1 -Alex > *I* don't have a problem with that. I don't think rdf:PlainLiteral made > sense in the first place. But the OWL and RIF people might care that one of > their documents no longer has any foundation in the RDF data model. > > And I *do* slightly care that they're squatting in the RDF namespace. Why > were they allowed to do that? This could be fixed as part of the update of > the rdf:PlainLiteral document. > > > Let's just > > include some "Historical Notes" in RDF 1.1 that explains what terms like > > "Plain Literal" meant, and how they should now be understood. > > Well, we currently have this: > > [[ > In earlier versions of RDF, literals with a language tag did not have a > datatype IRI, and simple literals could appear directly in the abstract > syntax. Simple literals and literals with a language tag were collectively > known as plain literals. > ]] > > http://dvcs.w3.org/hg/rdf/raw-file/default/rdf-concepts/index.html#dfn-plain-literal > > I would rather not clutter the spec too much with “Historical Notes” that > no one will care about five years down the road. The spec already had way > too many notes in 2004, and the need to point out every change between 2004 > and 1.1 doesn't help. > > Best, > Richard > > > > > Then, I > > think, existing systems will be fine. (Or... perhaps I'm missing > > something.) > > > > -- Sandro > > > >> Thanks > >> > >> Ivan > >> > >> On Nov 10, 2011, at 18:06 , RDF Working Group Issue Tracker wrote: > >> > >>> > >>> RDF-ISSUE-80 (rdf:PlainLiteral): Ask OWL and RIF WGs to update the > rdf:PlainLiteral spec [RDF General] > >>> > >>> http://www.w3.org/2011/rdf-wg/track/issues/80 > >>> > >>> Raised by: Richard Cyganiak > >>> On product: RDF General > >>> > >>> The ISSUE-12 and ISSUE-71 resolutions, which are now implemented in > the RDF Concepts ED [1], have removed the distinction between plain and > typed literals from the RDF abstract syntax. > >>> > >>> This has a major effect on the rdf:PlainLiteral spec [2]. Parts of it > are now obsolete, and the rest needs updating. > >>> > >>> RDF-WG should ask OWL-WG and RIF-WG to update the document. > >>> > >>> [1] http://dvcs.w3.org/hg/rdf/raw-file/default/rdf-concepts/index.html > >>> [2] http://www.w3.org/TR/rdf-plain-literal/ > >>> > >>> > >>> > >> > >> > >> ---- > >> Ivan Herman, W3C Semantic Web Activity Lead > >> Home: http://www.w3.org/People/Ivan/ > >> mobile: +31-641044153 > >> FOAF: http://www.ivan-herman.net/foaf.rdf > >> > >> > >> > >> > >> > >> > >> > > > > > > > > >
Received on Thursday, 17 November 2011 17:11:02 UTC