- From: Gregg Kellogg <gregg@greggkellogg.net>
- Date: Fri, 31 Aug 2012 09:55:58 -0400
- To: Oskar Welzl <lists@welzl.info>
- CC: Ivan Herman <ivan@w3.org>, W3C RDFa Community <public-rdfa@w3.org>, Thomas Baker <tom@tombaker.org>
On Aug 31, 2012, at 4:09 AM, "Oskar Welzl" <lists@welzl.info> wrote: > >> As far as I remember, the advice of DCMI (Tom, correct me if I am >> wrong) was to use only /dc/terms in future, [...] >> Personally, I would like to defer the choice to DCMI. > > The choice to prefer DCTerms for current/future use is one thing. > > The problem arises when you change existing RDFa to RDFa 1.1 and rely on > the initial context. In this case, the RDF output doesn't comply to the > DCMI recommendations. I came across this when I tried to convert the > RDFa 1.0 example in wikipedia [1] to RDFa 1.1, removing XML namespace > declarations while doing so. > > I expected the same RDF triples thanks to the initial context, but got > > <http://example.org/john-d/> > <http://purl.org/dc/terms/creator> "Jonathan Doe"@en; > > instead of the original > <http://example.org/john-d/> > <http://purl.org/dc/elements/1.1/creator> "Jonathan Doe"@en; > > Now dc:creator (or http://purl.org/dc/elements/1.1/creator) may be used > with literals, so the original RDFa 1.0 was OK. > OTOH, dcterms:creator, according to the user guide [2], must not be used > with literals. I had overrule the initial context and insert the > corresponding prefix into the HTML+RDFa 1.1 example to make it work. > > I don't know how many will change from existing markup to 1.1, but if > they do there'll be some more unwanted literals as dcterms:creators out > there. > > Maybe it'd be enough to point this out somewhere. I have no idea how > relevant it will become. One purpose of the initial context is as a fallback mechanism, in case a document is authored accidentally without appropriate prefix definitions. For example, this made many documents using OGP valid RDFa again. We had considered that a best practice would be for authors to not rely on prefixes defined in the initial context and recommend that authors define their own @prefix definitions. The initial context is a convenience method which doesn't replace the need for authors with specific vocabulary needs from being explicit. As new use of DCMI should use dc/terms anyway, providing a convenience mechanism makes sense, IMO. Gregg > Cheers > Oskar > > [1] http://en.wikipedia.org/wiki/RDFa#XHTML.2BRDFa_1.0_example > [2] > http://wiki.dublincore.org/index.php/User_Guide/Publishing_Metadata#dcterms:creator > > >> Cheer >> >> Ivan >> >> On Aug 29, 2012, at 20:42 , Oskar Welzl wrote: >> >>> I only noticed a few day ago that both dc: and dcterms: are used for >>> http://purl.org/dc/terms/ >>> in the initial context. In most examples or practical uses so far, I've >>> seen dc: being used for >>> http://purl.org/dc/elements/1.1/ >>> instead, making it easy to distinguish between the legacy vocabulary and >>> DCTerms. (Practical example: >>> http://wiki.dublincore.org/index.php/User_Guide/Publishing_Metadata) >>> >>> >>> What's the rationale behind this decision in the initial context? >>> >>> Oskar >>> >>> >> >> >> ---- >> 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 Friday, 31 August 2012 13:56:43 UTC