- From: Sebastian Heath <sebastian.heath@gmail.com>
- Date: Tue, 6 Nov 2012 10:37:10 -0500
- To: RDFa Working Group <public-rdfa-wg@w3.org>
Ivan wrote: > > Indeed, the difference between what you propose and what is currently there already, is whether non-registered prefixes would be completely disallowed in your scheme. > > Agreed. And it would be truly problematic for me if non-registered prefixes were to be disallowed. Take bibliographic information for example. Dublin Core (which I assume we all agree would make the cut, despite its poor historical choice to engender confusion of 'terms' and 'elements' and varieties of the latter) has some basic bibliographic vocabulary. "BibliographicResource" and "BibliographicCitation" are useful, along with creator, title, etc. But nobody could think that it's a complete vocab for describing bibliographic data. So other vocabs exist: http://bibliontology.com/ , http://purl.org/spar/cito , various renderings of MARC into RDF (e,g, http://www.bl.uk/bibliographic/datafree.html), the same for FRBR (http://vocab.org/frbr/core.html). Which of these will the W3 endorse with predefined prefixes? All? That seems silly. Have their sponsoring organizations agreed to the W3 patent principles? Or will the W3 pick and choose? That will be arbitrary. Or endorse none of these and abandon bibliographic data as a first class citizen on the web? That's totally deficient. As in, you can't seriously be asking me to stick in full URIs when using these vocabs? That will be error prone. And here's a (only half-serious) shot across the bow: Schema.org is only at an experimental phase and is relatively new. So I'm assuming we'll exclude it from any set of standardized prefixes? We have a well thought out, existing and working mechanism for allowing prefixes. Again, I move we keep it as is. -Sebastian
Received on Tuesday, 6 November 2012 15:37:42 UTC