- From: Mark Birbeck <mark.birbeck@webbackplane.com>
- Date: Sun, 28 Nov 2010 19:17:21 +0000
- To: "nathan@webr3.org" <nathan@webr3.org>
- Cc: RDFA Working Group <public-rdfa-wg@w3.org>
Yes, I would think so. Wording is pretty much a direct lift from RDFa though, which has all of this in. On Sunday, November 28, 2010, Nathan <nathan@webr3.org> wrote: > Cheers Mark, > > Thus I guess we'll be needing some base functionality and notes about resolving relative references in there (?) > > Best, > > Nathan > > Mark Birbeck wrote: > > Hi Nathan, > > RDF only uses IRIs, that's true, but RDF *serialisations* generally > use IRI references. > > Note that an 'IRI reference' is not just a relative IRI as I think you > are implying -- they can also be absolute. That's why 'IRI reference' > is usually used in specs where you want to allow both relative /and/ > absolute paths. In those situations using the definition for 'IRI' > wouldn't work, because that would then require a scheme and a path, > ruling out relative IRIs. > > I think it's important to allow relative IRIs in the API. Obviously > they have to be converted to absolute IRIs in order to obtain RDF, but > as with RDF serialisations they're an important and useful shortcut > for programmers. > > Regards, > > Mark > > On Sun, Nov 28, 2010 at 4:49 PM, Nathan <nathan@webr3.org> wrote: > > Hi All, > > Throughout the API documentation, it references "IRI References", however to > the best of my knowledge the API, and RDF uses only "IRI"s, and in fact IRI > References (../foo) aren't used at all. > > IRI = scheme ":" ihier-part [ "?" iquery ] [ "#" ifragment ] > > Correct? or? > > Best, > > Nathan > > > > > > > >
Received on Sunday, 28 November 2010 19:17:56 UTC