- From: Toby Inkster <tai@g5n.co.uk>
- Date: Tue, 26 Oct 2010 22:08:06 +0100
- To: Mark Birbeck <mark.birbeck@webbackplane.com>
- Cc: Shane McCarron <shane@aptest.com>, nathan@webr3.org, Ivan Herman <ivan@w3.org>, public-rdfa-wg@w3.org, Martin McEvoy <martin@weborganics.co.uk>
On Tue, 26 Oct 2010 18:24:07 +0100 Mark Birbeck <mark.birbeck@webbackplane.com> wrote: > I don't think we should go any further than the canonicalisation > algorithm. FWIW, RDFa 1.0 says to use XML exclusive c14n. However, test case 0101 in the RDFa 1.0 test suite calls for xml:lang attributes to be copied into the child elements of literal - that's a feature of xml c14n, but as I understand it, xml ec14n explicitly says not to do this. Personally I'd be happiest to go with using plain old ec14n. It's a stable specification. XMLLiterals are a rare enough use case that anyone doing particularly delicate things with them should be expected to handle making the literal as self-contained as possible themselves rather than relying on the parser to do so. -- Toby A Inkster <mailto:mail@tobyinkster.co.uk> <http://tobyinkster.co.uk>
Received on Tuesday, 26 October 2010 21:08:36 UTC