- From: Marc Hadley <marc.hadley@sun.com>
- Date: Fri, 10 Jan 2003 14:07:40 -0500
- To: Rich Salz <rsalz@datapower.com>
- Cc: Joseph Reagle <reagle@w3.org>, w3c-ietf-xmldsig@w3.org
On Friday, Jan 10, 2003, at 13:51 US/Eastern, Rich Salz wrote: >> Sounds like casting the soap-c14n spec as a transform is the best >> bet, I'll modify it accordingly. > > No, I disagree. If you make it a URI then it is easy for XSLT folks > to recognize that URI and call the built-in transform. It is also > easy for code-based folks to call their built-in subroutine. If you > make it a transform, then both folks will have to do some serious work > (e.g., canonicalize!) the transform to see if it's the special-case > applies. > I think we may be having a terminology problem here. What I meant by 'casting the soap-c14n spec as a transform' was to remove the 'and then do exclusive c14n' step from the spec text (leaving the textual description of what to do to the SOAP message infoset) and discuss it instead in terms of a DSIG transform. I'd still assign a URI to the resulting algorithm. An actual XSLT could be included in a non-normative appendix as a example implementation. Make sense ? Marc. -- Marc Hadley <marc.hadley@sun.com> Web Technologies and Standards, Sun Microsystems.
Received on Friday, 10 January 2003 14:07:19 UTC