• To: "Roger L. Costello" <costello@m...>
• Subject: Re: Canonicalizer that uses XML Schemas (rather than DTDs)?
• From: Elliotte Harold <elharo@m...>
• Date: Mon, 22 Nov 2004 19:47:37 -0500
• Cc: xml-dev@l...
• In-reply-to: <200411221948.iAMJmNu10704@s...>
• References: <200411221948.iAMJmNu10704@s...>
• User-agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.3) Gecko/20040910

Roger L. Costello wrote:


> Why would it matter that a canonicalizer tool utilize an
> XML Schema rather than a DTD?  Where would it make a difference?

Such a canonicalizer would be nonconformant to the XML canonicalization 
specification. It would not produce the specified result in many 
important cases. This would have downstream effects of breaking XML 
digital signatures.

The purpose of canonicalization is not simply to normalize a document by 
removing its DTD and applying default attribute values. Canonicalization 
is intended to reproducibly create byte-per-byte identical binary data 
stream from equal XML documents, for a very precise definition of 
"equal". Changing the definition of "equal" pulls the legs out from 
under canonicalization. :-(

-- 
Elliotte Rusty Harold  elharo@m...
XML in a Nutshell 3rd Edition Just Published!
http://www.cafeconleche.org/books/xian3/
http://www.amazon.com/exec/obidos/ISBN=0596007647/cafeaulaitA/ref=nosim

• Follow-Ups:
  • RE: Canonicalizer that uses XML Schemas (rather than DTDs)?
    • From: "Roger L. Costello" <costello@m...>

• References:
  • Canonicalizer that uses XML Schemas (rather than DTDs)?
    • From: "Roger L. Costello" <costello@m...>

• Prev by Date: Re: licensing ... [Re: binary XML API and scientific use cases [Re: [ANN] nux-1.0beta2 release
• Next by Date: Re: use cases: binary XML for scientifc computing
• Previous by thread: Re: Canonicalizer that uses XML Schemas (rather than DTDs)?
• Next by thread: RE: Canonicalizer that uses XML Schemas (rather than DTDs)?
• Index(es):
  • Date
  • Thread