W3C home > Mailing lists > Public > w3c-ietf-xmldsig@w3.org > January to March 2005

RE: Test Case with xml-dsig

From: John Boyer <JBoyer@PureEdge.com>
Date: Mon, 7 Feb 2005 15:29:42 -0800
Message-ID: <7874BFCCD289A645B5CE3935769F0B52750785@tigger.pureedge.com>
To: "Elliotte Harold" <elharo@metalab.unc.edu>
Cc: "Norman Walsh" <Norman.Walsh@Sun.COM>, "Joseph Reagle" <reagle@mit.edu>, "Gabe Wachob" <gwachob@wachob.com>, <public-xml-id@w3.org>, <w3c-ietf-xmldsig@w3.org>

Hi Elliotte,

Well, how about I say something controversial for a change!

C14N is the canonicalizer for XML 1.0.  XML 1.0 has a namespace.
There are two attributes in it, xml:lang and xml:space.  C14N 
does the right thing for the "collection of names identified by 
the URI reference" associated with the prefix xml.  

Now, the W3C wants to add more names to the collection of
names associated with the XML 1.0 namespace.  As we are now
finding, something breaks when you do that.

This is the whole point of namespace qualification.  We associated
the addition of new vocabulary to XML with an upgrade to the 
XML version, which would then use a different namespace.

In other words, xml:id should not break C14N for XML 1.0 *because* 
xml:id should not exist.  This new, ubiquitously available ID 
attribute belongs in another namespace, preferably an upgrade to XML.

John Boyer, Ph.D.
Senior Product Architect and Research Scientist
PureEdge Solutions Inc.


-----Original Message-----
From: Elliotte Harold [mailto:elharo@metalab.unc.edu]
Sent: Monday, February 07, 2005 3:05 PM
To: John Boyer
Cc: Norman Walsh; Joseph Reagle; Gabe Wachob; public-xml-id@w3.org;
w3c-ietf-xmldsig@w3.org
Subject: Re: Test Case with xml-dsig


John Boyer wrote:

> C14N isn't "just plain broken" with respect to xml:id.
> 
> C14N was produced years before xml:id and therefore 
> "does not support" xml:id.
> 
> Moreover, it cannot be modified to do so without also
> upgrading XPath, which also "does not support" xml:id.

The brokenness of C14N with respect to canonicalization has nothing to 
do with XPath. It has everything to do with the inheritance of all 
attributes in the XML namespace.


> Note that XML Schema can declare something to be an ID,
> and C14N "does not support" that either.

But it doesn't need to. Attribute type information, be it schema or DTD 
based, does not survive the canonicalization process. The mismatch 
between xml:id and C14N does not depend on the type of the xml:id 
attribute.

-- 
Elliotte Rusty Harold  elharo@metalab.unc.edu
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
Received on Monday, 7 February 2005 23:30:23 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Monday, 7 February 2005 23:30:24 GMT