W3C home > Mailing lists > Public > public-xmlsec@w3.org > April 2010

Re: Prefix rewriting considerations

From: Peter Saint-Andre <Peter.SaintAndre@webex.com>
Date: Tue, 13 Apr 2010 20:14:23 -0600
To: Meiko Jensen <Meiko.Jensen@rub.de>, <public-xmlsec@w3.org>
Message-ID: <C7EA811F.85FC%Peter.SaintAndre@webex.com>
On 4/13/10 12:41 PM, "Meiko Jensen" <Meiko.Jensen@rub.de> wrote:

> continuing with today's discussion on prefix rewriting I wanted to bring
> up the prefix-free canonicalization we published last year (ACM SWS
> 2009). The approach basically consists in rewriting every XML element to
> a representation like this: <localname xmlns="namespace-uri">
> This way, you don't need prefixes at all (besides for the attributes),
> the resulting XML is still well-formed, there is no "unsigned"
> information besides the prefix name, and you don't need doing
> digests+base64 stuff nor excessive search for predefined prefix
> mappings. 

FWIW, this is exactly what we've always done in Jabber/XMPP: no prefixes, no
namespaced attributes, apply the namespace of an element to all elements and
attributes below that element in the tree until and unless you come upon a
child element that is qualified by a different namespace. This is probably a
legacy of the time when Jabber was first defined because namespaces were not
well-understood back then (1998/1999), but it has worked quite well for us
since then. I never knew that this approach had a fancy name ("prefix-free
canonicalization")...

Peter
Received on Wednesday, 14 April 2010 02:15:44 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 14 April 2010 02:15:45 GMT