RE: [xml-dev] Rich & Elliotte were right, I was wrong

["Derek Denny-Brown" <derekdb@microsoft.com>]

I had heard of this (which is why I finished my rant that way, in fact),
but I wasn't sure if it was public.

I mostly agree with all their points _except_ the need for WXS datatype
normalization.  That should be distinct from everything else, and should
almost never be necessary.  You want to shred your UDDI information and
store it in tables/rows?  Great idea.  Just keep a bit copy of the
original around too... then you wouldn't need this.

Xml-DSig targeted messaging scenarios, where people might parse/filter
the message, but preserved the basic content.
SchemaCentricCanonicalization only seems necessary if you complete
parse/shred the original into a strongly-typed store (i.e. not generic
xml-text), _and_ dropping the original text.  How expensive can it be to
just keep the original around to?  Given that this is all related to
signing xml documents, I'm not sure how comfortable I am having
intermediates having that much ability to reformulate the signed
document and still consider it signed.  That just scares me.

-derek

> -----Original Message-----
> From: Robin Berjon [mailto:robin.berjon@expway.fr]
> Sent: Monday, March 15, 2004 10:41 AM
> To: Derek Denny-Brown
> Cc: xml-dev@lists.xml.org
> Subject: Re: [xml-dev] Rich & Elliotte were right, I was wrong
> 
> Derek Denny-Brown wrote:
> > I can't wait for people to ask for WXS normalization.
> 
> Wait no longer:
> 
>    http://uddi.org/pubs/SchemaCentricCanonicalization.htm
> 
> --
> Robin Berjon