Re: comments on the XML Canonical specification

John Boyer wrote:
> 
> Hi Lauren,
> 
> <lauren>
> It is harmful to some documents and specifications, but not all
> documents and applications. This doesn't mean that those applications
> that do not need such a dependency are wrong, which is what the current
> language implies. I don't want DOM applications that do treat the prefix
> as syntactic sugar (since those authors read the Namespaces Rec and
> implemented it, without regard to XPath etc) to be labelled as being
> wrong. So my suggestion, again, is to come up with some language that is
> neutral on this point.
> </lauren>
> 
> <john>Fair enough.  I can change the language further to a kind of 'There
> exist documents which are dependent...'.
> However, note that since there exist XSLT and XPath transforms in DSig's
> SignedInfo element, the DSig SignedInfo element is among those documents
> that would be harmed by namespace rewriting.
> </john>

This sounds reasonable. This would also be a flag to various
applications to be careful what they do with the DSig SignedInfo
element.

thanks,

Lauren