[jdom-interest] more on namespaces && CDATA (and xpath, oh my!)

[Jason Hunter]

> Been reading more standards (Canonical XML this time)...

That's the nice thing about XML.  As much reading as you have time for. 
:-)

> I think this W3C working draft explains that namespace declaratsions
> can't be shifted around by the underlying XML processor, regardless of
> semantics it thinks it understands.
> 
> Also, as an argument against CDATA, it says that the processor should
> replace all CDATA sections with their character content.

Canonical XML deals with the question of how to decide if two docs are
"logically equivalent".  It saying you ignore CDATA when testing doesn't
mean that we shouldn't have CDATA.

-jh-