I have earlier used the XSLT document() function to point to a URL
which returned a well-formed XML. The URL in my case was google news
feed (an XML stream). I applied XSLT templates on the XML returned by
the URL.

In many places around my site--which deals with sensitive data--my XSL
calls upon the document function to ask a server script to retrieve
and serve up database results in XML.

This script does not execute SQL arbitrarily, but based on its passed
variables. However, this server script is set to ignore anything out
of the house (originating from IP other than 192.168.0.7).

While this isn't XSL, per se, I'm sure someone out there has a similar
method for getting live XML via document(). Or maybe its just bad
practice in principle to rely on it so heavily?

Thoughts, please!

-S

--
Regards,
Mukul Gandhi