Steve Tinney wrote:

> If this is really, absolutely necessary, you could hack the text()
> template to detect 0xa0 and replace it with  .  The input to the
> next phase of XML processing you mentioned might not then be valid,
> though.  You might get somewhere by investigating the output encoding,
> but I'm not familiar enough with XT to say these days.  I've switched
> irrevocably to Saxon; you might look at that instead.
>

Right, but I don't want it to work on a browser.  I want it to be parseable
again as XML.

As for what you're saying as far as Saxon goes... I would assume that that
would have the same problems that I'm complaining about (whether or not I
want it to behave as the spec recommends :-).  Unless of course, Saxon would
support the iso-8859-1 (or whatever) spec that I have read seems to preserve
the &nbsp's as I want them preserved.

Thanks for your response.  I'll be looking into saxon.

Duke



 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list