RE: several messages

> -----Original Message-----
> From: Chris Lilley []

> I notice that XML Spy 4.0 for example assumes that all xml-stylesheet
> PIs are for XSL, and removes any existing CSS ones if it adds an XSL
> one, with the message 'replacing existing XSL style sheet"

Something I totally agree with. I think that having PIs for both XSL and
CSS just doesn't make sense. I believe that if the XSL/XML content
cannot be handled by the agent (or the human behind it prefers another
version) then that version of the document shouldn't end up in the
user's machine in the first place. Instead, the browser should be able
to know the user's preferences (and it's own capabilities) and ask for
alternative versions of the document.

Please note that this is just my opinion and I will not get involved in
a looong debate over it (although I would like some nice conversation to
take place). I am having a hard time with all these mailing lists
already (as most people here too I suppose).

Another thing that comes in mind is the user stylesheet. I mean, ok, the
user can have a CSS in his machine and tell the browser to render
everything with that. How about our case? The XML document will probably
not make any sense without some real transformation, but the user simply
cannot have an XSL of his own as with CSS. Maybe we need yet another
language that describes the XML document data in a way that every XSL
can make a real use of it (or is this my caffeine OD talking here)? I
don't think that either the XMLSchema or Infoset can help on this. We
need something that will help an agent know about what the 'result tree'
*fragments* should be. Am I making any sense here?

Kindest regards,


Received on Friday, 29 June 2001 08:55:56 UTC