Re: Structure of a transform

petermarshall   wrote:
> I have separated the information into two chunks, display xml
> and business xml.  The display xml describes the display (in
> html at the moment)

XHTML I hope.

> and contains some markup that indicates what business information
> needs to go where.  The business xml is pure business information.
> 
> The job of the transform is to insert the correct business info 
> (from the business xml) into the indicated spots in the display
> xml.  This means that the dhtml programmers can get to work
> without worrying about xslt (or xsp, the business xml source).
> The basic transform is now working to prove the technologies,
> Im just not convinced that the approach is correct.

This approach is a cornerstone of a very successful product I helped design
for my company. The static portions of sites are XHTML, and places where
something else (either more XHTML or XSLT-generated XHTML) will go are
denoted by elements or attributes in a unique namespace. I have several more
layers of information in other XML files to help fill in the info.

As for its correctness, I don't see anything incorrect about it. The only
thing is, ultimately it may be offering too much flexibility; it may be too
complex for the less tech-savvy HTML designers to deal with. Our customers
are pushing the deisngers to produce relatively simple site designs that do
not exploit the layers of XML I've developed. Most of the customers don't
care about reusing the XML or making their own XHTML, so I'm left feeling
like less formal approaches to intermingling data with HTML, like ASP, PHP,
CFML, etc., may have been less hassle.

-Mike


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