[thelist] doctype mess up

[Geoff Sheridan]

The xml prologue puts IE6 into quirks mode. That's the reason most 
people doing CSS and XHTML leave it out.
This means your CSS box models won't match, and I'm not sure which 
Box Model Hacks work in IE6 quirks mode.

Then Eike wrote:
>I'm doing the same thing (XHMTL and HTNL 4 doctype) for a client - 
>he needs well-formed Code for his Content Management System. The 
>Output of the CMS is tag soup (HTML 4.0 + proprietary tags), but it 
>needs well-formed code to transform it to XML and apply different 
>XSLT style sheets ( so all I have to do is plain HTML, which comes 
>easier to me than writing XML/XSLT).   Maybe Epicentric works in a 
>similar way.

Eh? I am seriously confused here. Are you saying that you write XHTML 
and it gets transformed to tag soup like so:

XHTML you wrote => CMS does its evil => HTML4 soup

Or are you saying that you give bits of both the HTML4 and XHTML 
doctypes to the SAME document, in a kind of franken-doctype, because 
the document contains both XHTML and HTML4?

XHTML you wrote + CMS output => Frankendoc

Because if it's the latter then surely it is simply invalid code. And 
if it's the former, it would at first have a valid XHTML doctype, and 
then when turned into tag soup it would have no doctype at all 
because there is little point having a doctype on tag soup.

Does your method use an XML prologue with a HTML4 doctype, Eike? If so, why?

g