> But why would you want to:
>
> XML -> XSL -> XSL:FO -> MIF -> Frame -> PDF
>
> When you could just go:
>
> XML -> MIF -> Frame -> PDF

You are misleading people with the latter. The XML -> MIF stage needs to be
done with something. In the former, it is done with XSLT + a tool for
FO2MIF. In the latter you would use a programming language like Python, Perl
or Omnimark.

Also, your example merges two orthogonal choices: XSLT versus
Python/Perl/Omnimark and direct MIF versus MIF via FO. I know where you
stand on the first issue, so factoring that out, the two choices are:

XML -> FO -> MIF

versus

XML -> MIF

The advantage of the former is that you can code a stylesheet that is
independent of the fact your are outputting MIF. You could then have a whole
range of backends: MIF, RTF, TeX, etc.

James Tauber


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