Re: Guidelines on marking up data using XML/XSL

["Kent Fitch" <kent.fitch@xxxxxxxxxxxx>]

"Frank Boumphrey" <bckman@xxxxxxxxxxxxx> wrote:

>You are really just repeating the tag v. attribute argument that has
been
>recently aired on the XML list.
>
>After numerous postings, and a review of the SGML postings going back
to
>1980, the consensus was, there is no correct or best way, so just do
what
>suits your fancy!!


Sorry, I didnt intend to rehash this old ground, and I've
updated my HTML page to point to a very good summary of that
debate by Robin Cover:
http://www.sil.org/sgml/elementsAndAttrs.html

What I was really wondering about but failed to make explicit
was how elements should be related/nested, and *in particular*
whether the XSL processing model favours some designs over
others.

Robin Cover's references contain good advice to people like
me with database and o-o background (rather than an SGML
background), as does his inheritance discussion summary at
http://www.sil.org/sgml/arch-inherit.html
but I still worry that I am missing something fundamental
about designing XML for XSL.

>I suspect though that if you benchmarked your code you would find that
the
>tag method is faster!!

Yes, but that is probably implementation dependent - functions
to access attributes could be very efficient depending on the
data structures used by the XSL processor.

Kent Fitch                           Ph: +61 2 6276 6711
ITS  CSIRO  Canberra  Australia      kent.fitch@xxxxxxxxxxxx



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