Walter Underwood wrote:
> 
> I agree that they mix presentation and structure, but I also
> feel that it is worthwhile to capture some common situations.
> That is, allow people to define link "roles", but start out
> with a few standard roles. This is analogous to including
> xml:lang in the XML spec.

These are two very different things: link types and anchor roles are
semantic, not behavioral. I have no problem with a few pre-defined roles
though I can't think of many common ones.

> I'm hunting down a copy of the PCTE rationale, since it has a
> nice description of the link roles in PCTE, and how they got
> to that design.

Good idea.

> Meanwhile, maybe I should write a NOTE proposing a PI analogous
> to the robots meta tag (<?robots index="yes" follow="no"?>).

Another good idea. Published layered conventions are better than
"builtins". Too many builtins turn out to be not very useful "standalone"
is a perfect example.

-- 
 Paul Prescod  - ISOGEN Consulting Engineer speaking for only himself
 http://itrc.uwaterloo.ca/~papresco

Earth will soon support only survivor species -- dandelions, roaches, 
lizards, thistles, crows, rats. Not to mention 10 billion humans.
	- Planet of the Weeds, Harper's Magazine, October 1998


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