Re: Can cascading work?

Scott E. Preece wrote:
> While I agree with Paul that cascading on top of an arbitrary stylesheet
> is unlikely to be useful, I don't agree that that makes cascading
> useless.  On the contrary, it still leaves cascading on top of a *known*
> stylesheet as a viable and useful capability.  In the open Internet,
> that capability will be of limited utility until a few common
> stylesheets come to dominate net publishing; I suspect that will take a
> couple of years, but I'm reasonably confident it will happen.  In
> intranets, however, the ability to have a corporate stylesheet and
> cascade divisional, group, and personal stylesheets on top of it would be
> immediately useful.

When we are talking about a known stylesheet that is designed to be
externally influence, I prefer the word "parameterized." The stylesheet
has certain input parameters: "default-font-size",
"generate-table-of-contents?", "black-and-white?" etc., and customizers
set those parameters.
 
> I'd like to see someone develop good tools for automatic generation of
> overlay stylesheets - a tool that would put up, for instance, all of the
> unique fonts and colors that it would use based on an input cascade,
> which the user could then manipulate for personal preferences and save
> as the personal stylesheet to use whenever that underlying cascade is
> seen.  

Right. I think that tool would be useful too, but I would just call the
output of that process the "user's stylesheet" and the input the
"author's stylesheet." There is no sense in which they are being
combined arbitrarily. The reader's stylesheet is just a modification of
the author's stylesheet.

> I think you could develop a tool that showed all the unique cases
> in a sample (greeked) document and allowed the user to click an element
> to change it, with a feasible amount of work.

Agreed. This is possible for either CSS or DSSSL. In DSSSL you can ask:
"What construction rule made this object?" and you could delete it or
modify it to suit your needs.

 Paul Prescod

Received on Friday, 25 April 1997 11:16:56 UTC