Re: Cascading and scripting (was: The concept of cascading)
Todd Fahrner wrote:
> What does it mean for this debate that the first "serious" mass-market
> browsers to implement CSS will also feature the ability to alter style
> sheets dynamically through scripting languages? I believe
> [youradhere]Script, in conjunction with CSS, can perform many of the
> selection/transformation tasks that Paul cites as unique DSSSL virtues.
Actually, I don't think I claimed any virtues to be *unique* to DSSSL.
> may even be possible to write quite complex scripts inside local (user)
> stylesheets that will effectively parameterize the system, regardless of
> whatever classes or tagsets are in use.
Yes. This is the question of standardized formats vs. downloadable
programs. Instead of standardizing math (or even HTML) we could just
require everyone to supply Java Applets that render their documents.
Anybody who wanted to do anything other than view the documents would be
screwed, but JavaSoft would be happy!
> I am making no value judgment about the elegance of one (xxML + CSS +
> scripting) relative to the other (xxML + DSSSL), but browser support for
> the former is halfway out the birth canal, and will be deployed in perhaps
> 100 million instances before it could be "superseded" - which will of
> course complicate any such supercession enormously.
That's true. I suspect that both will be implemented in IE 4.0 or 5.0
and perhaps Netscape 5.0 or 6.0 and authors will have to learn both and
choose. That's too bad. On the other hand, perhaps "Document Object
Model" scripts will become testing grounds for new ideas that can then
be incorporated in a rigorous manner into DSSSL. I would rather avoid
the bifurcation of effort and education, but I suppose that is what will
separate the professional publishers from the amateurs.