DOM, markup, and "dynamic html"

Have y'all seen the dynamic html demo at:
http://www.microsoft.com/workshop/author/dhtml/dhtmlovw.htm ?

You need IE4b (Win32) to see what's going on, though it degrades gracefully
elsewhere - sorta. The document describes its own functionality. It is
relevant to www-style because CSS attributes are being manipulated in

This is neat, but the source reveals that it relies on "new" markup in the
form of <div>s and id/classes, as well as hard-coded anchors. The former
compromises the generality of the markup, while the latter produces
confusing results in downlevel browsers ("links to what?").

Can some overworked browser developer - say, from Redmond - explain whether
it will be possible to produce this functionality without extra markup,
relying on the existing parent/child and other logical relationships of
generic structural HTML? I think this is the same question: will IE4
actually parse HTML?

Can scripts live in an external file, like CSS - or within a CSS file? If
so, imagine the power of a "collapse/expand" personal script/style sheet
that would allow one to drill through valid HTML documents, regardless of
whether the author might have imagined such treatment.

Or do we need to wait for XML support and a better-developed object model
in 5.0-browsers for this sort of thing? DSSSLists want to decry the
inelegance of this all? :^)

I can't check mail 'til Monday, just in case anybody tries to follow up
with me.

Todd Fahrner

The printed page transcends space and time. The printed page, the
infinitude of books, must be transcended. THE ELECTRO-LIBRARY.

--El Lissitzky, 1923