Re: more comments on HTML-Math proposal

>The current plugin interfaces seem to have so many limitations that they
>can't support decent rendering of math as part of a larger document (as
>it always is). As Dave said, some of the most basic issues (like font
>size and baseline info) may be fixed. Whether and how soon this happens
>remains to be seen. Meanwhile I think we have no choice but to design
>HTML-Math "optimistically" even though we know that some of its
>features will be handled worse or not at all by plugins as opposed to
>integrated browsers. Hopefully at least one browser will be given the
>ability to render HTML-Math decently, which ought to put some pressure
>on the other browser vendors to do so.

Many, if not all, of the pluggin problems can be solved by writing a
"content handler" for the Hot Java browser.  If I understand it
correctly, one can subclass the existing HTML content-handler and add
whatever features one needs.  I believe this approach gives the
programmer access to all the surrounding context information about
font size and baselines, etc, and should allow for things like math
within text within math anchors to be fully implemented.  Since one is
subclassing an existing content handler, one doesn't have to
completely redo the existing HTML processing within the pluggin.

Of course, I realize this is an esoteric solution with a small
audience.  On the other hand, as Burce points out, if one browser does
it right, the pressure on other browser makers to follow suit
increases.