W3C home > Mailing lists > Public > www-math@w3.org > May 2004

Re: MathML and CSS

From: Paul Libbrecht <paul@activemath.org>
Date: Mon, 17 May 2004 21:29:11 +0200
Message-Id: <79A3474C-A838-11D8-BB4F-000A95C50B1C@activemath.org>
To: www-math@w3.org

On 16-May-04, at 17:11 Uhr, Stan Devitt wrote:

> Two points related to content MathML.
> 1.  "Doing just enough" could include using a functional notation for 
> content MathML. The place where you get into trouble in trying to do 
> content MathML right now with CSS is in the tree re-structuring and 
> this would allow you to avoid that.
> We start with  writing out content expressions in parsable formats 
> like int( bvar(x) , sin(x) ),  or   plus(a,b,c)  .  (basically writing 
> things in a functional notation) then Content MathML is not totally 
> abandoned
> Then, as you add CSS functionality that allows you to restructure the 
> expression tree, you can go for something better. The key point is 
> that the semantics would need not be lost even from the word go.

Wow... you would expect CSS-rewriting to do as much as taking a second 
argument (in the syntax you mention) and putting it in index, say... 
and does this seem in reach, css-ers ?
(of course, removing the brackets, ....).

> And of course, while we are waiting for the ability to re-structure 
> trees, we still have the choice to do that restructuring server side 
> using XSLT or the like...

Oh sure. And we're doing this in ActiveMath for example...
This was the part of my mail saying that some DHTML would sort of allow 
us to reach great heights (with some, sadly, very special browsers, I 

We're doing it in HTML 4 presentation-target for example. And it allows 
the sub-term under the mouse to be highlighted (the subterm of the 
content tree)... but that's all "added" by the XSLT and the markup 
becomes very heavy...

It works in HTML-4, and could work in XHTML+MathML on Mozilla. But for 
it to work in MathPlayer we would need the latter to run some 
javascript, and DOM, and... is it really wished ?

It looks cleaner to be able to associate the semantic of something 
under the mouse from the markup synchronized than with some 
onMouseOvers, or ?

Received on Monday, 17 May 2004 15:29:36 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:27:35 UTC