RE: Math on the web without MathML

> It seems to me that David Carlisle's little XML document
> he sent in his last message makes his point very well

If the point was that matrices can not be formatted with CSS
or superscript can not be aligned properly then this is probably 
not the best example to illustrate the point or lack of such.

> David Carlisle provided a simple example that he feels can
> not be rendered well in CSS.  

Here is some of the possible ways to do this 

http://geocities.com/chavchan/temp/vectors.xml
http://geocities.com/chavchan/temp/vectors.pdf

PDF is produced by CSS formatter (Prince).

There are many other markup options for current example 
that would admit CSS formatting.

Generally speaking there are no problems with rendering 
of math layouts like fractions, matrices, over/under scripts, 
sub/superscripts and similar stuff using CSS. 

There are some layouts that require quite verbose markup,
which is not a problem in XSLT + CSS approach, but is a 
problem in pure CSS approch where we want to keep markup
as simple is possible, the problem however is likely to 
be solved when CSS support will grow slightly stronger.


> Please show that you know what you are
> taking about by showing how to do this in CSS so that we too can
> understand how it is done.  

http://geocities.com/chavchan/css/annotated.css

> he has run into the same limitations with CSS
> that David Carlisle ran into when trying to use CSS as a fallback for
> browsers that don't render MathML.

We do not have any serious limitations on CSS side so you can stop spreading myths.

> However,
> it's rendering (which only works in Opera), would not be considered
> acceptable by most people who author math.  Because it doesn't tag
> variables, numbers, or operators, it doesn't render variables in italics
> when appropriate, 

Feel free to use italic when appropriate, either using explicit markup,
or appropriate Unicode characters. No one prohibits you to do so, but it is not 
required by DTD as some people may not be partcularly interested in shaping.

> nor can it use appropriate spacing between operators.

Feel free to use appropriate spacing, there are plenty of space characters in Unicode.
You can use padding too. 

> square roots look poor, 

What's the problem, use power notations instead. But in any case there are several ways 
to address issue in CSS3 (SVG or other scalable images that can be added using generated content,
or borders with appropriate radia, or image borders), at the moment they does not work in browsers
and therefore are not included in current style sheets.

> parens are drawn as brackets,

Not necessarily, there are other options http://geocities.com/chavchan/css/annotated.css
but in some style sheets we prefer to use brackets of different style instead of parens and braces.
There is border radius property in CSS3 that can be employed.

> under/over braces are drawn like rotated brackets, 

Border radius will do the trick, if issue is so important for someone.

> integrals don't stretch

Many math editors don't stretch integrals and no one seem to complain about it,
it is a matter of style. It for someone it is important to see integrals stretched from the
left bottom to right top of browser window, SVG is one of the ways to go (SVG can be inserted from style sheet
using generated content).

> These are more than minor problems to most people.  

I assume these means stetching of some characters. The problem is not really serious, and is more or less solvable.


> he has run into the same limitations with CSS
> that David Carlisle ran into when trying to use CSS as a fallback for
> browsers that don't render MathML.

We tried to use CSS to render MathML and ISO 12083 in Opera (more precisely it is done via UserJS, but scripting is used only 
to transform MathML into something more CSS friendly, formatting is done exclusively via CSS) and did not notice any CSS related limitations (apart of the fact that MathML has to be transformed into something more CSS friendly before applying style sheets), also current UserJS does not position 
some indices properly (David's example works, but there are some test cases that fail) but it is not CSS related limitation and it is likely to 
be resolved when and if scripts that transform languages like MathML, ISO 12083 into XML + CSS will be updated.


-- 
_______________________________________________
Surf the Web in a faster, safer and easier way:
Download Opera 9 at http://www.opera.com

Powered by Outblaze

Received on Friday, 14 July 2006 12:49:28 UTC