Re: Math on the web without MathML (CSS 2.1 rendering for HTML and XML)

Bruce Miller said:
> juanrgonzaleza@canonicalscience.com wrote:
>> Bruce Miller wrote:
>>> While it certainly would have been a nicer world had MathML fit
>>> better into the CSS scheme, this criticism is, I think, extremely
>>> unfair.
>>>
>>> The CSS1 spec came out at the end of 1996, the MathML spec near the
>>> beginning of 1998; roughly a year apart, which means that
>>> most of the development was going on in parallel.  Decent
>>> implementation of CSS, of course, took even longer.
>>
>> Today, many browsers have decent implementation of CSS. Only Mozilla
>> (and w3c Amaya) has partial implementation of p-MathML and not of
>> content MathML.
>
> Yes, _today_ (but not then) _many_ (but not most, if you take market
> share into account) browsers have _decent_ (but neither complete,
> nor consistent) implementation of CSS.

Then _none_ browser had support for MathML. Instead promoting MathML
support, CSS would be improved and promoted _then_.

> ...
>
>> I understand that Lie said "a markup language that could be presented
>> using existing CSS properties."
>
> Exactly! whatever markup language you designed, anything
> that could have been presented using existing CSS properties (CSS1)
> wouldn't be recognizable as math.

Lie did _not_ say, "a markup language that would be restricted to using
existing CSS properties."

If then available CSS (version 1) properties were not sufficient, then the
MathML WG would present us: content MathML + CSS-Math module.

Moreover, the MathML WG also provided us redundant presentational markup
duplicating available CSS presentational features. Fortunately, this is
being deprecated in each new version of p-MathML.

What is the point to do exactly the same errors that presentational HTML
but powered to the 10 power now in presentational MathML?

Why <font> was harmfull in HTML and _eliminated_ from more rencent specs
but lots of inefficient <mstyle> are considered nice today?

I believe that the real problem then was that the MathML WG thought that
CSS was not winner, ignored lesson from the ISO 12083 + DSSSL world and
decided reinvent the wheel. Somewhat as today we see MathML folks claiming
that a CSS approach _cannot_ render matrices, when just ***can*** (it is
really amazing to see the surprise of Carlisle and others when they see
how a CSS approach can do they sure could not just some days ago!)

With a bit of work and better CSS implementations, the CSS approach can
render matrices better than p-MathML can.

> Why doesn't the same criticism apply to SVG?

Who said contrary? But this list is for Math not for SVG. At the SVG list
we could discuss about SVG strengths and weakness.


Juan R.

Center for CANONICAL |SCIENCE)

Received on Saturday, 15 July 2006 13:42:43 UTC