W3C home > Mailing lists > Public > www-math@w3.org > July 2006

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

From: <juanrgonzaleza@canonicalscience.com>
Date: Tue, 4 Jul 2006 03:17:07 -0700 (PDT)
Message-ID: <3151.217.124.88.235.1152008227.squirrel@webmail.canonicalscience.com>
To: <www-math@w3.org>

I have updated a post on Canonical Science Today with screenshoots of
mathematical formulae rendered with a Firefox 1.0 browser without the
special fonts over Windows XP.

[http://canonicalscience.blogspot.com/2006/07/rendering-mathematics-in-html-via-css.html]

I believe that the experiment proves that CSS rendering can be so good as
native presentational MathML 2.0 but without the limitations asociated to
the latter; see further comments about web, structure, and first class
rendering on the above source. The experiment works with XML, XHTML, and
HTML. It also will work in future HTML5 and XHTML 2.0.

In theory, i could copy duplicate CSS formulae in XSL-FO markup for
printing. I suspect that printing quality would be better than using
MathML in the printing engine but i have _not_ verified this point.

Of course there are some limitations for a CSS rendering (also in the
MathML side), but limitations are due to i) current CSS specification ii)
browsers support. This will change in a near future. In fact, i wait that
presentation MathML was of none utility in some few years (and completely
forgotten), since rendering thecniques based in CSS, SVG, or similar are
infinitely more powerful than a mathematical presentational markup. For
instance, via a full CSS or SVG (or even XSL-FO) approach we could draw
geometry, chemical reactions, biogeochemical cycles, Feynman diagrams,
etc. Chemical Markup language already uses SVG for rendering of chemical
objects. Could we use CSS for rendering chemistry?

All of above mathematical-scientific stuff cannot be rendered via
presentation MathML doing prefered the more general approach. Simply
compare the number of browsers supporting SVG and supporting MathML and
you can obtain an idea i am saying -MathML is younger but rejected by
browsers developers-.

There is a number of ways to improve CSS rendering, for instance using SVG
(maybe canvas?) content, GIFs (i disacourage this method but is used in
jsMath approach) and future CSS graphics improvements.

Of course, more complex formulae could be also rendered with CSS. For
example nested fractions are not limited to 2 levels. I have limited the
complexity of formulae simply because i am very busy those days.

In future postings, I will present examples of formulae are correctly
rendered with CSS thecnique but fail with MathML native browser and, also,
examples of MathML formulae (extracted from real world: blogs, academic
journals, databases) that in theory would be correctly rendered but in
practice -due to limitations on current implementations and tools- are
best rendered via CSS.

The source code (MathML and HTML+CSS) generating the screenshoots will be
posted after finished the experiment.


Juan R.

Center for CANONICAL |SCIENCE)
Received on Tuesday, 4 July 2006 10:17:17 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 20 February 2010 06:12:58 GMT