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

David Carlisle said:
>
>
>> Well I never could imagine that when you write "reasonable xml syntax"
>> you can mean XML, SGML, or even HTML. Sorry, I have not developed
>> telepathic capabilities yet.
>
> Clearly "any xml" includes xhtml and it's highly likely that any css
> stylesheet that you have that works for xhtml markup would work (or
> could be made to work) for the equivalent html markup, so in the context
> of css rendering, "any xml" is not a restriction, it was simply a
> suggestion (repeated below) that you show, given any markup, how to
> style some simple mathematical layouts using CSS alone.

Again you fail to understand i am saying. Unlike restricted MathML, the
CSS techniques to render mathematics can be used for HTML, XML, XHTML...
You can use CSS techniques using HTML 4, or using XHTML 1. You could use
CSS for old HTML markup is not available on next XHTML 2. You could use
CSS techniques also for SGML-like markup. For instance, the recently
proposed at the HTML5 mailing list

<frac>a<den>2</frac>

also works with CSS, even if is *not* valid xml. Ok?

>   Examples: - multiple-character identifiers are rendered roman
>
> mathml allows you to use any of the mathvariants here. Roman is the
> default,it's not forced. Defaulting to Roman seems perfectly
> reasonable and accords with common practice (which is another way of
> saying it's what TeX does)

However, prof. Ian Hutchinson (the father of one of the MathML tools)
cited as one of anyonances of MathML rendering at the beginning of this
year in a IAP talk about mathematics on the web

[http://web.mit.edu/violeta/www/IAP2006/]

>   - it is impossible to label aligned equations at the page edge
>
> MathML has facilities to do this (mlabeledtr) (which may not be fully
> implemented in all rendering engines) so this is a comment on the
> implementation, not the spec.

But wait, is posible or imposible as Hutchinson reported this year?

>   - often large braces or integrals are too big
>
> Subjective comment, and as Neil has said, MathML has facilities for fine
> tuning operator size if necessary.

And how many tools support that? But wait again, Subjective comment? were
not your criticism to CSS rendering based in subjective comments: "the +
is too high", "the integral sign is too bold".

>   - fonts are not as well chosen as TeX, e.g. italic v looks like
>
> Choice of fonts is again largely a feature of the renderer, not of the
> markup and would obviously apply equally to any markup system.

And are annoyances or not as said in the talk?

>> > either it relies on some new font technology or is producing math
>> display whose quality has been deemed unacceptable in the past.
>>
>> Or simply once again MathML folks ignore CSS...
>
> You've consistently failed to demonstrate how CSS can be used to produce
> large operators. many people on this list would be interested to see if
> you have a usable technique here (since we've tried and failed in the
> past) even CSS 3 mechanisms for inserting content don't appear to me to
> be enough, but if you have something that works, please show it.

Sorry but again I would say you that one can render matrices in CSS 2
without joining fragments not rely on special fonts. Please visit some of
links I cited.

Moreover, the research for a CSS rendering of math is not closed.

> To be explicit here's a document that displays in IE and
> Firefox/Netscape/Mozilla with large brackets and a superscripted matrix,
> with the baseline of the equation correctly aligned with the surrounding
> text. The markup seems perfectly natural to me and not a lot more than
> one would see for marking up html tables (the main thing being the extra
> mn level)

You are carefully choosed one of few MathML examples where its verbosity
over HTML is acceptable.

> Can you suggest an alternative markup scheme together with
> some CSS to style that markup that produces an equivalent rendered
> result?

This was replied, visit link I cited here for further information.
Moreover, I remark again that CSS techniques are rather independent of the
markup scheme choosed.

> David
>
>
>
>
>
> <?xml version="1.0"?>
> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1 plus MathML 2.0 plus SVG
> 1.1//EN" "http://www.w3.org/Math/DTD/mathml2/xhtml-math11-f.dtd">
> <html xmlns="http://www.w3.org/1999/xhtml">
>   <head>
>   </head>
>   <body>
>
>     <p>xxx
>     <math xmlns="http://www.w3.org/1998/Math/MathML">
>       <msup>
> 	<mfenced>
> 	  <mtable>
> 	    <mtr><mtd><mn>1</mn></mtd></mtr>
> 	    <mtr><mtd><mn>2</mn></mtd></mtr>
> 	    <mtr><mtd><mn>3</mn></mtd></mtr>
> 	  </mtable>
> 	</mfenced>
> 	<mi>T</mi>
>       </msup>
>       <mo>=</mo>
>       <mfenced>
> 	<mtable>
> 	  <mtr>
> 	    <mtd><mn>1</mn></mtd>
> 	    <mtd><mn>2</mn></mtd>
> 	    <mtd><mn>3</mn></mtd>
> 	  </mtr>
> 	</mtable>
>       </mfenced>
>     </math>
>     yyy</p>
>   </body>
> </html>
>
>

Of course, you choosed a simple prototipical example where aligment is
good. Somewhat as fraction 3/2 renders nice in your CSS approach but
{x^2}/5 would render a bit distorted, true?

I will use your example as basis for further comparisons of CSS and p-MathML.


Juan R.

Center for CANONICAL SCIENCE)

Received on Thursday, 13 July 2006 16:23:54 UTC