Re: Supporting MathML and SVG in text/html, and related topics

Boris Zbarsky said:
> juan@canonicalscience.com wrote:
>> I have seen W3C official stress and torture MathML pages being fully
>> rendered in some CSS 2.1 engines.
>
> Rendered, or rendered well?

That depend on user expectations, the CSS engine of her|his browser, and
what one understands by the relative concept of 'well'.

>> Are all renderings of roots and
>> stretchy parentheses from W3C list of MathML [1] software reasonable?
>
> I looked at the first 10 or so images that you linked to.  They don't
> exercise  this case, really.

I simply linked most of images accessible from official MathML software
list at W3C. Many render roots and some stretchy parentheses. Some just
render roots enclosed in stretchy parentheses.

In any case, you can play with the software for the rendering of your
particular case. Several tools are non-commercial.

> But I'm not sure why it matters whether _all_ existing MathML packages
> do a good  job of rendering.  The point is that they have the option of
> doing a good job.

My point when introduced CSS and MathML links was to *visually* show that
rendering quality ranking may be something like:

{best MathML engine}

...

{best CSS 2.1 engine}

...

{poor MathML engine}

...

{poor CSS 2.1 engine}

If someone was to say that CSS 2.1 rendering quality is unacceptable, then
all MathML tools listed in the W3C giving a rendering quality below the
best CSS 2.1 engine would be automatically eliminated. I ask for this by a
simple question of  consistency.

In the contrary case, we would find paradoxical cases as that of certain
MathML author who criticized a CSS rendering of a square root because the
radical sign was too bold, but he never criticized MathML tools are
rendering the radical sign more bold still!

Also rendering quality limits would not be in the eye of a few persons. I
know 'a hundred' of normal peoples for which the rendering quality given
by Word 97 was fine, still many TeX gurus were saying "Ouch!"

>> All MathML would be rendered with the future CSS math module.
>
> Do you have a reference on this?  Because I don't recall this being
> mentioned  anywhere.

Why do not ask CSS-MathML profile people directly? Sure they have more
recent info.


Juan R. González-Álvarez

Center for CANONICAL |SCIENCE)

Received on Friday, 18 April 2008 10:06:00 UTC