Re: [Filter Effects][css3-transforms] Using MathML for formulas

I really like the way wikipedia creates formulas:
http://en.wikipedia.org/wiki/Help:Displaying_a_formula. It's simple and the
output looks very good.
Maybe we can have texcv (http://en.wikipedia.org/wiki/Wikipedia:Texvc) on
the server when we send the HTML to the w3c server for post-processing.

Rik

On Fri, Apr 20, 2012 at 11:48 PM, Dirk Schulze <dschulze@adobe.com> wrote:

> Hi FX TF,
>
> I just saw this comment on the Filter Effects 1.0 specification [1] from
> Erik: "ED: Consider making the following in mathml"
>
> I heard a lot of comments about using MathML instead of text or image
> alternatives. The Filter Effects specification e.g uses pure text:
>
> <pre>    0  20  40 235 235
>  100 120 140 235 235
>  200 220 240 235 235
>  225 225 255 255 255
>  225 225 255 255 255</pre>
>
> <pre>| R' |     | a00 a01 a02 a03 a04 |   | R |
> | G' |     | a10 a11 a12 a13 a14 |   | G |
> | B' |  =  | a20 a21 a22 a23 a24 | * | B |
> | A' |     | a30 a31 a32 a33 a34 |   | A |
> | 1  |     |  0   0   0   0   1  |   | 1 |</pre>
>
> or
>
> <p class="filterformula">Norm(A<sub>x</sub>,A<sub>y</sub>,A<sub>z</sub>) =
>   sqrt(A<sub>x</sub>^2+A<sub>y</sub>^2+A<sub>z</sub>^2)</p>
>
> The first two examples are clearly not accessible on screen readers that
> linearize the content. The second example is a already linearized formula.
>
>
> In CSS3 Transforms[2] I use a different approach. I included an image of
> the formula but used TeX formula description to provide some kind of
> accessibility:
>
> <img alt="\begin{bmatrix} 1 &amp; \tan(\alpha) &amp; 0 &amp; 0 \\ 0 &amp;
> 1 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 1 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 1
> \end{bmatrix}" height="106" src="skewX.png" width="155">
>
>
> MathML exists for 14 years. Therefore it should be the first choice on web
> specifications. On the other hand we have still browsers which lack MathML
> support. Namely MS Internet Explorer and Google Chrome. But even existing
> implementations in browsers are not consistent [3].
>
> Even if viewers are not consistent, at least specifications should be.
> Ideally we use MathML and provide some fallback. Sadly I have no idea how
> it can be done without JS (which is not necessarily enabled and shouldn't
> be required to read a specification). That is the reason why I choose the
> image version with alternate text. Any preferences?
>
> Greetings,
> Dirk
>
> [1] https://dvcs.w3.org/hg/FXTF/raw-file/tip/filters/index.html
> [2] http://dev.w3.org/csswg/css3-transforms/#mathematical-description
> [3] http://www.w3.org/Math/testsuite/build/main/overview.html
>