[whatwg] Mathematics in HTML5

[juanrgonzaleza at canonicalscience.com]

Anne van Kesteren wrote:
>
> Quoting juanrgonzaleza at canonicalscience.com:
>>>> Since MathML does not fit into the WHATWG philosophy, I would
>>>> aknowledge information about your own solution to the problem of
>>>> mathematical markup on the web.
>>>
>>> Oh please, cut the crap. Did you miss the message from Ian saying how
>>> it could integrate?
>>
>> If by "integrate" you mean integrate, then yes I did miss it. Please
>> could you cite it for I can learn from.
>
> http://listserver.dreamhost.com/pipermail/whatwg-whatwg.org/2006-June/006518.html
>
> has most of the details. It's dicussed in related messages from him as
> well. Not too hard to find if you look in the   archives...
>
> (I was referring to your usage of "fit" when saying "integrate"...)
>

On that case you would be able to find why I cannot agree with Ian proposal.

--- Resume ---

1) One recovers all difficulties and errors of MathML desing.

2) One recovers a CSS, DOM, and even XML unfriendly markup language.

3) The main official guidelines of the WHATWG are violated.

4) Ian specific proposal is MathML incompatible and probably would add a
exponential difficulty to realistic implementation in browsers than
original MathML has been. Just for illustration let me do some
annotations:

What when 1 is not <mn>1</mn>?

What about 0xFFEF, MCMLXIX, and twenty one? are parsed to numbers or not?

What when + or = are not operators?

How would <mrow>d x</mrow> be parsed? And <mrow>dx</mrow> and <mrow>D
x</mrow>?

How is the problem of entities solved?

How is 3,14 parsed? And 3 , 14? And 3, 14?

How is supposed that we encode \dot{q} in Ian proposal?

How is parsed the string "maps to"?

<msqrt>- 1</msqrt>

and

<msqrt><mrow>- 1</mrow></msqrt>

are to be treated as equivalent or no?

"- 1" and "-1" are equivalent or no?

What about <mrow>a b</mrow> and <mrow>a <mi>b</mi></mrow>? And
<mrow>a<mi>b</mi></mrow>?

How would one deal with " T" vs. " T"?

Ian said "Maybe we could imply one of [&#x2062;] between pairs of terms in
<mrow>s that don't have any <mo>s." Then would <mrow>L ρ</mrow> work
as <mrow>L &#x2062; ρ</mrow>?

and <mrow> 5 </mrow>? Equivalent to <mn>5</mn> or to <mrow><mn>5</mn></mrow>?


Anne van Kesteren wrote:
>
> Quoting White Lynx <whitelynx at operamail.com>:
>> "Web application technologies SHOULD BE BASED ON technologies
>> authors are familiar with,
>> including HTML, CSS, DOM, AND JAVASCRIPT"
>
> As it would work with that, it's not really a problem.
>
>
>> "Basic Web application features SHOULD BE IMPLEMENTABLE using
>> behaviors, scripting, and style sheets IN IE6 TODAY"
>
> I guess that's possible.

No if choosing MathML or HTML-MathML (special parsing mode) for the spec.
Microsoft rejected native support for MathML even if initially they
claimed that were interested in native support of, at least,
presentational markup. I find logical that they will reject HTML-MathML
also.

>> "The core features of an XML vocabulary should require the use of
>> elements from ONLY ONE NAMESPACE."
>
> Is math really a core feature?

That could be debated. My own reply is affirmative. Some countries
consider that language (aka HTML text) and math (aka HTML math) are two
basic stuff may be learned in the classroom even until 16 old.

You can also count number of times mathematical equations, generic
graphics, sound, and text are used in Wikipedia enciclopedia. You can
notice that mathematics is a very popular stuff in the educative articles.

I think that core order would be text, graphics, math. Text is well-served
by HTML text, graphics are already covered by canvas, math would be next
step.

>> "IT IS VERY IMPORTANT that authors BE ABLE TO MOVE FROM AN HTML
>> ENVIRONMENT TO A CLEAN COMPOUND DOCUMENT ENVIRONMENT (typically first
>> simply by    moving to XHTML) IN A GRADUAL FASHION."
>
> That seems to be what Ian is proposing (for math), more or less.

I disagree!

c<sup>2</sup> ---> <msup>c 2</msup> ---> <msup><mi>c</mi><mn> 2</mn></msup>.

is not a gradual migration path, is it?

George and others proposal (and variations discussed)

c<sup>2</sup> ---> c<sup>2</sup>

is cheap, WHATWG fully compliant, and introduces a gradual migration path
for _both_ developers and authors.


Juan R.

Center for CANONICAL |SCIENCE)