[whatwg] Mathematics in HTML5

Ian Hickson wrote:
> So, while I applaud the re-use of ISO12083 here, it seems like an odd 
> choice. 
The choise to resuse ISO 12083 notations is not that important for us.
What is important is that ISO 12083 markup is good example of simple,
easy to learn and easy to use, human processable markup that at least theoretically
does things in right way (markup captures basic structure of document and mathematical formulae,
style languages like DSSSL take care of all presentation related tasks).
So what we want is not to follow ISO 12083 element naming notations, or
keep markup backwards compatible with ISO 12083, but to preserve basic
principles of human processable SGML based markup that admits separation 
of style and content. If the same principles would be shared by MathML we would use MathML.

> The resulting language seems to be just as verbose as MathML, so 
> why not just reuse MathML, which already has a Web presence? 
What is valuable in MathML that we could reuse? It is not suitable for
authoring so we need separate markup (parsing rules) for authoring and
it is not suitable for rendering in CSS rendering engine unless transformed 
in something different. So if not authoring and rendering stages 
then on what stage we would use it? Are you sure that MathML has any real web presence?

> Also, your 
> version of the vocabulary doesn't seem quite the same as ISO12083, 
The reason for this is that XML is not quite the same as SGML and CSS is not quite the same
as DSSSL so one had to redesign some parts of markup to ensure that it can be used in XML + CSS
framework. 
> which 
> means we would be introducing yet another mathematical markup language, in 
> a space which already has many such languages.
Yes.
> Why not just re-use MathML? 
Verbosity + incompatibility with CSS.

> MathML already has defined semantics, a 
If we are talking about presentational MathML then similar sematics was defined 10-15
years ago in ISO 12083 and AAP Math DTDs.
If we are talking about content MathML then current proposal is orthogonal to content MathML,
that currently is transformed to presentational MathML in order to be viewed in browsers.

> defined interpretation based on DOMs, defined DOM interfaces, and an 
> implementation in at least one browser.
DOM interface does not add any actual functionality to MathML, the same can be done using
DOM Core (and should be done as language specific interface is unlikely to work).



-- 
_______________________________________________
Surf the Web in a faster, safer and easier way:
Download Opera 8 at http://www.opera.com

Powered by Outblaze

Received on Thursday, 8 June 2006 00:54:02 UTC