Re: [Texmacs-dev] Standart compliance in HTML export

[david]

On Mon, Jan 06, 2003 at 07:21:04PM +0100, Joris van der Hoeven wrote:
> 
> I rather propose to make it possible to customize
> the behaviour of the output filters in the user preferences.
> This gives optimal flexibility, which is also needed because
> "standards" are a very fuzzy concept anyway in the Html and *ml world.

That does not seem fuzzy to me. There are W3C recommandations (i.e.
"the standard") and a number of more or less "common" applications
(e.g. RDF) and (mis)features (e.g. "wild" HTML).

Well... that is clean until the W3C start including restricted
patented technology in its recommandations.

> I do not see very well how we can make use of CSS2 yet...

In the context of HTML-4, one could use 'style' tags to embbed style
information in the document (not very structural though). It is also
possible to use the head/style element to embed a stylesheet inside a
document, the conversion would then handle this in a similar way it
handles the LaTeX preamble.

> I propose to use Html 4 plus MathML for mathematics.
> If certain tags have natural translations in proprietary
> extensions, then this should become optional.

Excuse me if I have missed something.

AFAIK HTML4+MathML is *not* an option. If you want to use MathML, you
have to use XHTML1.1+MathML, which is significantly different from
HTML4 because there must be a support for namespaces. Actually, that
would be my preferred solution.

Then we can handle extensions properly. For example we could use a
TeXmacs namespace (e.g. 'www.texmacs.org/xhtml-ext/1.0', as 'tm'
prefix) to include tags like tm:multicol.

So I will go for XHTML1.1+MathML+extensions.

-- 
David Allouche         | GNU TeXmacs -- Writing is a pleasure
Free software engineer |    http://www.texmacs.org
   http://ddaa.net     |    http://alqua.com/tmresources
   address@hidden  |    address@hidden
TeXmacs is NOT a LaTeX front-end and is unrelated to emacs.