- From: William F Hammond <hammond@csc.albany.edu>
- Date: Mon, 03 Nov 2014 21:33:01 -0500
- To: David Carlisle <davidc@nag.co.uk>
- Cc: Peter Krautzberger <peter.krautzberger@mathjax.org>, "www-math@w3.org" <www-math@w3.org>
> <mi jats:foo="bar"> is valid to the MathML3 schema (as long
> as jats: namespace is declared somewhere) it's just that you
> don't want to let attributes with colons in their name
> anywhere near an HTML parser, ...
I've never thought that xml-namespaces are of much real
value for xml document instances (as opposed to xml
electronic data instances).
In particular, I've understood the banishment of namespaces
from HTML 5 to represent revulsion toward xml-namespaces of
those who produce documents.
I think it should be possible to structure the world of HTML
so that automatic generators (from various profiles of LaTeX
or DocBook or ...) to HTML5 can switch easily between the
text/html and the application/xhtml+xml serializations.
This would mean, in particular, not really using
xml-namespaces with the latter serialization, but keeping
within a reasonable framework for validation. Maybe one
could introduce for MathML a value something like html's
"style". For example, instead of <mi jats:foo="bar">
something like <mi mmdata='ns: jats; foo1: bar1; foo2:
"bar2a bar2b"; foo3: bar3;'> which would, with mmdata added
to the definition of html as an unspecified cdata attribute,
be able to pass html validation. As with css a separate,
less essential, and perhaps not universal, validator could
check mmdata values.
-- Bill
Received on Tuesday, 4 November 2014 02:33:27 UTC