Re: text/html for xml extensions of XHTML

Ian Hickson <ian@hixie.ch> wrote:

>> And most XHTML publishers want their documents to be visible in the
>> vast number of browsers which do not understand XML. We cannot change
>> these older browsers, but it should be relatively easy to change
>> browsers that do understand XML.
> But you can already do that. That's no problem, and works fine with
> sending all text/html through HTML parsers. What people are arguing is
> that text/html should be sent through _XML_ parsers on modern UAs, so that
> namespaces can be processed... which immediately means that the document
> would not work on older browsers, so the argument falls apart.

Sorry if my wording wasn't clear, but that's what I meant. XHTML is XML, so
it should be parsed as such by browsers that understand that. Such browsers
should complain when a document isn't well-formed, or when the
HTML-namespaced portions are invalid.

What I don't understand is why this wouldn't work on older browsers. Most of
the browsers I have seen are rather lenient and process a document
containing (sorry, don't know MathML):

<p>The n<m:sub>x</m:sub>n<m:power>2</m:power>...</p>

just fine. Sure, the math would come out wrong, but one can see a
mediator[1] parsing the MathML and replacing it with images, or someone
reading the majority of the document and ignoring the mathematical formulas.
Also, sense could be made of such a document by using XSL or CSS.

I think that as more namespaces are used in XHTML, this kind of thing should
be encouraged.

[1] http://lfw.org/

-- 
[ Aaron Swartz | me@aaronsw.com | http://www.aaronsw.com ]

Received on Wednesday, 2 May 2001 10:19:38 UTC