W3C home > Mailing lists > Public > www-qa@w3.org > August 2005

Re: XHTML 1.0 Content-Negotiation

From: Henri Sivonen <hsivonen@iki.fi>
Date: Sat, 20 Aug 2005 09:37:51 +0300
Message-Id: <452c4d607b21722b474940594e9cb270@iki.fi>
Cc: www-qa@w3.org
To: Dominique HazaŽl-Massieux <dom@w3.org>


On Aug 19, 2005, at 16:07, Dominique HazaŽl-Massieux wrote:

> Le vendredi 19 aoŻt 2005 ŗ 15:58 +0300, Henri Sivonen a ťcrit :
>> On Aug 19, 2005, at 13:46, Dominique HazaŽl-Massieux wrote:
>>> Depending on the feedback I get, I'll update
>>> http://www.w3.org/2003/01/xhtml-mimetype/content-negotiation
>>> accordingly.
>>
>> I think the first and foremost problem with that document is that it
>> fails to establish a practical motivation for serving XHTML 1.0 using
>> different Content-Types to different browsers.
>
> I disagree; this document is about how, not why.

I think good how documents also summarize why--especially when the how 
provides people enough rope for shooting themselves in the foot.

> (As a matter of fact, this document is a complement to another article:
> http://webstandards.org/learn/askw3c/sep2003.html which does speak a 
> bit
> about the why)

That document asks the question "Which MIME type should XHTML be served 
with?" taking it for granted that XHTML is used without discussing 
whether it makes sense to use XHTML. Also, the document misrepresents 
browser issues with application/xhtml+xml as if there were no issues an 
author should be aware of in browsers other than IE.

The WaSP's "Ask the W3C" is more political than practical. It presumes 
that whatever the W3C has done recently needs to be advocated and is 
good for adoption right now and then tries to answer the questions in 
such a way that this political agenda is served.

Let's consider the answer to "HTML versus XHTML" question 
(http://webstandards.org/learn/askw3c/oct2003.html).
  * The document mentions several low-level syntactic details of HTML as 
issues without mentioning that the resulting parse tree does not 
suffer.
  * It conveniently omits the fact that there are some higher-level 
restrictions enforced in the HTML 4.01 DTDs that are not enforced in 
the XHTML DTDs.
  * It claims "in strict XHTML, all inline elements must be contained in 
a block element" implying this wasn't the case for HTML 4.01 Strict.
  * It talks about XSL-readiness conveniently ignoring the fact that 
tools exist for feeding a SAX pipeline from an HTML 4.01 document as if 
an XHTML document was parsed. John Cowan is distributing TSaxon--a 
version of Saxon that includes his TagSoup parser.
  * It perpetuates the future proofing myth. Switching to Appendix C now 
(what people who read the answer are likely to do if they do something) 
does not future proof anything. Swithing to real XHTML (ie. served as 
application/xhtml+xml) comes with issues 
(http://www.mozilla.org/docs/web-developer/faq.html#xhtmldiff) that are 
not mentioned in political XHTML advocacy.

-- 
Henri Sivonen
hsivonen@iki.fi
http://hsivonen.iki.fi/
Received on Saturday, 20 August 2005 06:38:16 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 6 December 2009 12:14:01 GMT