W3C home > Mailing lists > Public > public-xhtml2@w3.org > December 2008

Re: XHTML 1.x mime type document

From: Tina Holmboe <tina@greytower.net>
Date: Tue, 23 Dec 2008 17:52:14 +0100 (CET)
To: Dean Edridge <dean@dean.org.nz>
cc: public-xhtml2@w3.org
Message-ID: <tkrat.78ee8e6397642478@greytower.net>

On 24 Dec, Dean Edridge wrote:

> text/xml is being deprecated isn't it? I've seen several reports of
> problems with text/xml [1] [2] [3]

  The XHTML WG cannot "deprecate" text/xml. The Note in question 
  does not encourage people to use this type.

> text/html is the mime type for HTML, if you try to use XHTML 1.x and 
> label it as text/html it becomes invalid HTML4. Trying to use XHTML but 

  No. It is quite possible to write an XHTML 1.* document, label it as
  text/html, and make it valid. 

  As long as the document in question match the syntactic rules set
  forth by the DTD specified, the document is syntactically valid. The
  content-type has no impact on this.

  If you write a correct XHTML document and validate against a HTML 4.01
  DTD, then, yes, you WILL have errors on the xml:lang and xmlns
  attributes. This, however, is not what we say.

  What we DO say is that delivering an XHTML DOCTYPE a certain way will
  work just fine in HTML user agents.

> sending it as text/html also gives people the false impression that they 
> have transitioned from HTML to XHTML when they haven't. I don't think 
> it's in the best interest of XHTML to tell people that their documents 
> are valid XHTML when using the text/html mime type.

  With the current situation regarding browser support being as it is,
  we judged this a good, if interim, solution to aid migration.

>> Also, DO include a space before the trailing / and  >
> I think you'll find that the space is not needed on modern text/html 
> parsers, anyway, the syntax is invalid in HTML which is what text/html is.

  There is really no harm done in including support for "older"
  text/html parsers; not that they are particularly concerned with
  strict syntax.

  In addition I fear you are mistaken. "/>" is syntactically valid in
  HTML, but has a different meaning. It is generally not supported by
  UAs. *

>> A.16. The Named Character Reference &apos;
>> DO use &#39; to specify an escaped apostrophe. DO NOT use &apos;.
>> Rationale: The entity &apos; is not defined in HTML 4.
> You'll need to mention that since XML parsers are not required to fetch 
> external entities; there's only five named entities that you can use in 
> XHTML (&lt;, &gt;, &amp;, &quot; and &apos;). [4]

  This is correct, although not in the context of the appendix. In this
  case the HTML UA knows about the entities in question - except for
  &apos; which is not defined for HTML 4.

  We will consider, in plenum, adding a clarification.

> There's some other things that I'm concerned about:

  For formal questions regarding group scope, please refer to the XHTML
  2 WG and HTML WG charters.

  Merry Christmas.

 * http://blog.charlvn.za.net/2007/11/w3c-validator-on-net-enabling-start-tag.html

 - Tina Holmboe       siteSifter                  Greytower Technologies
            http://www.sitesifter.co.uk          http://www.greytower.net
      Website Quality and Accessibility Testing
Received on Tuesday, 23 December 2008 16:52:53 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:30:31 UTC