David Dorward wrote:

Jukka K. Korpela wrote:

But the W3C makes a big noise about it! See http://www.w3.org main page right now. They have created a working draft for XHTML 1.1 Second Edition.

Which says that XHTML 1.1 SHOULD be served as text/html... and references a document which says it SHOULD NOT be served as text/html to support that.

The other major change is that it adds a Schema to validate your X(HT)ML against. All XHTML 1.1 documents must continue to carry a DOCTYPE though, and conform to the DTD.

Overall, I think I am for allowing XHTML 1.1 to be served as text/html, provided it meets the compatibility guidelines in XHTML 1.0 Appendix C.

David, on your web page you write:

Additionally, as far as I know, nothing added in XHTML 1.1 (i.e. Ruby annotation) is supported by legacy user agents. So there seems little point in allowing it to be served as text/html.

As it happens, Ruby annotation is only supported by Internet Explorer 5.0+ for both Windows and Macintosh -- precisely the sort of browsers that benefit from allowing XHTML 1.1 to be served as text/html.

(With a bit of CSS trickery, you can fake Ruby support in other browsers.)

Yes, it would be nice if Internet Explorer supported XHTML properly, and sending it as application/xhtml+xml serves as a useful stick to beat Microsoft with, to encourage them to add proper XHTML support to their browser.

You could argue that sending XHTML as text/html prevents you from taking advantage of browsers' stricter parsing methods. But browser makers are free to apply these strict parsing methods to XHTML documents served as text/html -- there is nothing in any specification that even suggests that they should not.

Allowing XHTML 1.1 to be served as text/html finally gives the CJK community a standards-compliant way of serving ruby annotated text to a user agent that supports ruby annotated text.