"Jesper Tverskov" <jesper.tverskov@mail.tele.dk> wrote in message 
news:000001c50f67$3f5af1a0$440bc650@tversdatg7y7vv...
> 1) First I test if an http accept-header exists. If it does not, as is
> the case of Google, I use XHTML 1.1 and mime-type application/xhtml+xml
> and I include the xml declaration at the top of the document.

Google sends an accept header, so your google analysis here is flawed.

> 2) If an http accept-header exists, I test if it contains the string
> "application/xhtml+xml". If it does, as is the case for Mozilla/Firefox,
> Opera, Amaya, etc. I use XHTML 1.1 and mime-type application/xhtml+xml
> and I include the xml declaration at the top of the document (same as
> case 1).

This is incorrect, the presence of application/xhtml+xml in the accept 
header does not mean that the user can support that mime-type, please adhere 
to the relevant standards 
http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html  and 3.9

Jim.