- From: Steven Pemberton <steven.pemberton@cwi.nl>
- Date: Mon, 29 Sep 2008 16:49:55 +0200
- To: "tina@greytower.net" <tina@greytower.net>
- Cc: "XHTML WG" <public-xhtml2@w3.org>
Hi Tina, I had an action item to get you examples of delivering XHTML to IE while using application/xhtml+xml. Sorry it took so long, I had to do some digging, and use a different machine, since I have a filter on my main machine that accepts application/xhtml+xml anyway. So, I first discovered that it was possible by mistake. I couldn't work out why I could see http://www.w3.org/International/tests/sec-ruby-markup-1.html even though it claimed it was being sent as application/xhtml+xml (and that checked out in Opera, where you can request that sort of information). Other examples of application/html+xml documents working in IE are at http://www.w3.org/MarkUp/Test/xhtml-print/current/ All the tests are served as application/xhtml+xml. But try http://www.w3.org/International/tests/sec-ruby-markup-1 (which just serves the same file) and IE will give you the download dialogue, and modern browsers will display it. So the .html (and .htm in the xhtml-print tests) do make a difference to IE. So my conclusion: it is done with URL sniffing and content sniffing (which is why you can serve html as text/plain as well if the urge takes you). I think if the URL matches then IE will sniff the content, and for IE that means (I seem to remember) looking for <title> in the first 256 (512?) characters of the file. For the URL sniffing, it used to be possible to just append "?.html" to a URL, but that doesn't seem to work anymore. See also: http://blogs.msdn.com/ie/archive/2005/02/01/364581.aspx http://blogs.msdn.com/ie/archive/2008/07/02/ie8-security-part-v-comprehensive-protection.aspx Hope this helps, Steven
Received on Monday, 29 September 2008 14:50:47 UTC