- From: Frank Ellermann <nobody@xyzzy.claranet.de>
- Date: Wed, 25 Apr 2007 18:42:32 +0200
- To: www-validator@w3.org
Sierk Bornemann wrote: > How do you differ between a browser, which is XHTML-capable and > one which is not? A radical approach could be "don't", let the browser figure it out: http://purl.net/xyzzy/src/xhtmlxml.cmd (please ignore two typos) It works like this: a file whatever.html is sent to my browser with Content-Type: application/xhtml+xml My Wilbur browser has no clue what it might be and looks into its "MIME helper" table. It finds the script mentioned above, downloads the file, and tells my script "do!". The script renames the file to a common name, that simplifies any cleanup to one temporary file. After that it starts a helper application "netscdde.exe", which in turn tells my browser to display the temporary html file. It's a local file at this point, no http Content-Type, therefore my browser thinks that it's good old HTML 3.2 and displays it (as far as possible, if the file doesn't come with a <base> I can give up or patch it) Frank
Received on Wednesday, 25 April 2007 16:50:33 UTC