Because unless the HTTP response is a redirect to the www. url (or vice versa) it's technically a document available at a distinct URL. For most servers, the tw

Because unless the HTTP response is a redirect to the www. url (or vice versa) it's technically a document available at a distinct URL. For most servers, the two hostnames point to the same server (and vhost) but it's conceivable that they might not. You've got a similar problem with arbitrary appended query strings. I guess the clean way to handle this would be to detect very similar URLs and ask the user if it's the same document.