I am working on an implementation of the Mentalis Proxy server. I have a question about distinguishing requests. Is there any way to tell the difference between a request that is made in order to load a resource for a page (e.g. an image) and a request that is made because the user clicked a link on a page after the page was loaded in the browser? I would like to prevent navigation to some sites while allowing resources from the blocked sites to be referenced and displayed from other sites. I can use the existence of a referer header to indicate that the user tried to navigate to a site by typing a url in the address line. But I can't tell whether or not to grant a request when the referer header is present because I can't tell whether or not the request is the result of clicking a link on a page. All suggestions will be appreciated. Thanks in advance.

Russ