HTTP is really simple. The way it works is that the browser sends a request to the HTTP server for a certain page (like: GET/index.htm) along with some useful info about itself and the server replies with sending a header (like: HTTP1.0 OK\n Content-type: ... \nContent-length: ...\nSome info about server...) and the whole file as-is, and immediatley terminates the connection. When the browser recives the page it scans for anything else it might need to download (images, frames, stuff) and sends an identical request to the server again repeating the same process (like: GET /images/image01.jpg).
As for links is concered, it's the responsibility of the browser to follow all the links and not the server, so you can have a page with the images that have sources on a completely different server, but it's the browsers responsibility to connect to the right server and fetch the required data.
So it's wierd that the proxy doesn't work. Maybe a bug?
Another thought occurs: what about cookies? How are they treated?
Hope this helps even slightly...
|