Urls simply point to resources. Just because a url is /foo what about that means you need to be on a page/script named foo? You just need to know that the url points to the resource you want.
An awesome example if this is on Github's site. Goto https://github.com/trq/proem, click the lib directory then drill down to wherever. The url changes without an entire page refresh. But guess what? You can now bookmark something like https://github.com/trq/proem/blob/develop/lib/Proem/Application.php. Even though you are really on https://github.com/trq/proem, the resource is easy to locate.