I don't get it: An URL is something that really "locates" a ressource, i.e. it holds enough information to say "man, go here, then there, and you have it!" ...while an URI can say "the file is called foobar.foo ... have fun finding it!"
Or am I wrong? Can the URL loading system handle URI's? How would that look in example? For me it doesn't make any sense that it could handle anything other than really URLs that speficy exactly where *what* is.