Would be rather interesting if you supported a named directory for local content such as "www" and if someone specified a resource without a schema (HTTP/HTTPS) you would try to load it from this www...