From Requirements for String Identity Matching and String Indexing ( 1998-07-10) Web addresses, with various components; pivot point for much of the WWW. How to encode arbitrary bytes into a restricted set of characters (using %HH escapes) is well defined, but which character encoding to use to encode arbitrary characters into bytes is not defined. In most cases, e.g. in proxies, comparisons are strictly binary. Without some specification for uniform normalization, some characters cannot reliably be used.

A space-separated list of URIs as defined above.