discriminating ascii, unicode and raw byte strings in XML-RPC

[Skip Montanaro]

Just kind of curious, how are people distinguishing between ASCII strings,
Unicode strings, and raw byte strings (strings used to transport arbitrary
data, like database blobs or images) when transporting them via XML-RPC?
Its <string> tag is only supposed to transport ASCII strings.  Base 64
encoding Unicode and raw strings of bytes works, but doesn't distinguish
between the two at the other end.  Prepend "r" or "u" perhaps (no XML-RPC
changes required, but requires a protocol between the client and server)?
Have people established conventions or extended XML-RPC to support Unicode
directly?  I should mention that I do have to interoperate with Perl, so
Python-only hacks (like pickle.dumps) aren't ideal solutions.

Thx,

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html