Will the built-in String support Unicode?

If so, does this mean that String extensions coded in 'C' must cope with
the Unicode Strings?

My concern is that Strings in Ruby currently function efficiently as
byte arrays.  Also, there are many applications for which Unicode
support is inappropriate.  Take the example of machine<->machine
interfaces, for instance.

How will Unicode support affect these applications?

Will there by some way to instanciate a "classic" String object?
Will such objects support the current set of string methods?

- brent