#9 Get unicode strings, using protocol version 6

[Noam Yorav-Raphael]

CDDB.py is currently using protocol version 5, which
returns results in Latin1 encoding. The small change
that I post here makes it use protocol version 6, and
decode the result using utf8. It means that the return
values of read() and query() are now unicode strings
instead of regular strings.

For me it works perfectly, and I can now read the
titles of my Hebrew discs, instead of getting many
question marks.

[Noam Yorav-Raphael]