I'll take a look at the socketserver issue, but probably not today.

In the meantime, can you take a look at this port of the tests to py3k
when you get a chance?  The tests pass without changes to imaplib, but
the imaplib code is a bit different in py3k.  I'd like your opinion on
whether the fact that _get_line unconditionally chomps the last two
bytes is a bug waiting to happen in some other context, even though it
doesn't happen in this particular one in py3k.