utf-8 input and output support

[jquast]

Per this hack,

https://github.com/ianepperson/telnetsrvlib/blob/master/telnetsrv/telnetsrvlib.py#L785

It is actually possible to write fancy characters to most modern telnet clients, by way of encoding as utf-8. Decoding is more difficult, as bytes come in asynchronously. They can, however, still be decoded by using an "incrementaldecoder()" instance available from the codecs module. An example of this can be found here:

https://github.com/jquast/x84/blob/master/x84/blessings.py#L382

(Note this is not the final interface submitted upstream to the official 'blessings' project)

[jquast]

RFC's may require implementing negotiation for BINARY to support utf-8, due to the 127-256 and other magic bits used in utf-8 encoded bytestreams.