Floating point operations needed

[Textmode]

Presently the only types available in TMVM are Ints. and unsigned single-byte Ints at that. support for floating-point operations would greatly improve the usefulness of the machine.

[Textmode]

The new bitfield lib should enable this to be possible now. 8b microfloats, here we come!
(actually, in truth I'm not looking forward to this one)

[Textmode]

Made an initial effort...argh... ;_;