I am porting a (minimal) CBOR codec to micropython. In the process I hit on two deficiencies wrt CPython: * `struct` should support the `e` format code for half-float (16 bit) numbers. * `int.bit_count` would really help with encoding integers, esp. bigints.