Perhaps I misunderstood the intended behaviour of the ascii() builtin in 
the future_builtins module, but the following behaviour is unexpected:

  >>> from __future__ import unicode_literals
  >>> from future_builtins import ascii

  >>> test = 'hello'

  >>> ascii(test)
  "u'hello'"

Surely the leading 'u' should not be there?

After all, this is a future builtin we are importing -- when all 
'strings' are unicode by default?

This is a side effect of ascii returning a python 2.x unicode object.
The repr still gives a leading "u". The point of ascii is that it
returns a unicode object with only ascii characters. See PEP 3138 for
more information.