see a better way; the 'Binary' class is poorly specified so it's unclear
what behavior is relied upon.

(And if there are, they will now get an error when used.  So fix them!)

- Replace all asserts by ValuleErrors or TypeErrors as appropriate.
- Add _checkReadable, _checkWritable methods; these check self.closed too.
- Add a test that everything exported by io.py exists, and is either
  an exception or an IOBase instance (except for the open function).
- Default buffering to 1 if isatty() (I had to tweak this to enforce
  the *default* bit -- GvR).

PEP 3101: Removed _formatter_xxx routines from sysmodule, and made them unicode methods instead (per GvR suggestion).

Stop polluting namespace.

  "{0:.{precision}s}".format('hello world', precision=5)
I pretty sure it's because of the 'precision' keyword.
Still need to investigate further.

Issue1028

Hopefully this works with VC7.  If not, it should be pretty close.

I skipped the adding of libraries because it was harder and I'm less
confident about getting it right.  Also I'm not sure why they are required
now and weren't required before.

Applying without testing since I don't have Windows.  It
seems to make sense from a cursory review.

This makes the failing tests (test_unicodedata and, on OSX, test_hashlib) pass.

XXX However, I'm not sure that this is the right thing to do;
this behavior means that Unicode strings are automatically hashed as their
UTF-8 encoding.  Is that what we want?