-----------------
2014-04-23 3.2.6
-----------------

* Fixes issue #21291: Popen.wait() is now thread safe so that multiple
  threads may be calling wait() or poll() on a Popen instance at the same time
  without losing the Popen.returncode value.
* Fixes issue #14396: Handle the odd rare case of waitpid returning 0 when not
  expected in Popen.wait().
* Fixes issue #16962: Use getdents64 instead of the obsolete getdents syscall
  on Linux.  Some architectures do not implement the latter.

-----------------
2013-12-10 3.2.5
-----------------

* Fixes issue #15798: subprocess.Popen() no longer fails if file
  descriptor 0, 1 or 2 is closed.
* Fixes issue #18763: close_fd file descriptors are now closed after
  any preexec_fn call.

-----------------
2013-06-15 3.2.5rc1
-----------------

* Fixes issue #16650 - Don't reference ECHILD from outside the local scope.
* Unittests no longer spew any test data for human verification to stdout.
* Remove a bare print to stdout that could have happened if the child process
  wrote garbage to its pre-exec error pipe.
* Fixes issue #16327 - the subprocess module no longer leaks file descriptors
  used for stdin/stdout/stderr pipes to the child when the fork() fails.  It
  also no longer potentially double closes these pipe fds.
* Correct the Python version check around use of imp_module to specify 2.6.3
  as the minimum version that exists in. Why is anyone using such an old 2.6?
* Fixes Issue #16114: The subprocess module no longer provides a misleading
  error message stating that args[0] did not exist when either the cwd or
  executable keyword arguments specified a path that did not exist.
* Add more Popen cwd tests.
* Handle errno.ECHILD in poll.
* Don't leak a reference to the gc module on capi use error.
* Check return value to avoid a crash if the capi were misused.
* Check result of PyObject_IsTrue().
* Adds test_universal_newlines_communicate_input_none.
* Most everything above consists of backports.  See the hg logs for their
  upstream hg.python.org cpython revision numbers.

----------------
2012-06-10 3.2.3
----------------

* Fixes the references to the 'surrogateescape' unicode encoding error
  handler that does not exist in Python 2.x.  'strict' is used so that
  a UnicodeEncodeError exception is raised in these situations.  These
  MAY occur if your sys.getfilesystemencoding() is not UTF-8 and
  attempt to use a non-ascii executable, args or env values.  Prior to
  this change, those would result in a hard to debug LookupError for
  surrogateescape.
* Issue #15000: Support the "unique" x32 architecture in _posixsubprocess.c.
* Fix a compilation problem when O_CLOEXEC is not defined.

------------------
2012-02-18 3.2.3b1
------------------

This release brings in the last year and a half's worth of bugfixes and
improvements to Python 3.2's subprocess module:

Off the top of my head, some major bugfix highlights include:
 * Timeout support on the APIs.
 * close_fds=True is now the default (as it is in 3.2) and performs much faster.
 * Fixed EINTR handling.
 * Fixed SIGCHLD handling.
 * Fixed several race conditions.
 * Many more bug fixes too numerous to list.

You can grep out the full list of improvements related to subprocess in:
 http://hg.python.org/cpython/file/9ce5d456138b/Misc/NEWS

-------------
2010-06 3.2.0
-------------

This was the first release. Roughly equivalent to Python 3.2.0a1.
