the problems pointed in the issue, I removed the confusing Windows/Linux
distinction. It serves no real goal in the documentation, and is probably
wrong anyway since for Windows the WINFUNCTYPE constructor should be used.
In addition, the "look, this is faster on Linux" comment is misleading since
it's not explained. The outcome may just be an artifact of qsort implementation
for this particular input, and may change between C runtime version releases.

consistent to the one raised by the Python module (the 'code' attribute
was added).

In addition, the exception is now documented.

Added a test to check that ParseError has the required attributes, and
threw away the equivalent doctest which is no longer required.

  config.{guess,sub}. Don't use uname calls for cross builds.

This allows generators that are using yield from to be seen by debuggers. It
also kills the f_yieldfrom field on frame objects.

Patch mostly from Mark Shannon with a few tweaks by me.

  config.{guess,sub}.

port from 3.2 - Fix the urllib closing issue which hangs on particular ftp urls/ftp servers. closes issue11199

when configured with --(en|dis)able-ipv6 for cross builds, don't fail the configury due to the missing buggy-getaddrinfo check.

when configured --with(out)-computed-gotos for a cross, use this value instead of defaulting to no.

when using gcc, use a compilation test for the cross build check for long long format.

(sched) when run() is invoked with blocking=False return the deadline of the next scheduled call in the scheduler; this use case was suggested in http://bugs.python.org/issue1641#msg149453

Patch by Dylan Sarber.

(Not sure why hg told me it merged pyexpat.c too, it showed no changes)

Patch by Dylan Sarber.

Issue #13248:  Removed docs for two deprecated unittest features.  To be conservative, the code is left in for one more release.

using --with-system-expat working when the system expat does not have
salted hash support.

using --with-system-expat working when the system expat does not have
salted hash support.

timeout in queue and threading modules to not be affected of system time
update.

 * On Mac OS X, time.steady() now uses mach_absolute_time(), a monotonic clock
 * Optimistic change: bet that CLOCK_MONOTONIC and CLOCK_REALTIME are available
   when clock_gettime() is available
 * Rewrite time.steady() documentation