current versions of OpenBSD and NetBSD.  Patch by A. Jesse Jiryu Davis.

installing IDLE 2.7 on OS X: default configuration settings are
no longer installed from OS X specific copies.

* httplib.HTTP (deprecated and does not work) → HTTPConnection
* Server (deprecated) → ServerProxy

* Make it more obvious gettarinfo() is based on stat(), and that non-ordinary
  files may need special care
* Filename taken from fileobj.name; suggest dummy arcname as a workaround
* Indicate TarInfo may be used directly, not just via gettarinfo()

* Add headings for each concrete and mix-in class and list methods and
  attributes under them
* Fix class and method cross references
* Changed RequestHandler to BaseRequestHandler and added class heading
* Pull out Stream/DatagramRequestHandler definitions
* Reordered the request handler setup(), handle(), finish() methods
* Document constructor parameters for the server classes

Based on patch by Aviv Palivoda.

Patch from Chi Hsuan Yen.

versions 10.5 or higher.  Original patch by A. Jesse Jiryu Davis.

Linux PID namespaces enabled.

imported module with the empty one.

The original form is incorrect grammar and feels awkward, even though the
meaning is clear.

not have a math vocabulary.

Fixed the documented about buffer overflow error for "es#" and "et#" format
units.

* There was a link pointing to the section on the string.Formatter class (and
  multiple links in Python 3), when the section on the common format string
  syntax is probably more appropriate
* Fix references to various format() functions and methods
* Nested replacement fields may contain conversions and format specifiers,
  and this is tested in Python 3; see Issue #19729 for instance

Based on patches by Brian Brazil and Daniel Rocco.

Reported by Denis Akhiyarov on docs@

Fixed a crash when unpickle the functools.partial object with wrong state.
Fixed a leak in failed functools.partial constructor.
"args" and "keywords" attributes of functools.partial have now always types
tuple and dict correspondingly.

Based on patch by Aviv Palivoda.

Test test_wrong_cert() runs a server that rejects the client's certificate,
so ECONNRESET is reasonable in addition to SSLError. On the other hand, the
other three tests don't even need to run a server because they are just
testing the parsing of invalid certificate files.

This should fix intermittent failures on Windows where ECONNRESET was not
being caught.

Testing for a non-existing certificate file is already done in test_errors().
The wrongcert.pem test was originally testing behaviour with a mismatched
certificate.

Also add some more tests. Based on patch by Sye van der Veen.