skip to navigation
skip to content

Python Wiki

Python Insider Blog

Python 2 or 3?

Help Fund Python

[Python resources in languages other than English]

Non-English Resources

Add an event to this calendar.

Times are shown in UTC/GMT.

Add an event to this calendar.

PEP: 291
Title: Backward Compatibility for the Python 2 Standard Library
Version: 83f2a13bc1e3
Last-Modified:  2010-11-30 21:05:26 +0000 (Tue, 30 Nov 2010)
Author: Neal Norwitz <nnorwitz at gmail.com>
Status: Final
Type: Informational
Created: 06-Jun-2002
Python-Version: 2.3
Post-History: 

Abstract

    This PEP describes the packages and modules in the Python 2
    standard library which should remain backward compatible with
    previous versions of Python.  If a package is not listed here,
    then it need only remain compatible with the version of Python it
    is distributed with.

    This PEP has no bearing on the Python 3 standard library.


Rationale

    Authors have various reasons why packages and modules should
    continue to work with previous versions of Python.  In order to
    maintain backward compatibility for these modules while moving the
    rest of the standard library forward, it is necessary to know
    which modules can be modified and which should use old and
    possibly deprecated features.

    Generally, authors should attempt to keep changes backward
    compatible with the previous released version of Python in order
    to make bug fixes easier to backport.

    In addition to a package or module being listed in this PEP, 
    authors must add a comment at the top of each file documenting
    the compatibility requirement.

    When a major version of Python is released, a Subversion branch is
    created for continued maintenance and bug fix releases.  A package
    version on a branch may have a different compatibility requirement
    than the same package on the trunk (i.e. current bleeding-edge
    development).  Where appropriate, these branch compatibilities are
    listed below.


Features to Avoid

    The following list contains common features to avoid in order
    to maintain backward compatibility with each version of Python.
    This list is not complete!  It is only meant as a general guide.

    Note that the features below were implemented in the version
    following the one listed.  For example, features listed next to
    1.5.2 were implemented in 2.0.

        Version    Features to Avoid
        -------    -----------------
          1.5.2    string methods, Unicode, list comprehensions, 
                   augmented assignment (eg, +=), zip(), import x as y,
                   dict.setdefault(), print >> f,
                   calling f(*args, **kw), plus all features below

          2.0      nested scopes, rich comparisons,
                   function attributes, plus all features below

          2.1      use of object or new-style classes, iterators, 
                   using generators, nested scopes, or //
                   without from __future__ import ... statement,
                   isinstance(X, TYP) where TYP is a tuple of types,
                   plus all features below

          2.2      bool, True, False, basestring, enumerate(),
                   {}.pop(), PendingDeprecationWarning,
                   Universal Newlines, plus all features below
                   plus all features below

          2.3      generator expressions, multi-line imports,
                   decorators, int/long unification, set/frozenset,
                   reversed(), sorted(), "".rsplit(),
                   plus all features below

          2.4      with statement, conditional expressions,
                   combined try/except/finally, relative imports,
                   yield expressions or generator.throw/send/close(),
                   plus all features below

          2.5      with statement without from __future__ import,
                   io module, str.format(), except as,
                   bytes, b'' literals, property.setter/deleter


Backward Compatible Packages, Modules, and Tools

    Package/Module     Maintainer(s)          Python Version     Notes
    --------------     -------------          --------------     -----
    2to3               Benjamin Peterson           2.5
    bsddb              Greg Smith                  2.1
                       Barry Warsaw
    compiler           Jeremy Hylton               2.1
    ctypes             Thomas Heller               2.3
    decimal            Raymond Hettinger           2.3           [2]
    distutils          Tarek Ziade                 2.3
    email              Barry Warsaw                2.1 / 2.3     [1]
    modulefinder       Thomas Heller               2.2
                       Just van Rossum
    pkgutil            Phillip Eby                 2.3
    platform           Marc-Andre Lemburg          1.5.2
    pybench            Marc-Andre Lemburg          1.5.2         [3]
    sre                Fredrik Lundh               2.1
    subprocess         Peter Astrand               2.2
    wsgiref            Phillip J. Eby              2.1
    xml (PyXML)        Martin v. Loewis            2.0
    xmlrpclib          Fredrik Lundh               2.1

    Tool                         Maintainer(s)   Python Version
    ----                         -------------   --------------
    None


    Notes
    -----

    [1] The email package version 2 was distributed with Python up to
        Python 2.3, and this must remain Python 2.1 compatible.  email
        package version 3 will be distributed with Python 2.4 and will
        need to remain compatible only with Python 2.3.

    [2] Specification updates will be treated as bugfixes and backported.
        Python 2.3 compatibility will be kept for at least Python 2.4.
        The decision will be revisited for Python 2.5 and not changed
        unless compelling advantages arise.

    [3] pybench lives under the Tools/ directory. Compatibility with
        older Python versions is needed in order to be able to compare
        performance between Python versions. New features may still
        be used in new tests, which may then be configured to fail 
        gracefully on import by the tool in older Python versions.


Copyright

    This document has been placed in the public domain.