python-peps / pep-3099.txt

PEP: 3099
Title: Things that will Not Change in Python 3000
Version: $Revision$
Last-Modified: $Date$
Author: Georg Brandl <>
Status: Final
Type: Process
Content-Type: text/x-rst
Created: 04-Apr-2006


Some ideas are just bad.  While some thoughts on Python evolution are
constructive, some go against the basic tenets of Python so
egregiously that it would be like asking someone to run in a circle:
it gets you nowhere, even for Python 3000, where extraordinary
proposals are allowed.  This PEP tries to list all BDFL pronouncements
on Python 3000 that refer to changes that will not happen and new
features that will not be introduced, sorted by topics, along with
a short explanation or a reference to the relevant thread on the
python-3000 mailing list.

If you think you should suggest any of the listed ideas it would be
better to just step away from the computer, go outside, and enjoy
yourself.  Being active outdoors by napping in a nice patch of grass
is more productive than bringing up a beating-a-dead-horse idea and
having people tell you how dead the idea is.  Consider yourself warned.

Core language

* Python 3000 will not be case-insensitive.

* Python 3000 will not be a rewrite from scratch.

   It will also not use C++ or another language different from C
   as implementation language.  Rather, there will be a gradual
   transmogrification of the codebase.  There's an excellent essay
   by Joel Spolsky explaining why:

* ``self`` will not become implicit.

   Having ``self`` be explicit is a *good thing*.  It makes the code
   clear by removing ambiguity about how a variable resolves.  It also
   makes the difference between functions and methods small.

   Thread: "Draft proposal: Implicit self in Python 3.0"

* ``lambda`` will not be renamed.

   At one point lambda was slated for removal in Python 3000.
   Unfortunately no one was able to come up with a better way of
   providing anonymous functions.  And so lambda is here to stay.

   But it is here to stay as-is.  Adding support for statements is a
   non-starter.  It would require allowing multi-line lambda
   expressions which would mean a multi-line expression could suddenly
   exist.  That would allow for multi-line arguments to function
   calls, for instance.  That is just plain ugly.

   Thread: "genexp syntax / lambda",

* Python will not have programmable syntax.

   Thread: "It's a statement! It's a function! It's BOTH!",

* There won't be a syntax for ``zip()``-style parallel iteration.

   Thread: "Parallel iteration syntax",

* Strings will stay iterable.

   Thread: "Making strings non-iterable",

* There will be no syntax to sort the result of a generator expression
  or list comprehension.  ``sorted()`` covers all use cases.

   Thread: "Adding sorting to generator comprehension",

* Slices and extended slices won't go away (even if the __getslice__
  and __setslice__ APIs may be replaced) nor will they return views
  for the standard object types.

   Thread: Future of slices

* It will not be forbidden to reuse a loop variable inside the loop's

   Thread: elimination of scope bleeding of iteration variables

* The parser won't be more complex than LL(1).

   Simple is better than complex.  This idea extends to the parser.
   Restricting Python's grammar to an LL(1) parser is a blessing,
   not a curse.  It puts us in handcuffs that prevent us from going
   overboard and ending up with funky grammar rules like some other
   dynamic languages that will go unnamed, such as Perl.

* No braces.

   This is so obvious that it doesn't need a reference to a mailing
   list. Do ``from __future__ import braces`` to get a definitive
   answer on this subject.

* No more backticks.

   Backticks (\`) will no longer be used as shorthand for ``repr`` --
   but that doesn't mean they are available for other uses.  Even
   ignoring the backwards compatibility confusion, the character
   itself causes too many problems (in some fonts, on some keyboards,
   when typesetting a book, etc). 

   Thread: "new operators via backquoting",

* Referencing the global name ``foo`` will not be spelled ````.
  The ``global`` statement will stay.

   Threads: "replace globals() and global statement with global builtin
   "Explicit Lexical Scoping (pre-PEP?)",

* There will be no alternative binding operators such as ``:=``.

   Thread: "Explicit Lexical Scoping (pre-PEP?)",

* We won't be removing container literals.
  That is, {expr: expr, ...}, [expr, ...] and (expr, ...) will stay.

   Thread: "No Container Literals",

* The ``else`` clause in ``while`` and ``for`` loops will not change
  semantics, or be removed.

  Thread: "for/except/else syntax"


* ``zip()`` won't grow keyword arguments or other mechanisms to prevent
  it from stopping at the end of the shortest sequence.

   Thread: "have zip() raise exception for sequences of different lengths",

* ``hash()`` won't become an attribute since attributes should be cheap
  to compute, which isn't necessarily the case for a hash.

   Thread: "hash as attribute/property",

Standard types

* Iterating over a dictionary will continue to yield the keys.

   Thread: "Iterating over a dict",
   Thread: have iter(mapping) generate (key, value) pairs

* There will be no ``frozenlist`` type.

   Thread: "Immutable lists",

* ``int`` will not support subscripts yielding a range.

   Thread: "xrange vs. int.__getslice__",

Coding style

* The (recommended) maximum line width will remain 80 characters,
  for both C and Python code.

   Thread: "C style guide",

Interactive Interpreter

* The interpreter prompt (``>>>``) will not change. It gives Guido warm
  fuzzy feelings.

   Thread: "Low-hanging fruit: change interpreter prompt?",


This document has been placed in the public domain.

   Local Variables:
   mode: indented-text
   indent-tabs-mode: nil
   sentence-end-double-space: t
   fill-column: 70
   coding: utf-8