sum()'s second argument is returned if the first argument is empty, otherwise

it is ignored.

pep-3000.txt

@@ -90,14 +90,15 @@ Built-ins
 * Introduce ``trunc()``, which would call the ``__trunc__()`` method on its
   argument; suggested use if for objects like float where calling ``__int__()``
   has data loss, but an integral representation is still desired [8]_
+* The second argument to ``sum()`` will returned if the first argument is
+  empty, otherwise it is ignored [10]_
 
 To be removed:
 
 * ``apply()``: use ``f(*args, **kw)`` instead [2]_
 * ``buffer()``: must die (use a bytes() type instead) [2]_
 * ``callable()``: just call the object and catch the exception [2]_
-* ``compile()``: put in ``sys`` (or perhaps in a module of its own)
-  [2]_
+* ``compile()``: put in ``sys`` (or perhaps in a module of its own) [2]_
 * ``coerce()``: no longer needed [2]_
 * ``execfile()``, ``reload()``: use ``exec()`` [2]_
 * ``input()``: use ``eval(sys.stdin.readline())`` [2]_
@@ -143,13 +144,16 @@ References
 .. [7] python-dev email ("Let's get rid of unbound methods")
    http://mail.python.org/pipermail/python-dev/2005-January/050625.html
 
-.. [8] python-dev email( "Fixing _PyEval_SliceIndex so that integer-like
+.. [8] python-dev email ("Fixing _PyEval_SliceIndex so that integer-like
    objects can be used")
    http://mail.python.org/pipermail/python-dev/2005-February/051674.html
 
 .. [9] Guido's blog ("The fate of reduce() in Python 3000")
    http://www.artima.com/weblogs/viewpost.jsp?thread=98196
 
+.. [10] python-dev email ("Rationale for sum()'s design?")
+   http://mail.python.org/pipermail/python-dev/2005-March/052183.html
+
 
 Copyright
 =========