]> granicus.if.org Git - python/log
python
17 years agoIssue #1996: float.as_integer_ratio() should return fraction in lowest terms.
Raymond Hettinger [Fri, 1 Feb 2008 21:30:23 +0000 (21:30 +0000)]
Issue #1996: float.as_integer_ratio() should return fraction in lowest terms.

17 years agoUpdate for latest sphinx latex writer.
Georg Brandl [Fri, 1 Feb 2008 20:45:33 +0000 (20:45 +0000)]
Update for latest sphinx latex writer.

17 years agoWording nit.
Georg Brandl [Fri, 1 Feb 2008 19:24:01 +0000 (19:24 +0000)]
Wording nit.

17 years agoMore int -> pid_t.
Christian Heimes [Fri, 1 Feb 2008 18:49:26 +0000 (18:49 +0000)]
More int -> pid_t.

17 years agoRename batch file.
Georg Brandl [Fri, 1 Feb 2008 18:08:09 +0000 (18:08 +0000)]
Rename batch file.

17 years agoAdd link checker builder, written for GHOP by Thomas Lamb.
Georg Brandl [Fri, 1 Feb 2008 15:50:15 +0000 (15:50 +0000)]
Add link checker builder, written for GHOP by Thomas Lamb.

17 years agoUpdate IPv6 RFC number.
Georg Brandl [Fri, 1 Feb 2008 11:59:08 +0000 (11:59 +0000)]
Update IPv6 RFC number.

17 years agoFix refleak
Neal Norwitz [Fri, 1 Feb 2008 07:22:59 +0000 (07:22 +0000)]
Fix refleak

17 years agoRoll back r60248. It's useful to encourage users not to change Rational
Jeffrey Yasskin [Fri, 1 Feb 2008 07:05:46 +0000 (07:05 +0000)]
Roll back r60248. It's useful to encourage users not to change Rational
instances.

17 years agoMove __builtins__.trunc() to math.trunc() per
Jeffrey Yasskin [Fri, 1 Feb 2008 06:22:46 +0000 (06:22 +0000)]
Move __builtins__.trunc() to math.trunc() per
http://mail.python.org/pipermail/python-dev/2008-January/076626.html and issue
1965.

17 years agoFixed bug #1983: Return from fork() is pid_t, not int
Christian Heimes [Thu, 31 Jan 2008 23:08:23 +0000 (23:08 +0000)]
Fixed bug #1983: Return from fork() is pid_t, not int

17 years agoIssue #1678380. Fix a bug that identifies 0j and -0j when they appear
Mark Dickinson [Thu, 31 Jan 2008 22:17:37 +0000 (22:17 +0000)]
Issue #1678380.  Fix a bug that identifies 0j and -0j when they appear
in the same code unit. The fix is essentially the same as the fix for a
previous bug identifying 0. and -0.

17 years agoMinor wordsmithing on docstring
Raymond Hettinger [Thu, 31 Jan 2008 22:07:16 +0000 (22:07 +0000)]
Minor wordsmithing on docstring

17 years agoRemove unused to-be-magic methods from Rational per issue 1968. Do not port
Jeffrey Yasskin [Thu, 31 Jan 2008 17:45:59 +0000 (17:45 +0000)]
Remove unused to-be-magic methods from Rational per issue 1968. Do not port
this patch to py3k.

17 years agoAdded more documentation on how mixed-mode arithmetic should be implemented. I
Jeffrey Yasskin [Thu, 31 Jan 2008 07:44:11 +0000 (07:44 +0000)]
Added more documentation on how mixed-mode arithmetic should be implemented. I
also noticed and fixed a bug in Rational's forward operators (they were
claiming all instances of numbers.Rational instead of just the concrete types).

17 years agoMissing return
Raymond Hettinger [Thu, 31 Jan 2008 01:42:11 +0000 (01:42 +0000)]
Missing return

17 years agoFix defect in __ixor__ which would get the wrong
Raymond Hettinger [Thu, 31 Jan 2008 01:38:15 +0000 (01:38 +0000)]
Fix defect in __ixor__ which would get the wrong
answer if the input iterable had a duplicate element
(two calls to toggle() reverse each other).  Borrow
the correct code from sets.py.

17 years agoBug #1234: Fixed semaphore errors on AIX 5.2
Christian Heimes [Wed, 30 Jan 2008 22:54:18 +0000 (22:54 +0000)]
Bug #1234: Fixed semaphore errors on AIX 5.2

17 years agoUpdate itertool recipes
Raymond Hettinger [Wed, 30 Jan 2008 22:17:31 +0000 (22:17 +0000)]
Update itertool recipes

17 years agoThe previous change was causing a segfault after multiple calls to Py_Initialize...
Christian Heimes [Wed, 30 Jan 2008 18:58:29 +0000 (18:58 +0000)]
The previous change was causing a segfault after multiple calls to Py_Initialize() and Py_Finalize().

17 years agoFixed some references leaks in sys.
Christian Heimes [Wed, 30 Jan 2008 17:21:22 +0000 (17:21 +0000)]
Fixed some references leaks in sys.

17 years agoRemoved unused var
Christian Heimes [Wed, 30 Jan 2008 11:46:00 +0000 (11:46 +0000)]
Removed unused var

17 years agoPatch #1970 by Antoine Pitrou: Speedup unicode whitespace and linebreak detection...
Christian Heimes [Wed, 30 Jan 2008 11:32:37 +0000 (11:32 +0000)]
Patch #1970 by Antoine Pitrou: Speedup unicode whitespace and linebreak detection. The speedup is about 25% for split() (571 / 457 usec) and 35% (175 / 127 usec )for splitlines()

17 years agoDemonstrate new except/as syntax.
Raymond Hettinger [Wed, 30 Jan 2008 00:51:58 +0000 (00:51 +0000)]
Demonstrate new except/as syntax.

17 years agoMutableSets support a remove() method.
Raymond Hettinger [Wed, 30 Jan 2008 00:08:31 +0000 (00:08 +0000)]
MutableSets support a remove() method.

17 years agoAdd isdisjoint() to the Set/MutableSet ABCs.
Raymond Hettinger [Wed, 30 Jan 2008 00:01:07 +0000 (00:01 +0000)]
Add isdisjoint() to the Set/MutableSet ABCs.

17 years agoCallMethod is faster with a NULL third-argument than with an empty format string.
Raymond Hettinger [Tue, 29 Jan 2008 19:52:09 +0000 (19:52 +0000)]
CallMethod is faster with a NULL third-argument than with an empty format string.

17 years agoIgnore .pyc and .pyo files.
Brett Cannon [Tue, 29 Jan 2008 04:20:56 +0000 (04:20 +0000)]
Ignore .pyc and .pyo files.

17 years agoUpdate Vim syntax highlighting to specify what revision was used to generate
Brett Cannon [Tue, 29 Jan 2008 04:18:04 +0000 (04:18 +0000)]
Update Vim syntax highlighting to specify what revision was used to generate
the file.

17 years agoFix the reindent rule to use $(BUILDPYTHON).
Brett Cannon [Tue, 29 Jan 2008 04:13:07 +0000 (04:13 +0000)]
Fix the reindent rule to use $(BUILDPYTHON).

17 years agoDisable use of BerkeleyDB 4.6.x to see what the odd platform buildbots
Gregory P. Smith [Mon, 28 Jan 2008 23:21:00 +0000 (23:21 +0000)]
Disable use of BerkeleyDB 4.6.x to see what the odd platform buildbots
think.  In particular, neal norwitz has traced an Ubuntu sparc64 crash
to the Lib/test/bsddb/test_basics.py test when opening a db with DB_THREAD.

17 years agoRemoved unnecessary conditional (spotted by Neal Norwitz).
Raymond Hettinger [Mon, 28 Jan 2008 21:51:25 +0000 (21:51 +0000)]
Removed unnecessary conditional (spotted by Neal Norwitz).

17 years agoFactor-out common code with a new macro
Raymond Hettinger [Mon, 28 Jan 2008 21:48:07 +0000 (21:48 +0000)]
Factor-out common code with a new macro

17 years agoFactor-out common code with a new macro
Raymond Hettinger [Mon, 28 Jan 2008 21:47:42 +0000 (21:47 +0000)]
Factor-out common code with a new macro

17 years agoLet marshal built-up sets and frozensets one element at a time (without creating...
Raymond Hettinger [Mon, 28 Jan 2008 21:34:30 +0000 (21:34 +0000)]
Let marshal built-up sets and frozensets one element at a time (without creating an intermediate tuple).

17 years agoMake PySet_Add() work with frozensets.
Raymond Hettinger [Mon, 28 Jan 2008 20:34:33 +0000 (20:34 +0000)]
Make PySet_Add() work with frozensets.
Works like PyTuple_SetItem() to build-up values in a brand new frozenset.
Also, PyFrozenSet_New() is now guaranteed to produce a distinct new frozenset.

17 years agoRevert rev. 59925, it breaks comtypes (I need to further examine this).
Thomas Heller [Mon, 28 Jan 2008 08:44:13 +0000 (08:44 +0000)]
Revert rev. 59925, it breaks comtypes (I need to further examine this).

17 years agostatic PyObject* variables should use PyString_InternFromString() instead of PyObject...
Christian Heimes [Mon, 28 Jan 2008 02:07:53 +0000 (02:07 +0000)]
static PyObject* variables should use PyString_InternFromString() instead of PyObject_FromString() to store a python string in a function level static var.

17 years agoAdded clear cache methods to clear the internal type lookup cache for ref leak test...
Christian Heimes [Sun, 27 Jan 2008 23:34:59 +0000 (23:34 +0000)]
Added clear cache methods to clear the internal type lookup cache for ref leak test runs.

17 years agoMoved Rational._binary_float_to_ratio() to float.as_integer_ratio() because
Jeffrey Yasskin [Sun, 27 Jan 2008 23:08:46 +0000 (23:08 +0000)]
Moved Rational._binary_float_to_ratio() to float.as_integer_ratio() because
it's useful outside of rational numbers.

This is my first C code that had to do anything significant. Please be more
careful when looking over it.

17 years agoAdd refcounting extension to build config.
Georg Brandl [Sun, 27 Jan 2008 20:25:12 +0000 (20:25 +0000)]
Add refcounting extension to build config.

17 years agosocket.error doesn't have a headers attribute like ProtocolError.
Neal Norwitz [Sun, 27 Jan 2008 20:03:13 +0000 (20:03 +0000)]
socket.error doesn't have a headers attribute like ProtocolError.
Handle that situation where we catch socket.errors.

17 years agoChange isbasestring function as discussed on the cvs list a while ago
Christian Heimes [Sun, 27 Jan 2008 19:01:45 +0000 (19:01 +0000)]
Change isbasestring function as discussed on the cvs list a while ago

17 years agoTry to handle socket.errors properly in is_unavailable
Neal Norwitz [Sun, 27 Jan 2008 18:19:04 +0000 (18:19 +0000)]
Try to handle socket.errors properly in is_unavailable

17 years agoRemove effectless expression statement.
Georg Brandl [Sun, 27 Jan 2008 18:14:43 +0000 (18:14 +0000)]
Remove effectless expression statement.

17 years agoUpdate the comment and remove the close. If we close we can't flush anymore.
Neal Norwitz [Sun, 27 Jan 2008 18:09:48 +0000 (18:09 +0000)]
Update the comment and remove the close.  If we close we can't flush anymore.
We might still need to close after the for loop if flushing 6! times still
doesn't cause the signal/exception.

17 years agoPrint periodic "still working" messages since this suite is slow.
Neal Norwitz [Sun, 27 Jan 2008 17:13:07 +0000 (17:13 +0000)]
Print periodic "still working" messages since this suite is slow.

17 years agoReduce buffer size since we do not need 1k
Neal Norwitz [Sun, 27 Jan 2008 17:12:15 +0000 (17:12 +0000)]
Reduce buffer size since we do not need 1k

17 years agoCatch socket errors that are often the cause of transient failures.
Neal Norwitz [Sun, 27 Jan 2008 17:11:11 +0000 (17:11 +0000)]
Catch socket errors that are often the cause of transient failures.
Many of these exceptions are due to resource unavailable, so the
existing code should be able to handle many more spurious errors.

17 years agoRetry connection in case it fails to reduce flakiness
Neal Norwitz [Sun, 27 Jan 2008 17:10:58 +0000 (17:10 +0000)]
Retry connection in case it fails to reduce flakiness

17 years agoAdd a tiny sleep and additional flush to force the file to really be synced.
Neal Norwitz [Sun, 27 Jan 2008 17:10:50 +0000 (17:10 +0000)]
Add a tiny sleep and additional flush to force the file to really be synced.

17 years agoUpdate to newer version of ffi. Fixes crashes and test failures of longdouble
Neal Norwitz [Sun, 27 Jan 2008 17:10:43 +0000 (17:10 +0000)]
Update to newer version of ffi.  Fixes crashes and test failures of longdouble

17 years agoInitialize variable to prevent warning on some platform/config.
Neal Norwitz [Sun, 27 Jan 2008 17:10:35 +0000 (17:10 +0000)]
Initialize variable to prevent warning on some platform/config.

17 years agoAdd assertion that we do not blow out newl
Neal Norwitz [Sun, 27 Jan 2008 17:10:29 +0000 (17:10 +0000)]
Add assertion that we do not blow out newl

17 years agoWhitespace cleanup
Neal Norwitz [Sun, 27 Jan 2008 17:10:14 +0000 (17:10 +0000)]
Whitespace cleanup

17 years agoRemoved an unnecessary and confusing paragraph from the namedtuple docs.
Raymond Hettinger [Sun, 27 Jan 2008 10:47:55 +0000 (10:47 +0000)]
Removed an unnecessary and confusing paragraph from the namedtuple docs.

17 years agoDocstring nit
Raymond Hettinger [Sun, 27 Jan 2008 10:13:57 +0000 (10:13 +0000)]
Docstring nit

17 years agoAdd stdarg include for va_list to get this to compile on cygwin
Neal Norwitz [Sun, 27 Jan 2008 07:41:33 +0000 (07:41 +0000)]
Add stdarg include for va_list to get this to compile on cygwin

17 years agoUse int for the sign rather than a char. char can be signed or unsigned.
Neal Norwitz [Sun, 27 Jan 2008 07:37:38 +0000 (07:37 +0000)]
Use int for the sign rather than a char.  char can be signed or unsigned.
It's system dependent.  This might fix the problem with test_rfind failing.

17 years agoMostly reformat. Also set an error and return NULL if neither MS_WINDOWS
Neal Norwitz [Sun, 27 Jan 2008 07:36:03 +0000 (07:36 +0000)]
Mostly reformat.  Also set an error and return NULL if neither MS_WINDOWS
nor UNIX is defined.  This may have caused problems on cygwin.

17 years agoMake rational.gcd() public and allow Rational to take decimal strings, per
Jeffrey Yasskin [Sun, 27 Jan 2008 05:40:35 +0000 (05:40 +0000)]
Make rational.gcd() public and allow Rational to take decimal strings, per
Raymond's advice.

17 years agoTry to prevent this test from being flaky. We might need a sleep in here
Neal Norwitz [Sun, 27 Jan 2008 05:02:34 +0000 (05:02 +0000)]
Try to prevent this test from being flaky.  We might need a sleep in here
which isn't as bad as it sounds.  The close() *should* raise an exception,
so if it didn't we should give more time to sync and really raise it.

Will backport.

17 years agoEliminate the sleeps that assume the server will start in .5 seconds.
Neal Norwitz [Sun, 27 Jan 2008 01:44:05 +0000 (01:44 +0000)]
Eliminate the sleeps that assume the server will start in .5 seconds.
This should make the test less flaky.  It also speeds up the test
by about 75% on my box (20+ seconds -> ~4 seconds).

17 years agoConsistently use tempfile.tempdir for the db_home directory.
Neal Norwitz [Sat, 26 Jan 2008 23:14:17 +0000 (23:14 +0000)]
Consistently use tempfile.tempdir for the db_home directory.

17 years agoOn some systems (e.g., Ubuntu on hppa) the flush()
Neal Norwitz [Sat, 26 Jan 2008 23:13:46 +0000 (23:13 +0000)]
On some systems (e.g., Ubuntu on hppa) the flush()
doesn't cause the exception, but the close() does.

Will backport.

17 years agoReduce the race condition by signalling when the server is ready
Neal Norwitz [Sat, 26 Jan 2008 21:21:59 +0000 (21:21 +0000)]
Reduce the race condition by signalling when the server is ready
and not trying to connect before.

17 years agoAlways try to remove the test file even if close raises an exception
Neal Norwitz [Sat, 26 Jan 2008 21:02:45 +0000 (21:02 +0000)]
Always try to remove the test file even if close raises an exception

17 years agoCleanup the code a bit. test_rfind is failing on PPC and PPC64 buildbots,
Neal Norwitz [Sat, 26 Jan 2008 20:24:36 +0000 (20:24 +0000)]
Cleanup the code a bit.  test_rfind is failing on PPC and PPC64 buildbots,
this might fix the problem.

17 years agoUpdate docs for new callpack params added in r60188
Gregory P. Smith [Sat, 26 Jan 2008 18:51:05 +0000 (18:51 +0000)]
Update docs for new callpack params added in r60188

17 years agoUnit test fix from Giampaolo Rodola, #1938
Christian Heimes [Sat, 26 Jan 2008 16:43:35 +0000 (16:43 +0000)]
Unit test fix from Giampaolo Rodola, #1938

17 years agoMove C API entries to the corresponding section.
Georg Brandl [Sat, 26 Jan 2008 14:19:22 +0000 (14:19 +0000)]
Move C API entries to the corresponding section.

17 years ago#1473257: add generator.gi_code attribute that refers to
Georg Brandl [Sat, 26 Jan 2008 14:14:20 +0000 (14:14 +0000)]
#1473257: add generator.gi_code attribute that refers to
the original code object backing the generator. Patch by Collin Winter.

17 years ago#1940: make it possible to use curses.filter() before curses.initscr()
Georg Brandl [Sat, 26 Jan 2008 14:03:47 +0000 (14:03 +0000)]
#1940: make it possible to use curses.filter() before curses.initscr()
as the documentation says.

17 years agoClarify "b" mode under Unix.
Georg Brandl [Sat, 26 Jan 2008 14:02:38 +0000 (14:02 +0000)]
Clarify "b" mode under Unix.

17 years agoAdd some items
Andrew M. Kuchling [Sat, 26 Jan 2008 13:50:51 +0000 (13:50 +0000)]
Add some items

17 years agoFix markup again.
Georg Brandl [Sat, 26 Jan 2008 13:41:21 +0000 (13:41 +0000)]
Fix markup again.

17 years agoSlashes allowed on Windows.
Georg Brandl [Sat, 26 Jan 2008 11:02:22 +0000 (11:02 +0000)]
Slashes allowed on Windows.

17 years agoAdd missing things in re docstring.
Georg Brandl [Sat, 26 Jan 2008 11:00:18 +0000 (11:00 +0000)]
Add missing things in re docstring.

17 years ago#1934: fix os.path.isabs docs.
Georg Brandl [Sat, 26 Jan 2008 09:43:35 +0000 (09:43 +0000)]
#1934: fix os.path.isabs docs.

17 years agoRevert PySet_Add() changes.
Raymond Hettinger [Sat, 26 Jan 2008 09:31:11 +0000 (09:31 +0000)]
Revert PySet_Add() changes.

17 years agoUpdate test code for change to PySet_Add().
Raymond Hettinger [Sat, 26 Jan 2008 08:41:13 +0000 (08:41 +0000)]
Update test code for change to PySet_Add().

17 years agoLet marshal build-up sets and frozensets one element at a time.
Raymond Hettinger [Sat, 26 Jan 2008 08:37:28 +0000 (08:37 +0000)]
Let marshal build-up sets and frozensets one element at a time.
Saves the unnecessary creation of a tuple as intermediate container.

17 years agoThe OS X buildbot had errors with the unavailable exceptions disabled. Restore it.
Neal Norwitz [Sat, 26 Jan 2008 08:26:00 +0000 (08:26 +0000)]
The OS X buildbot had errors with the unavailable exceptions disabled.  Restore it.

17 years agoMake PySet_Add() work with frozensets. Works like PyTuple_SetItem() to build-up...
Raymond Hettinger [Sat, 26 Jan 2008 08:19:06 +0000 (08:19 +0000)]
Make PySet_Add() work with frozensets.  Works like PyTuple_SetItem() to build-up values in a brand new frozenset.

17 years agoFix exception in tearDown on ppc buildbot. If there's no directory,
Neal Norwitz [Sat, 26 Jan 2008 07:38:03 +0000 (07:38 +0000)]
Fix exception in tearDown on ppc buildbot.  If there's no directory,
that shouldn't cause the test to fail.  Just like it setUp.

17 years agoUse a condition variable (threading.Event) rather than sleeps and checking a
Neal Norwitz [Sat, 26 Jan 2008 07:26:12 +0000 (07:26 +0000)]
Use a condition variable (threading.Event) rather than sleeps and checking a
global to determine when the server is ready to be used.  This slows the test
down, but should make it correct.  There was a race condition before where the
server could have assigned a port, yet it wasn't ready to serve requests.  If
the client sent a request before the server was completely ready, it would get
an exception.  There was machinery to try to handle this condition.  All of
that should be unnecessary and removed if this change works.  A NOTE was
added as a comment about what needs to be fixed.

The buildbots will tell us if there are more errors or
if this test is now stable.

17 years agoPrevent this test from failing if there are transient network problems
Neal Norwitz [Sat, 26 Jan 2008 05:54:48 +0000 (05:54 +0000)]
Prevent this test from failing if there are transient network problems
by retrying the host for up to 3 times.

17 years agoMake 'testall' work again when building in a separate directory.
Thomas Wouters [Sat, 26 Jan 2008 01:47:05 +0000 (01:47 +0000)]
Make 'testall' work again when building in a separate directory.
test_distutils still fails when doing that.

17 years agoUse the right (portable) definition of the max of a Py_ssize_t.
Thomas Wouters [Fri, 25 Jan 2008 21:09:34 +0000 (21:09 +0000)]
Use the right (portable) definition of the max of a Py_ssize_t.

17 years agoRewrite the list_inline_repeat overflow check slightly differently.
Guido van Rossum [Fri, 25 Jan 2008 19:50:26 +0000 (19:50 +0000)]
Rewrite the list_inline_repeat overflow check slightly differently.

17 years agoChanges 54857 and 54840 broke code and were reverted in Py2.5 just before
Raymond Hettinger [Fri, 25 Jan 2008 19:24:46 +0000 (19:24 +0000)]
Changes 54857 and 54840 broke code and were reverted in Py2.5 just before
it was released, but that reversion never made it to the Py2.6 head.

17 years agoAdded the Python core headers Include/*.h and pyconfig.h as dependencies for the...
Christian Heimes [Fri, 25 Jan 2008 15:52:11 +0000 (15:52 +0000)]
Added the Python core headers Include/*.h and pyconfig.h as dependencies for the extensions in Modules/
It forces a rebuild of all extensions when a header files has been modified

17 years agosetup.py doesn't pick up changes to a header file
Christian Heimes [Fri, 25 Jan 2008 14:54:23 +0000 (14:54 +0000)]
setup.py doesn't pick up changes to a header file

17 years agoBackport of several functions from Python 3.0 to 2.6 including PyUnicode_FromString...
Christian Heimes [Fri, 25 Jan 2008 12:18:43 +0000 (12:18 +0000)]
Backport of several functions from Python 3.0 to 2.6 including PyUnicode_FromString, PyUnicode_Format and PyLong_From/AsSsize_t. The functions are partly required for the backport of the bytearray type and _fileio module. They should also make it easier to port C to 3.0.
First chapter of the Python 3.0 io framework back port: _fileio
The next step depends on a working bytearray type which itself depends on a backport of the nwe buffer API.

17 years agoAdd prototypes to get the mathmodule.c to compile on OSF1 5.1 (Tru64)
Neal Norwitz [Fri, 25 Jan 2008 08:04:16 +0000 (08:04 +0000)]
Add prototypes to get the mathmodule.c to compile on OSF1 5.1 (Tru64)
and eliminate a compiler warning in floatobject.c.  There might be
a better way to go about this, but it should be good enough for now.

17 years agoMake the test more robust by trying to reconnect up to 3 times
Neal Norwitz [Fri, 25 Jan 2008 06:37:23 +0000 (06:37 +0000)]
Make the test more robust by trying to reconnect up to 3 times
in case there were transient failures.  This will hopefully silence
the buildbots for this test.  As we find other tests that have a problem,
we can fix with a similar strategy assuming it is successful.  It worked
on my box in a loop for 10+ runs where it would have an exception otherwise.

17 years agoMore design notes
Raymond Hettinger [Fri, 25 Jan 2008 01:46:33 +0000 (01:46 +0000)]
More design notes

17 years agoFix-up signature for approximation.
Raymond Hettinger [Fri, 25 Jan 2008 01:23:38 +0000 (01:23 +0000)]
Fix-up signature for approximation.

17 years agoAdd one other review comment.
Raymond Hettinger [Fri, 25 Jan 2008 01:13:12 +0000 (01:13 +0000)]
Add one other review comment.

17 years agoMark todos and review comments.
Raymond Hettinger [Fri, 25 Jan 2008 00:33:45 +0000 (00:33 +0000)]
Mark todos and review comments.