On 1/28/2014 8:09 AM, Terry J. Reedy wrote:
>
> Terry J. Reedy added the comment:
>
> On 1/28/2014 4:48 AM, Stefan Krah wrote:
>>
>> Stefan Krah added the comment:
>>
>> I think test_idle is failing on many build slaves following this
>> commit.
>
> The two failures I saw on the first 4 3.x bots with Idle failures are
> the result of changes in the list docstrings. I presume these are
> related to use of Argument Clinic. I knew and documented that such
> failure was a possibility.
>
> On my Win 7 machine
>
> Jan 20, 32-bit repository build just before I added the tests that now fail.
>   >>> list.__doc__
> "list() -> new empty list\nlist(iterable) -> new list initialized from
> iterable's items"
>   >>> list.__init__.__doc__
> 'x.__init__(...) initializes x; see help(type(x)) for signature'

In rev88792, UTC 13:00 Larry fixed this mismatch by patching 
test_calltips.py to match the new string.

> Jan 26, 64-bit installed 3.4.0b3:
>   >>> list.__doc__
> "list(iterable) -> new list initialized from iterable's items"
>   >>> list.__init__.__doc__
> 'Initializes self.  See help(type(self)) for accurate signature.'

and fixed fixed the first-line deletion somewhere in changes to about 30 
other files.

Until I know that there is another test somewhere that docstrings are 
not somehow mangled, I will leave some tests here that they are as expected.

> I will re-compile and adjust the tests by the end of the day.

I recompiled but will do something else instead.

Terry