[python-committers] Time for 3.4.9 and 3.5.6

[Antoine Pitrou]

OTOH, we're talking about a doc fix for a very old version.  Surely
people who rely on the documented opcodes there have already fixed their
software to deal with the discrepancy.

Regards

Antoine.


Le 08/07/2018 à 18:44, Serhiy Storchaka a écrit :
> 08.07.18 13:02, Larry Hastings пише:
>> On 07/08/2018 01:31 AM, Serhiy Storchaka wrote:
>>> I suggest to merge a 3.5 specific documentation fix for 
>>> CALL_FUNCTION_VAR and CALL_FUNCTION_VAR_KW opcodes. There were 
>>> undocumented changes in 3.5, and third-party projects which implement 
>>> interpreting or generating these opcodes do it incorrectly. The 
>>> behavior change was subtle, simple tests will not catch it.
>>>
>>> https://bugs.python.org/issue33216
>>> https://github.com/python/cpython/pull/6365
>>
>> I'm happy to consider documentation changes at any time.  As far as this 
>> change goes: if you look at the PR you'll see I'm the one who closed 
>> it.  While I acknowledge that the current documentation is imprecise, 
>> this PR made the docs inaccurate, and therefore, worse.
> 
> The current documentation is not just imprecise, it is wrong. My English 
> is so bad that I don't understand what is wrong with the proposed PR. To 
> me, it adds correct information. Would be nice if you or other native 
> English speaker suggest better wording. This is not a minor 
> documentation change, this is a fix of a single official documentation 
> about CPython bytecode.
> 
> I'm sorry for reminding about this issue. I thought you forgot about it. 
> This issue was open for 3.5 and marked as "release blocker", but you 
> have not mentioned it.
> 
> _______________________________________________
> python-committers mailing list
> python-committers at python.org
> https://mail.python.org/mailman/listinfo/python-committers
> Code of Conduct: https://www.python.org/psf/codeofconduct/
>