Home Original page

[python-committers] Why?

Nick Coghlan ncoghlan at gmail.com
Thu Jul 31 16:15:49 CEST 2008 
Brett Cannon wrote:
> On Wed, Jul 30, 2008 at 10:41 AM, Jesse Noller <jnoller at gmail.com> wrote:
>> On Wed, Jul 30, 2008 at 1:38 PM, Brett Cannon <brett at python.org> wrote:
>>> On Wed, Jul 30, 2008 at 6:41 AM, M.-A. Lemburg <mal at egenix.com> wrote:
>>>> On 2008-07-30 14:54, Benjamin Peterson wrote:
>>>>> Hi everyone,
>>>>> Sorry if this is a naive question, but I was just wondering what the
>>>>> purpose of this mailing list would be. I don't think there are often
>>>>> python related topics that are only applicable to committers. What
>>>>> sort of things will we discuss here?
>>>> I suppose the list would be useful for notices related to
>>>> a frozen trunk or other admin notices. Filtering those from
>>>> python-dev is not very reliable.
>>>>
>>> MAL hit it right on the head. Guido's idea was to have a list where
>>> you should not have to mentally filter out threads because it has
>>> deteriorated into some competition over what color to paint the
>>> bikeshed (see the whole unittest method name fiasco for a good
>>> example). Because we all ignore various threads on python-dev there is
>>> a chance that something important might get missed. This lowers that
>>> chance by making sure this list is low-volume and high-quality in
>>> terms of what all committers need to be aware of.
>>>
>>> -Brett
>> This might also be a good place to request code-reviews for sensitive commits
>>
> 
> [re-sending; mouse didn't make it all the way over to "reply to all"
> the first time]
> 
> Possibly, although the archive is public, so this is not exactly
> secretive. But yes, if you have some code that needs a committer's
> review (which is practically anything that isn't trivial at this
> point), this list is probably a good place to make the request.

I took Jesse's 'sensitive' there to be more in the sense of "folks this 
patch is really, really important, but also runs the risk of breaking 
things if there is something I haven't thought of". For example, if this 
list had existed before the 2nd beta, I might have pinged it for some 
additional review of the changes to the implementation of __hash__ 
inheritance, simply because the handling of slot inheritance in 
typeobject.c is a fairly complex piece of code (I'm still not completely 
certain why some of the things I tried when fixing the hash inheritance 
bug didn't work the way I expected, which tells me I don't understand 
that code as well as I would like).

Obviously, requesting reviews directly from this is something to be used 
with restraint (since too much traffic would drown out the main 
announcements to do with the tree being frozen and unfrozen), but 
developers that can't exercise appropriate restraint probably aren't 
going to keep their commit privileges for too long ;)

Making the announcement of the tree status on this list should probably 
be added to the release PEP though.

Cheers,
Nick.

--
New computer, signature returning soon :)
More information about the python-committers mailing list