[Python-Dev] [Idle-dev] IDLE contributors and committers

[Tal Einat]

On Sat, Jul 17, 2010 at 2:47 PM, Steve Holden wrote:
> On 7/17/2010 7:33 AM, Antoine Pitrou wrote:
>>
>> Hello,
>>
>> On Sun, 11 Jul 2010 02:05:22 +0300
>> Tal Einat <taleinat at gmail.com> wrote:
>>>
>>> I would like to propose removing IDLE from the standard library.
>>>
>>> I have been using IDLE since 2002 and have been doing my best to help
>>> maintain and further develop IDLE since 2005.
>>
>> I haven't seen any conclusive statement or action to this thread.
>> Without being an IDLE user myself (for good reason), I think that if
>> IDLE should stay, active contributors such as Tal should be given commit
>> access and enough free rein to maintain and improve it.
>>
>> Otherwise there's no reason to continue claiming that IDLE is important
>> while discouraging such people's contributions. The current situation,
>> where several core developers support IDLE's continued inclusion but
>> none actually cares for the issues and patches in the tracker, is
>> schizophrenic.
>>
> +1
>
> There's no reason why Tal should be obstructed in his goal of making
> IDLE at least acceptable again. It's fairly obvious thaat there aren't
> any committers who have both the inclination /and/ the time to do this,
> so adding Tal (and other interested parties) as a developer makes a lot
> of sense.

I would certainly accept this as a first step. Although I now use IDLE
much less than I have in previous years, I would be willing to put in
some time towards fixing the major current issues and integrating the
few polished enhancements.

However, in the long run just allowing "heavy" contributors such as
myself commit rights won't be enough. There's definitely a need for
one or more active maintainers of IDLE who can take care of incoming
bug reports and patches. We may hope that at least one serious
contributor who is given commit rights will take up this position
naturally, but perhaps a more active approach would be beneficial?

I also think that there is a need for a guiding hand for IDLE, as
Guido is for Python. It took a bit of time until I "got" the goals and
principles of IDLE (e.g. easy to learn, minimal and obvious interface)
by having KBK explain them in detail and explain the drawbacks of
certain proposed changes. Having some kind of central authority is
especially important in order to keep IDLE on track because the active
development of IDLE is slow and done by various contributors -- there
is currently no central group of active developers making such
decisions. This doesn't have to be one person who also takes care of
bugs, patches and testing, it could be someone who is just readily
available via the idle-dev mailing list and keeps up with development
of IDLE.

Going along these lines of thought, I reach my original conclusion:
IDLE is somewhat a project of its own. Perhaps considering IDLE a
daughter-project of Python is appropriate, and continuing to develop
it as part of the Python codebase could be reasonable, if more active
maintainers can be found. I certainly support continuing to package it
as part of the standard distribution.