[Python-Dev] discourage patch reviews to the list?

["Martin v. Löwis"]

Brett C. wrote:
> All valid points, but I also don't want people to suddenly start posting 
> one-liners or bug posts.

I agree that keeping the noise level low is desirable; I hope this will
come out naturally when we start commenting on high-noise remarks.
For example, I would have no problems telling somebody who says
"me too" on a feature request that he should go away and come back
with an implementation of the requested feature. I would still apply
the "standard" conventions of python-dev: that you should be fairly
knowledgable about the things you are talking about before posting.

> I guess it comes down to a signal-to-noise ratio and if the level of 
> signal we are currently getting will hold.  If we say it is okay for 
> people to send in patch reviews *only* and not notifications of new 
> patches, bug reports, or bug reviews, then I can handle it.

People do tend to notify about patches from time to time, especially
when they are committers, and want to weigh in their reputation to
advance peer review of the proposed changes. Other people who notify
about new patches they made will continue to get my "5 for 1" offer
which actually triggered this new interest in contributing-by-reviewing.

Another reason not to post patches to python-dev is message size for
modem users although I'm doubtful how valid this rationale is these
days, given ADSL, spam, HTML mails, and everything...

> And neither do I.  I just don't want a ton of random emails on 
> python-dev that really belong in the SF tracker instead.  Reason why we 
> don't tend to take direct bug reports in email unless there is a 
> question over semantics.

I certainly don't want to see random comments on python-dev, either
(and I do see random comments come in bursts, and have to choose
to ignore entire threads because of that. I don't have to
write python-dev summaries, though :-)

I disagree with the primary reason not to take bug reports on
python-dev, however: bug reports in email get lost if not
immediately processed; usage of a tracker is necessary to
actually "keep track". So this kind of bug management is the
primary reason for the tracker, not that we want to keep
random users out of python-dev (although this is a convenient
side effect).

Regards,
Martin