incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Greg Stein <gst...@gmail.com>
Subject Re: Third-party dependencies in CouchDB
Date Mon, 10 Aug 2009 23:52:01 GMT
On Sun, Aug 9, 2009 at 23:15, Curt Arnold<carnold@apache.org> wrote:
>> Noah Slater | 9 Aug 12:42 wrote:
>>
>> I think this is a poor summary. Your thread was taken seriously, and
>> people responded, but as far as I knew, discussion was still ongoing. The
>> way you've worded it could lead people into thinking that you were ignored.
>
> Did not mean to imply that it was ignored, just that the suspect commit was
> not reverted until the issue of the third-party code was resolved.

I was just kind of browsing this thread, and this particular item
struck me as something that I can contribute to...

If a question comes up about some code, then the answer is *not* to
revert the code immediately. A question merely begins a discussion. It
does not impose an action upon the Project. If that were the case,
then any willy-nilly question could cause reverts
left-right-and-sideways.

A hard question about a code is a lesser form of a full veto. In the
veto case, we do not request a revert. Instead, the policy is "deal
with the issue before the next release." If some code has an
outstanding veto, then it cannot be released. It must be resolved as
"keep in" or it must be reverted. But we provide plenty of time for
that to resolve.

In this case, it isn't even at the veto level, so it certainly does
not rise to the level of "revert it until we agree".

There may occasionally be times when code is improperly/accidentally
committed, and the *PMC* takes action to get it out of source control.
But a query from a single committer, even with a veto in hand, is
never really cause to revert-first. Follow through on discussion. At
the end, it may turn out that the code should be left in, so any
(premature) reversion is just a lot of make-work, consternation, and
ill-will.

In this particular case, it sounds like the proper attributions and
copyright notices were added to the NOTICE file, so all is (now) well
and proper.

"Rules to revert" are not the answer. Awareness and discussion are the
right way to go. If those don't work, then a community problem exists,
and escalation is possibly the right answer. Before escalating
anything, I'd suggest you rope in your local ASF Member for
consultation.

Cheers,
-g

---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org


Mime
View raw message