incubator-sanselan-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Carsten Ziegeler <cziege...@apache.org>
Subject Re: Ready...Set...
Date Mon, 24 Sep 2007 09:50:46 GMT
Craig L Russell wrote:
> On Sep 22, 2007, at 8:19 AM, Charles Matthew Chen wrote:
> 
> 
> No, and in fact there isn't always a need for a formal vote if there is
> consensus. The question is how to gauge consensus, and sometimes the
> easiest way to gauge it is to have a vote. If there is a vote on
> process, majority rules. Not everyone has to vote.
> 
Yes, there is always a thin line here, we have votes, lazy consensus,
commit-then-review etc. Now, if the person who does something/wants to
do something, thinks that its just a small thing (simple patch, bug fix
etc), she commits. With the peer review by other committers, possible
problems can be detected and will be fixed either by reverting or
adapting - that's not a big issue usually (if things are handled in a
friendly manner)
If you think that its more a substantial thing, a vote is better. A vote
can either be a "real vote" which means majority rules, or lazy
consensus applies which means if noone objects, you can go ahead. This
is usually done by sending an email to this list like "I want to do xyz.
If noone objects I'll do the changes tomorrow" etc. Of course, you
should give people enough time to think about the changes. :)

How these things are handled often vary from community to community. If
the community has a long experience and people know each other for
years, there are usually less votes. If a community is new, like we are,
its probably better to discuss things more often and call for votes. We
can then see how this works over time and adapt.

Carsten
-- 
Carsten Ziegeler
cziegeler@apache.org

Mime
View raw message