Improving how we work with external contributors: Make double-reviews optional in review trial

[Andrew Bartlett]

This came up for me a couple of months ago, where it was pointed out to
me that our review trial, which I've been a part of (at least as regards
my own patches) requires that two team members to review any patch,
which isn't itself authored by a team member.

I think this creates a unnecessary barrier between team members and
non-team members, has not been universally applied even in the voluntary
trial, and I would like to propose that we change 
https://wiki.samba.org/index.php/CodeReview to say:

proposed future policy 
The ultimate policy we (or some of us) would like to establish for
mandatory code code review is this:

      * Each commit in master should have been reviewed by a samba team
        member. For this purpose, involvement in patch creation is
        considered as reviewing.

I wrote at length about my concerns here before, and aside from Micheal
Adam, I didn't get any response.  In short, the reason I propose this is
to encourage those who are not members of the team, by avoiding barriers
to their participation, and to simplify the process for team members who
push these changes, by not requiring the co-ordination of both two team
members and the outside contributor.

My previous comments are here:
https://lists.samba.org/archive/samba-technical/2013-June/092980.html

Certainly I've almost ceased actively applying patches from external
contributors - with other calls on my time, the additional burden here
means I've left to others what I was previously actively merging. 

None of our review policy excuses us as team members from asking for
further review of changes that we are not sure of, but the same applies
to patches by team members.  

Thanks,

Andrew Bartlett

-- 
Andrew Bartlett
http://samba.org/~abartlet/
Authentication Developer, Samba Team           http://samba.org
Samba Developer, Catalyst IT                   http://catalyst.net.nz