[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Proposal: mediators



Dear all,

I would like to propose a system for mediation that can be used before
or during expulsion processes, and for other purposes.

Please note that while I use the word "mediation", there may be a better
one.  Thus, it is not needed to reply to me that what I propose is not a
real mediation if this is the case.  Please focus on  what I propose and
not on how I named it !

### Goal

The goal will be to help all parties or their proxies to prepare
fact-driven synthethic position statements that can be presented to
others, in particular to the decision makers, both to help them to take
their decision and to help announce it in a way that is the most
resilient to misunderstanding and the criticisms that it triggers.  It
is also hoped that the mediation process may also sometimes help both
parties to settle their disagreement, although it will not be considered
a failure when settlement is not reached.

### Who

I think that the mediator should not be doing this task on a regular
basis, to avoid burn-out, to avoid being in contact with too many
private information, and to avoid being perceived as partial.  In order
to ensure that the mediators are available and have demonstrated some
williness to communicate with others, I propose to select them randomly
among the Developers who have been the application maintainer of at
least one person.  A call would be valid one day; in case of no answer,
another mediator would be called, and so on (better strategies using
further filtering could be done if it does not look realistic that a
volunteer could be found after 10 calls).  The call would be initiated
by the decision makers, when they feel that this mediation process would
help them.

### How

The mediator will contact both sides separately and will never connect
them nor request or suggest direct communication between them.  The
mediator will minimise the volume of communication, aiming at addressing
each issue only once.  The mediator will ask each party to sort the
points they want to make in order of relevance (I hope that this process
will help to reduce the number of points), summarise them, and ask if
the summary is clear and accurate (I think that when somebody can
recognise one's views in the words of another person, it is a good
indication that there is mutual understanding).  The mediator will then
present the views of each side to the other side, and ask if there is
agreement, disagreement, and recommend to avoid lengthy rebuttals.  The
mediation is only part of the conflict resolution, and it is already a
great achievement to agree to disagree.  Lastly, even when clear action
points emerge from the discussion, the mediator will never propose an
implementation, leaving that choice to the decision makers.

### Privacy

Communication will be written, encrypted and kept confidential.  It may
be given to the decision makers if needed and must be destroyed after a
reasonable delay (advice welcome).  The mediator will never disclose the
contents of the messages and information exchanged.

### Limitations

To avoid any legal consequences and to avoid damage that could be caused
by improper consideration of a victim's suffering, this mediation system
should never be used when the facts being discussed are so grave that
they could be punished by a justice court.

### Summary

In light of the expulsion being discussed at the moment, my impression
is that such a process could, or even still can, be useful to underline
what are the points considered most important by each side, and which
action of the other side can solve them (I am not going to speculate
here).  I also hope that this proposal may be more broadly useful,
especially for the issues at the inteface between behaviour and
technique (typical examples are when an NMU are a commit revert trigger
a latent conflict).

### Next step

I personally would prefer to avoid long point-to-point discussions on
the weakest parts of what I wrote above.  How about reacting on what you
liked (if you liked some part), and just sending the rest to /dev/null ?

Thank you for reading so far, and have a nice day !

Charles

PS: please pardon my English, it might betray my thoughts.

-- 
Charles Plessy
Debian Med packaging team,
http://www.debian.org/devel/debian-med
Akano, Uruma, Okinawa, Japan


Reply to: