[Top][All Lists]

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

Announcing the GNU Kind Communication Guidelines

From: Richard Stallman
Subject: Announcing the GNU Kind Communication Guidelines
Date: Mon, 22 Oct 2018 01:16:46 -0400

Announcing the GNU Kind Communication Guidelines

The GNU Kind Communication Guidelines, initial version, have been
published in  On
behalf of the GNU Project, I ask all GNU contributors to make their
best efforts to follow these guidelines in GNU Project discuaaions.

In August, a discussion started among GNU package maintainers about
the problem that GNU development often pushes women away.1  Clearly this is
not a good thing.2

Some maintainers advocated adopting a "code of conduct" with strict
rules.  Some other free software projects have done this, generating
some resistance.3  Several GNU package maintainers responded that they
would quit immediately.  I myself did not like the punitive spirit of
that approach, and decided against it.

I did not, however, wish to make that an excuse to ignore the problem.
So I decided to try a different approach: to guide participants to
encourage and help each other to avoid harsh patterns of
communication.  I identified various patterns of our conversation
(which is almost entirely textual, not vocal) that seem likely to
chase women away -- and some men, too.  Some patterns came from events
that happened in the discussion itself.  Then I wrote suggestions for
how to avoid them and how to help others avoid them.  I received
feedback from many of the participants, including some women.  I
practiced some of these suggestions personally and found that they had
a good effect.  That list is now the GNU Kind Communication

The current version not set in stone; I welcome comments and
suggestions for future revision.

The difference between kind communication guidelines and a code of
conduct is a matter of the basic overall approach.

A code of conduct states rules, with punishments for anyone that
violates them.  It is the heavy-handed way of teaching people to
behave differently, and since it only comes into action when people do
something against the rules, it doesn't try to teach people to do
better than what the rules require.  To be sure, the appointed
maintainer(s) of a GNU package can, if necessary, tell a contributor
to go away; but we do not want to need to have recourse to that.

The idea of the GNU Kind Communication Guidelines is to start guiding
people towards kinder communication at a point well before one would
even think of saying, "You are breaking the rules."  The way we do
this, rather than ordering people to be kind or else, is try to help
people learn to make their communication more kind.

I hope that kind communication guidelines will provide a kinder
and less strict way of leading a project's discussions to be calmer,
more welcoming to all participants of good will, and more effective.

1. I read that the fraction of women in the free software community
overall is around 3%, whereas in the software field overall it is over

2. I disagree with making "diversity" a goal.  If the developers in a
specific free software project do not include demographic D, I don't
think that the lack of them as a problem that requires action; there
is no need to scramble desperately to recruit some Ds.  Rather, the
problem is that if we make demographic D feel unwelcome, we lose out
on possible contributors.  And very likely also others that are not in
demographic D.

There is a kind of diversity that would benefit many free software
projects: diversity of users in regard to skill levels and kinds of
usage.  However, that is not what people usually mean by "diversity".

3. I'm not involved in those projects, even if in some cases I use the
software they release, so I am not directly concerned about whatever
internal arrangements they make.  They are pertinent here only as
more-or-less comparable situations.

Dr Richard Stallman
President, Free Software Foundation (,
Internet Hall-of-Famer (

reply via email to

[Prev in Thread] Current Thread [Next in Thread]