[FFmpeg-devel] IRC meeting on Saturday 2015-09-12, UTC 15:00

Nicolas George george at nsup.org
Fri Sep 11 14:46:52 CEST 2015

Le quintidi 25 fructidor, an CCXXIII, Michael Niedermayer a écrit :
> I have a few problems with using the UN security council as
> comparission

I agree that this comparison is not very good, but I do not think that your
fist mail about vetoes and unanimous consensus address the issue.

Discussing the best option and trying to all agree is already how the
project work. There is no need to codify it by calling objections vetoes.

The problem happens when the project does not manage to reach unanimous
consensus. The more people are involved in the project, the more likely it
is to happen. If we are talking about merging, remember why the fork
happened in the first place: because people did not agree. If the same
people come together again, they will disagree again.

The project needs a way of making a decision when people do not agree.

The process of making that not-unanimous decision must be codified, so that
people who do not like the outcome can not contest it.

And the process needs to be codified while we mostly agree, because we need
to agree on the process. Otherwise, people could try to propose a process
that favours the outcome they prefer, or be accused of doing so.

At the very least, the project probably needs some kind of voting process to
make important and controversial decisions. It needs to express who can
vote, and how much weight each person has, how to call for a vote, how to
vote (delay, secret or public ballot), how to combine the ballots into a
result (if it is not a yes/no question, see Condorcet's paradoxes and
Arrows's theorem).

Of course, most of it can be imitated from other projects who already have
that kind of procedure.

And of course, we should all bear in mind that this is only a last resort,
when we have utterly failed to agree.

Another point: disagreements of that kind happen because of non-technical
considerations, basically for reasons of taste. There is not one side right
and the other side wrong, there are just people who give different weight to
different kind of annoyances. This can happen for small things as well as
big ones, and we do not want to call for a vote every time someone wants to
add a workaround for a broken compiler while someone else wants to tell the
users to get a working one.

Therefore, I believe a lightweight decision making process would be useful
as well. And the simplest solution would be to just have someone make the


  Nicolas George
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20150911/e7c2b8db/attachment.sig>

More information about the ffmpeg-devel mailing list