Not that it makes much difference to 'further discussion', but:
On Tue, Mar 24, 2009 at 04:10:49PM -0700, Lucas Nussbaum wrote:
> PROPOSAL START
> =====================================================================
> General Resolutions are an important framework within the Debian
> Project. While over those years, some problems have arised during the
> discussion and/or voting of some resolutions, there is no evidence that
> changing the number of sponsors (seconds) for GR proposals or amendments
> will help solve those problems. Instead, by making it harder to propose
> general resolutions or amendments, it might make it harder to improve
> imperfect resolutions, or to add valuable options to a ballot.
>
> Therefore the Debian project reaffirms the current requirements for the
> sponsoring (seconding) of GR proposals or amendments, and for overruling
> of delegates.
> =====================================================================
> PROPOSAL END
I second this proposal.
I have felt more positive about the whole idea of increasing sponsorship
requirements in the past, but as time has gone on, I now feel that it
does not really make much sense anymore.
The main argument for increasing the sponsorship requirement seems to be
"there is too much politics in Debian, so let's make it harder to start
a vote". That argument is flawed, for multiple reasons:
- Making it harder to vote isn't going to change people's opinions on
controversial subjects. If a person feels that his opinion is in line
with the project's core ideals, there are other people who feel
differently, but not enough people who care enough to second a vote on
the subject, then this will *not* cause them to stop talking about the
subject.
Me, I prefer to have a vote on a subject (so that a consensus is
known, even if not necessarily liked by everyone) rather than have to
listen to people talking about the same subject over and over and over
and over and over and over and over again. I'm quite sure most people
in the project agree on that bit, since we've kicked people out for
precisely that reason.
- Like it or not, in some regards Debian actually *is* a political
organization. We've come up with a number of documents that have
helped shape the Open Source/Free Software landscape, such as the
DFSG/Open Source definition, our Social Contract (which Gentoo based
theirs on[1]); we've taken a position on matters such as the GFDL and
the Mozilla trademark policy which were not always in line with the
rest of the FLOSS world, but which have since come to be accepted by
many people unrelated to Debian. Ignoring that fact because we don't
like voting as much as we like hacking isn't going to change it, nor
make it go away.
- There is no proof that any change in sponsorship requirements is going
to actually improve things. I find this kind of thing to be extremely
dangerous; it has the potential to ruin our entire GR procedure (if it
does turn out to be too high a number in practice) for a stated
benefit that it may even fail to produce.
Additionally, there is the point of amendments; while it could perhaps
be a good thing to discourage starting the GR procedure, the same most
certainly is not true for amendments. If a vote has been taken,
reverting it is going to be almost impossible[2]; so it had better be
clear and complete to start with. Making it harder to propose amendments
is counter to that goal.
A suggestion was made in the thread that would increase the seconders
requirement to start a vote, but that would not increase the requirement
for amendments. This, I believe, will also not work.
It requires us to make a clear distinction between a 'GR proposal' and
an 'amendment'. Currently, they're mostly the same thing. If we make it
clearly separate things, where it would be easier to do an amendment
than it is to do a vote, that would make it attractive to craft what is
really a GR proposal in such a way that it could be seen as an amendment
to an already-existing vote. If the text of a proposed amendment only
tangentially relates to the original text which it claims to amend, is
it then still an amendment or should it be considered a vote in its own
right?
That doesn't even have to be done on purpose; sometimes you could think
in good faith that a particular statement is related to an
already-existing vote, but then find that you're alone with that
opinion. What's going to have to be done with that? These may seem
trivial questions for now, but they're going to be very hard questions
to answer once we're actually talking about a proposed vote and emotions
are running high.
Moreover, I find that the Debian GR procedure actually is a great way to
build consensus. The GFDL debate is an excellent example of that; it
took us two votes (if you don't count 2004_004) and several years, but
we've gone from a general feeling of "The GFDL is from GNU, so it must
be free" to a stated and clear consensus of "The GFDL has some parts
which may make texts covered by it non-free if used", in part thanks to
the General Resolutions that were held; these GRs forced us to
reevaluate our positions, and did invalidate claims stating that "the
silent majority" agreed with one or the other argument.
In short, I believe that this proposed vote is going to cause much,
much more problems than it is trying to solve, and that the chosen
numbers are too arbitrary for me to place any confidence in them. If the
proposers change their proposal so that it clarifies which problems it
is going to solve, and how it is going to do so, /and/ manage to come
up with some actual proof of those claims, I might be persuaded to their
position. For now, however, I'm not even going to place that proposal
above NOTA.
[1] http://www.gentoo.org/main/en/contract.xml
[2] See for example vote 2004_004; this was a direct result of the
changes to the social contract in vote 2004_003. There were several
options on the ballot then, including one to completely rescind vote
003, with the stated intent among some of the seconders (at the very
least, myself) to perhaps restart the GR, with a better title, after
rescinding the changes. This proposal did not reach its required
majority.
--
<Lo-lan-do> Home is where you have to wash the dishes.
-- #debian-devel, Freenode, 2004-09-22
Attachment:
signature.asc
Description: Digital signature