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

Re: RFD: amendment of Debian Social Contract

On Tue, Nov 11, 2003 at 08:26:38PM -0500, Branden Robinson wrote:

> > AIUI, making this explicit and adding competing alternatives can only
> > prejudice the original proposal if the success of that proposal
> > depended on an ambiguous meaning.

> That doesn't follow at all, as far as I can tell.  As I said, one can
> legitimately feel that the Social Contract should be changed as proposed
> without holding any particular view on the removal of non-free from the
> FTP archive.

> Is the proposal to change the Social Contract and "punt" any more
> dependent on an ambiguous meaning if the ballot looks like this?:

> [   ] Change social contract, keep non-free forever
> [   ] Change social contract, remove non-free in 1 week
> [   ] Change social contract, remove non-free in 1 month
> [   ] Change social contract, remove non-free in 1 year
> [   ] Change social contract, remove non-free in 5 years
> [   ] Change social contract, remove non-free in 10 years
> [   ] Change social contract, remove non-free in 100 years

Er, why would the ballot look like this?  Where on this proposed ballot
is the option that you actually favor, the one which defers the
decision?  If your objections are based entirely on the fear that a
ballot can be hijacked so that your original proposal no longer appears
on it, then I have no answer for you.  If you cannot muster the support
of the half-dozen developers required for a consensus-building wording
to appear on the ballot alongside these other options, then I don't
think the problem is with the voting system.

Either people think it's acceptable to modify the Social Contract in the
absence of explicit specification of non-free's fate, or they do not.
If they do, they vote this option above the default option, and their
vote contributes to that option's quorum and supermajority requirements.
If they do not, then they vote this option below the default option.
Assuming sincere voters, the presence or absence of other options on the
ballot should not compromise this one option's ability to pass its
supermajority requirement.

Now, your own comments during the last DPL candidate debate suggest that
you might hold the belief that a sound voting strategy is to always vote
'further discussion' above all options other than your first choice.
Numerous examples were presented during the discussion preceding the
voting GR to show why this voting strategy is ineffective.  If your
first choice is a more radical position, and your sincere second choice
represents an acceptable, moderate compromise, refusing to rank your
second choice above the default option will only serve to hinder the
acceptance of a compromise -- it does NOT improve the chances of your
first choice achieving quorum or supermajority, and if there are other
popular ballot options that you sincerely consider unacceptable, you
have correspondingly increased the chances of one of those options
winning by refusing to endorse a moderate position that would be
acceptable to both sides of the debate.

> I think there *is* a strategic advantage to seeing to it that the ballot
> looks this way, for those who have no intention of ranking any of the
> above options above "further discussion" -- because this way they can
> divide the "pro-change-Social-Contract" bloc.

Again, this strategy only works if there are voters for whom none of the
above options represent their true position on the question.  In such a
case, it is a responsibility of those voters to propose, and acquire
seconds on, a ballot option that they DO agree with.  If a supermajority
of voters agree with a given phrasing, but none took the time to write
it up for consideration, this again is not a failing of the voting
system -- it is a failing of the electorate.

> > Someone who definitely wants to change the social contract and has
> > only a mild preference about whether or not to keep non-free (for the
> > moment) would rank both options above the default option, and
> > "splitting the vote" would have no effect on supermajority
> > requirements.

> Ah, that's a good scenario to raise; thanks for bringing it up.

> I offer this fresh new strategy for dividing and conquering through
> sabotage of consensus:

> [   ] Change social contract, venerate RMS as deity
> [   ] Change social contract, condemn RMS as diabolical menace
> [   ] Further Discussion

> Some of our developers may violently agree with one of the first 2 
> options and violently disagree with the other, and some developers may 
> feel that Debian has no place expressing opinions as a Project about a 
> particular person.

> Now, as the original proposer I might be tempted to see to it that:

> [   ] Change social contract (express no opinion about RMS)

> appears on the ballot, but this doesn't help as much as it should.  I 
> have to alienate the people who would support my proposition but for 
> this crazy noise about RMS.  They may passionately feel one way or the 
> other about RMS, and may just as passionately feel that the entire 
> Project should issue a position statment sharing their opinion -- but 
> the simple fact is that their opinion of RMS, or the Project's 
> decision to issue a statement about him, simply doesn't matter.

So you're arguing that, by making it explicit that the third ballot
option expresses no opinion about RMS, you will alienate voters who
don't wish to express that they don't wish the project to express an
opinion about RMS?  This is an unusual view of the electorate, one that
I have certainly never considered.  Do you have any reason to believe
that a non-negligible number of developers would feel this way in the
case of the Social Contract+RMS vote or (more germanely) the Social
Contract+non-free vote?

> So while I would not argue that keeping or dropping non-free has no
> conceivable relevance to whether or not we strike clause 5 from the
> Social Contract, I *would* say that reasonable people can agree that the
> clause should be stricken, and at the same disagree as to what we should
> actually do about non-free.  That's orthogonality; that means the
> options should not be presented as alternatives to each other.

It's orthogonal because you've considered only a subset of voters -- the
ones who agree with you that removing this text from the Social Contract
is more important than the actual fate of non-free.  Reasonable people
can also hold that their position on the amending of the Social Contract
*is contingent upon* the consequences for non-free.  It is therefore
important that they have the opportunity to express this position when
voting.  The voting system should be a means of *expressing* consensus,
not of *engineering* consensus where none actually exists.

Steve Langasek
postmodern programmer

Attachment: pgpSStTYjzajd.pgp
Description: PGP signature

Reply to: