debate
It's clear that Bruce and I have different views about the way (for
example) that the DFSG was adopted. Us arguing about it now is not
going to help - and that's the real problem with this as a
decisonmaking mechanism: there's nothing clear for either Bruce or I
to point to. His key argument is:
> I logged your objections at the time they were received, but you had
> very little support.
I disagree with this claim, but of course neither of us can prove it
either way. If there had been a formal vote then one of us would have
known that the issue had been decided.
> I did hold a confidence vote a while back in which the developers
> rejected a shift to a "Roman Senate" form of government by a large
> margin.
Here again we see the problem of the `complex question'. This style
of `approval voting' doesn't allow the developers real input, because
they're always given the choice between agreeing with the project
leader's suggestion or effectively passing a motion of no confidence
in the leader.
Since passing motions of no confidence in the leader is almost always
worse than whatever individual issue is at stake the leader has a
great deal of concentrated power.
Even with the best will in the world (which I'm sure Bruce does have)
it's not possible to accurately determine the consensus of the
developers by reading debian-devel - in order for this to work we'd
have to make it OK to post `me too' postings and to repeat oneself
endlessly, resulting in whichever group `shouts loudest' getting their
way.
> OK. First, I want to make the point that I am a bazaar-method manager
> (see the oft-cited "Cathedral and the Bazaar" paper by Eric Raymond).
> This means that my goal is to help the project reach a consensus, and
> to only step in when that consensus is not happening. I am in place to
> protect the project from the perils of design-by-comittee, while
> allowing consensus management wherever it will work.
I don't disagree with consensus management - in fact, I feel that my
proposals (which I'll be making a more detailed posting of in another
posting) would make it IMO easier to determine consensus on these kind
of questions, and build a firmer mandate when decisions are made.
The problem with the current situation is that if someone disagrees
with the leader regarding the nature of the consensus they have no way
of testing their support.
> The alternative [to the way the DFSG were agreed on] would have
> been the developers voting on the document line-by-line, which would
> not have been a compromise with each other at all, and would not
> necessarily have given us a usable result. The document doesn't hold
> together if you start deleting pieces.
I totally reject the argument that if you allow people to vote on
amendments to a document then the document must become incoherent.
The amendments clearly need to be well-written as well as
well-supported.
I don't propose to have the developers vote on anything `line by
line'. I'll explain in more detail what I do propose.
Ian.
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-devel-request@lists.debian.org .
Trouble? e-mail to templin@bucknell.edu .
Reply to: