Re: Debian's custom use of Condorcet and later-no-harm
Thue Janus Kristensen <thuejk@gmail.com> writes:
> So in the init system vote example with my rule modification, D, U and
> FD would end up in the Schwartz set, Bdale would choose D, and the final
> result would then be FD, because D doesn't beat FD. So this rule change
> means that U cannot win unfairly due to strategic voting for FD. That FD
> wins is fair, given that it reflects the actual votes which were cast.
This change would also fix a different problem that came up during the
debate, namely one of the problems with the 2:1 majority required for a TC
override. Currently, if we have a general project vote on something on
which the TC has already ruled, you can get situations where overriding
the TC is the Condorcet winner but doesn't achieve supermajority over FD,
and a different option is chosen as the winner even though it's beaten by
the override because it doesn't require a supermajority and beats FD.
The change suggested here would convert that to an FD result, which makes
a great deal more sense to me. If the Condorcet winner can't win because
of super-majority requirements, having the vote outcome be FD instead of
choosing an otherwise losing option seems saner than the current result.
(I believe we should remove the 2:1 supermajority requirement for TC
overrides regardless, but even if we do there are other supermajority
scenarios where the same thing could happen.)
--
Russ Allbery (rra@debian.org) <http://www.eyrie.org/~eagle/>
Reply to: