On Nov 19, Raul Miller wrote: > Here's some thoughts about how we might implement supermajority: > > [1] The simplest: discard supermajority entirely. Nothing special is > required to override "important decisions". This has some elegantly > simple mathematical properties but I don't know of any other argument > for it.
Simple, perhaps, but I'm not willing to risk the project on the assumption that this would have no negative effects. Supermajority has the nice property that avoids the recurrent flip-flop problem in a close voting decision (where Group A wins, then Group B wins, then Group A wins again, ad nauseum). > [2] Discard the election if the winner doesn't meet supermajority. > Perhaps, if the ballot has nondefault options with differing supermajority > ratios, immediately hold another election using only the options with > a lesser supermajority ratio. I'm leaning towards this at the moment. How does this differ from [2]? Does that mean the winner (if a supermajority applies to it) has to appear in first place on a supermajority of ballots? My gut feeling: require any supermajority option to be assented to separately on the ballot; a supermajority option only takes effect if it both wins the Condorcet vote and receives a supermajority on the assent question. (I think aj proposed something similar to this.) i.e. 1. [1] [Y] AMEND CONSTITUTION to allow modification of "foundation documents". 2. [3] [N] AMEND CONSTITUTION to include the existing Social Contract and Debian Free Software Guidelines within it. 3. [2] FURTHER DISCUSSION. So, if the Condorcet winner is option 1, this counts as an assent to the amendment; if the winner is option 3, this counts as a dissent. Tally assents and dissents and determine their ratio. The voting system should reject as spoiled any ballot that doesn't assent or dissent. Chris -- Chris Lawrence <[EMAIL PROTECTED]> - http://www.lordsutch.com/chris/ Computer Systems Manager, Physics and Astronomy, Univ. of Mississippi 125B Lewis Hall - 662-915-5765 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]