On Mon, Dec 16, 2002 at 01:27:17AM +1100, Clinton Mead wrote: > Here, A wins, even though a superminority of people prefer the status quo. > In my opinion, supermajority requirements are set to make sure major > changes need very popular support, to ensure that major changes only > happen when absolutely needed.
You're welcome to your opinion, of course. I've claimed and continue to claim that supermajority requirements are there to allow a superminority to *block* a change, and nothing else. Granted, this is mildly counter intuitive, but after some thought it makes sense. In traditional votes, it's obviously what happens when a vote requires a supermajority, after all the only possibilities are to let the change pass or to block it. In our votes, though, there are more options: we can make that change, we can make some other change, we can continue doing what we're currently doing, or we can discuss some more and reconsider our options. Traditionally, working out which option is to be voted on is done in advance using meeting rules: someone makes a proposal, then it gets amended multiple times, and finally it gets voted on. > Again in my opinion, if a superminority of voters support no change > rather than major change, then no change should take place. No. If all your superminorities are willing to accept a change, and the majority of people want to make that change, it should be made. It's not particularly meaningful to talk about "supporting" an option with our voting scheme, by the way -- only to say which options you prefer to which other options. It's quite reasonable to say that a vote "ABD" indicates support for *both* A and B, in which case you have unanimous support for all your options. > People could say that the "40 BAD" voters should of voted "40 BDA". But > this means they would of had to vote insincerely. No, it doesn't. It means they shouldn't be exercising their supermajority powers, since they actually find A to be an acceptable outcome. If they wish to insincerely claim that they find A unacceptable in order to strategically assist their preferred option, well, that's kind of petty, really. You can reread the, what, Y2k -vote archives for the discussion of this same issue. Cheers, aj -- Anthony Towns <[EMAIL PROTECTED]> <http://azure.humbug.org.au/~aj/> I don't speak for anyone save myself. GPG signed mail preferred. ``Australian Linux Lovefest Heads West'' -- linux.conf.au, Perth W.A., 22nd-25th January 2003
