On Fri, Nov 07, 2003 at 03:27:14PM +1000, Anthony Towns wrote: > Sure there is; people might legitimately want to vote: > [ ] Change social contract, remove non-free > [ ] Change social contract, don't remove non-free > [ ] Don't change social contract, don't remove non-free > [ ] Further Discussion > > 100 x 1234 : Do everything we can to remove non-free > 100 x 3124 : Keep non-free, but preferably deprioritise it > 150 x 2413 : Keep non-free, but take a stance either way > The developers' preference is then: > > A beats B, 250 to 100 > B beats C, 200 to 150 > C beats A, 250 to 100 > A, C beats D, 350 to 0 > B beats D, 200 to 150 As I said, if we split the vote by considering the social contract first, then whether to remove non-free, we presumably decide first to remove the social contract (A+B beats C and D) then to remove non-free. However if we vote in the opposite order, for example: [ ] Remove non-free, once social contract is changed. [ ] Keep non-free. [ ] Further Discussion [ ] Change social contract [ ] Keep social contract as is. [ ] Further discussion. Then we find that non-free is kept (B+C beats A), and the social contract is not changed (fails supermajority). That is, given the *same opinions* by the developer body, the outcome of the GR procedure will vary depending on the order of the votes, if we split related issues into separate ballots. The reason why we should vote on related considerations is that otherwise the outcome can be determined by the machinations of the secretary in deciding how the vote is conducted. That is also why we spent so long ensuring our voting system can deal with many options, even ones that are closely related. Cheers, aj -- Anthony Towns <aj@humbug.org.au> <http://azure.humbug.org.au/~aj/> I don't speak for anyone save myself. GPG signed mail preferred. Australian DMCA (the Digital Agenda Amendments) Under Review! -- http://azure.humbug.org.au/~aj/blog/copyright/digitalagenda
Attachment:
pgpkbZlxFcm_0.pgp
Description: PGP signature