On Sat, Nov 01, 2003 at 06:04:55PM -0500, Anthony DeRobertis wrote: > On Nov 1, 2003, at 10:27, Anthony Towns wrote: > >On Sat, Nov 01, 2003 at 09:47:04AM -0500, Anthony DeRobertis wrote: > >>I think I just realized something... Due to the supermajority > >>requirements, given my favorite ballot: > >> A: strike SC 5 > >> B: trivial > >> C: strike SC 5 + trivial > >> D: further discussion > >>If my true preference is CABD, I should vote CADB or even CDAB. I > >>should > >>do this because A.6.3 makes my vote for D count against B (and A) > >>three > >>times. > >Well, no you shouldn't, because you're increasingly likely to end up > >with the default option winning, which is what you claim to want > >_least_. > No, it doesn't. My preferred option still has just as many votes over > the default option. That's only relevant in the limited range of cases where the other options are eliminated either by having your preferred option beat them, or by having the default option beat them. You don't know that at the time you vote however; you also have to protect yourself against a bunch of other people voting D above C and eliminating C from consideration (whether sincerely or not). If so, the only remaining choice will be amongst D, A and B; and if it comes down to your vote, you will ensure that D wins. Which, by assumption, is not what you want. > I think I failed to get my point across. Let's say that we expect a > close outcome between C and B, but with B slightly ahead. ...and B barely making its supermajority requirement, and C comfortably making its supermajority requirement. Certainly, you can vote like this; but if you're thinking like this, it's unlikely that you really would prefer the options you're trying to vote against to "further discussion". To take a more concrete example, I voted [1432] in the recent vote; because I didn't really think either of the non-supermajority options were good ones. You'll note though, that most people didn't vote like this (although enough did to make those options fail supermajority), and that the option that didn't fail supermajority was put in first place by 148/242 voters (61%). > The main > thing I want is to strike SC 5, but B is still a good result. So my > honest preference is CABD. Uh, not quite. All you've said there is that your preferences are CAB. The question is whether you want C enough that you'd prefer to have nothing happen and keep considering the issue than not have C. If that's the case, then please, go ahead and vote CDAB. If that's not the case, but you're completely sure that you're not going to get into a situation where C fails its supermajority _and_ you think it's possible that A might almost fail its supermajority _and_ A might beat C, then you might also want to vote CDAB. If you're not sure of all those things, voting your true preferences is the most likely to get you the result you want. > All three other options are better than the > status quo, so I rank them as such in a sincere vote. Uh, that's not exactly what the default option is. The default option is "leave the issue unresolved, continue discussions, eventually propose a new vote." If someone thinks the "status quo" is worth preserving explicitly, that should be proposed as an alternative on the ballot, eg: [ ] Drop non-free [ ] Support non-free better [ ] Continue supporting non-free at the current level [ ] Further Discussion > Now, I realize that under A.6.3, B and A need to both independently get > thrice the votes of the converse. So, wanting C above those two, I > decide to give the converse a vote. I vote CDAB. That isn't sincere, > but it's smart. Not really; since there's likely to be a bunch of people voting ADCB too. If there are enough for you to both succeed in your goals of thwarting your opponent, either B or further discussion will end up winning (probably further discussion). In otherwords, congratulations: you've both been so smart you've made it impossible for either of your first two choices to win. > Further discussion == do nothing this vote, preserve the status quo. > Even if further discussion loses, I can always post to -project, > -devel, -vote, whatever to drum up support. Certainly. And if further discussion loses you can be told "read the archives, and accept the project's decision on this topic, which has already been made." > >(Note that this has been discussed copiously in the lead up to the > >voting > >GR; and that we've had the GR on it, which has passed) > I know. I was part of those discussions after all. Don't recall this > particular issue coming up, though. Mmm. It was discussed in detail, and iirc came up repeatedly. See, eg, the threads from May: http://lists.debian.org/debian-vote/2003/debian-vote-200305/msg00245.html http://lists.debian.org/debian-vote/2003/debian-vote-200305/msg00195.html http://lists.debian.org/debian-vote/2003/debian-vote-200305/msg00203.html Actually, I also notice: http://lists.debian.org/debian-vote/2003/debian-vote-200305/msg00205.html ...that you were around at that time. 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:
pgpUI5PW83dih.pgp
Description: PGP signature