On Fri, 27 Mar 2009, comex wrote: > Unfortunately, I'm not sure that I can unilaterally avoid this sort of > error in the future. If someone else can make that guarantee, let > them be Rulekeepor; otherwise, Goethe, there is always the chance of > an accidental, hard-to-catch difference between the ratified voting > results and the published ruleset. Proposals that explicitly mention > rule text might fail because the text is different; proposals that > number paragraphs might delete important material. Of course, I'll > always try my utmost to ensure the correctness of the ruleset, but I'm > not convinced that ratification is such a bad idea.
I know there's always a chance of error. This is a perfect example, though, where ratifying an incorrect ruleset would clearly have been the *wrong* choice. At the moment, all we have is "Rulekeepor left out some text accidentally... the text is really still in there". If we'd have ratified it in, we would have ratified an incorrect state of the ruleset with a (possibly serious) bug. And if the bug was due to the (true) accidental action of a proposal, that's everyone's fault and Agora is about living with those. What does our Hero say, "the only cure for a bad proposal is a negative vote"? I *do* think if you'd mentioned it publicly when you first noticed the error, we would have all gone back and looked to trace the error and resolved it almost immediately. I think the proposal process in combination with FLR annotations is still pretty easy to use to trace diffs. I'd rather have each ruleset publication have a slight error from the "true state" then reconstruct the true state from proposals whenever an error is suspected, then actually set one of the mistaken texts as the true state too often. -Goethe
