[Haskell-cafe] Re: The Haskell re-branding exercise
Sterling Clover wrote: IP based limitations are a terrible idea. Multiple users can be and often are behind the same IP if they're in some sort of intranet, be it corporate, academic, or simply multiple home computers. Mail-based authentication can be screwed with, sure, but it's also very easy to notice this (as opposed to ip nonsense) through simply eyeballing the results. There's no general everywhere way to prevent vote fraud. However, if we make it even require a mild bit of thought, that should be sufficient in this case, as there won't be enough votes to prevent some sort of rough eyeball-based check of the results, and if there are, then that's a sign of fraud for sure! Furthermore, there's very little incentive for someone to go the extra mile here, as we're voting for a haskell logo, and not, e.g., giving away ten thousand dollars. Furthermore, since I assume we'll only be presenting reasonable logos, there's not even some room for pranksters to stage a write-in of some gag slogan. I suggest we do voting by email, and restrict voting to those who have ever posted on haskell-cafe before 1 Jan 2009. We could then have an auto-confirmation scheme similar to mailing list sign-up where the confirmation message is sent back to the originator to confirm their identity, containing a verification link to click on. I realise there are flaws in this, but it seems to be (a) cheap to implement and participate in, and (b) good enough. Cheers, Simon ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Re: The Haskell re-branding exercise
Simon Marlow wrote: I suggest we do voting by email, and restrict voting to those who have ever posted on haskell-cafe before 1 Jan 2009. We could then have an auto-confirmation scheme similar to mailing list sign-up where the confirmation message is sent back to the originator to confirm their identity, containing a verification link to click on. I realise there are flaws in this, but it seems to be (a) cheap to implement and participate in, and (b) good enough. That sounds better than my Haskell Wiki verification method. Richard. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Re: The Haskell re-branding exercise
marlowsd: Sterling Clover wrote: IP based limitations are a terrible idea. Multiple users can be and often are behind the same IP if they're in some sort of intranet, be it corporate, academic, or simply multiple home computers. Mail-based authentication can be screwed with, sure, but it's also very easy to notice this (as opposed to ip nonsense) through simply eyeballing the results. There's no general everywhere way to prevent vote fraud. However, if we make it even require a mild bit of thought, that should be sufficient in this case, as there won't be enough votes to prevent some sort of rough eyeball-based check of the results, and if there are, then that's a sign of fraud for sure! Furthermore, there's very little incentive for someone to go the extra mile here, as we're voting for a haskell logo, and not, e.g., giving away ten thousand dollars. Furthermore, since I assume we'll only be presenting reasonable logos, there's not even some room for pranksters to stage a write-in of some gag slogan. I suggest we do voting by email, and restrict voting to those who have ever posted on haskell-cafe before 1 Jan 2009. We could then have an auto-confirmation scheme similar to mailing list sign-up where the confirmation message is sent back to the originator to confirm their identity, containing a verification link to click on. I realise there are flaws in this, but it seems to be (a) cheap to implement and participate in, and (b) good enough. Seems good enough. Who's going to tally the votes? ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
