Matt Blaze@brianhonan @mikeylikestech I guess I just don’t understand elections, security, or large scale systems well enough for my imagination to see past all the problems and requirements.
@brianhonan @mikeylikestech Just to be clear about why lotteries are an especially poor example here: They have OPPOSITE requirements. In a lottery, you want the winner to be able to prove they picked the winning numbers (a reasonably tractable problem). In a civil election, you want to make sure no one can find out how anyone voted, and that there's no way for a voter to even prove how they voted (a very different, much harder, problem).
Tazor@mattblaze @brianhonan @mikeylikestech
Yes, this difference in requirements is something I run into often.
In Denmark we have a lot of digital systems for banking, public services etc. And people often ask "if we can do these things, why can't we do digital elections?"
Fundamentally different requirements and objectives.
@Tazor @brianhonan @mikeylikestech Right. Elections have an almost uniquely difficult set of requirements, differing in critical ways from almost any other high-stakes system.
Banking? Different, and much easier. Nuclear command and control? Hard, but for different reasons than elections. Real time control of avionics? Ditto. Etc.
@mattblaze @brianhonan @mikeylikestech
When digital elections were discussed in Denmark the politicians were talking about giving a company an opportunity to do test elections and we were many IT professionals asking them to just present their concept on paper. Because if we could tear their concept apart on paper why do the actual trials?
Luckily the proposal was voted down.