4 ms·

I don't like how the article implies that if these systems passed some "cybersecurity verification", they'd be safe against a nation-state attacker, or internal

• nine_k 7y ago
  Frankly, low-tech systems with more human involvement and cross-checks are much safer in this regard, because if the sheer amount of natural intelligence involved (and the amount of corruption required to interfere). This is why they are also slow and expensive.
  • deogeo 7y ago
    > This is why they are also slow and expensive. They're not all that expensive - data is inconclusive, but looks like they're anywhere from twice as expensive as e-voting [1] to cheaper [2]. And they're not all that slow either - only takes about two days to get results, about a week to make them official. Which is much, much faster than necessary, since the time it takes to form a coalition, or take over governing, is longer. So it's entirely unclear what problem e-voting is meant to solve, other than enriching voting machine vendors. [1] https://www.researchgate.net/publication/327977453_How_Much_Does_an_e-Vote_Cost_Cost_Comparison_per_Vote_in_Multichannel_Elections_in_Estonia https://www.researchgate.net/publication/327977453_How_Much_... [2] https://www.wired.com/2008/04/the-cost-of-e-v/ https://www.wired.com/2008/04/the-cost-of-e-v/
    • tmm84 7y ago
      The problem, at least what I remember, was that each state had a different voting style. Some had punches, some had squares you filled in by hand. The chance to make mistakes and the like was there (crossing out a square or double punching) and some voting booths/voting locations only allow straight votes (all x or y party only). Voting machines were designed to make those mistakes/situations less likely.