Adam Jacob Muller wrote:
At a recent family gathering I spent about an hour trying to explain to various people why "open source" voting machines are more secure.Everyone perceived "open" as being able to go in and change votes...The fact that I was trying to explain the open source model for the first time did not help...
Therein lies the issue. Understanding the (possible) benefits of open-source voting machines, and how computerized voting systems might or might not be reliable and verifiable has two big problems:
i) it's obscure ii) it's boringIt's obscure because at the least you need a grasp of various concepts of computers and software to understand the terminology, let alone decide on the relative merits of different approaches. It's boring because people who don't know those things on the whole really don't want to, especially given faith that "someone else is checking" and that elections "don't get tampered with in the West" (etc.)
Paper votes are slow to count and may be spoiled. Ballot boxes may be lost. But the basics can be grasped by just about anyone, and from there much of the detail understood. It's a piece of paper, somehow marked to indicate preference. Those pieces of paper are counted, and that count decides who won (whether it's first past the post, STV, ATV or whatever). Even the complicated stuff is understandable. That's why the obvious compromise is a paper audit trail: the machines can count the votes very quickly, but if there's a problem you can do it the old-fashioned way, and everyone can understand the old-fashioned way.
Craig.