> since electronic elections are not verifiable in any meaningful way

As I once saw someone point out here, the crucial thing is that they be verifiable by the average person.

Imagine a technically perfect electronic voting system, immune to tampering, preserving anonymity, etc. It's provably perfect - but only to the handful people who can understand the proof.

Average people could not have confidence in such a system, so it undermines democracy. (This may also be a barrier to using vote counting systems better but more complex than "first past the post" unless they can be explained clearly.)

"Put your paper ballot in that bin everyone can see and we'll all count them together at the end of the day" is a process that anyone can see is legitimate. So that's what we should do.

I think a lot of the verification problems for electronic voting could be solved by simplification and using old 'outdated' technology. You don't need a full fledged PC to tally votes, you need the computing power of a 30 year old calculator that could be built on a 2-foot wide board with large and simple traces. Something somebody at home with a multi-meter could probe around and test.

We need the equivalent of a moped to do what we need done, but are being sold on F1 racecars which end up being built by equivalent of rednecks in a scrapyard to line the pockets of the scrapyard owners.

> I think a lot of the verification problems for electronic voting could be solved by simplification and using old 'outdated' technology. You don't need a full fledged PC to tally votes, you need the computing power of a 30 year old calculator that could be built on a 2-foot wide board with large and simple traces. Something somebody at home with a multi-meter could probe around and test.

Yeah, no thanks. We had that in The Netherlands. The NEDAP voting machines were running on m68k machines. With very weak cryptography. In mid '00s.

What we need to do, is keep this last part of technology out of our democratic process. When people count ballots, there are many eyes, and you do not know of each other which candidate or party you represent. Hence, the system works. It isn't broken. Why fix it with a worse solution? A solution not everyone understands? Yet everyone age ~5 and higher understands manual counting.

I've volunteered 3 times thus far (end of may is gonna be my 4th time) as ballot/vote counter in my country (NL). I can highly recommend to just try it once, to learn what it's like. Its also (mostly) fun to work in a diverse team with people of different backgrounds and ages. You never know who you're gonna get beforehand. One footnote: do not discuss politics while counting ballots. Its inappropriate (since you're supposed to be neutral and not know each other's preference) and pointless (won't affect the outcome anymore in any way).

> I think a lot of the verification problems for electronic voting could be solved by simplification and using old 'outdated' technology. You don't need a full fledged PC to tally votes, you need the computing power of a 30 year old calculator that could be built on a 2-foot wide board with large and simple traces. Something somebody at home with a multi-meter could probe around and test.

Even that's too complicated to be trustworthy. That's why I think optical scan ballots are the way to go: even a totally unskilled, untrained regular person would feel fairly confident that they could count them.

IMHO, all the technology investment should be put into processing and validating ballots after they've been filled out. Personally, I like the idea of a voter being able to feed their optical-scan ballot into a machine at the voting booth that will let them verify it will be read exactly as they intend. Then they carry it over to the ballot box to be tallied by a separate machine there.

> IMHO, all the technology investment should be put into processing and validating ballots after they've been filled out

I think the opposite. When filling a ballot, the voter can use a machine with a touchscreen. The person with poor sight can use huge fonts, the blind can use a screen reader, etc. The output from this is a printed paper form, filled out with perfectly legibility. No hanging chads, ambiguous marks, etc.

At this point, the voter (or an assistant) can verify the form just as if a human had filled it out for them.

From there, these perfectly-filled forms can be counted the old-fashioned way with many witnesses.

I think you're making quite the leap to state that average people could not have confidence in the system, people seem to have confidence in things that they can't prove all the time.

That's true. Fiat money is a good example. But in this case, losers have incentive to claim the system doesn't work. You want the average person to be able to refute that claim, or else suspicion will run rampant.

> As I once saw someone point out here, the crucial thing is that they be verifiable by the average person.

But that undermines the purpose of moving to a black-box system.

No, seriously. Every cash register can spit out a receipt; the idea behind audit trails has been around since Hammurabi. The fact that voting systems have none for something as important as the foundation of democracy betrays their purpose.

Every successive society insists on subjugating the plebes by imposing some new barrier to understanding the rules that govern them.

* You can't read the written word of God because you're illiterate. You have to come to church.

* Oh, but you learned to read? Well, only the priests can actually talk to him. You still have to come to church and talk to God through the priest.

* (Who the hell does this Martin Luther guy think he is?)

* Now we have complex tax and legal codes only accountants/lawyers can understand and negotiate. What started as a goat theft escalates to 20 other charges you don't understand the nuances of to successfully defend yourself in court.

* We have voting systems only the administrators have visibility into. Does your vote matter? Was it even recorded in the first place, or did it get wiped out by convenient Russian floating-point rounding errors?

* Our understanding of the world around us is no longer driven by dogma, but personalized algorithms administered by a handful of corporate gatekeepers hell-bent on driving engagement and consumption. If the Faceless Oracle of Google doesn't want you to see it, you'll never discover it organically.

* Freedom of movement is being replaced with self-driving cars, powered by math and code only its engineers (supposedly) understand.

* Soon our personal assets will be governed by the blockchain, another technology understood and exploited by its administrators (and their scammer buddies).

It's a classic hustler and cult-leader tactic. The less someone understands, the easier it is to fleece them.

Mathematical proof is not same as "mathematical proof + implementation detail". Beware of side channel attacks.

The first line of your post is not necessary, and against HN guidelines. Your post doesn't need such either as it stands on its own. It only weakens your argument as an appeal to pity. Don't do that please.

It’s totally necessary because HN is predictable as clock work to downvote when you insinuate the liberal standard has flaws, the biggest of which is the inability to be critical of themselves without instinctively reacting to inflict pain (downvoting)

The post is about electronic voting, which the first line cleverly alludes to. You missed that in your rush to be the HN police.

With all due respect, that appeal to pity only weakens the argument of his post is read from a persuasive writing perspective. However, in viewing this post in an exchange of ideas, it carries necessary information. OP cites the unpopularity not as an appeal to pity but as an expression of ideas. To require all posts adhere to a framework of persuasion is unnecessary speech censorship.

The first time I remember people making comments of this nature on tech sites was close to 20 years ago, when in the aftermath of the 2000 elections, some said electronic voting was the answer, but the companies promoting such solutions seemed universally slimey and not security conscious.

I don't think anything has changed since those days.

I think part of the problem has got to be that non-experts cannot judge how good a security solution is, so local governments are ill equipped to view it with a skeptical eye. There might also be in some way a downside of federalism here - if the states pooled more resources perhaps they would be able to solve the problem once instead of 50 times.

It begs asking at which election would this matter spring to true light. The last election was so crowded with attention baiting headlines that the real conversation wasnt had.

Then again, in the bush election the judiciary decided floridas votes and not a peep was heard. American poltical machinery is far from perfect

I've been proud of my senator, Ron Wyden, for continuing to focus on this.

Oregon's voting system seems ideal to me. I highly recommend it to anybody in other states who would like to see how voting should be done.

Obligatory XCKD Cartoon: https://xkcd.com/2030/