Live data from Hacker News

Colorado scrambles to change voting-system passwords after accidental leak

arstechnica.com

391–400 of 682 posts

Re: Colorado scrambles to change voting-system passwords after accidental leak

#391

Earlier quoted context omitted.

Are you saying that the only check possible is looking at it while its in your hand?

The problem stated was that the marker machine lies 1 out of 15 entries. The paper would contain an incorrect selection occasionally. So, yeah, it would require no one noticing during the act.

Indeed, and the math is the same. If out of 3 million voters, just 1000 double-check the printout, they will detect a 1/100 flip with probability 99.996%.

Re: Colorado scrambles to change voting-system passwords after accidental leak

#392

Earlier quoted context omitted.

I mean, if you're willing to spend that much, and it'll be very expensive, then sure. It's just technophobia - machines are going to be more accurate than a human (who also can make a mistake!).

The disconnect is that in most of the world we only vote for one or two candidates on a ballot. In America you vote for everything from the president to the dog catcher on one ballot. While I think of it, the USA and UK should both stop holding votes on working days. That is nuts! Do what Australia does and vote on a Saturday and make it compulsory.

Believe me, we've been aware that this is a non-bug feature for a long time.

The Tuesday law was passed in 1845. Instead of changing it, many legislators are pushing in the opposite direction: trying to selectively suppress their opponents' votes further. If it hurts them more than us, it's a worthy goal!

Re: Colorado scrambles to change voting-system passwords after accidental leak

#393
post #364

Earlier quoted context omitted.

> First of all zero-knowledge proofs allow you to verify stuff without being able to prove it to others I doubt it, and I suspect if you try to point at a specific system to implement you will find that that none exist even in theory. I can verify I voted with zero knowledge, yes. But I can't verify who I voted for . So I can put candidate A into the machine, it switches to candidate B and we can all prove I voted in…

If everyone got a unique prime number and a running total vote product was available, I always thought this would be a neat solution. Still susceptible to the goon-with-a-wrench technique I think

BLS signatures can provide similar properties

Crypto (by which I always mean cryptography) can help secure a lot of things that normally are just “trust in a middleman”

https://news.ycombinator.com/item?id=29306829

Re: Colorado scrambles to change voting-system passwords after accidental leak

#394

Earlier quoted context omitted.

It is important that the voting system have credibility for everyone - regardless of party. Has anyone done a ground up exercise of rethinking the process and the involved technologies from a cybersecurity standpoint? It would be great to offer voters verification of their votes while maintaining secrecy. But right now I feel like we are stuck, with one half the country having doubts about the process and the other h…

> Has anyone done a ground up exercise of rethinking the process and the involved technologies from a cybersecurity standpoint? Chesterton’s fence.

I take it that you mean that before you tear down this system, understand why it is the way it is. And, yeah, sure. I don't think that invalidates reimagining the solution from what may be new/updated first principles.

Re: Colorado scrambles to change voting-system passwords after accidental leak

#395
post #386

Earlier quoted context omitted.

It starts with being able to tell that the information was encoded correctly when I submitted it. Tell me this: what is the advantage of a barcode, over a scantron-esque system where I can see which item I chose because a dot is filled in? The scantron-esque system is still efficiently machine readable; we've had scantron since I was a kid. The difference is, I can verify with my own two eyes that the information is…

Pretty sure GP is saying a scantron-style one can still be flipped or offset at the destination. They use position on the ballot, not OCR, to determine what the vote is.

At least the source would be correct so 50% less chance of cheating i.e. the cheating did at least not occur while producing the vote.

Re: Colorado scrambles to change voting-system passwords after accidental leak

#396

Earlier quoted context omitted.

In my locale there is a header on the physical ballot that contains a bunch of barcodes, presumably to make your votes machine readable. It then prints the votes in text below. I absolutely hate that fact. I am a human, I cannot read barcodes without a computer. Therefore, I cannot tell if the important part of what was recorded is correct. Not sure if Colorado's are the same...

The value of absolute transparency is why nothing will beat paper ballots written and marked in plain English counted by hand with anyone and everyone who cares about election integrity watching the process.

And we should dye the thumb of those that already voted.

Re: Colorado scrambles to change voting-system passwords after accidental leak

#397
post #382

Earlier quoted context omitted.

That wouldn't help for this case. Even a PhD in cryptography and computer science doesn't help you in any way be convinced that a particular machine is securely counting your votes. If you want to be convinced of that, you have to audit the code and the hardware specs and the network code and everything in between to ensure that the system: (a) implements the claimed algorithms, (b) does so correctly and free of side…

This is wrong. You don’t need a Ph D or inspect code to know that your vote is included in a Merkle tree. And you can verify that the vote total matches what is in the Merkle tree for your district, and the national Merkle tree of districts. You can also verify that each voter was issued a unique token, which went through a mixer. About the only thing you can’t verify is that the agency giving out the token hasn’t be…

So you agree that you can't verify that the system has one and only vote tabulated for every person that actually voted, or that the vote they intended is the one that got counted. So, you agree that you can't trust the results of this election.

Furthermore, if you check and find out that your own vote was incorrectly counted, you can't actually do anything about it, unless voter anonimity is not guaranteed: if you can't prove to an outside party what your real vote was, you can't pursue any legal action, you just know for yourself that the vote was rigged. And if you can prove to an outside party what you voted, that opens up a whole host of other attacks.

So no, this is not even close to an acceptable solution.

I'll also note that the Bush V Gore election issues were not caused by hand counting, but by machine counting as well. So, they should be taken as further proof that simple ballots and manual counts are the right way to conduct an election.

Re: Colorado scrambles to change voting-system passwords after accidental leak

#398

I think all US folks reading this should volunteer at their county's Registrar of Voters (or equivalent agency for their county). Spend one election working at a polling place, and another election working at the RoV HQ. See what it's like to go through the training, and what things are like on Election Day, and in the days leading up (for places that allow early voting, drop-off, etc.).

To what end? Local to me, an “trained” election volunteer was still questioning voter’s citizenship at the polls. I’d say this was a fluke if the GOP hadn’t spent the last umpteen months pushing all this non-citizen voting nonsense. https://wapo.st/3AsIvnf

Except non-citizens have voted. And the Democrats found Virgina over removing confirmed non-citizens from the voter rolls. Why would anyone support keeping illegible voters on the rolls? We all know why.

Re: Colorado scrambles to change voting-system passwords after accidental leak

#399

Earlier quoted context omitted.

Virginia purged 1600 non-citizens from their voter rolls and a Chinese student actually voted in Michigan. Clearly requiring citizenship to register as a voter is not sufficient. Poll volunteers should be verifying citizenship.

One person voting who is not allowed is as bad (in terms of fidelity of the vote to legal voters' intentions) as one person being kept from voting who is allowed. There is copious evidence that these purges, and the atmosphere of fear they create, cause far more harm than they prevent.

Atmosphere of fear? Hypobolic nonsense.

Re: Colorado scrambles to change voting-system passwords after accidental leak

#400
post #305

Earlier quoted context omitted.

Maybe the best answer would be to just create a system where only people who are legally allowed to vote, and those that aren't allowed can't, and the provenance of any given ballot is very clear and secure.

Just make a system that works, why hadn’t anyone else thought of that?

Nobody wants that. The more mayhem, the easier it is to cheat. Same reason the U.S. has such a complicated tax code.
Post reply on HN