/tech/ - Tech

Technology.

catalog
Mode: Reply
Name
E-mail
Subject
Message

Max message length: 8192

Files

Max file size: 80.00 MB

Max files: 5

Password

(used to delete files and postings)

Misc

Remember to follow the rules


(38.57 KB 700x470 AccuVote TSX.jpeg)
Designing transparent and secure election systems with computers Comrade 07/09/2020 (Thu) 23:00:09 No. 3173
Many countries around the world, after some initial experiments, have completely dumped the idea of running their election systems with computer hardware and returned to hand-counted paper ballots. One look at the cartoonish hodgepodge of election machines with a million security holes across the United States all making use of unauditable proprietary software and hardware and manufactured by private companies mired by a history of corruption and scandals. One look at all that would be enough to give any reasonable person pause to reconsider the entire idea of electronic voting. Is it possible to design an electronic voting/counting system that fulfills some basic expectations of security and transparency? I and many other computer security experts would argue that it is not and never will be due to some fundamental aspects of computers. But let's not let that spoil our fun. How would you design electronic voting systems to be secure and transparent? What would the hardware be like? What would the software be like?
We should probably start by dumping x86 and ARM shit and going with an OpenRISC processor. Simplicity and openness in hardware are key to making audits easy and trustworthy.
I feel like it really shouldn't be that hard with some simple end to end encryption and identification tbh. I just think the deep state is afraid to make things too democratic
>>3175 How do you verify an encrypted ballot while preserving the secret ballot?
>>3176 PGP or GPG maybe?
There is an idea called Three-Ballot for physical ballots with receipts. Maybe the value of 3B is more educational than practical, here it is: You vote yes or no in a referendum (also works for approving or disapproving a candidate in an approval-voting election). You vote with three ballots. If you want to vote yes, you vote yes twice and no once. If you want to vote no, you vote no twice and yes once. The ballots all have IDs, you select a copy of one of the filled out ballots to take home. All the filled out ballots and their IDs get published. You can check the ballot you got a copy of wasn't tempered with. It is not possible to prove whether you voted yes or no in effect, since everybody voted yes and no at least once. If there is some device in the voting station tracking which ballots are getting copied and which are not, this doesn't work. But the whole copying procedure can be very simple and mechanical. So… did reading this inspire you in some way?
Before you start reading textbooks on cryptography and reliability engineering, answer these two questions: 1. In what ways are current pen-and-paper election systems are broken? 2. How would an ideal digital election system fix this issues? I am not convinced that the main issues with elections are related to the tools used for voting.
In fact, I wouldn't be surprised if electronic voting was originally invented as a deliberate distraction from real issues.
(806.92 KB 500x281 puff puff.png)
>>3192 >>3193 >Is it possible to design an electronic voting/counting system that fulfills some basic expectations of security and transparency? I and many other computer security experts would argue that it is not and never will be due to some fundamental aspects of computers. But let's not let that spoil our fun. Now think up how you would make the most secure and accountable computer system you can.
Yes. Use Ethereum smart contracts. Public, auditable & immutable, while still having mechanisms for privacy. Look into projects like Kleros or Daostack for examples of collective organisation or voting mechanics.
>>3192 It's better for direct democracy. Its simply not feasible to have tons & tons of properly organised paper votes for things like workers councils making small decisions. Some electronic system could offer a lot of benefits. This was an interesting read: http://mason.gmu.edu/~rhanson/futarchy.html It's a little libertarian, but the ideas are interesting.
>>3177 This work work fine too. A pseudonymous but verifiable cryptographic signature is a solved problem. Look into zero knowledge proofs as well: https://research.kudelskisecurity.com/2018/11/05/e-voting-crypto-protocols/
>>3197 >Its simply not feasible to have tons & tons of properly organised paper votes for things like workers councils making small decisions. Why not? It's being done right now.
The fundamental issue isn't that its an unsolveable problem, its that you still have to trust the organisation running the election more than you do for a paper ballot.
>>3174 >copyleft openrisc >not permissively-licensed risc-v meme ISA giga based
>>3192 >1. In what ways are current pen-and-paper election systems are broken? Insecure, unreliable, centralized, slow, expensive, inflexible. >2. How would an ideal digital election system fix this issues? Public-private keys are a simple centralized solution, blockchains are a more elaborate decentralized solution. >>3219 >its that you still have to trust the organisation running the election more than you do for a paper ballot Not true. Crypto systems can be completely decentralized.

Delete
Report

no cookies?