4 ms·

The main problem here is that an ordinary voter should be able to check the integrity of a voting system. With paper votes, (which we still have in the UK) an o

by ajb 12y ago

The main problem here is that an ordinary voter should be able to check the integrity of a voting system. With paper votes, (which we still have in the UK) an ordinary voter can do this - they can watch the ballot boxes being carried about, count people voting, and watch the votes being counted. With electronic voting, an ordinary person has to take it on faith. The more clever technology you put in, the less it can solve this problem - the harder it is for an ordinary person to check that the vote is being counted correctly. A better route is the development of voter-verification, such as Rivest's ThreeBallot: https://en.wikipedia.org/wiki/ThreeBallot https://en.wikipedia.org/wiki/ThreeBallot He didn't include any crypto, even though he's one of the world's most famous crypto experts, precisely because that would make it opaque to anyone who isn't a cryptographer.

atmosx 12y ago
This is a bitcoin-based system. The integrity of the keychain can be in danger only if a 51% attack takes place (creates a larger keychain). That's highly unlikely IMHO in a national election system, where any geek could setup a node not to mention political parties, organizations, etc. There is a public ledger. The votes are easily verifiable by anyone, on the fly, all you have to do is a client with a nice GUI for the user. The user doesn't care about the crypto, all he wants to know is that his vote counted and that 40 years later, it's verifiable and private. Bitcoin can do all those things. Actually it solved the most important problems related to e-voting IMHO. ps. The only not-solved problem is selling votes. With this sort of scheme the buyer could request a proof of vote (e.g. a picture, or something more sophisticated like your user/pass). But this is already happening in Greece and I guess elsewhere too. I'm not sure though if an e-voting scheme would allow this to take a much larger scale. Technology can overcome the lack of social/political education, unfortunately.
- kaoD 12y ago
  Why not use a non-bitcoin-based system and completely avoid the 51% attack? https://www.youtube.com/watch?v=ZDnShu5V99s https://www.youtube.com/watch?v=ZDnShu5V99s I'm not sure if this is state of the art[0] since the talk is 7 years old. This scheme provides proof to the voter, but not enough so he can prove his vote to someone else, avoiding vote selling altogether. The biggest challenge IMHO is trust in the system. [0] I'd appreciate criticism or any other progress in the field.
  - atmosx 12y ago
    I've seen the video you posted. It's a 2007 video, otherwise I'm sure the author would have mentioned the Bitcoin protocol time and again, because it solves all the problems mentioned + it's decentralized. Decentralization is a key-aspect here. Votes count equally for all citizens. So every citizen has the opportunity to operate a node for the time (48hours) that the elections will take place, in order to inhibit possible corruption. Given the No of organizations, individuals (e.g. political activists, freedom believes, anarchists, you-name-it, conservatives, liberals, you-name-it) and geeks (who will do it for fun, get an RPi, setup a node for the elections, measure traffic, do a JS/CSS3 front-end to display traffic, real-time results etc.) The government (formed by all parties) should be able to setup an X number of nodes that would allow the elections to take place anyway, without relying to third parties. But third parties will join to ensure reliability of the network. A government could block the access of nodes from IP netmasks that are located outside the country's IP blocks, to avoid interference of foreign governments[1]. Now, I'm not 100% about the interface and the interaction which is equally important: * Should the voters use PKI? Should an account be linked with their IDs? * Should we use a dedicated machine with touch screen and accessibility options in a private chamber? * Should we use electronic signatures generated by a special 'card' which every citizen should have one? * etc Estonia has an electronic card[2] for example that could turn handy into situations like these. There many ways of deployment with pros/cons to be considered on a country by country scenario. Bitcoin protocol solves the double voting, the possibility to re-count votes, the possibility (which is nowhere today) for citizens to review their votes even years later. Also we should not underestimate the ability to vote from the comfort of your home, using any device: Imagine how greatly that would improve the % of people participating in the elections for example. [1] I can't think of what the NSA could come with in this scenario. Would make an interesting discussion. [2] http://en.wikipedia.org/wiki/Estonian_ID_card http://en.wikipedia.org/wiki/Estonian_ID_card
    - nebulous1 12y ago
      I don't see the point in using a Bitcoin-esque protocol for a voting system (ie a hashing based blockchain). In fact the existence of a double spent vote could simply invalidate both (or all) votes; the user attempted to cheat, no need to decide which is valid. Otherwise vote acceptance could be based on the votes of the ids themselves rather than by hashing power. You say decentralization is a key aspect, but the distribution of whatever underlying identity the users are voting with still has to be centralized. Bitcoin has opened people up to these sorts of ideas, but that doesn't mean it has to form the basis for everything in the future.
ntoshev 12y ago
The Wikipedia entry is quite unclear. The paper itself states at this point it's a proposal of academic and not practical interest, since a vote buying attack has been discovered. http://people.csail.mit.edu/rivest/Rivest-TheThreeBallotVotingSystem.pdf http://people.csail.mit.edu/rivest/Rivest-TheThreeBallotVoti...
- ajb 12y ago
  Yes, I should have made it clear that ThreeBallot is not itself a contender. The point is though, that lack of crypto is a desirable property of a voting system.

by ajb 12y ago

atmosx 12y ago

This is a bitcoin-based system. The integrity of the keychain can be in danger only if a 51% attack takes place (creates a larger keychain). That's highly unlikely IMHO in a national election system, where any geek could setup a node not to mention political parties, organizations, etc. There is a public ledger. The votes are easily verifiable by anyone, on the fly, all you have to do is a client with a nice GUI for the user. The user doesn't care about the crypto, all he wants to know is that his vote counted and that 40 years later, it's verifiable and private. Bitcoin can do all those things. Actually it solved the most important problems related to e-voting IMHO. ps. The only not-solved problem is selling votes. With this sort of scheme the buyer could request a proof of vote (e.g. a picture, or something more sophisticated like your user/pass). But this is already happening in Greece and I guess elsewhere too. I'm not sure though if an e-voting scheme would allow this to take a much larger scale. Technology can overcome the lack of social/political education, unfortunately.

kaoD 12y ago
Why not use a non-bitcoin-based system and completely avoid the 51% attack? https://www.youtube.com/watch?v=ZDnShu5V99s https://www.youtube.com/watch?v=ZDnShu5V99s I'm not sure if this is state of the art[0] since the talk is 7 years old. This scheme provides proof to the voter, but not enough so he can prove his vote to someone else, avoiding vote selling altogether. The biggest challenge IMHO is trust in the system. [0] I'd appreciate criticism or any other progress in the field.
- atmosx 12y ago
  I've seen the video you posted. It's a 2007 video, otherwise I'm sure the author would have mentioned the Bitcoin protocol time and again, because it solves all the problems mentioned + it's decentralized. Decentralization is a key-aspect here. Votes count equally for all citizens. So every citizen has the opportunity to operate a node for the time (48hours) that the elections will take place, in order to inhibit possible corruption. Given the No of organizations, individuals (e.g. political activists, freedom believes, anarchists, you-name-it, conservatives, liberals, you-name-it) and geeks (who will do it for fun, get an RPi, setup a node for the elections, measure traffic, do a JS/CSS3 front-end to display traffic, real-time results etc.) The government (formed by all parties) should be able to setup an X number of nodes that would allow the elections to take place anyway, without relying to third parties. But third parties will join to ensure reliability of the network. A government could block the access of nodes from IP netmasks that are located outside the country's IP blocks, to avoid interference of foreign governments[1]. Now, I'm not 100% about the interface and the interaction which is equally important: * Should the voters use PKI? Should an account be linked with their IDs? * Should we use a dedicated machine with touch screen and accessibility options in a private chamber? * Should we use electronic signatures generated by a special 'card' which every citizen should have one? * etc Estonia has an electronic card[2] for example that could turn handy into situations like these. There many ways of deployment with pros/cons to be considered on a country by country scenario. Bitcoin protocol solves the double voting, the possibility to re-count votes, the possibility (which is nowhere today) for citizens to review their votes even years later. Also we should not underestimate the ability to vote from the comfort of your home, using any device: Imagine how greatly that would improve the % of people participating in the elections for example. [1] I can't think of what the NSA could come with in this scenario. Would make an interesting discussion. [2] http://en.wikipedia.org/wiki/Estonian_ID_card http://en.wikipedia.org/wiki/Estonian_ID_card
  - nebulous1 12y ago
    I don't see the point in using a Bitcoin-esque protocol for a voting system (ie a hashing based blockchain). In fact the existence of a double spent vote could simply invalidate both (or all) votes; the user attempted to cheat, no need to decide which is valid. Otherwise vote acceptance could be based on the votes of the ids themselves rather than by hashing power. You say decentralization is a key aspect, but the distribution of whatever underlying identity the users are voting with still has to be centralized. Bitcoin has opened people up to these sorts of ideas, but that doesn't mean it has to form the basis for everything in the future.

ntoshev 12y ago

The Wikipedia entry is quite unclear. The paper itself states at this point it's a proposal of academic and not practical interest, since a vote buying attack has been discovered. http://people.csail.mit.edu/rivest/Rivest-TheThreeBallotVotingSystem.pdf http://people.csail.mit.edu/rivest/Rivest-TheThreeBallotVoti...

ajb 12y ago
Yes, I should have made it clear that ThreeBallot is not itself a contender. The point is though, that lack of crypto is a desirable property of a voting system.