US Voting Machines Standards Open To Public

Online Voting writes "The U.S. Election Assistance Commission has published new voting systems testing and certification standards for 190 days of public comment. For all the critics of electronic voting, this is your opportunity to improve the process. This will be the second version of the federal voting system standards (the first version is the VVSG 05). To learn more about these Voluntary Voting System Standards see this FAQ."

Comments

How about

[SamP2]

- Printed voting receipt
- All code open source, all architecture fully documented and publicly available
- No person-vote information recorded in database (database lists people as "voted" or "not voted", as soon as person enters a vote it changes to "voted" and won't allow another vote, while a separate database increments a counter for a particular candidate. These two databases are NOT linked together.
- No timestamps to ensure manual matchmaking between people and votes are not possible.

Ah hell. I could come up with lots of other reasonable suggestions, but its not like any of this will ever be implemented.

Re:How about

[Conspiracy_Of_Doves]

Yes, print the voting receipt, but don't let the person take it with them. They can see it in the machine to verify that was who they voted for, but it stays in the polling place in case a manual recount is needed.

Re:How about

[megaditto]

Receipt is a great idea.
For one, you could get a discount on your union dues with a Democrat on your voter receipt.
Or you could use it to secure your job (since your boss won't fire you if he can see you voted Republican).
Or you could sell it to the highest bidder: exchage your Billary/Osama receipt for a $20 gift card (for example). Buying votes otherwise is a real pain: people take your money but can still vote for the other guy if you don't watch them.

Re:How about

[mithras invictus]

No, the receipt should be the ballot, not the other way around. One machine is meant to help the voter produce a human and machine readable vote, the voter can check the produced ballot unassisted and decide whether or not to submit it.

I certainly much better now!

[e9th]

From the EAC's FAQ:

Q: Will the source code be available to the public? A: No.

big problem

[ILuvRamen]

Has anyone else noticed that more money and time and effort has been spent trying to make and use good, fair, electronic voting machines than it would have taken to just keep using paper ballots and have them counted like usual? Isn't the point to save money and time and make it more efficient? I think another point was to make elections less riggable and more accurate but Diebold killed that dumb idea behind a long time ago lol.

Re:big problem

[mOdQuArK!]

The main advantages of using voting machines is that they can be used to print out a nice, clean ballot which can be easily counted (no misaligned filling-out of ovals or odd marks, don't worry about #2 pencils or color of pens, no hanging chads, the ballot contains only the selected choices so no "they really meant this choice!" type of counting, etc).

They're also good at providing alternative interfaces for the disabled (sound or braille) while still printing out a nice, clean ballot.

The only reason for COUNTING machines is for speed though, and since there's no easy way to make sure the counting machines haven't been compromised, we shouldn't depend on them at ALL except maybe for "preliminary results". For the final official result, we should still stick to the hand counting votes (especially if we have nice, clean, easily-readable ballots).

Re:big problem

[zcat_NZ]

You missed another advantage. Since the printed ballot is in a consistent (and preferably standard) format, those votes can be optically counted by a tallying machine built by a completely different vendor. If the preliminary count and independent OCR count agree within some agreed margin (we'll allow for misreading a vote or two per million, OCR isn't perfect). Then we can have a final, trustworthy election result within minutes of the closing of the polls. Accurate, trustworthy, _and_ fast. Wouldn't that be nice!

Re:big problem

[bVork]

Sounds like the problem is with your country's implementation of paper ballots, and not the general idea itself. Here in Canada, voting takes maybe half an hour at most. You show up, verify your identity, get your ballot, go behind a screen and put an X in the circle next to the candidate, fold it up, hand it to the person working the box, watch them place the ballot in the box, go home.

To supervise the whole thing, we require people from multiple parties to be present at the polling station. It's hard to fiddle with something when it has to be verified by two (or more) opposing people at the same time.

I don't understand your references to multiple ballots. Is each party on a separate ballot or something? Why in the world would it be done like that?

I for one....

[edwardpickman]

prefer our Diebold Overlords. It takes all the guesswork out of the voting process. There's something comforting knowing the outcome of an election months before the day.

What a bunch of crap

[rastoboy29]

Too bad neither of the "major" political parties has the country's interests at heart, or we would have real, open standards for the machines themselves, and not just a voluntary fucking testing process.

Problems, not solutions

[michaelmalak]

You've violated the golden rule of specifying requirements:

- Printed voting receipt

The requirement is:

1. Individual vote verifiable by the voter's unassisted eye at the time of voting as to the vote selection and whether or not it has been tampered.
2. All votes verifiable by auditors' unassisted eyes after voting is complete as to the vote selections and whether or not they have been tampered.

Re:Sweet

[thatskinnyguy]

Several generations of my family have worked for Diebold. They're a fixture in the community of Canton, Ohio. They're really good at physical security. Hell. They make most of the bank vaults and ATMs that you see.

But when it comes to voting machines, the only thing that separates the voting machines from their other products is strong bias. Tamper with an ATM at the factory, sure some FDIC bank will lose a few thousand dollars but the one doing the tampering gains nothing. Tampering with a voting machine, the perpetrator stands to influence an election in ways they see fit.

Vote counting research

[slashqwerty]

Dear grahamsz,

In response to your question, "Is there really a concern that some competing software vendor will copy their 'tally up the votes' routine", we here at Diebold take great pride in the quality of our product. Our "tally up the votes"TM routine is a prized trade secret developed through extensive research and experimentation. If our competitors could simply copy our unique technique for counting votes they could develop the same product without incurring the significant costs of researching how to count.

I'm sure you can appreciate the sensitive technical know-how at the core of our product. Only a few vendors have discovered the secret to counting votes. If this knowledge became public anyone could count see how we count votes which would take away our incentive to create a much valued product which serves to protect democracy.

God Bless America,
Tom Swidarski
CEO of Diebold, Inc.

Software independence is required.

[zestyping]

For those of you who have wanted voter-verifiable paper records, the new VVSG says:

Software independence means that an undetected error or fault in the voting system's software is not capable of causing an undetectable change in election results. All voting systems must be software independent in order to conform to the VVSG.

See section 2.4 for a discussion of "software independence." The draft guidelines present "independent voter-verifiable records" (IVVR) as one method of achieving "software independence," though it leaves the door open for other innovative ways of achieving the same goal (such as end-to-end cryptographic verification).

I definitely recommend reading the guidelines. There's a lot of stuff in there.