Black-Box voting hits slashdot

[Augspies]

See what the geeks have to say.

http://slashdot.org/articles/03/07/08/1949200.shtml?tid=103&tid=126&tid=99


Any computer data can be quickly and easily changed. The best solution I can think of is to print out two paper receipts for each vote, one to go to the election commission (for manual recounts) and one to go to the voter. Each receipt would contain a random code which the voter could then type in on a web site to verify their choices have not been changed. Of course, most people wouldn't bother to verify, but it only takes one person to catch vote fraud.

[htuttle]

1. I had that same idea awhile back

It seemed to be a good compromise between security, verification and privacy.

[Maudlin]

4. Wasn't that the problem though?

The Diebold machines registered all the votes correctly, but then it kept a second set of books that was used to add up the totals. It was this second book that was reporting wrong information, while a spot check would use the first book and come up correct.

The idea above doesn't seem to help with that.

[Augspies]

6. Right, but if the second book had the same code

and a different vote then you can check against the first paper.

This of course requires a paper trail. Without it nothing works.

Jeremy

[DEMActivist]

14. Ahhhh, but here's the thing....

When you use GEMS to "spot check" the first set of books, the first thing you have to do is "Reset the Election" to Recount mode, resetting the database files.

[jimmynochad]

2. How much do I have to pay you to sit over your shoulder

and watch you confirm your vote? Corzine spent over $10 per registered voter to get elected. He could have saved his money by paying only those that voted for him and proved it with their receipts.

I love receipts, just have to engineer around the voter being allowed to touch it.

[haymaker]

3. You're point being?

Jimmynochad? What are you saying?

[jimmynochad]

5. The two receipt concept

You will always be shot down with the two receipt concept. If the voter can confirm their vote with a paper proof. Then as a candidate I can offer you money for that proof prior to the election. The highest bidder wins my vote.

You have to follow the "Mercuri Method" where the voter can review the paper under glass and then it drops into the box. Then the paper is randomly inspected against the electronic record. If the records don't jive, then recount the whole election. If paper records don't jive with electronic records, open up code inside each machine, compare to escrowed code, all in the open.

If Bev is right then the precinct totals could match but the overall totals could be forged so I would suggest having all precinct totals posted at each polling place prior to any consolidation.

[Capn Sunshine]

7. ITS ALSO FRONT PAGE ON "WHAT REALLY HAPPENED"

http://www.whatreallyhappened.com/

!

[ex_jew]

8. Statistical voting is the only answer

All of the many schemes proposed to improve the situation run up against the great if not insuperable difficulties involved in counting 100 million ballots. Why not learn from our experience with the jury system ? We don't make everyone cast a vote on someone's guilt or innocence. Instead, we select citizens at random and put them in charge of studying the facts and arriving at a conclusion. We would easily do the same thing with electoral processes, by randomly selecting voters from the pool of registered voters, publicizing their names, and allowing only them to vote, possibly by absentee means. It's just so much easier to verify the processing of 1000 or 2000 votes than 10 million. There's a huge savings because no electronic machines are needed, and there's a likelihood that participation rates will be very high.

While I'd like to believe that Bev Harris etc. are performing a useul servive, asking the populace to evaluate the reliability of particular computer programs is completely pointless ! We need a simple system that anyone can understand

[bitchkitty]

9. That's not a good idea -

at all. One person, one vote. Not 1 million people, one vote.

[ex_jew]

11. Well, if you like the present unverifiable mess...

then you're right. I'd prefer to learn from the difficulties we've faced rather than hiding behind worn-out platitudes. But I suspect I'm in the minority. Perhaps we should ask EVERY DAMN PERSON in America if that's true. Or else we could use a statistical sample. Which would you recommend ?

One man one vote was a nice slogan when some people were prevented from voting. That's not the problem now. Instead our problem is that our vaunted electronic technology has a VERY HARD TIME dealing with millions and millions of tiny pieces of data. We could get EXACTLY the same results with a randomly-selected sample, instead of heading down the rathole of manually inspecting 40,000 pieces of code for one voting machine manufacturer. How do we even know that we're looking at the code that's running.

[htuttle]

12. Oh bull...

A verifiable, accountable, secure voting system wouldn't be a whole lot more complex than a bank's accounting system. You wouldn't recommend using a 'sampling' technique to track your bank balance, would you?

The issue I have with the current electronic voting systems is that they are a magnitude LESS secure than any bank (or even any honest business) would allow for their accounting system.

[ex_jew]

13. Not more complex than a bank's accounting system ?

You make it sound like a very simple matter. I happen to have worked on bank accounting systems and I know that the effort to guard against fraud and error in all the places where they might lurk is no place for amateurs.

Here's a for-instance. I recently went on-site a regional bank you've probably never heard of called SouthTrust in Birmingham, Alabama. The product I was working on involved scanning every executable on the bank's mainframe, so I happen to know that there were over 1 MILLION of them. Are you going to blandly assert that nowhere in this group of one million programs could someone have "tweaked" things ? If so, your degree of trust in technology goes way beyond mine.

One little thing that makes bank accounting childish simple compared to counting the votes is the principle of anonymity, another holdover from particular historical problems. And before you decide that low-paid officials with a potential interest in fudging the numbers could be prevented from doing so once and all, tell me how you plan to secure absenteee ballots? All this talk about voting machines COMPLETELY IGNORES the fact that absentee balloting is growin by leaps and bounds.

[jimmynochad]

15. About Absentee voting - some details and stuff

Yes there are places that have relaxed absentee voting rules to get the vote counts up but one other force on the side of voting machine companies is to increase "early voting". Sometimes this is also refered as absentee voting or no-fault voting. It means voting on a touch screen up to three weeks before an election.

Imagine voting 15 days before an election and then find out that your favorite candidate was found guilty of some crime 7 days before the election!

Also, some states like Vermont and New Hampshire and some mid-west states are already trying to figure out how to get the HAVA money without buying touch screens. This is in a backlash against the technology and also the hidden costs of touch screens. My favorite cost story is how a county in CA was offered Diebold machines at $3500 each but when the state reviewed the contract, the true cost was well over $7000! One item not included in the price of the machine? The power cord! Sheesh.

The anonymity you speak of is actually very crucial in this whole discussion. If a secret vote was not a requirement, then we probably would not know who Bev Harris is.

[bitchkitty]

16. One person, one vote

is not a worn out platitude. It's the only reason I care about politics.

[Robb]

10. Darn those slashdotters

You know, I submitted the story as soon as I got Bev's mail.

They rejected it! Peer-reviewed, my fanny... :)

(On edit: Should shut mah mouth and consider myself one of the "many others" that submitted the story...)