if there is a paper trail, who is to say that someone can't tamper with that?
Of course it's possible to tamper with a paper trail... anything is possible given enough time and effort. The point is that it's much more difficult to do, especially if you don't want to leave evidence of tampering. With a computer you can make any changes you want in a matter of seconds... with a paper trail, it would take you minutes or hours to surreptitiously alter a significant number of votes.
It's not relevant whether something is possible... what's relevant is whether it's practical.
I didn't have anything to verify what lit up on the panel was what was going to be tallied by the election commission.
... and because of that, not only do you have no way of knowing whether your vote was counted, your local electoral officials have no way of knowing either. Without a paper trail, all they can do is say "well, the computer said such-and-such". What if the computer is wrong/lying? Lying is easy for a computer, it's much harder for paper to do.
People buy votes already. You can look at almost any election and you'll see that the party that spends more, wins. It's not a perfect correlation but it's very close.
Um, no. Buying a vote is when you give somebody money in return for their vote. Advertising is not buying votes, because you are not giving money to people in exchange for their votes.
And how about voting by mail? Anybody can watch you vote and pay you for what you do. Or eliminate the middleman and just hand over your signed, blank ballot for cash, to let them fill in and mail for you. Did you know that over 45% of California's votes this election will be by mail? Did you know that 100% of Oregon's votes are by mail? How could that be legal, if as you say the Constitution requires perfect anonymity?
That's a very good point. Maybe having 100% vote-by-mail is a bad idea. I certainly wouldn't be happy if I didn't have the option of voting in a polling center where my privacy could be guaranteed.
It's time to recognize that vote selling is not something that can be forbidden by technological means.
True, we can't stop people from giving people money in return for a promise to vote one way or another. But what we can do is make sure that the vote-buyer can't verify that his money was well-spent. As long as there is no way for the vote-seller to prove to the vote-buyer that he actually did vote the way the vote-buyer wanted him to, there is little incentive to bribe voters. And that's a good thing.
Eliminating this boogeyman from the discussion will allow voting systems to be created that are far simpler, easier to use, and more reliable than anything we have had so far.
Sorry, this boogeyman is real. Human nature being what it is, we need to protect voters from both bribery and extortion, or we have no hope of obtaining honest results.
I don't understand why an open voting system wouldn't work.
Because it suffers from the same problems that a close system suffers from: it's very difficult to prove that a voting machine is working correctly and has not been tampered with. Showing people pages of well-written source code might make them feel better, but they can only verify the logic that they are reading, not the logic that the machine is actually using.
I'm all for open source everything, but it's not going to fix this problem. Even with lots of fancy checksums and hashcodes, you still end up basically taking the computer's own word for it that it's doing what you expected it to do.
Jeremi: "Hey computers. Are you all running the official, many-eyeballed correct voting software?" Good Computer: "I sure am! See, hashcode 123456789 checks out". Evil Hacked Computer: "I sure am! See, hashcode 123456789 checks out".
[...]but rather, "Will this program generate so much interest that it might actually motivate someonen to subcribe to our channel?".
You forgot the other half of the equation: "Will this program (and our other programs) generate enough interest to keep our current subscribers from cancelling?"
Attracting customers is one thing... keeping them is also important.
Yeah, I can - I can point to his creating an automated script that let anyone create a boarding pass with a false name on it
Not good enough. Please point to somebody who actually created such a boarding pass and tried to use it to gain unauthorized access to an airport's secure areas.
Every screaming liberal here on Slashdot treats random allegations about Diebold to be indubitably true, unquestionably the reason for their 2004 disappointment
Hi there. I'm a liberal, occasionally screaming, and I don't think that. I do think, however, that the fact that it's impossible for anyone to demonstrate otherwise is a disgrace. How can any election ever be trusted if we can't verify that the results weren't tampered with? If you don't think it's a problem, I imagine the issue will be made clearer for you the next time your guy unexpectedly loses an election you thought he was going to win.
But here's an allegation that a devout socialist and an avowed enemy of the United States might have an ownership interest in a compnay that makes electronic voting machines, and not only do some folks dispute the allegation off the bat, but many or most of the same screaming liberals actually cheer for the idea of stolen elections as long as the "right" folks win.
I just read the entire list of comments and didn't see anyone posting anything that said that. Granted, I only read the +2 comments and above, so maybe I missed some trolls. Perhaps you can point to some of these posts of which you speak?
On the other hand, maybe you are just so convinced of the hypocrisy of those damn libruls that you see what you expect to see no matter what?
All your post proves is that the Diebold CEO is a right winger. Bet he's never written a line of code in his life.
Does it really matter what Diebold's politics happen to be at the moment? The point is that we shouldn't have to be having this discussion. Our electoral system should be set up in such a way that its integrity shouldn't depend on the honesty of any one company or individual. If things were done correctly, then the voting machine companies could be owned/run by Lucifer himself and we'd still be guaranteed free and fair elections.
what matters is whether you conspired to break that security.
Okay... but it appears he only "conspired" to point out that the security was already broken. It's not like he actually took a counterfeit boarding pass into an airport, or even encouraged anyone else to do so.
When the kid yells out "the emperor has no clothes", do you arrest the kid for public nudity? Or do you find the emperor something to wear?
No, they are saying he's lying by presenting a fake boarding pass to TSA agents
... well, he didn't do that...
or making it easy for other people to do so.
I think part of his point is that it was already easy for other people to do so. Not that pointing out the obvious will probably help him much from his cell in Gitmo...
He did create a giant national wildlife preserve in most of American Pacific Island territory to protect thousands of square miles of coral reef and thousands of rare fish species. [...] That's notable and should appease to a few liberals. Just a few, like the tree huggers.
Unless I missed something in biology class, there aren't very many trees that grow in coral reefs... so while the fish kissers might like it, I doubt the tree huggers will be terribly impressed.
In other news,/.ers continue with wierd offf the wall fantasys..
No, no, no, you're doing it all wrong. A curmudgeonly rant must always start with "If man were meant to $(ACTIVITY), God would have given him $(POWER)". Then you should follow that up by complaining about some perceived shortcoming of "kids these days", and conclude with a demand that they get off your lawn.
Simple checksums are subject to forgery, yes. But to this date there are no known collisions of SHA-1 checksums, and as far as I know SHA-256 doesn't even have such a weakness.
I think you are missing my point. It doesn't matter how secure your checksum algorithm is. Even if your checksum algorithm is absolutely perfect, what is there to stop Mr. Evil Hacker from overwriting the machine's checksum routing with his own, that looks like this:
for (int i=0; i2000000000; i++) {/* look busy for a second */}// The correct checksum value, generated from the correct firmware by Mr. Evil
const char * CORRECT_CHECKSUM_FOR_V4.63c = "23939867201032945234";
printf("The checksum is %s, checksum matches! Nothing to see here...\n", CORRECT_CHECKSUM_FOR_V4.63c); }
You are exceedingly naive if you think reality is the same thing as the rules.
I never claimed that the non-elecronic solution was 100% foolproof... only that the electronic solution suffers from additional vulnerabilities that the current solution does not.
We can run off any "independent observers" and do what we like with the ballot boxes. We can even pretend to be representatives of the major parties. We were all appointed by the local board of elections, whose members were all hand-selected for their willingness to "go with the program" by the local political boss.
Of course. If you are willing and able to break the law with impunity, you can do all of those things and more. But at least people will know that you broke the law ("running off independent observers" is a violation of the law), and with any luck the ensuing publicity will land you in jail (or more likely, the high likelihood of being caught will dissuade you from trying such shennanigans in the first place).
I don't think any voting system could be made to work properly in cases where all of the election workers and the local government are dishonest. So one has to assume that in most cases, at least some of the pollworkers are honest, or give up on democracy altogether. At least with the current system, cheating requires the subversion of the entire group of poll workers. That's better than a system where a single individual could botch the results without anyone else's help and without any creating any evidence of wrongdoing.
not really since you can (and should in this case) have say 20 different machines and then pick three machines ( compare md5 sums) and then pick another 3 machines (compare sha hashes) then compare to your target machine. if you do this right then the chances of getting this wrong are about state lotto grade
But what hardware and software are you using to compute these checksums? If you are running the checksum computation on the (potentially) compromised machines, then the checksum computation routine itself is suspect. It wouldn't matter (AFAICT) how many machines you sampled, or how many different algorithms you used, you'd still have the same problem: a compromised machine could be programmed to act the same as the uncompromised machines, so there would be no (practical) way for you tell the difference.
What you'd really need is some sort of "known good" machine that could run the checksum computation, but even then it would need to first acquire the checksum function's input data from.... the suspect machine. Which of course could be programmed to return the bits of the "correct" binaries instead of the actual, compromised code it was running. Maybe there is a foolproof solution to this, but I don't see it.
AMD is pushing multitasking, a model of parallel processing that will never do desktop users much good beyond a small handful of processors.
I'll bet you a beer that in 10 or 15 years you'll look back on the above statement and admit that you were completely wrong. You may be right that current apps, and even current types of apps, will receive limited benefit from dozens of processors, but what you're missing is that massive parallelism will enable new types of application that are barely imagined now.
Perhaps you remember the famous (apocyphal?) quote, "640K of RAM ought to be enough for anybody". I suspect that the author of that statement thought that because all apps at the time ran in 80x40 monochrome text mode, and what text-mode app could possibly need so much RAM? He didn't forsee the migration to GUI-based apps that was made practical by the availability of large amounts of RAM.
Until a large, dispersed group of people break into a large number of these machines and rig the elections so that "Homer J Simpson" is the presedential victor in multiple states, we aren't going to see the government persue a real alternative to these proprietary magic voting machines.
If you really want to stir up a hornet's nest, hack the machines so that write-in candidate "Osama B. Laden" wins a surprise victory in several states. That oughtta keep the pundits talking for a while...;^)
See, the problem with your antiquated Canadian 'system' is that you often end up electing officials based on the will of the voters. That's no way to run the most powerful country in the world -- leave it up to those yahoos? They'd probably just go and do something stupid, like electing the wrong guy. Who would save us from terrists, gays, and health care then? Hmmm? Hmmm??
The problem in American is, everybody is so sure that we're the best democracy ever, that nobody bothers to check to see if that's the really case. People are able to overlook a lot, if seeing it would mean seeing their beloved country in a less-than-positive light.
Exactly. You would make sure that a clean version of the code "leaks", which shows no evidence of any tampering whatsoever.
You're assuming that (a) lots of people will review the leaked code, and (b) not finding any problems in it will allay peoples' suspicions. While I could certainly see faulty code being treated as a 'smoking gun', I doubt it would work the other way... if nothing is found, the effect will be neutral at best.
That may be a solution, release the code after the election, and run checksums on all the meachines.
Anybody tampering with the software could just tamper with the checksum-calculating routine to make sure it returns the expected "correct" checksum, no?
So how is this any different from a traditional low tech ballot box? If you allow someone unrestricted and unsupervized access to a box full of ballots its security breaks pretty fast too.
You're right, and that's why nobody has "unrestricted and unsupervized access" to the ballot box once it contains ballots. It is kept locked and in full public view during the election, and the ballots are carefully supervised (by at least two poll workers, usually more) at all times afterwards.
The difference is, with the Diebold-style systems the "ballot box" is also a security hazard when it's empty. If you want an analogy, you'd have to imagine a ballot box that could be programmed before the election to create or destroy ballots during the election.... a device that would not be easy to implement in plastic;^)
What is "animal style"? Is that better than "the works"?
"Animal style" means you have to hide in tall grass, sneak up on the cow, tear its throat out, then use your teeth to pull steaming hunks of raw meat off of its still-warm carcass. It is indeed better than 'the works', if you're into that sort of thing.
I don't believe there are any non extreme ways of getting rid of the damn thing. It has its little claws dug in deep and you have to bash it repeatedly on its ugly little head with a crowbar before it finally lets go (spewing gore everywhere). [...] Reminded me of the headcrabs in HL2.
Are you referring to the SpamThru trojan, the Symantec Anti-Virus Suite, or Windows itself? It's hard to tell....
Slashdot Mirror
Of course it's possible to tamper with a paper trail... anything is possible given enough time and effort. The point is that it's much more difficult to do, especially if you don't want to leave evidence of tampering. With a computer you can make any changes you want in a matter of seconds... with a paper trail, it would take you minutes or hours to surreptitiously alter a significant number of votes.
It's not relevant whether something is possible... what's relevant is whether it's practical.
I didn't have anything to verify what lit up on the panel was what was going to be tallied by the election commission.
Um, no. Buying a vote is when you give somebody money in return for their vote. Advertising is not buying votes, because you are not giving money to people in exchange for their votes.
And how about voting by mail? Anybody can watch you vote and pay you for what you do. Or eliminate the middleman and just hand over your signed, blank ballot for cash, to let them fill in and mail for you. Did you know that over 45% of California's votes this election will be by mail? Did you know that 100% of Oregon's votes are by mail? How could that be legal, if as you say the Constitution requires perfect anonymity?
That's a very good point. Maybe having 100% vote-by-mail is a bad idea.
I certainly wouldn't be happy if I didn't have the option of voting in a polling center where my privacy could be guaranteed.
It's time to recognize that vote selling is not something that can be forbidden by technological means.
True, we can't stop people from giving people money in return for a promise to vote one way or another. But what we can do is make sure that the vote-buyer can't verify that his money was well-spent. As long as there is no way for the vote-seller to prove to the vote-buyer that he actually did vote the way the vote-buyer wanted him to, there is little incentive to bribe voters. And that's a good thing.
Eliminating this boogeyman from the discussion will allow voting systems to be created that are far simpler, easier to use, and more reliable than anything we have had so far.
Sorry, this boogeyman is real. Human nature being what it is, we need to protect voters from both bribery and extortion, or we have no hope of obtaining honest results.
Because it suffers from the same problems that a close system suffers from: it's very difficult to prove that a voting machine is working correctly and has not been tampered with. Showing people pages of well-written source code might make them feel better, but they can only verify the logic that they are reading, not the logic that the machine is actually using.
I'm all for open source everything, but it's not going to fix this problem. Even with lots of fancy checksums and hashcodes, you still end up basically taking the computer's own word for it that it's doing what you expected it to do.
Jeremi: "Hey computers. Are you all running the official, many-eyeballed correct voting software?"
Good Computer: "I sure am! See, hashcode 123456789 checks out".
Evil Hacked Computer: "I sure am! See, hashcode 123456789 checks out".
You forgot the other half of the equation: "Will this program (and our other programs) generate enough interest to keep our current subscribers from cancelling?"
Attracting customers is one thing... keeping them is also important.
Not good enough. Please point to somebody who actually created such a boarding pass and tried to use it to gain unauthorized access to an airport's secure areas.
Hi there. I'm a liberal, occasionally screaming, and I don't think that. I do think, however, that the fact that it's impossible for anyone to demonstrate otherwise is a disgrace. How can any election ever be trusted if we can't verify that the results weren't tampered with? If you don't think it's a problem, I imagine the issue will be made clearer for you the next time your guy unexpectedly loses an election you thought he was going to win.
But here's an allegation that a devout socialist and an avowed enemy of the United States might have an ownership interest in a compnay that makes electronic voting machines, and not only do some folks dispute the allegation off the bat, but many or most
of the same screaming liberals actually cheer for the idea of stolen elections as long as the "right" folks win.
I just read the entire list of comments and didn't see anyone posting anything that said that. Granted, I only read the +2 comments and above, so maybe I missed some trolls. Perhaps you can point to some of these posts of which you speak?
On the other hand, maybe you are just so convinced of the hypocrisy of those damn libruls that you see what you expect to see no matter what?
Does it really matter what Diebold's politics happen to be at the moment? The point is that we shouldn't have to be having this discussion. Our electoral system should be set up in such a way that its integrity shouldn't depend on the honesty of any one company or individual. If things were done correctly, then the voting machine companies could be owned/run by Lucifer himself and we'd still be guaranteed free and fair elections.
An interesting analogy... can you point to any of these 'moths'? Or is releasing imaginary moths a crime now?
Okay... but it appears he only "conspired" to point out that the security was already broken. It's not like he actually took a counterfeit boarding pass into an airport, or even encouraged anyone else to do so.
When the kid yells out "the emperor has no clothes", do you arrest the kid for public nudity? Or do you find the emperor something to wear?
or making it easy for other people to do so.
I think part of his point is that it was already easy for other people to do so. Not that pointing out the obvious will probably help him much from his cell in Gitmo...
Unless I missed something in biology class, there aren't very many trees that grow in coral reefs... so while the fish kissers might like it, I doubt the tree huggers will be terribly impressed.
No, no, no, you're doing it all wrong. A curmudgeonly rant must always start with "If man were meant to $(ACTIVITY), God would have given him $(POWER)". Then you should follow that up by complaining about some perceived shortcoming of "kids these days", and conclude with a demand that they get off your lawn.
I think you are missing my point. It doesn't matter how secure your checksum algorithm is. Even if your checksum algorithm is absolutely perfect, what is there to stop Mr. Evil Hacker from overwriting the machine's checksum routing with his own, that looks like this:
ChecksumValue CalculateAndPrintSuperSecureSHAChecksumForVotingS
{
printf("Calculating checksum for voting software, please wait....\n");
for (int i=0; i2000000000; i++) {/* look busy for a second */}
const char * CORRECT_CHECKSUM_FOR_V4.63c = "23939867201032945234";
printf("The checksum is %s, checksum matches! Nothing to see here...\n", CORRECT_CHECKSUM_FOR_V4.63c);
}
I never claimed that the non-elecronic solution was 100% foolproof... only that the electronic solution suffers from additional vulnerabilities that the current solution does not.
We can run off any "independent observers" and do what we like with the ballot boxes. We can even pretend to be representatives of the major parties. We were all appointed by the local board of elections, whose members were all hand-selected for their willingness to "go with the program" by the local political boss.
Of course. If you are willing and able to break the law with impunity, you can do all of those things and more. But at least people will know that you broke the law ("running off independent observers" is a violation of the law), and with any luck the ensuing publicity will land you in jail (or more likely, the high likelihood of being caught will dissuade you from trying such shennanigans in the first place).
I don't think any voting system could be made to work properly in cases where all of the election workers and the local government are dishonest. So one has to assume that in most cases, at least some of the pollworkers are honest, or give up on democracy altogether. At least with the current system, cheating requires the subversion of the entire group of poll workers. That's better than a system where a single individual could botch the results without anyone else's help and without any creating any evidence of wrongdoing.
But what hardware and software are you using to compute these checksums? If you are running the checksum computation on the (potentially) compromised machines, then the checksum computation routine itself is suspect. It wouldn't matter (AFAICT) how many machines you sampled, or how many different algorithms you used, you'd still have the same problem: a compromised machine could be programmed to act the same as the uncompromised machines, so there would be no (practical) way for you tell the difference.
What you'd really need is some sort of "known good" machine that could run the checksum computation, but even then it would need to first acquire the checksum function's input data from.... the suspect machine. Which of course could be programmed to return the bits of the "correct" binaries instead of the actual, compromised code it was running. Maybe there is a foolproof solution to this, but I don't see it.
I'll bet you a beer that in 10 or 15 years you'll look back on the above statement and admit that you were completely wrong. You may be right that current apps, and even current types of apps, will receive limited benefit from dozens of processors, but what you're missing is that massive parallelism will enable new types of application that are barely imagined now.
Perhaps you remember the famous (apocyphal?) quote, "640K of RAM ought to be enough for anybody". I suspect that the author of that statement thought that because all apps at the time ran in 80x40 monochrome text mode, and what text-mode app could possibly need so much RAM? He didn't forsee the migration to GUI-based apps that was made practical by the availability of large amounts of RAM.
If you really want to stir up a hornet's nest, hack the machines so that write-in candidate "Osama B. Laden" wins a surprise victory in several states. That oughtta keep the pundits talking for a while...
See, the problem with your antiquated Canadian 'system' is that you often end up electing officials based on the will of the voters. That's no way to run the most powerful country in the world -- leave it up to those yahoos? They'd probably just go and do something stupid, like electing the wrong guy. Who would save us from terrists, gays, and health care then? Hmmm? Hmmm??
The problem in American is, everybody is so sure that we're the best democracy ever, that nobody bothers to check to see if that's the really case. People are able to overlook a lot, if seeing it would mean seeing their beloved country in a less-than-positive light.
You're assuming that (a) lots of people will review the leaked code, and (b) not finding any problems in it will allay peoples' suspicions. While I could certainly see faulty code being treated as a 'smoking gun', I doubt it would work the other way... if nothing is found, the effect will be neutral at best.
Anybody tampering with the software could just tamper with the checksum-calculating routine to make sure it returns the expected "correct" checksum, no?
Woohoo! And so dawns a bright new day for GNU/America!
You're right, and that's why nobody has "unrestricted and unsupervized access" to the ballot box once it contains ballots. It is kept locked and in full public view during the election, and the ballots are carefully supervised (by at least two poll workers, usually more) at all times afterwards.
The difference is, with the Diebold-style systems the "ballot box" is also a security hazard when it's empty. If you want an analogy, you'd have to imagine a ballot box that could be programmed before the election to create or destroy ballots during the election.... a device that would not be easy to implement in plastic
"Animal style" means you have to hide in tall grass, sneak up on the cow, tear its throat out, then use your teeth to pull steaming hunks of raw meat off of its still-warm carcass. It is indeed better than 'the works', if you're into that sort of thing.
Are you referring to the SpamThru trojan, the Symantec Anti-Virus Suite, or Windows itself? It's hard to tell....