Slashdot Mirror
NY Times Endorses Open-Source Election Software
jdauerbach writes "On its editorial page today, the New York Times called for election system reform, saying among other things that 'Congress should impose much more rigorous safeguards, including a requirement that all computer code be made public. It should require that all electronic machines produce a voter-verified paper trail.'"
While I don't disagree in the least with the spirit of the concept of making the system(s) open source, it should be noted that, contrary to popular belief, Diebold asserts that its systems have been scrutinized, including at a source code level, by independent authorities, and that there is also a paper record:
http://www.securityfocus.com/archive/1/375954
I don't know if the paper record is "voter verified", or what mechanism it uses, but there is apparently a paper record nonetheless.
Notwithstanding Diebold's CEO's extremely inappropriate campaign comments, I really do think they're trying to put out the best electronic voting systems they can, but are suffering from the same problems that any large, proprietary system suffers from when it languishes in the comfort of large government-guaranteed long-term contracts: namely, inattention to the details that need to be addressed, that sometimes get lost in not seeing the forest for the trees.
Perhaps opening the source to these critical systems and having it overseen by an independent election agency would be an idea worth considering...
How do we know that the code that is actually on the machines we're voting with is the same as the public code? Even if the public code is compiled and built, then tested to see if it's the same binary instructions as what's going on the mass-produced machines, how do we know that each, individual machine that actually ends up at the voting booth won't be rigged? Who's to say that some dishonest, partisan fuck won't change it at the last minute?
I think Badnarik's solution is the best. Get rid of the official ballots and let everyone bring their own ballot with them so that they can vote for whoever they want, not whoever the ruling government wants to let them choose from. And naysays... believe it or not, but that system is probably less prone to corruption than what we have today.
Click here or a puppy gets stomped!
There are two kinds of paper trails. One is a readable ballot that must be submitted into the ballot box, and the other is a sort of receipt to let you know whom you voted for.
The first kind is acceptable, and I believe the open voting consortium has this idea correct: the machine should print out a barcode, that can then be verified by another scanning machine. This barcode must then be submitted into the ballot box.
The second kind is flawed for two reasons. First, there is no way to verify that what the computer printed is actually what's recorded on the bar code, or what has been submitted electronically. Second, and more importantly, it provides an easy way for proving whom you voted for. I could tell all of my employees to bring in their receipts, and those who vote for candidate A will receive benefits. Yes, this is illegal, but we shouldn't make it any easier.
what's a sig?
I was going to post this very argument, and here you've said exactly what I wanted to say.
So instead of just saying me too, let me add my perspective as an American who now lives in Germany. The way they run elections here was a real revelation to me. After a lifetime in a culture that is fascinated with high-tech solutions, and where high-tech is uncritically assumed to be better, I was amazed to see that a simple solution was clearly superior.
Voters are handed a piece of paper with the names of the candidates. They take it behind a privacy barrier and mark an 'X' in circles next to their candidates' names. Then they fold up the paper, seal it in an envelope, and drop the envelope through a slit in a box. Then at 6 PM, the envelopes are dumped out of the box and the votes are counted and re-counted by hand. Anyone who wants to can witness the counting.
With this system, a fiasco such as Florida in 2000 (or in a number of states in 2004, as I predict) simply cannot happen. The are far fewer possibilities for error, and the credibility of the result is much greater.
The problem in the US is cultural. The very idea that a low-tech solution could be better simply doesn't cross our minds. For some things in life, we really are better off with more computers and machinery, but for elections, we should just dump them all on the trash heap, all they do is compound mistakes.
Always keep a sapphire in your mind
NIST did a great job with the AES competition (to develop and standardize a new block cipher to replace the aging DES) - why don't they have a competition to standardize a electronic voting machine platform? There's no reason this shouldn't be done on a national basis.
I think that if we as a community put enough pressure on NIST, they'll do it. And since NIST is a non-partisan body, there's no good reason for congress to not support a design that is sponsored by NIST.
Such a process would promote both openness of participation and review of designs. The winning design could then be standardized and vendors could simply implement them to spec.
Wish we had that here in Venezuela las august.
The voting machines here for the presidential referendum produced a paper trail.Suddenly when there was a doubt of the transparenncy of the whole process (because the voting machines were black boxes, noone knew what the code on them did) the government refused to count the papers from each machine.
Instead, they performed an "audit" where a member of the national electoral council on TV announced that a certain number of boxes would be chosen at random...by another computer running who knows what code on it and after the program was done "generating" the number of the boxes to be audited he proceeded to open a Word document with the numbers on it.
Of course, when the audit was done nothing was found amiss.
Transparent indeed...