For those interested in politics...

[danaeris]

Some 46,000 New Yorkers are registered to vote in both the city and Florida

Also, for those who were inclined for some reason to trust the e-voting machines...

Not only is the code for e-voting machines being kept secret, but the three companies that certify the nation's voting technologies operate in secrecy and refuse to discuss flaws. What's more, they were *hired by the companies* which made the e-voting machines. Can you say conflict of interest?

Comments

[dicedork.livejournal.com]

I'd wait to see this verified in a more mainstream media outlet before chickens are counted. It seems HIGHLY suspect that forty-six *THOUSAND* people are planning to go from NY to Flordia to vote on election night. Might just be a case of same names or soemthing.

The voting machine thing? That scares the hell out of me.

[cos]

My best guess: Most of them are probably students or snowbirds who re-registered in the state they intend to vote in and whose old registrations in the other state are still on record. Most of them probably haven't even thought of the possibility of voting in both places in the same election. There are probably a handful who are intending to double-vote.

[dicedork.livejournal.com]

Or that. And that's not so much corruption as oops.

[tunape.livejournal.com]

What I don't understand about voting machines is why they don't print up a receipt of who you voted for... The data can be transmitted and tallied quickly, but then you leave a paper trail of what happened. Also, the person in the voting booth can double check that what is printed is what they actually entered.

It's nearly impossible to create a 100% reliable system, but there's nothing wrong with leaving a trail of each step taken so that manual intervention can take place. it sounds more like a conspiracy to oust voting all together, and blaming it on technology.

bleh.

[nathanjw]

This is what a lot of lobbying has gone for (and against). It does improve the situation with electronic voting machines, but from the technologist angle, it's one more thing to fail - it can run out of paper, jam, stuff like that.

Why is anyone opposed to it? Because it costs more money. Either from the perspective of local electoral boards, who have already blown their budgets on these things, or from the perspective of the manufacturers, who don't see much profit in a re-fit operation.

(All this is assuming that the receipt is something that stays at the election place, one way or another. Take-away recipts are not useful, and in fact, actively harmful, since they permit all sorts of vote-buying and vote-coercing schemes).

I'm pretty fond of Cambridge's system, which is paper ballots, marked by hand (with ink), and generally counted by machine. The recount is straightforward. The issues it doesn't address that the all-electronic ones do are those of language and disability accomodation, but I tend to think that there are solutions to both of those that are less radical than all-electronic machines.

[auros]

Yeah, the Luddite solution -- going to large-bubble scantrons, with the instructions and candidates' names printed right next to the places you mark -- is starting to look pretty attractive.

Have you seen the Diebold ATMs that have gotten installed at a lot of Wells Fargos lately? Every now and then you run into one which has crashed to its WindowsXP desktop, and you can futz around with the filesystem because the touchscreen can control the mouse pointer. I've heard also that the ATM network had some problems a month or so ago when a worm got into a backend machine and propagated onto a bunch of the individual ATMs. I'm starting to seriously consider closing my accounts and taking my business elsewhere -- though I don't know that the other banks are any saner. :-P

[nathanjw]

Optical scanning is hardly Luddite; you can still get the fast results that we seem so addicted to these days. The truly Luddite solution is to hand-count everything in the first place, and to use separate ballots for each race. I've been told that the British system is something like that. You really have to stop and ponder exactly what it is we think we're accomplishing with all this complicated voting tech.

Wells Fargo? I think you've confused me with someone on the west coast :) The banks I tend to patronize around here seem to be a couple of generations behind in the whiz-bang-iness of their ATMs, which is probably just fine. ATM problems don't worry me as much as voting problems because banking transactions are logged, identifiably, which permits unrolling mistaken transactions. The identifiability is key to why broken or hacked ATMs are less of a risk than broken or hacked voting machines (and a good thing to remember whenever someone says "we have electronic banking, why not electronic voting?").

[auros]

The Aussies use true hand counting, IIRC.

And yeah, older ATMs are almost certainly more secure -- they use OS2/Warp, for which nobody ever bothered to develop viruses, worms, or tools for scriptkiddies.

I wouldn't feel too secure about ATMs in general, though -- identity theft issues have come up before with them. Identifiability is nice, until somebody else knows how to fake "being you", at which point it becomes problematic, since the people they bilk will want to collect from you.

That doesn't rectify the problem.

[etherial.livejournal.com]

Sure, the receipt can display exactly who you voted for, but there's no way for you the voter to verify that that is the data that's saved. That's the problem with electronic voting, all the actual votes are stored in computers and are completely unverifiable.

A better system uses a computer screen to help you fill out your ballot, prints it out, and then you submit the paper ballot into a counting machine. A hand recount remains possible.

Re: That doesn't rectify the problem.

[auros]

If you exit-poll every station, or do handcount or scantron a sample from every station, and check whether the results from that match up (within a reasonable margin) with the machines' results, you can detect fraud pretty easily, and correct for it by doing a full hand/scan-count.

Re: That doesn't rectify the problem.

[cos]

(much-belated comment as I go through some old comment emails...)

you can detect fraud pretty easily, and correct for it by doing a full hand/scan-count.

I think you missed the point of etherial's comment, though: if machines print "receipts" for voters, you can't do a full hand count. What are you going to hand count?

Receipts aren't the solution. Paper (or plastic, or other physical format) ballots are the solution. They must remain, locked up at the polling place, as the ballots of record. You can use machines to count them. You can even use machines to generate them (as etherial suggested in the second paragraph). But they must be physical, secure, recountable ballots.

Apparently you misunderstand the purpose of the receipt.

[auros]

The voter doesn't take it home. The voter verifies that the receipt reflects his or her intentions, and then drops the receipt in a ballot box, which is kept secure exactly as a scantron ballot would be.

Re: Apparently you misunderstand the purpose of the receipt.

[cos]

That's not a receipt. That's either a "paper trail" or, with the proper rules, a "ballot of record". What I campaign for is the latter.

"receipts" are things people take with them. They serve no purpose in voting. If you're talking about something the voter doesn't take with them, it is most decidedly not a receipt, and calling it that will just cause confusion.

I work on voting reform a lot. The confusion does exist. Lots of people do actually think we should have voting machines that produce "receipts" (which the voter takes with them), and don't understand why that doesn't solve the problem. Paper audit "trails" are only a partial solution. Real solutions involve a physical "ballot" ("ballot of record"). These terms are important. They have real meaning.

[auros]

And I've worked at the polls.

The county is calling them receipts.

So whatever.