Electronic Voting Machines Never Make Mistakes

We at DU need to ralize this point, the way a computer is programmed it never makes mistakes. Never. If a computer is programmed to add one vote to a candidate it will do exactly that every time you repeat the process, doesn't matter if its once, twice, or a billion times.

So when you hear of counties having 100,000 votes when there is only 20,000 registered voters in that county the machine didn't screw up, somebody tampered with it.

This is what I can't stand about our media. I do appreciate they are finally, less than 2 weeks before a national election, reporting on how inaccurate voting on these machines is but they are still missing the main point. If the vote comes out wrong it is not the fault of the computer, it is the fault of someone that messed with the computer. The questions they should be asking is who did this, not why it happened.

I'm not saying votes aren't stolen. But you can not say that a piece of software wont bungle votes by accident. I have gotten weird and bizarre errors that are virtually impossible to replicate in software I've written myself, and would never have expected to get looking at the code. Software is incredibly complex. The bigger it is, or the heavier the usage, the more likely a bug is to pop up.

if software screws up you will not get a real number as the bytes get screwed up. So instead of having a real number show up you would get a bunch of letters and numbers assuming you could even read it. If you know a little programming assume the following:

while ($vote => 1){
++$john_smith;
}


Every time you do the above the loops goes through and adds one, it will never ever add more than one. The only way it would add more than one is if you do the following:

while ($vote => 100){
++$john_smith;
}

That would give John Smith 100 votes each time he gets voted for. Software does not mistake numbers, I have never seen that happen on tested code.

Rarely in practice are programs of the 'hello world' variety. The mistakes don't occur in the basic math, they occur in the logic itself.

Here's an example I ran into:

Software X allows users to place into a queue payments to be applied against an insurance claim. As each payment is entered an entry in the payment table is placed, marked as paid and tied to an accounting ledger entry. The payment entry and the ledger entry are handled by two separate procedures, each returning a value, or false if the entry failed.

Upon the user marking as finished their entries, the software attempts to reconcile certain financial balances, and if it can't it rolls back both the payments and ledger entries. Unfortuneatly, there was a bug in one of the balancing routines that only occurred during a very rare set of data circumstances that would result in the payment record being recorded successfully and the ledger entry not being recorded. The sub-routine failed in such a way that the calling procedures were not aware of the failure.

As a result a payment would be recorded and go out on a check that was not in the ledger and would not show up in a manual view of the single claim, but would skew the total tallied on the report.

These things happen.

The blue screen of death is proof of it.

variables which wouldn't be used in counting integer votes. Not trying to quibble, but if I knew the architecture of the machines (which 'button' creates which input, etc.) I would bet I could write all the necessary code in a couple hundred lines. It's not even skateboard science...
;-)

...That makes much more sense where the error is a mistake of deliberate all voting tabulation software/programs should be transparent and subject to open independently authorized audits. That at least will keep the hackers and criminals and incompetents in check by independent oversight.

errors would work for the other guy once in a while. The machines tilt toward GOP and that's no error.

....and those responsible arrested and charged? I agree with that. But, you must find evidence that the program controlling the counting process is in fact coded to do that. Just because results like that happen, without the actual code instructing the tabulation process to do that there is no admissible evidence. Find the coded instruction. Just one line, somewhere, anywhere. It would be simple if the source code was available and transparent. But the vote tabulation companies are protected from showing their source code.

All it takes in one line or two to make it so the virus destroys itself.

...written and never erased? I know that many program routines have these automatically built into them so that each step could be traced at any point during the program run. Just set these up in a way that these must run until a separate instruction purges them. That instruction would be controlled by a key code which only authorized people have. Those tracer files then could be kept in backup archives, just like paperwork is maintained in storehouses indefinitely. Redundancy is the key to security. Tabulation files are farly simple to track, so the log files would be relatively small by comparison.

Everyplace which electronically tabulates turns on a turnstiles for attendance figures is nothing more than a tabulating machine, just like an e-vote machine. Toll road booths, sports stadiums, theme parks, amusement park, etc, this is not rocket science. I would think that touch screen voting machines are no different.

But that is not what the vote stealing bastards did.

They coulda produced a computer system that was damn near perfect at counting votes. They woulda if they wanted votes to be accuarately counted. And they shoulda.

But then, they wouldn't be in power NOW, would they?

If I wanted to hack an election, I'd simply store a victory for the guy I'd want to win. Then I'd just make the machine not do ANYTHING when someone enters a vote.

hours working on. Does the rule "Save and save often" have any bearing? Computers are not infallible but your point of most computer errors are man made is a valid one.

and will they fight what we didn't fight in 2004?