(Last updated November 5th, 2004)

After 2000, a large chunk of the nation rushed to upgrade their voting systems to computerized ones.

Supposedly, we're supposed to rely on the fact that voter fraud is illegal as a defense against it. And as a substitute for actually taking technical measures to either secure the machines or render it possible to double-check them.

I'm already hearing about voting irregularities in the 2004 election. Calls on election day reminding people to vote "tomorrow". San Francisco is calling in ES&S troubleshooters to help them figure out why the shiny new IRV system is coming up with numbers that don't match up in four of the races. Ohio had people standing in line for hours and voting machines rebooting.

And the people selling the vote-counting machines say their software has to be secret.

Well, here is the binary for MY super secret voting program.
Two copies of it. Both copies have a few bugs and incompletenesses that shouldn't be in production software.
One of them has a backdoor in it that permits a person familiar with the program to engage in a little simple election fraud.

From just the binaries (now with Win32 as well as Linux/i386 copies), can you tell the difference?

tell me if you can.

The source will be posted in a few days for further analysis. Or maybe weeks.

Oh, BTW: to view the totals, run `./vote.[extension] --show`.