A major breakthrough considering a history of nondisclosure by other election providers including Sequoia themselves, as well as ES&S.
But what's that mean, publicly disclosed source code? Simply this, Sequoia's computer programs for their new Frontier system are open for the public and election integrity advocates to analyze and dissect. Maybe we'll find some bugs. Maybe not. But the next time an election produces cockamamie results, at least with this Sequoia system we won't have to wait for the courts to order Sequoia to hand over their code. Or will we? Will the code we're given access to be the same code running our elections?
More importantly, will publicly disclosed source code of any or even all election systems guarantee our votes are counted? No. While this level of transparency is an important breakthrough, it will not ensure fair and accurate elections.
1. First and foremost, there will never be a perfect national election. Considering the sheer numbers of computers and humans and ballot configurations and peripherals. . . It's simply not going to happen. Not ever. Somewhere in some jurisdiction(s) something's going to hiccup. Perhaps it will be 37 votes. Perhaps 537. Either way if it's enough to change the outcome of an election and spot-on-accurate results cannot be determined by any other means, the election should be sent straight back to the voters who voted it. Politically charged courts should not be deciding how you and I voted.
We cannot guarantee fair and accurate elections in this country until uniform election laws identify invalid election results for what they are and send a bad election back to the people. The Election Assistance Commission (EAC) and the Help America Vote Act (HAVA) must recognize this major omission. Why should we expect less from our election systems than we do from any other business sector? A bill or statement or report that can't be read or that is incorrect, gets sent back to the source for corrections.
2. By mistake or intent, the code you see may not be the code running the election. While it is more likely that uniform code can be systematically deployed across all computers in a scanner based election system, there is no fool-proof guarantee. Touchscreen based systems present a much greater challenge, since literally thousands of touchscreens can be required to support the voters in one jurisdiction at 100 voters per touchscreen.
3. Poorly designed ballots can adversely affect voter intent.
4. Optical scanners are only as good as the pre-printed ballot forms and the ballots hand-marked by voters. Ballot positioning and alignment, ink, and quality of paper are some of the factors that can cause an undetected number votes to be misread and miscounted.
5. Programming code can be influenced by outside factors. For example, the date and time of day can influence a computer's behavior.
6. Communication of the voting machine (scanner or touchscreen) with the central computer can be incorrectly defined and interrupted.
Even the most perfect computer program code cannot prevent the failure of a computer's individual electronic parts and peripherals. Computers will fail and humans will make mistakes. Only when these unalterable facts are accepted and accounted for can we achieve stable elections in this country.
For more see:
First Voting Machine Company to Publish Source Code, Kim Zetter
Election Defense Alliance Opens Public Review of Sequoia Voting System Source Code, Jim March
