In the late-breaking "E-Vote Firm's Bill Comes Due" (11/11), California has announced that it will not continue with the certification of Diebold voting machines until the company pays for an audit of all the company's voting machines used in the state. The halt in Diebold certification is causing problems for counties that have selected Diebold, but California election officials would "rather err on the side of inconvenience and delay."

So starts this extensive New York Times article, which runs through a host of other problems Diebold has faced lately (the Hopkins/Rice study [239 KB, PDF], the FTP-software discovery, the leaked e-mails, and the Maryland/SAIC study (See "Did I say that?" below). Over all, a good "go to" article for anyone just getting into the Diebold e-vote issue.

 Other Diebold Election Systems in the News:

• Electronic Voting Firm Sued Over Threats ~ 8:52 PM 11/4/03 - AP Online
• Voting machines face two tests ~ 9:33 AM 11/4/03 - CBS MarketWatch
• Diebold Warns on Electronic Voting Papers ~ 10:21 PM 10/29/03 - AP Online
• 
  Reality check in the election booth ~ 10/3/03 - Maryland Gazette ~ Amazing!
• Diebold Election Systems Tabulates Accurate, Secure Votes In Boston ~ 10:32 AM 10/3/03 - Press Release

Let me explain further my feelings on the Maryland Risk Assessment Report [Body (1.25 MB, PDF), Appendix B (response to the Hopkins/Rice study, 700 KB, PDF)]. While this was a good study in that it highlighted many important vulnerabilities, it missed two critical issues:

1. It addressed security vulnerabilities as if Maryland election workers were honest. Now I am not suggesting that they are not, but it is a far more trivial problem to design and build a secure system when you assume employee honesty. It is much harder when you have to provide for the possibility of employee dishonesty. This is especially troublesome in voting systems because a large influx of temporary workers are hired only for a particular election.

2. It failed to address the fact that this software simply shouldn't have all of these problems!. Let me explain. You're sitting in front of your computer. It's a pain in the ass. Software doesn't always work right. It crashes, sometimes with no apparent cause. But this is just what computers are like, right?

   No, they're not! If computers were like that, they never would have gotten this far. But it is apparent that the authors of this report think that they are, and that is simply a bad assumption. Here's the difference:

   • Your home computer: It runs on an operating system (O/S) that can run on many different computers. Each of those computers has a different set of input/output (I/O) devices attached, and more can be added or deleted per your needs. Your O/S will run thousands of different programs, each with different needs. It can run several dozen of these simultaneously (multi-tasking). It handles dozens of different communication protocols, some local (your printer, your DVD, etc.) and some remote (your home network, the internet, etc.) And all of this stuff goes on simultaneously. Pretty amazing. And consider also that you can buy one of these for less than $1,000, the O/S and your programs being a small fraction of that purchase price. And you wonder why your home computer is tempermental at times? Easy. You get what you pay for.

   • Now contrast that with a touch pad voting computer: First, hundreds and perhaps thousands of identical touch pads are purchased at a single time. Each of these runs at most four programs: Receive the ballot, record the vote, (perhaps) transmit the results, and (perhaps) run in test mode. And none of these run simultaneously. Communications protocols? Five: Run the display, accept input, read/write hard drive, read/write votes, and (perhaps) communicate externally. Compared to your home computer, this is a walk in the park. And the price tag? Millions of dollars. So why so many problems? There is only one answer to this: incompetence.

   Clearly, touchpad voting computers should not be suffering all sorts of problems. The application is easy (in computer terms) and the price is high. And yet they do.

   So here are the key points of this situation. Maryland's purchasers of these computers have expectations of them that are consistent with their own experience with their home computers. But the authors of this study are clearly "computer professionals". As such, they should have written this study from the perspective of their own experience and expectations, and not from the perspective of the experience and expectations of their client. But they did not. Why?
   

What a way to run an election. As the Washington Post, suggests, "Why leap?"

• Two parishes use 3,992 iVotronic voting machine by ES&S. (iVotronic machines were in use during the troublesome Miami-Dade County elections in 2002, although equipment quality is not believed to have been a source of those problems.)

• Sixty-two parishes remain on an assortment of older manual voting machines (4,272 machines total).

"The point of elections is to test ideas and hold officials accountable. This process is short-circuited when like-minded voters are so concentrated in districts as to render the outcome a certainty. Lack of competition amplifies ideological differences and further polarizes U.S. politics, because Republican officeholders need not answer to Democratic constituents and Democratic officeholders can ignore Republican voters."