July 1, 2004

Here I go again

I just called my county voter's office and volunteered to be an election official for the November polls.

Why?

Earlier today, I remembered getting all worked up over the Diebold Debacle and figured it'd make a nice brain-break over lunch to skim the electronic version of Black Box Voting.

The writing is a little too conspiracy-theory for me. It's clear that much of what is normal in software occurs at Diebold and other companies that develop voting software. Mistakes happen. Products are shipped in a hurry and QA isn't what it should be. Programmers make ascerbic comments in the code. Also, it's shocking but true that local government has a hard time upgrading technology in its voting processes and will fumble quite a bit.

This may be why most tech people don't support e-voting. The average standards of the software industry are not high enough to inspire tech geek confidence in using software for important crap like Fly-by-wire, medical software and individual rights, like voting. This isn't to say that it can't be done properly. It's just that if you transport typical software engineers into a mission critical product, ummm, most of the other software geeks won't want to be early adopters.

The actions of the Diebold programmers don't appear malicious to someone who's worked with software engineers under tight deadlines with bugs to fix and product managers who can't allow the product version to roll. This is a point that I think Ms. Harris misses in her book, and one that made me view the rest of her writing with cautious suspiscion.

However, the allegations of abuse are more serious. The lack of certification, horrid Access "database" design (no primary keys!), last minute upgrades on-site at elections, apparent tampering or removal of votes and denials after the fact are huge problems. The fact that major parts of the Diebold code were written by a convicted felon who served 2.5 years for multiple counts of embezzlement (and one of his jail buddies, a convicted cocaine trafficker managed a division of Diebold)--well, clearly, there's an accountability hole in this process somewhere. What I can glean from the book indicates that there's the potential to alter voting tallies in most versions of Diebold systems that have been deployed in recent years and the reports from election sites make it seem likely that tampering occurred.

But, before we all jump on the grand conspiracy cruise, it seems like CA Secretary of State, Kevin Shelly is on it. Santa Clara County met the re-certification requirements, so electronic voting will occur in November. I figure if there is a conspiracy, I should at least go participate in the process and be able to report back about how bad it appears to the poll workers. Should be an interesting experience, I imagine.

Regardless, I'll be using a paper ballot.

No comments: