Michael Shnayerson, Vanity Fair: The Election 4/04

[AtLiberty]

I recently discovered this link and figured there were others here who hadn't read it... I know, I know, I know...there are many references to Bev Harris. For historical purposes, I feel that shouldn't automatically disqualify it from reading here. What I found curious was that such an article was published 18 months ago in "Vanity Fair" and didn't produce a ripple...

The Election
Vanity Fair 4/04
by Michael Shnayerson

*snip*

As a subcontractor, says Rob Behler, he worked on the Diebold machines for around 30 days before a "difference of opinion" with the senior project manager led to his dismissal. So perhaps he was, as he says Professor Brit Williams characterizes him, a disgruntled employee. And as he gives his account to Vanity Fair, almost everything he says is at odds with Williams's version, starting with the talk of "one patch." (Was he the Rob of "rob-georgia?" Harris says the file predates his arrival.)

"Very simply, Brit Williams did not work for Diebold, so he has no idea what patches Diebold did or didn't do," Behler says. "I worked for Diebold, and I didn't seek Brit's approval for anything, because I was told to avoid him, period."

Williams, Behler clarifies, was doing "acceptance testing" on the thousands of machines arriving in various Georgia counties. Behler's job was to update the machines, then boot them up and be sure they could pass that test. "And they didn't," he says. "They had stacks and stacks of machines on pallets that had failed, bombed out. I went down to the DeKalb County warehouse, one of the larger ones, with Greg Loe, who was second in command under Bob Urosevich, he and I together, because he wanted to see the machines. I had explained, 'Don't expect a lot — they're broke, man. They do crazy crap, and they don't do the same crazy crap twice'." According to Behler, about 25 percent of the machines — not 3 to 4 percent failed.

Behler says that at the company's direction he assembled two "SWAT" teams of five or six people each to go in vans to the various county warehouses and debug and update the machines before Williams and his team got to them. "Wherever they were headed," Behler says, "we'd get ahead of them and they try to lower the failure rate to more acceptable levels."

Even with the frantic debugging and updating, Behler says, the failure rate was about 15 percent while he was there. "And here's the really scary thing: you could test the machine and it would test fine, then you'd turn it off, power it up again, and it would fail." When they couldn't fix a machine, Behler says, his SWAT team would move it out of the warehouse and into the van. "So swap that machine, swap the barcode numbers, and replace it with a machine we believed worked."

*snip*

http://www.ejfi.org/Voting/Voting-18.htm