Posted on 03/25/2004 11:37:37 AM PST by calcowgirl
The president of Diebold Election Systems pledged Wednesday to fix multiple problems plaguing its oldest touchscreen-voting customer on the West Coast.
Officials in Alameda County, purchaser of a $12.7 million Diebold electronic-voting system in May 2002, reserved judgment on whether the McKinney, Texas, firm is capable of delivering on its promises.
"I got the feeling very much that they were committed to making everything right. Of course, our goal is that the presidential election will run smoothly," said county voting Registrar Brad Clark. "We'll see how they follow through on our discussions."
Diebold Election Systems President Bob Urosevich brushed aside press questions after a two-hour meeting in which Alameda County officials laid out a litany of problems with Diebold equipment and printing services.
The worst -- mass failure of voter-card encoders in the March 2 primary -- alienated an unknown number of voters from the polls and fueled calls by two state lawmakers for a statewide moratorium on electronic voting.
This week, after months of study, California's three affiliates of the American Civil Liberties Union cautiously endorsed new requirements that electronic-voting systems in the state produce paper backup records by July 2006. But the ACLU found no problem with the use of touchscreen voting machines in the November general election, in concert with added security procedures.
Partway into the 2004 elections, Alameda County has few practical alternatives to Diebold, Clark said. Switching to a paper-based optical scan voting system with comparably low error rates could cost $4 million, he said.
Diebold's voting system also inexplicably gave thousands of Democratic votes in the Oct. 7 recall election to a Southern California socialist. The firm has failed to obtain timely state approval of hardware and software.
For Super Tuesday, Diebold supplied scantily tested devices that failed in 200 Alameda County polling places and more than 560 in San Diego County.
The devices, a kind of voter-card encoder called the PCM-500, eventually were to revamp the polling place. Diebold and the counties planned as early as 2005 to program voter-registration lists into the PCM-500s, eliminating the need for paper pollbooks and making polling places virtually paperless.
On Wednesday, Diebold representatives said they didn't know they had to have such "peripheral devices" tested and certified for an election until late December.
State officials say that's not true.
"Elections officials were very clear in October in letting Diebold representatives know that anything related to their systems had to first be certified," said Doug Stone, spokesman for the California Secretary of State's Office.
The encoders were essential to Diebold's voting system. They were the sole method by which poll workers were trained to create voter cards that, in turn, called up each voter's electronic ballot. As a practical matter, no encoders meant no ballots and so no electronic voting.
"We're going to do everything possible to address the concerns and do everything we can to provide the best election possible" in November, said Diebold spokesman David Bear.
Electronic-voting critics are keenly watching Alameda County's demands on Diebold. No other U.S. county is known to have pursued a contract complaint with an e-voting vendor.
"It seems obvious to me that as a local election official, the thing I would hate the most is a messed-up election," said Stanford computer-science professor David Dill, a critic of electronic voting and founder of VerifiedVoting.org.
The encoder failures on Super Tuesday cast doubt on vendor promotion of electronic voting as fully tested, certified and bombproof.
"Diebold has taken a pretty hard line in saying there are no problems," Dill said. "These problems weren't hypothetical. Voters saw them."
Diebold officials came back Jan. 8 without the tests but seeking conditional certification of the encoders for use in the March 2 primary.> The lab, Ciber Inc, cautioned in its report that the test does NOT provide any assurance that the device would pass a certification or qualification test based on the (Federal Elections Commission) Standards (a more thorough functional test, a source code review and a documentation review would be required). Security features for this device were not evaluated. Hardware testing or source code review was not performed.
But San Diego, Alameda and several other counties said they had no practical alternative to using the new encoders, and state election officials granted a conditional certification.
I can think of a relatively foolproof way to make electronic voting machines. Have them print a continuous paper tape of the votes in UPC encrypted format, which would be read at a central location. At the same time, give the voter a printed receipt with a CRC number that could be used online to verify that his vote was counted without alteration. This could be done without revealing the votes, so union bosses couldn't stand over a worker's shoulder while he checks.
this depends on software that works, but once written, verified and burned into Read-only-memory, would be as foolproof as possible.
Amen!
There is nothing to stop them from maintaining two sets of records in your plan, the one you query and the one they count.
Fuggedaboutit. Paper ballots.
If you researched further, I believe that 'bad feeling' would turn into severe pain.
There are lots of connected folks behind the curtain... be wary of the wizard(s)!
Paper ballots... nothing less.
Scarier than sKerry: Inside A U.S. Election Vote Counting Program. And that article is from last July!
PAPER BALLOTS!!! AUDIT TRAIL!!! ACCOUNTABILITY!!!
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.