Entries tagged as sequoia
Tuesday, October 14. 2008
The good news. San Francisco gets to use their new voting machines. Results should be fairly quick. The San Francisco Department of Elections received good news Tuesday: Secretary of State Debra Bowen approved its optical scan voting machines for ranked-choice races in next month's election. The really really bad news. San Francisco should NOT use their new voting machines. Results may be quicker than counting by hand, but there is reason to believe the results can't be trusted. John Gideon, Co Director of Voters Unite delivers the bad news, (which is actually admitted in the Secretary of State's report): http://www.sos.ca.gov/elections/voting_systems/sequoia/system40/2008-10-14_system40_approval.pdf
The approval is only for San Francisco in November. The report also says
that a Source Code Review found that previously reported security
architecture issues remain issues in version 4.0. They also found no
effective mechanism to protect the integrity of data transferred between
components and a potential vulnerability for SQL injection attacks, and
a user can exploit a system weakness to gain access to the database
without going through WinEDS. They found several other vulnerabilities.
Item #25 on Page 8 is especially puzzling. "By order of the Secretary of
State, voting systems approved for use in California shall comply with
all applicable state and federal requirements, including, but not
limited to, those voting system requirements as set forth in the
California Elections Code and the Help America Vote Act of 2002 and
those requirements incorporated by reference in the Help America Vote
Act of 2002. Further voting systems shall also comply with all state and
federal voting system guidelines, standards, regulations and requirement
that derive authority from or are promulgated pursuant to and in
furtherance of the California Elections Code and the Help American Vote
Act of 2002 or other applicable state or federal law when appropriate."
The Sequoia System 4.0 cannot be certified as meeting the above
requirements. Testing at the federal level had not even started as of
May 19. In fact they were still preparing the test plan at that time.
HAVA requires accessibility yet the state has waived that by allowing it
not to be tested. HAVA requires accuracy yet the state has ignored that
requirement by ignoring the fact that iBeta has not yet tested for
accuracy of the system.
Since May when the state of Washington began testing for provisional
certification the system version has changed from WinEDS 4.0.108 to
4.0.116 and this, again, is without a majority of testing taking place.
How many more version changes are going to be made before Nov. 4?
--
John Gideon
Co-Executive Director
VotersUnite.Org
www.votersunite.org
So, San Franciscans, still happy about those voting machines? Recent reports of problems (serious ones) with Sequoia machines October 3. 2008 NJ Judge Suppresses Report on Voting Machines' Security New Jersey Superior Court Judge Linda Feinberge has prohibited the scheduled release of a report on the security and accuracy of the Sequoia AVC Advantage voting machine. October 2. 2008 Test Shows Palm Beach Voting Machines Not Accurate! Tests run yesterday in Palm Beach County Florida show there is a real problem with their voting machines. The optical scan machines, made by Sequoia, can't count the same ballots the same way two times in a row! September 22. 2008 New York. Sequoia's Sinking Ship - voting machines still not certified Even Sequoia's disabled accessible optical scanner is easily hacked - the old fashioned way by physical stuffing through a gap in the ballot box. On August 18, Rady Ananda showed how you could stuff the Sequoia optical scanners like a Thanksgiving turkey. September 14. 2008 Pierce County Instant Runoff Voting System has new bug, says WA SOS - may affect San Francisco This email from the Secretary of State of Washington outlines another bug in the new IRV voting system for Pierce County. It affects "rank choice voting" (IRV/Instant Runoff) only.
Thursday, October 2. 2008
Tests run yesterday in Palm Beach County Florida show there is a real problem with their voting machines. The optical scan machines, made by Sequoia, can't count the same ballots the same way two times in a row! They get a different count each time. Washington DC, another customer of Sequoia, is also having trouble with their voting machines, high speed optical scanners. Earlier this month, "DC's machines somehow managed to inflate the vote totals in some races by more than 100 percent, making up thousands of write-in votes and adding thousands of votes to the totals of candidates on the ballot." according to Government Computer News. Pierce County Washington found problems with their Sequoia precinct optical scanners and could not allow them to be used to count their Instant Runoff Voting ballots. San Francisco uses the same machines for IRV. What can be done to protect the voters in these districts, come November? Palm Beach County's ballot-counting machines off by dozens in tests Rejected under-votes and over-votes tallied; totals off by dozens By Brian Haas | South Florida Sun-Sentinel October 2, 2008 Palm Beach County's high-speed ballot counting machines couldn't count the same ballots the same way twice in tests performed Wednesday evening. As part of a challenge to a disputed judicial race, elections officials tested two Sequoia 400-C counting machines to see how they handled ballots they previously rejected as either over- or under-votes. Wednesday's tests were simple. Election workers took 262 ballots previously rejected by the machines as over- or under-votes in the judicial race and ran them through two machines. All of them should have been rejected again in the tests.
That didn't happen.
On the first two tests of 160 ballots, the machines accepted three ballots as good votes. On tests on 102 more ballots that should have been rejected, the machines first accepted 13 ballots as good votes and then 90 on a second run....more at the link
Sequoia officials did not attend the testing. It would be too embarrassing and they would be asked to explain everything.
Related:
Continue reading "Test Shows Palm Beach Voting Machines Not Accurate!"
Wednesday, September 24. 2008
In John Gideon's Daily News today, he says the Palm Beach County Florida nightmare is FINALLY over. I can't say it better than John, and frankly it was getting WAY too confusing, so here's John: "The election nightmare may be over in Palm Beach Co Florida. After seven
recounts, machine and hand, the election for a county circuit court seat
has been certified. The losing candidate, of course, is considering
legal actions. Still not explained is how ballots were lost then found
and how the machines just couldn’t seem to get an accurate count.
Sequoia should have some explaining to do beyond just “No, no, no, it
wasn’t us”.
Here's the latest media report: Recount gives Abramson Palm Beach County Circuit Court seat Canvassing board certifies judicial race after third recount By Brian Haas | South Florida Sun-Sentinel September 24, 2008 Our long electoral nightmare is over.
Twenty-nine days after Palm Beach County voters actually went to the polls, members of the election Canvassing Board wiped their hands clean of a disputed judicial race and certified the election.
Challenger William Abramson unseated Circuit Judge Richard Wennet by 61 votes — the same margin Wennet was once winning by after one of three sets of recounts the election has endured. The board ended up not being able to account for only four votes, a far cry from the 3,500 that were thought to be lost in the days following the Aug. 26 primary.
"I'm overwhelmed," Abramson, a lawyer, said by phone Tuesday night. "This is why we're the greatest country ... We were faced with adversity and we all rose to the challenge."
I was getting overwhelmed too. We tried to keep up with the Palm Beach counting and recounting fiasco here, but it got more and more frustrating: September 12. 2008 Palm Beach Elections Goes from Missing Ballots to Having too Many This could all change, but here's the latest update on the Palm Beach Elections mess: September 12, 2008 Palm Beach Fiasco: Officials Say Ballots Not Trashed Palm Beach County Florida still can't account for 3,400 missing ballots. September 10. 2008 Florida Voters Coalition Calls on State to Secure Palm Beach Elections The Florida Voters Coalition, The League of Women Voters, and Democratic and Republican leaders call on the State of Florida to take emergency action to secure November’s General Election. September 4. 2008 Palm Beach Election Desperation Part III - 2500 ballots missing now Well, things have just gotten worse, perhaps coming to a head. The new number of missing ballots as of today is 2,500. September 4, 2008 Palm Beach Election Fiasco - 2+2 now = 3 1/2: 2,700 ballots found overnight I guess someone checked the "lost and found" department last night, and discovered 2,700 of the 3,478 missing ballots we discussed yesterday... September 3. 2008 Palm Beach Elections: 2+2 = 3 or The ballots are missing, so start finding answers In Palm Beach County Florida, 2 + 2 = 3 if you are counting votes, and it gets worse from there. The editors of the Palm Beach Post want to know where the 3,478 missing votes went... September 2. 2008 Oops - 3,400 votes disappear in Palm Beach County Fl recount Maybe Florida elecions are just plain cursed. Even with brand new paper ballot based machines in Palm Beach County Florida, they can't count votes. In the initial contest, the two candidates in the judicial contest were a mere 17 votes apart...
So we should all thank John Gideon of www.VotersUnite.org for providing the best election integrity related news content in the United States. You can receive the Daily News for free by emailing DVN@votersunite.org
Sunday, September 14. 2008
This email from the Secretary of State of Washington outlines another bug in the new IRV voting system for Pierce County. It affects "rank choice voting" (IRV/Instant Runoff) only. Pierce County Washington adopted "instant runoff voting" in November 2006. In May 2008, officials were backed into a corner and ended up getting permission to use uncertified IRV software, violating their own state's laws.
Pierce county has Sequoia "Insight" optical scanners and DREs. The tabulating software, WinEDS 4.0.34 is still under test by iBeta. There is not even a test plan presented to the EAC for this system at this time. Washington State provisionally certified this software for use even after finding that there were problems with the use of the Insight machines when used in the RCV process. San Francisco is also using this system, and currently it is not certified in California to count IRV. San Francisco officials still don't know how they will count their "ranked choice" voting in November 2008.
Here's the email (Thanks to John Gideon of www.VotersUnite.org for sharing)
Dear Team,
It has come to our attention that there is a very rare occurrence of a problem with the Rank Choice Voting results loading with Pierce County’s provisionally certified software.
Continue reading "Pierce County Instant Runoff Voting System has new bug, says WA SOS - may affect San Francisco"
Thursday, September 11. 2008
D.C. Election Glitch Blamed On Equipment
Washington Post
D.C. election officials blamed a defective computer memory cartridge yesterday for producing what appeared to be thousands of write-in votes that officials say did not exist. Ah, the return of the "glitch".
As I have stated before, "glitch" is a word the press likes to use anytime they discuss computer failures. The word conveys a problem, but doesn't assign blame to anyone. Things just mysteriously happen.
I wonder if they would refer to the collapse of the bridge in Minnesota as the result of a "glitch".
No, of course not. The bridge collapsed because people failed to perform the proper maintenance on it, and allowed serious structural failures to go untreated.
Same with computers. Recovering from a hardware failure is the acid test of a well written program. In this case, the software failed the test, since it allowed phantom votes to register.
"It was determined that one defective cartridge caused vote totals to be duplicated into multiple races on the summary report issued by our office. The Board immediately caught and addressed this error, as is reflected in the last unofficial results report issued on Election Night," Murphy said in the statement. This is the hallmark of bad software. A good programmer anticipates errors, especially hardware failures, they even have a name for it, they call it "error-trapping".
He refused to answer questions from reporters, and no members of the election board appeared. Of course, this assumes that the cause of the problem is being accurately reported to the press, which is a risky bet most of the time, especially since election boards seldom have their own programmers of techs on staff. The usual routine is to repeat whatever the vendor told them happened, and not answer questions which would reveal that they are simply reading lines from a script and have no clue what they are talking about.
I am not the only one who is skeptical.
"That press release is a model of obfuscation," said Henry E. Brady, a professor of public policy at the University of California at Berkeley who has studied voting systems extensively, including in the deadlocked 2000 presidential contest.
The explanation that a defective cartridge caused tallying errors across multiple races "is what throws me off," Brady said. "It is hard to know what that means. I'm having trouble figuring out how that happens."
|
