ULTRA26 # 1
08-05-2007, 03:27 PM
Smokin,
Here is some source for you. Please read it
Black Box Voting 164
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
13
Security Breaches
San Diego County and the states of Maryland, Arizona and Ohio
planned to buy new voting machines, and Diebold planned to sell
them. All told, these contracts were worth over a quarter of a billion
dollars. Despite all the information that had come out, officials were
barreling right ahead with their purchase orders.
Such confidence must be supported by a powerful factual underpinning,
but so far I havent been able to find it. Could someone
please share the secret decoder ring with us, so we, too, can see why
these machines should be trusted?
Election officials explained that all this criticism was just so much
hooey; they trusted the machines and those computer scientists didnt
know what they were talking about. Diebold announced, after the
SAIC report gave it a failing grade, that the report (yes, the same
one) said its voting system gave voters an unprecented level of security.
Er I guess you could call it unprecedented.
Many election officials are still giving Diebolds encryption scheme
a clean bill of health, but Im not sure many of them can spell the
word algorithm, much less explain it. Why do we allow election
officials to pronounce opinions on computer programming anyway?
I have yet to see any of Diebolds programmers answer a single
question about these software flaws. Public-relations team, yes.
Software engineers? Total silence. I would like to hear from principal
Black Box Voting © 2004 Bev Harris
Rights reserved. ISBN 1-890916-90-0. Paperback version available at
www.Amazon.com
This free internet version is available at www.BlackBoxVoting.org
Security Breaches 165
This free internet version is available at www.BlackBoxVoting.org
engineer Ken Clark, who wrote this about altering the audit log in
Access:
From: Nel Finberg, 16 Oct 2001: Jennifer Price at Metamor (about to be
Ciber) [Independent Testing Authority ITA certifier] has indicated that she
can access the GEMS Access database and alter the Audit log without entering
a password. What is the position of our development staff on this
issue? Can we justify this? Or should this be anathema?
From: Ken Clark, 18 Oct 2001 RE: alteration of Audit Log in Access; Its a
tough question, and it has a lot to do with perception. Of course everyone
knows perception is reality. Right now you can open GEMS .mdb file with
MS-Access, and alter its contents. That includes the audit log.
This isnt anything new ... Ive threatened to put a password on the .mdb
before when dealers/customers/support have done stupid things with the
GEMS database structure using Access. Being able to end-run the database
has admittedly got people out of a bind though. Jane (I think it was
Jane) did some fancy footwork on the .mdb file in Gaston recently. I know
our dealers do it. King County is famous for it. Thats why weve never put
a password on the file before ... Back to perception though, if you dont
bring this up you might skate through Metamor.
There might be some clever crypto techniques to make it even harder
to change the log ... Were talking big changes here though, and at the
moment largely theoretical ones ... Bottom line on Metamor is to find out
what it is going to take to make them happy. You can try the old standard
of the NT password gains access to the operating system, and that after
that point all bets are off ... This is all about Florida, and we have had VTS
certified in Florida under the status quo for nearly ten years.
I sense a loosing [sic] battle here though. The changes to put a password
on the .mdb file are not trivial and probably not even backward compatible,
but well do it if that is what it is going to take. Ken
Nels reply: For now Metamor accepts the requirement to restrict the server
password to authorized staff in the jurisdiction, and that it should be the
responsibility of the jurisdiction to restrict knowledge of this password. So
no action is necessary in this matter, at this time.
We are leaning heavily on local election officials to set up security.
Black Box Voting 166
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
Setting aside the references to doing end runs around the voting
system, four examples show that local officials have been unable to
restrict access to authorized staff and approved software:
1. A San Luis Obispo County, California, vote database popped
up on the Diebold Web site during the March 2002 primary, tallied
hours before the polls closed. Election officials cant explain how it
got there.
2. A cell phone was used to transfer a vote database in Marin County,
California. This is insecure and was never approved by anyone.
3. In November 2000, an unexplained replacement vote database
in Volusia County, Florida, overwrote the original votes, causing TV
networks to erroneously call the election for George W. Bush.
4. Voting software in 10 states, certified by the ITA and NASED
and escrowed by the secretary of state, was replaced by unauthorized
versions which came from five men in Canada.
SLO County Mystery Tally
A vote tabulation saved at 3:31 p.m., five hours before poll closing
for the March 5, 2002, San Luis Obispo County primary (SLO
County to the locals) was found on the Diebold FTP site. SLO County
Clerk-Recorder Julie Rodewald says that she doesnt know who put
that file on the FTP site, and only two people have access to the
GEMS computer the Deputy Registrar of Voters and Rodewald
herself. 1
The SLO file contains votes from a real election. It also contains
a problem for Diebold, because in California it is illegal to tabulate
votes before the polls close. According to California law, counties
are allowed to begin counting mail-in and absentee ballots prior to
election day, but results may not be posted before the polls close at
8 p.m.
We dont release those results. In fact, we dont even print results.
We dont know what the results are until 8 p.m., Rodewald said.
This file contains an audit log which documents GEMS activities
step by step for months leading up to the election, stopping precisely
at 3:31 p.m. on March 5, 2002.
The votes in the file correspond with the final vote tally, which
Security Breaches 167
This free internet version is available at www.BlackBoxVoting.org
can be found on the San Luis Obispo County Web site for that election
but only about 40 percent of the votes had come in by 3:31 p.m.
Computer programmer Jim March discovered that this file contained
real votes.
Maybe the clock was off? It was for a different time zone? When
it said 3:31 it was 8:31? Checking the date and clock is part of the
election procedures, marked important. But more than that, after
the polls closed there were more votes.
Was it for training? No one does training during an election.
How do the votes correspond to the final vote tally? The vote distribution
parallels that of the final tally.
The SLO vote file was assigned a password and placed on a Dieboldowned
FTP site. The password was: Sophia. Sophia Lee was a
Diebold project manager. Was she there that day? Yes. Did Sophia
put that file on the Diebold site?
Shes saying she did not post (the data) on election day, Rodewald
said. She said its something she never would have done.
Did Rodewald give Sophia access to the GEMS computer and the
vote database? Rodewald says that neither she nor any of her staff
put that file on the Diebold site, nor does she know how it got there.
Only the deputy (registrar of voters) and myself have access to
the computer on election day or any day, Rodewald said.
The large file, which was on the GEMS computer, takes time to
upload to an FTP site even with a fast Internet connection. Rodewald
said that the GEMS computer does not connect to the Internet.
Somehow this file made its way from the secure, inaccessible, lockedin-
a-room, not-connected computer onto the Diebold company FTP
site. Diebold denied that the information was posted on Election Day.
Diebold is trying to track down when the information was posted,
said Deborah Seiler, western regional representative for Diebold. (If
Diebold was trying to find out when it was posted, why did Diebold
state that it was not posted on a particular day?)
Rodewald says that the votes in the SLO file were absentee votes,
which were counted before March 5. She says they are not votes
cast at the polling place, which is reassuring, because the only way
polling place votes could be on a Diebold company site at 3:31 in
the afternoon is if the machines had an E.T. moment and decided to
phone home while the election was in progress.
Black Box Voting 168
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
However, the absentee explanation doesnt exactly correspond with
the tags in the file. And it doesnt explain why a partially-voted backup
file would be parked on a Diebold FTP site. 2 Why should Diebold
take any election vote file and keep it on a company site?
Perhaps because just 21 hours before the election, the software
wasnt working correctly.
From: Sophia Lee, 4 Mar 2002: SLO County Cards cast for precincts
in multiple vote centers are incorrect.
Sophias memo is a bug report showing that the software is
miscounting the day before the election. It references GEMSReport
1-17-21, a version that does not have a NASED certification number.
Whether or not anything unscrupulous is involved with this file,
it seems that unauthorized access was allowed into the system on
Election Day, and also that machines were using uncertified software
which wasnt working properly.
Transferring votes by cell phone
On October 8, 2003, I spoke with Marc Carrel, assistant secretary
of state for policy and planning for the state of California. I
asked Carrel about a set of memos indicating that Diebold has used
cell phones to transfer vote results.
Thats not certified. he said. True. Not in California, they
havent, he said, after a stunned pause. Yes, they have. In Marin
and Tulare counties, according to the Diebold memos.
I passed the memos to an investigative writer named Tom Flocco
(www.tomflocco.com). In his blog he wrote:
Diebold sales representative Steve Knecht wrote on April 12, 2000 that
We are using cell phones in Tulare and Marin, while also introducing a
rather curious, unfamiliar electronic election official called a rover: Rovers
are the ones who are given the cell phone with the modem for end of
night totals upload, not the precinct worker, at least in these two locations.
Guy Lancaster, Diebold software programmer, wrote on April 12, 2000,
regarding cell phones: I know of no written instructions, leading us to wonder
if there were rules and traceable documentation, or why cell phones were
Security Breaches 169
This free internet version is available at www.BlackBoxVoting.org
being used in the first place. ...
[Diebold sales representative Juan Rivera wrote] Also, we did not
have to dial the phone manually; the AccuVote did that just as if it was
connected to the wall jack. ... So now we have private cell-phones, laptop
computers and rovers, ostensibly uncertified by any government
authority.
On April 17, 2000, Guy Lancaster wrote more about the Diebold AccuVote
internal modem: We use whats called blind dialing (ATX0) which means
that itll dial with nothing plugged into it. Thus if the AV wont work without
this Dial Tone Emulator, then its doing something in addition to providing a
dial tone.
Dr. David Dills webmaster, Greg Dinger, arranged for a friend
to assist as an official pollworker. Shortly after the 2003 recall election,
Dinger filed this report:
OK, I have some news, At the end of the day, the head of the scanner
was removed from the base. It was connected to some sort of cell phone for transmitting
the results. ... It was wireless. ...
During the transmission process, errors occurred. The phone apparently reported
that a ballot was stuck in the reader. The precinct folks confirmed that
this was not the case. There was a phone call placed to some support number
which turned out to be a bad number. The lead precinct worker happened to have
another phone number, reached some unidentified (to my friend) person, and eventually
resolved the issue after a lengthy delay. ...
... The precinct leader was provided a cordless phone of some sort. At the
end of the day, she pulled the scanner out of the base and moved it to a table. Then
the phone was attached (as I understand it) with a short cable. I do not believe
the unit was built into the scanner, nor was it connected during the day. 3
Yeesh. A well-financed operation can penetrate the voting system
with the right equipment and the correct information. Cell phones
connect to the access tower with the strongest signal. It is relatively
easy, though not inexpensive, to set up a rogue access tower. If you
do, this cell phone will automatically communicate with you. You
could connect the call to your own GEMS server, load the real results,
modify them and then call up the real GEMS server to upload
your results.
Black Box Voting 170
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
Volusia County, Florida:
John Ellis, hired as an analyst for Fox TV News, knew exactly
the margin needed to call the 2000 presidential race for George W.
Bush. He was privy to the Voter News Service data, and he spoke
several times during the evening to his two cousins: Jeb and George
W. Bush.
At 2:09 a.m., the required vote margin appeared from Volusia
County, Florida. At 2:10 a.m., this margin was enhanced by a 4,000-
vote bump in Brevard County, and at 2:16 a.m., Ellis called the race
for George W. Bush. Within four minutes, NBC, CBS and ABC
followed suit. 5
Precisely the right margin appeared on a Volusia County machine
(Global Election Systems, now Diebold), amplified by a Diebold/
Global Election Systems machine in Brevard County. Unfortunately,
these vote totals were incorrect and soon disappeared, along with a
card no. 3 which helped to create them.
If Al Gore had publicly conceded on election night, would we ever
have learned that these votes were bogus? Would there have been a
recount, and could the Help America Vote Act have passed, triggering
the rush to touch-screen machines?
Well never know, but thanks to an internal CBS report and a memo
written by Talbot Iredale, vice president of research and development
at Diebold Election Systems, we now know that the unexplained
replacement of a set of votes on a Diebold optical-scan machine in Volusia
County triggered a premature private concession from Al Gore to George
W. Bush and resulted in TV networks erroneously calling the election
for Bush instead of deeming it too close to call. The final offical
tally showed Gore losing by 527 votes, though the hand recount stopped
by the Supreme Court later gave the election to Gore.
Fox News Network, 29 November 2000: Brit Hume, host: ... It seems a
broken computer modem and a faulty memory card were culprits in the
erroneous election-night call of George W. Bush as the Florida winner ...
computers with a bad memory card caused it to appear for a time that Al
Gore had lost more than 16,000 votes, which seemed to put George W.
Bush up by 50,000 at that stage in the night, an insurmountable margin.
Every network saw that as a basis for calling the state for Mr. Bush. ... 6
Security Breaches 171
This free internet version is available at www.BlackBoxVoting.org
Was it a bad memory card that produced the 16,000-vote spread?
Or is there another explanation? And is it true that these 16,000 mystery
votes caused the networks to call the election for Bush?
Lets look at the symptoms of a bad memory card. A memory card,
as youll recall, is like a floppy disk. If you have worked with computers
for any length of time, you know that a disk can go bad. When
it does, which of the following is most likely:
a) In an Excel spreadsheet that you saved on the bad disk, is it
likely to read a column of numbers correctly the first time: 1005,
2109, 3000 ... but the second time, replace one of the numbers like
this: 1005, 2109, 16,022 ... ?
b) Or is it more likely that the bad disk will do one of the following
things: Fail to read the file at all; crash your computer; give
you an error message; or make weird humming and whirring noises
while your computer attempts unsuccessfully to read the disk?
For most of us, the answer is b). But according to news reports,
the official explanation from Global Election Systems was that a bad
memory card reported votes correctly in every race except the presidential
race, where it changed Gores total to minus 16,022.
This kind of explanation gets my nose twitching. Really? Is that
what a bad memory card does? If so, how many bad memory
cards have been out there changing vote totals, unbeknownst to
voters?
If the symptom of a corrupted memory card was arbitrary votechanging,
as explained to the media in Volusia County, wed be in
real trouble according to Diebold sales representative Steve Knecht
in a March 24, 2000, memo: Cards were corrupted throughout California
at a rate exceeding our normal 1 in 100 that weve been seeing.
Marin is now up to 8 cards corrupted out of 114.
With these numbers, wed better hope that the symptoms do not
include randomly changing the vote totals.
According to an exchange between principal engineer Ken Clark
and Donna Daloisio, who was systems administrator for Supervisor
of Elections Gertrude Walker in St. Lucie County, Florida, the following
symptoms typify a corrupt memory card:
When beginning to upload results the following message appears:
Please re-insert memory card. If you take the memory card out
and put it back in, you are likely to see this error: Pct data error
Black Box Voting 172
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
OK to continue? If you say yes, this message appears again: Please
re-insert memory card.
When Daloisio described these symptoms, principal engineer Ken
Clark shot back this diagnosis: Garden variety corrupt memory card.
The Diebold memos reveal that the story given to the media about
Volusia Countys sudden vote discrepancy isnt quite the whole story.
On January 17, 2001, Volusia County employee Lana Hires asked
the technical staff at Global Election Systems for help. She was being
put on the hot seat over Al Gores strange tally of negative 16,022 votes.
I need some answers! she wrote. Our department is being audited
by the County. I have been waiting for someone to give me an explanation
as to why Precinct 216 gave Al Gore a minus 16022 when it was
uploaded. Will someone please explain this so that I have the information
to give the auditor instead of standing here looking dumb. 7
Global Election Systems John McLaurin tossed the hot potato to
Sophia Lee and Talbot Iredale. Sophia and Tab may be able to shed
some light here, keeping in mind that the boogie man may me [sic]
reading our mail.* Do we know how this could occur? 8
Talbot Iredale, senior vice president for research and development,
explains: Only the presidential totals were incorrect. Iredale then
hits us with this bombshell: 9
The problem precinct had two memcory [sic] cards uploaded. The second
one is the one I believe caused the problem. They were uploaded on the
same port approx. 1 hour apart. As far as I know there should only have
been one memory card uploaded.
Where did this second card come from? Iredale then gives a cursory
nod to the official explanation given to the media:
Corrupt memory card. This is the most likely explaination [sic] for the problem
but since I know nothing about the second memory card I have no ability
to confirm the probability of this.
Again, where did the second card come from?
Invalid read from good memory card. This is unlikely since the candidates
*Thats a damn curious remark!
Security Breaches 173
This free internet version is available at www.BlackBoxVoting.org
results for the race are not all read at the same time and the corruption
was limited to a single race. There is a possiblilty that a section of the memory
card was bad but since I do not know anything more about the second
memory card I cannot validate this.
Theres that pesky second card again. He then suggests that perhaps
the second card might have been well another way to say
this would be election tampering, I guess:
Invalid memory card (i.e. one that should not have been uploaded). There
is always the possiblity that the second memory card or second upload
came from an un-authorised source.
So, who is investigating this unauthorized source?
If this problem is to be properly answered we need to determine where
the second memory card is or whether it even exists.
But it turns out that this second card did exist:
I do know that there were two uploads from two different memory cards
(copy 0 (master) and copy 3).
There were two uploads from two different cards.
The votes were uploaded on the same port about 1 hour apart.
Only one memory card was supposed to have been uploaded.
Copy 0 uploaded some votes.
Copy 3 replaced the votes from Copy 0 with its own.
Iredale believes the second one is the one that caused the problem.
The problem: 16,022 negative votes for Al Gore
We know that the problem was noticed and corrected. An election
worker noticed Gores votes literally falling off the tally, and
the number of votes in Precinct 216 was totally out of whack. Eventually,
a manual recount was done. No harm, no foul?
That depends on how you look at things. I found a report called
CBS News Coverage of Election Night 2000: Investigation, Analysis,
Recommendations prepared for CBS News. 10
Black Box Voting 174
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
It would be easy to dismiss the bizarre events of Election Night
2000 as an aberration, as something that will never happen again,
the report begins. ... But, this election exposed flaws in the American
voting system, imperfections mirrored in televisions coverage of the
election results.
Yes. This election exposed flaws, but the imperfections were not
really quite mirrored in televisions coverage of the results. A more
apt metaphor would be that the imperfections exposed the tip of an
iceberg and then, with the HAVA bill, everyone in America decided
to buy a ticket on the Titanic.
It is, as one of the computer scientists Ive talked with likes to
say, like The Amazing Randi. Dont look there look here! An
illusion. Ridicule the dangling chads. Voter News Service blew it.
Dont worry, we caught that crazy error of minus 16,022 votes, it
made no difference. Well give you the Help America Vote Act (HAVA)
and promise $3.8 billion (much of which may never materialize) to
prevent this fiasco from ever happening again.
Look over here: Chads are bad. Look over there: Lets vote on a
black box!
Dont look there: No one paid much attention to the optical-scan
machines, which, we now know from Greg Palasts research, used
different settings depending on whether you were in a minority district
or an affluent suburb. White? Suburban? Set the machine to
provide an error message if the ballot was overvoted, so the voter
can correct it. Minority? Poor? Accidental overvotes discarded, thank
you. Back that up with statistics, of course: Too dumb to vote. 11
While we fixated on a butterfly ballot, no one asked about the
GEMS program, or demanded to see card number 3 from Volusia
County, or asked who made this card and how it got past all the
election procedures and physical security, or whether any other counties
had a card number 3.
Here is a chronology of how the election was called for Bush.
You decide whether card number 3 made a difference: 12
7:00 PM: CBS News estimate, based upon exit-poll interviews,
shows Gore leading Bush by 6.6 percentage points.
7:40 PM: Voter News Service (VNS) projects Florida for Gore.
7:48 PM: NBC projects Florida for Gore.
7:50 PM: CBS projects Florida for Gore.
Security Breaches 175
This free internet version is available at www.BlackBoxVoting.org
7:52:32 PM: VNS calls Florida for Gore.
8:10 PM: CBS News analysts recheck the Florida race and feel
even more confident about the call for Gore.
9:00 PM: A member of the CBS News Decision Team notices a
change in one of the Florida computations. One of the estimates, the
one based solely on tabulated county votes (voting-machine results
rather than exit polls) is now showing a Bush lead.
9:07 PM: VNS reports vote data from Duval County that put Gore
in the lead. This was then deemed to be an error.
9:38 PM: VNS deletes the Duval County vote from the system.
Gores total in Florida is reduced by 40,000 votes.
10:00 PM: CBS withdraws the Florida call for Gore.
10:16 PM: VNS retracts its Florida call for Gore.
At some point between 10:16 p.m. and 1:12 a.m., Bush took the lead.
1:12 AM: Associated Press, which collects its numbers separately
from VNS, shows the Bush lead dropping. VNS differs. Correspondent
Ed Bradley begins warning people in the CBS studio of irregularities.
1:43AM: Bradley points out that more than 30 percent of the vote is
still uncounted in Dade and Broward counties, Democratic strongholds.
1:48 AM: Bradley does the math: Bush is ahead by 38,000 votes.
And still out there, about 5 percent of the vote is still out, 270,000
votes. So thats a big chunk of votes. Bradley seeks more information
from the AP wire and from CBS News correspondent Byron Pitts.
2:00 AM: According to VNS, Bush leads by 29,000 votes. Heavily
Democratic counties have not weighed in yet. Ed Bradley is talking
about the AP reports, but CBS is not using that information.
2:09 AM: VNS adds Volusia Countys incorrect numbers to its tabulated
vote. This change increases Bushs VNS lead to 51,000 votes.
2:09:32 AM: Bradley sounds an alarm, but no one pays attention:
Among the votes that arent counted are Volusia County. Traditionally
theyre
one of the last counties to come in. Thats an
area that has 260,000 registered voters. Many of them are black and
most of them are Democrat.
2:10 AM: Brevard County omits 4,000 votes for Gore.
Bushs lead in the VNS count includes 16,000 negative votes for
Gore and unspecified other voting problems such that Bushs lead
appeared to increase by 20,000 votes in Volusia (plus the 4,000 missing
from Brevard).
Black Box Voting 176
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
According to the CBS News report: These 24,000 votes would
have nearly eliminated the 30,000-vote final Bush margin the CBS
News Decision Desk has estimated. There would have been no call
if these errors had not been in the system.
2:16 AM: John Ellis calls Florida for Bush.
2:16 AM: NBC calls Florida for Bush.
2:16 AM: The AP lead for Bush drops by 17,000 votes, to 30,000.
This 17,000-vote drop, occurring in only four minutes, is a Volusia
County correction. But VNS does not use the correction, and no one
at CBS is listening to Ed Bradley or watching the AP wire.
2:16:17 AM: Dan Rather talks with Bradley about the large number
of votes still out in Volusia County.
2:17:52 AM: CBS calls Florida for Bush.
2:20 AM: ABC calls Florida for Bush.
2:47 AM: The AP reports that Bushs lead has dropped to 13,934.
2:48 AM: VNS still shows the Bush lead at 55,449.
2:51 AM: VNS corrects part of its Volusia error, and Bushs lead
drops to 39,606.
2:52 AM: The AP reports the Bush lead down to 11,090.
2:55 AM: Palm Beach County weighs in with a large number of
votes, and VNS reports the Bush lead down to 9,163.
3:00 AM: Rather preps viewers for a Gore concession speech: We
havent heard yet from either Al Gore or from the triumphant Governor
Bush. We do expect to hear from them in the forthcoming minutes.
3:103:15 AM:* Al Gore telephones Mr. Bush to concede.
3:10 AM: CBS begins investigating the VNS numbers. It also,
finally, begins watching numbers from the AP. CBS also looks at
the Florida Secretary of States Web site. The three sets of numbers
dont match, but all of them indicate the race is much closer.
3:32 AM: From 3 a.m. until now, there is much talk about the
expected Gore concession speech.
3:30-3:45 AM:** Gore boards a motorcade for a 10-minute journey
to War Memorial Plaza in Nashville, Tennessee, to deliver a concession
speech to the nation.
3:40 AM: Bushs VNS lead drops to 6,060 votes.
* At this point I have drawn the timeline from three sources: CBS report, an
Agence France-Presse 13 (*) report, and Dow Jones News 14 (**). The events
reported after 3 a.m. sometimes differ by a few minutes between these reports.
Security Breaches 177
This free internet version is available at www.BlackBoxVoting.org
Around this time, Gore Campaign Chairman William Daley places
a call to CBS News President Andrew Heyward. Daley asks whether
CBS is thinking about pulling back its call for Bush. Heyward wants
to know what Gore is planning to do.
According to the CBS report, Daley says, Ill get right back to
you, hangs up and does not call back.
When the lead is down to 6,000 votes, Daley asks whether CBS
is thinking about pulling its call for Bush. The answer is they want
to know what Gore is planning to do. Is it just me, or does this
response bother you?
3:48 AM: Rather says, Now the situation at the moment is, nobody
knows for a fact who has won Florida. Far be it from me to
question one of our esteemed leaders [CBS management], but somebody
needs to begin explaining why Florida has now not been pulled
back to the undecided category. ... A senior Gore aide is quoted by
Reuters as confirming that Gore has withdrawn [his] concession in
the U.S. President race.
3:45-3:55 AM:* Two blocks away from the plaza, Gore field director
Michael Whouley pages traveling chief of staff Michael Feldman
to tell him the official Florida tally now shows Bush up by just 6,000
votes, with many ballots left to be counted. By the time the Gore
motorcade reaches the plaza, according to Agence France-Presse, he
is down by fewer than 1,000 votes. Gore did not, then, give the speech
he had planned to give. Instead he consults with his staff.
3:57 AM: According to CBS, the Bush margin has narrowed to
fewer than 2,000 votes. CBS News President Heyward orders that
CBS News retract the call for Bush.
By 4:05 AM: The other networks have rescinded the call for Bush.
4:10 AM: According to CBS, Bushs lead drops to 1,831 votes,
which is roughly where it remains until the first recount.
4:30-4:45 AM:** Gore makes a second telephone call to Bush to
retract his concession, saying that he is waiting for all the results
from Florida.
5:05 AM:** A Florida election official announces a recount, with
the two candidates separated by a few hundred votes.
According to the CBS report, the call for Bush was based en-
Black Box Voting 178
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
tirely on the tabulated county vote. There were several data errors
that were responsible for that mistake. The most egregious of the data
errors has been well documented. Vote reports from Volusia County.
Four thousand votes for Gore were omitted from the tabulation in
Brevard County, and in Volusia, 4,000 votes were erroneously counted
for Bush and 16,022 negative votes were recorded for Gore.
The mistakes ... which originated with the counties, were critical,
says the report. They incorrectly increased Bushs lead in the
tabulated vote from about 27,000 to more than 51,000. Had it not
been for these errors, the CBS News call for Bush at 2:17:52 AM
would not have been made.
If you strip away the partisan rancor over the 2000 election, you
are left with the undeniable fact that a presidential candidate conceded
the election to his opponent based on results from a second
memory card (card #3) that mysteriously appears, subtracts 16,022
votes, then just as mysteriously disappears. If this isnt disturbing
enough, consider these three points:
1) We dont know whether this was an isolated incident. It may have
occurred elsewhere, but in smaller, less spectacular totals.
2) The errors were correctable because paper ballots existed and a
100 percent audit was done.
3) The fact that negative votes could be applied to a candidates
total demonstrates such a fundamentally flawed software model
that it calls into question the competence and integrity of the programmers,
the company and the certification process itself.
The Diebold Memos
During the middle of the night on Friday, September 5, 2003, a
set of memos leaked into my FTP server. They originally came from
a person with inside access to the Diebold server who used an employee
I.D. number to obtain and copy them.15 The memos were first leaked
to David Allen on July 29, but, because of technical problems, he
says he was unable to extract them from their compressed .tar file.
On Saturday, September 6, I downloaded this 15,000-document
tarball, found a utility to open it and started reading. I read 7,000
memos and made 300 pages of notes, divided into five categories,
Security Breaches 179
This free internet version is available at www.BlackBoxVoting.org
and didnt come up for air until Monday, September 8.
What I found was not good.
While the certified version of the voting software sat in escrow at
the secretary of states office, unauthorized versions were being put
on the FTP site and, from there, downloaded and installed, overwriting
the approved software. The memos documented this.
I therefore did three things. I made a copy of everything and put
it into the hands of someone I trust. I burned the memos onto a CD
and met face-to-face with a U.S. congressman,16 who asked questions
for 30 minutes, and then took the memos to Washington. And I
selected 24 memos that describe the practice of substituting unexamined
software for the approved version, and posted them on my Web site. 17
While writing up my notes on the memos, I discovered a curious
thing. I wanted to find out whether the software they were uploading
(and using in elections) was certified or not. But for some reason,
links to the official certification list had been pulled off both NASED
and The Election Centers Web sites. I was able to locate an out-ofdate
version but could find nothing current.
The more I hunted, the odder it looked. If this whole system is
based on certification, why is it so hard to find out which versions
are certified? It should be a simple matter to compare the NASED
certification number with the version number used in elections.
Andy Stephenson, a researcher who worked with me on this project,
called R. Doug Lewis of The Election Center to ask for the certification
list. Lewis hung up on him. I called the Washington State Elections
Division and was told someone would have to call me back. No
one did. Linda Franz, of Citizens for Voting Integrity-Washington, found
one document through an obscure link on the Johnson County, Kansas,
Web site. By searching with an Internet tool that archives old Web pages,
I found three reports, which had been pulled off the Web.
Looking at every ITA-tested, NASED-authorized version number
available leads to only one conclusion: Diebold has been putting unauthorized
software into our voting systems.
When Georgia Secretary of State Cathy Cox says of ballottampering,
It would take a conspiracy beyond belief, of all these
different poll workers
I dont see how this could happen in the
real world, shes dead wrong. If you can slip uncertified software
into voting machines, it takes only one person, working alone at night.
Black Box Voting 180
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
The Diebold memos made their first public appearance on
BlackBoxVoting.org in the form of 24 memos with a commentary
about Diebolds use of uncertified software. Everyone yawned except
Diebold, which issued a cease-and-desist.
You might be yawning now. Someone used version 1.18.14 instead
of version 1.17.23. Ho hum.
Except for this: A programmer never changes a version number
unless he changes the underlying computer code. If the versions are
never submitted for certification, only the individual who programmed
the change knows for sure what he put in there.
Certification is the foundation of the voting industry. Remove it,
and the whole house of cards tumbles.
Maybe theyll say it was just to fix bugs, one reporter suggested.
Sure. And theyll say it was just to add features or to create a new
report format. And that might all be true, but before you breathe a
sigh of relief, let me give you a taste of just how out of control this
problem has been:
As far as we know, some guy from Russia could be controlling
the outcome of elections in the United States, Lynn Landes wrote.
Lynn, meet Dmitry Papushin, some guy from Russia. He is one of
five Diebold programmers who have been putting programs on the FTP
site. Take a look at his memos. What hes doing here is placing uncertified
software versions on a Web site, and people are using it.
Underlines represent versions that were never certified, or the implementation
of poor security procedures.
18 Jan 2000 memo from Dmitry Papushin: GEMS 1-14-5 is ready.
January 25, 2000 memo from Steve Knecht: Will all future 1.14x versions
be compatible with 1.14.5 if we burn mem cards in San Luis Obispo now?
15 January 2003 memo from Dmitry Papushin: Ballot Station 4-3-14 for
Windows CE and Windows NT are ready.
10 February 2003 memo from Cathy Smothers: Can anyone send me
the BS CE 4.3.14 .ins file? I have a demo tomorrow and I need this to
upgrade the TS units.
From 1999 to 2003, Papushin uploaded more voting-system software
Security Breaches 181
This free internet version is available at www.BlackBoxVoting.org
onto the unprotected Diebold Web site than any other programmer.
Papushin has been a keeper of the passwords and the king of singlesentence
memos. He knows the voting-system programming intimately
and has uploaded computer code that programs your smart card,
captures your votes at the polling place and accumulates and reports
them at the county.
His programming skills and his ability to distribute programs to
techs and county officials make him a tempting target for bribery.
We assume that Dmitry Papushin has integrity and ironclad ethics.
But to deter the unscrupulous from making inappropriate solitications
to programmers like Papushin, we need to enforce regulations which
require that only authorized software be used, and we need fraud-deterring
audit procedures.
* * * * *
To examiners of the Diebold files, Ken Clark has become somewhat
famous for his blunt writing and ethical shortcuts. Clarks comments
in the touch-screen source code are quite a hoot, though not inspiring
of confidence in the touch-screen system:
the BOOL beeped flag is a hack so we dont beep twice. This is really a
result of the key handling being gorped. (WriteIn.cpp,v)
this is completely screwed up. the iIndex calculations are incorrectly based
on nybbles for some unknown reason, and so the offsets are incorrect. This
works only because the offsets are also incorrect when the card is read.
(VoterCard.cpp,v)
Reserve place in hell for person who renamed CRace and friends to
CRaceKey. (BufferedSocket.cpp,v)
Add and comment out code to work around bogus -1 in ballot level IDs.
(BallotRstDlg.cpp,v)
this is a sick hack to parse out a jurisdiction from a multi-line election
title. The jurisdiction field should be eliminiated altogether and this code
removed. This whole section is fairly broken wrt GEMS. GEMS doesnt
store the election information in any kind of multilingual sense, let alone
rich text. For now just stuff the english into all languages. (BuildElecDlg.cpp,v)
Deep magic is not working? Tried input of 6 and got back 1 (CIssue.CPP,v)
The scaling stuff is complete voodo.[sic] Trust me or rewrite it to make
more sense. (TextCell.cpp,v)
Black Box Voting 182
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
I justify the label by saying the existing code was crap structurally to begin
with. (Votercard.cpp,v )
In a July 1999 memo, Ian Piper wrote, What is GEMS written
in? Clark replied, GEMS is written in my office.
So Clark programs the GEMS system that accumulates and reports
votes from polling places. We have weak and sometimes unenforced
procedures for comparing polling-place results with the county
tabulations, so the GEMS program is a tempting target.
Clark repeatedly advises field technicians to skirt U.S. election
law pertaining to using only certified software versions.
From: Cathi Smothers, June 05, 2000, to Ken Clark: How do I know which
version of GEMS (i.e. 1.16.3, 1.16.4, etc.) to use?
From: Ken Clark, 5 Jun 2000: ... Baring any certification issues, the latest
stable release is what you want to upgrade accounts to ... Right now 1.16.latest
is considered stable, 1.16.4 being the current release by my mail ... Its fair to
say the nature of this company and business make this process fairly informal,
perhaps more so than I would like. Testing releases go out to customers when
they shouldnt, and new features get added to stable branches when they shouldnt
... Sometimes a bug slips into a stable branch, in which case its better to ship
a version you trust, or wait for it to get corrected ... The DLL files shipped on
the GEMS CD get updated from time-to-time as well ...
From: Ken Clark, 6 Jul 1999: I hate more than anyone else in the company
to bring up a certification issue with this, but a number of jurisdictions require
a system test before every election ... That is why the AccuVote displayes
[sic] the silly ***System Test Passed*** message on boot up instead of memory
test passed, which is all it actually tests. No argument from me that it is
pointless. You could probably get away with a batch file that prints system
test passed for all I know.
From: Ken Clark, 7 Jan 2000: *Any* testing we can do on 1.14 is a good
idea. With the risk of sounding alarmist, 1.14 really needs more testing. Even
though much of GEMS looks the same from the outside, the guts changed
substantially between 1.11 and 1.14. Thats why you see all kinds of things completely
unrelated to shadow races broken in the early 1.14 releases.
Security Breaches 183
This free internet version is available at www.BlackBoxVoting.org
From: Steve Knecht, 14 Jan 2000: Is it the intention of development staff
that California March election will be run on some version of 1.14 or will we
end up in the 1.15 range ...
(Answer from Ken Clark, 14 Jan 2000): Needless to say, the changes were
extensive. The paint is still wet ...
* * * * *
We know nothing at all about Whitman Lee, another Vancouver
programmer who uploads software to the FTP site. His memos consist
of fixing bugs and uploading replacement software, which he did three
dozen times over a 24-month time period.
From: Whitman Lee
GEMS 1-5-3 is ready to download.
GEMS has evolved so many times that it breaks some of the pre-election
reports...
Here is the latest changes since 1.5.8.
AVTS-3-4-1.zip is up. Here is this short upgrade instructions.
The password for ATTemplate-3-4-1.zip is msd8sdh3isohr.
GEMS-Reports-1-9-6.zip is ready for download.
GEMS 1.11.2 is ready.
The fix will be in GEMS-1-18-9.
* * * * *
Guy Lancasters specialty is programming the optical-scan system.
He has also arranged for uncertified software to enter our optical-
scan machines. A lot of it, apparently.
From: Guy Lancaster, 27 Jan 1999: For those romantics that fell in love
with 1.94f, the latest flavor incorporates all the changes made since 1.94f<.
This includes the changes in 1.94q, r, s, t*, and u. Pass your orders on to
McKinney. * Note: The PC 1.94t release was preempted by 1.94u.
From: Guy Lancaster, 25 Feb 1999: Our latest 1.94 releases have been
rushed out to fix a bug that slipped out with 1.94u and 1.94f>. Namely,
1.94u and 1.94f> fail to detect unvoted ballots and therefore cannot return
blank ballots ...
Black Box Voting 184
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
From Ken Clark, July 2, 2002: You have to be careful when talking about
1.94f firmware. There is a symbol after each f that corresponds to the
actual release of the firmware. They very probably have a 194f that is in
reality r, s, t etc.
From Don Bizmaier, Support services specialist, July 2, 2002: I am not
sure where they came from ... but Jeff Co [Jefferson County] KY uses an
"S" chip in the Absentee and Mail in AVOS [Diebold optical scan machine]
to ignore sequence nimbers [sic].
* * * * *
Talbot Tab Iredale, senior V.P. for research and development,
was hired when the company was still called North American Professional
Technologies.18 He oversees the programming. You will find his work
in the most sensitive sections of the touch-screen source code, and
he also programs the Windows CE operating system.19 According to
Rob Behler, Iredale wrote the Georgia Windows CE patches.
From Ian S. Piper 12 Jul 2002 re: Windows CE changes: Upgrading from
WinCE 3.0 (June 7th edition) to WinCE 3.0 (July 5th edition.) When upgrading
from the June 7th edition of WinCE 3.0 to the July 5th edition of
WinCE 3.0 (were ignoring the July 2nd and July 4th editions), the settings
should remain in the internal Flash memory ...
From Tari Runyan Tue, 2 Jul 2002: Is it necessary to upgrade at this point
- Early voting starts in 1 month and I am hesitant due to tight timelines ...
(reply from Talbot Iredale 16 Aug 2002): Yes, it is recommended to upgrad
[sic] all units to this version.
From: Rodney D Turner , 31 Aug 2000: The computer for LA has GEMS
1-16-9 and the AVTS units have 3-13-1-4. The computer for Alameda has
GEMS 1-16-10 and GEMS 1-16-9 ( there is a short-cut on the desktop for
GEMS 1-16-9) the AVTS units have 3-13-1-4.
From: Talbot Iredale, 31 Aug 2000 Re: Software for Los Angelas [sic], CA
Jeff and Rodney, LA and Alameda will need a revised version of GEMS
and maybe BallotStation to support the import/export that they require. I
am working on it now but I am certain there will be more changes.
Security Breaches 185
This free internet version is available at www.BlackBoxVoting.org
From: Larry Dix, 31 Aug 2000 RE: Software for Los Angelas [sic], CA Tab
Would you be willing to venture an outside guess as to when the revised
GEMS version will be ready. This really becomes an issue since I need to
coordinate staff to be onsite. Is this also the case for Alameda? Coordination
of time and staff is everything on these 2 installs.
From: Ken Clark, 31 Aug 2000 Subject: RE: GEMS-1-17-1: Is this a testing
release or not?
From: Talbot Iredale, 31 Aug 2000 Subject: Re: GEMS-1-17-1: This is no
more of a test release than 1.16.9 was though I would not be surprised if
we have to make more changes to fully support LA and Alameda.
From: Talbot Iredale, 29 Oct 2002: ... We have found a serious bug in
GEMS 1-18-14 ... We will be releasing a GEMS 1-18-15 that fixes this
bug within the next 2 days. Please ensure that all accounts that are using
GEMS 1-18-14 upgrade to GEMS 1-18-15* prior to the election.
Here are a couple odd memos pertaining to an uncertified version
number popping up on the screen in Florida:
From: Greg Forsythe, 17 Feb 2000: Just received a call from Beverly Hill,
Alachua County [Florida] ... She is at the SA screen and the version is
1.92-15 ... This copy has 1.92-14. 1.92-14 is certified, 1.92-15 is not.
SOLUTION REQUIRED!
From: Greg Forsythe, 17 Feb 2000: ... Solution might be to make the copy
the official database showing the correct version.
From: Nel Finberg, 17 Feb 2000: The problem has been fixed.
From: Nel Finberg, 17 Feb 2000: ... It would be a good idea to get rid of
the original diskette in order to remove the perception of version conflicts.
From: Don Vopalensky, 12 Sep 2002: Ken,Texas now requires ITA certification
... ITA certifications, state certifications, and time constraints play
a big part in what needs to be done, and sometimes that means putting
fixes or additions back into versions that are already in use ...
Black Box Voting 186
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
Nel Finberg September 25, 2002: What will be run in Texas will depend
on the outcome of the Texas certification decision, wont it?
From: Ken 25 Sep 2002: Hard to say. It never has in the past.
Tari Runyan, July 15, 2002: this bug affects Co [Colorado] - primary Aug
13 and Ga [Georgia] Primary Aug 20 are we proposing to upgrade again
this close to an election?
From Ken - That would be up to you.
Jeff Hallmark, October 08, 2002: RE: Tippacanoe, IN upgrade to 4-3 first
then downgrade to 4-1-11, this is quite fast if one sets up 3 or 4 machines
at a time. no backdoor humm..
Sue Page, October 16, 2002: - Maryland ... We had some units that we
downgraded from 4.1.11 to 4.1.6.
From Tyler to Ken Clark, 15 Feb 1999: ... But then again, with regards to
the entire NASED certification process, I can never quite get a handle on
the relationship between ostensible and reality.... :-)
Unknown programmers were putting unlooked-at code into our software.
Probably, these programmers are honest, but we really have no
idea. Regardless, it is idiotic to put seven million votes into the hands
of a few unknown people without even doing a backBlack Box Voting
© 2004 Bev Harris Bookstore & Library Edition
Rights reserved to Talion Publishing ISBN 1-890916-90-0. To purchase paperback
copies of this book, request the book at your local bookstore or library, or
call 425-228-7131, fax 425-228-3965, or e-mail talion@ix.netcom.com.Black Box
Voting © 2004 Bev Harris Bookstore & Library Edition
Rights reserved to Talion Publishing ISBN 1-890916-90-0. To purchase paperback
copies of this book, request the book at your local bookstore or library, or
call 425-228-7131, fax 425-228-3965, or e-mail talion@ix.netcom.com.ground
check.
This problem isnt limited to Diebold, and it isnt just the vendors.
County officials may or may not know the implications of using uncertified
version numbers, but state officials absolutely know better. Washington
State elections director David Elliott said on the Dave Ross radio show
that, in Washington State, voting software is certified by ITAs like Wyle.
However, Washington has been using versions never certified by an ITA.
Security Breaches 187
This free internet version is available at www.BlackBoxVoting.org
Dozens of us went looking for the most recent ITA and NASED
certification numbers, and all of us came up empty-handed. Finally,
after asking Elliott for the NASED certification list in front of members
of the Washington legislature (I did this while giving public testimony,
a breach of protocol which mortified the legislators and resulted in
a reprimand),21 I was able to obtain a supposedly complete set of
NASED certification documents.
I say supposedly because someone took a pen and wrote a version
number on it that wasnt there originally. In another case, a version
number was overwritten in pen but its approval number was not
overwritten, resulting in a mismatch.
I wondered why former Washington Secretary of State Ralph Munro,
who is now chairman of VoteHere, signed off on GEMS version 1.11.2.
(Or did he?) Why did current Secretary of State Sam Reed sign off on
GEMS 1.18.18. (Or did he?) Whos minding the store here, guys?
Here is an e-mail from Dean Logan, former Washington state
elections director, now director of King County Records, Elections
and Licensing Services:
Tuesday, November 25, 2003 From Dean Logan to Andy Stephenson: King
County currently operates GEMS Software version 1.18.18. This version
of the software was installed in August 2003 and was used for the September
16, 2003 Primary and the November 4, 2003 General Election. GEMS
Version 1.18.18 was certified by the Washington Secretary of State on August
12, 2003 ... From 1999 through August 2003, King County operated
GEMS Software version 1.11.2 ... No additional software patches or upgrades
have been installed or in use in King County.
(Neither 1.18.18 nor 1.11.2 had a NASED number.)
Wed, 03 Dec 2003 From Bev Harris to David Elliott: As I mentioned to
you after the meeting, I was astonished when Andy Stephenson, Democratic
candidate for Secretary of State, discovered that King County, Washington,
has been using uncertified software for four years ... In the case of
King County -- perhaps there is some old, obscure NASED document that
has never been published, which goes out of order numerically, and which
is for some reason missing from the discovery documents obtained for a
recent citizen lawsuit?
Black Box Voting 188
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
You did tell me today that you have the complete set and that you will
provide it. I really appreciate your help. Because you sat on the NASED
committee for so many years and according to your testimony today, even
helped to write the standards, your expertise is quite valuable and therefore
I would like to get your signature with the documents indicating that
what you provide does represent the complete set of version numbers ...
Dec 4, 2003 From David Elliott to Bev Harris: ... I can not offer a complete
set of the NASED information although they serve as a continuous log of
the work done by the ITA's. The last should provide all that came before it.
I have contacted the FEC about whether there is an update to the list past
6/5/03 and I was told that they are working on it.
Dec 4, 2003 From Bev Harris to David Elliott: Thank you for responding
so promptly ... I did not get your response until too late to pick up the documents.
But that's okay, because the answer is insufficient and perhaps we
can rectify that so that when I pick them up next Tuesday, we will have
what we need. Your response raises several questions:
(I have contacted the FEC about whether there is an update to the
list past 6/5/03 and I was told that they are working on it.)
This is not an acceptable answer. In fact, I can't tell whether or not any
versions were certified after 6/5/03 or not. Are there any new versions after
the June 2003 report, or not? ... I would assume that NASED would
have some system to inform the secretary of state when they certify new
versions ... why do you have to call the FEC for it?
I will arrange to come to Olympia by 4:30 p.m. on Tuesday, and at that
time I really do need ALL the NASED certified version numbers.
Dec. 8 2003 from David Elliott to Bev Harris: ... I am providing you with
copies of the lists that I have received. ... Concerning the Diebold certifications:
The optical scan reader was originally certified as the Global CF-
1 in 1991. In 1992 software called Vote Tally System VTS version 1.81 was
certified and the hardware was re-named the Accu-vote. The VTS software
was rewritten and renamed GEMS and was re-certified in 1998. GEMS
was re-examined again in 2001 as a part of the certification of the Accuvote
touch screen (DRE) finalized in September of 2002 as version 1.17.17.
The most recent examination was completed in July and August of 2003
for GEMS version 1.18.18.
Security Breaches 189
This free internet version is available at www.BlackBoxVoting.org
OK, now we get into some problems. GEMS 1.18.18 was used in
two King County elections in violation of Washington state law, since
it had not been used in elections elsewhere first. According to Dean
Logan, the previous version was GEMS 1.11.2, and there were no
upgrades or patches, but this cannot have been certified in 1998, as
Elliott implies, because it wasnt released until 1999. 22 And if only
1.11.2 and 1.18.18 were used, and there were no upgrades or patches,
why did Sophia Lee reference GEMS 1.17.16 in King County? 23
King County has been using unauthorized software for six years.
A whole bunch of citizens on the BlackBoxVoting.org forums went
after all the state certifications, and mostly got stonewalled, but we
did document the following problems:
Chelan County Washington is using GEMS version 1.17.21. Not certified.
King County Washington GEMS 1.11.2, 1.17.16, 1.18.18. Not certified.
Placer County used GEMS version 1.17.20. Not certified.
El Paso County Texas used GEMS version 1.17.21. Not certified.
Johnson County Kansas GEMS version 1.18.14. Not certified.
Alameda County California used GEMS version 1.18.13.9. Not certified.
Alameda County California also used GEMS version 1.18.14. Not certified.
Yavapai County Arizona optical scan version 2.0.11. Not certified.
Pima County Arizona optical scan version 2.0.11. Not certified.
City of Tucson Arizona optical scan version 1.94y. Not certified.
Johnson County Kansas optical scan version 2.00g. Not certified.
Yavapai County Arizona touch screen firmware 4.3.11. Not certified.
Los Angeles County California touch screen version 4.3.8. Not certified.
El Paso touch screen firmware version 4.3.9* Not certified at the time
Alameda County California touch screen firmware 4.3.11. Not certified.
Johnson County Kansas touch screen firmware 4.3.11. Not certified.
On December 16, 2003, the State of California held a certification
hearing for the Diebold TSx system. An 18-inch-thick pile of
copies containing the material you have just read was marched into
the middle of the meeting, and California Secretary of State Kevin
Shelley made an unusual personal appearance to express his displeasure.
The results of an audit ordered by the California voting-systems
panel were equally disturbing:
Black Box Voting 190
Black Box Voting © 2004 Bev Harris ISBN 1-890916-90-0
This free internet version is available at www.BlackBoxVoting.org
According to WiredNews, At least five California counties were
using versions of software or firmware that were different from what
Diebold had indicated they were using. All counties were using
uncertified software ... Marc Carrel, assistant secretary of state for
policy and planning, said he was disgusted by the situation and worried
that it could call into question any close races that might have occurred
in three counties that used federally unqualified software.
... And Im frustrated ... that were not going further today, he said.
There certainly needs to be something done to this vendor. 24
* * * * *
While California was busy clearing up its unauthorized software,
Washington state officials were denying that the problem existed.
Post-Intelligencer reporter Wyatt Buchanan got this answer when
he asked for a response to the allegations made in this chapter:
Elections officials said the two [Harris and Stephenson] had never
contacted them about their concerns. 25
Well no, I suppose not, unless you count six e-mails, a 20-minute
conversation with assistant elections director Paul Miller, a confrontation
with David Elliott in front of 12 witnesses outside Senate Room
1 in the Washington state capitol, and testimony before members of
the Washington state legislature with Elliot present. Buchanan ran a
correction, citing incorrect information given by the office of the secretary
of state.
Elliott, the Washington state elections director, told Tacoma News
Tribune reporter Aaron Corvin on December 17 that all software
used in elections has been certified by the state. 26
All of it?
OK. Stephenson and I will be picking up the signature pages for
those certifications.
* * * * *
After issuing a cease-and-desist for publishing 24 memos
documenting the use of unauthorized software, after a California audit
revealed that 100 percent of Diebolds machines used uncertified
programs and California threatened to revoke the companys right
to do business in the state, after dozens of journalists including two
documentary film producers and the AP wire covered this story, Diebold
Security Breaches 191
This free internet version is available at www.BlackBoxVoting.org
Election Systems President Bob Urosevich released a news statement:
Diebold Election Systems, Inc. (DESI), is announcing a complete restructuring
of the way the company handles qualification and certification processes
for its software, hardware and firmware. 27
Im sure someone will slap me for saying this, but Im reminded of
the sociopath who discovers his conscience as soon as he goes to trial.
It wasnt just Diebold. On the Mike Webb radio show, Snohomish
County (Washington) Auditor Bob Terwilliger admitted that one of
his staff members wrote a program modification for his Sequoia touchscreen
software. He said it was for the WinEDS central tabulating
software (the Sequoia equivalent of GEMS; this software was also
found on an unprotected Internet site during the fall of 2003).
What was this patch for? Just to help run some reports. But it has
to access the votes to do that. It just uses the database. But the database
consists of the votes. I trust the person who made the changes. 28
Lets suppose everyone in every jurisdiction is honest. Thats not
likely, but lets pretend. Here is an incomplete list of the bugs fixed
on just one release of the Diebold software:
From: Whitman Lee, 23 Jan 2003
- Items are not being recorded in the Audit Log.
- Connect to Data Base security crack.
- Ballots printed from Ballot Viewer do not correctly reflect selections.
- Non-administrator users can assign themselves administrator privileges.
- Races marked Not Counted can be downloaded and appear in reports.
- Regional users can unset the elections status.
- No error message is displayed if Confirm password and Password fields
dont match.
- Loading a file from backup results in the backup copy being deleted.
- Central Count drops a batch if a race on the ballot has headers.
- LA County Export: Election Id incorrect for Nov. 2002 election.
- Incorrect IP address displayed on the AccuVote-TS Server console.
- Resetting election on non-Host database gives Internal Error.
- Summary Reports Cards Cast totals are incorrect
- Incorrectly enabled functions and settings when election status is Set
for Election.
Its bugs like these that get elections wrong. Would you rather
Powered by vBulletin® Version 4.2.5 Copyright © 2024 vBulletin Solutions Inc. All rights reserved.