Avi Rubin - New Diebold Feature Discovered - Reboot and Lose

[WillYourVoteBCounted]

New Diebold Feature Just Discovered -If you reboot the Diebold touchscreen,
the memory pack loses the votes.

These same problems in Maryland appear to have happened in Gaston County NC.

Read Avi Rubin's report on the failure of Diebold memory packs to upload votes in
Maryland's recent primary:
http://avi-rubin.blogspot.com/2006/10/following-is-exce...

Here is a synopsis of what an election judge reported to Avi:

Republican Chief Judge in the September 12 primary in Montgomery County, Maryland reports:

One of the eight touch screen voting machines used on election day recorded no votes,
even though 55 voters were logged onto the machine.

The problem as explained by Diebold and the county BoE:

That particular touch screen machine had been rebooted, and
the memory card had been removed and reinserted into the machine.

The election judge was told that this was a "security" feature!

The Board of Elections, with the help of Diebold was able to recover the “missing” votes from the hard drive.

Gaston County North Carolina used the Diebold TS from 1998 to 2004 with serious problems reported. (I don't have data for 2000)

In both elections Gaston also had problems with the memory packs not recording the votes, and the central tabulator. The DESI technicians had to "manually" retrieve the votes. The central tabulators lacked the capacity or ability to receive vote totals.

In November 2004, in Gaston County NC, (a Diebold TS county) half of the precincts in Gaston County did not balance.

State to probe Gaston election
19 November 2004

The Observer reported Thursday that the numbers did not balance in more than half the precincts in Gaston County. Gaston elections officials said they had been unaware of the problem.

http://www.votersunite.org/article.asp?id=3915

Legal Documents Regarding the 2004 Diebold Problems
Correspondence between the NC State Board of Elections General Counsel and DESI's attorney
reveal many serious problems with the Diebold machines in the 2004 election

Two malfunctions of particular note:

15 Memory packs that didn't work, affecting 11,945 votes

There was a problem with 11,945 votes from Dallas, NC. Seven days after the election the County Board of Elections noticed that 11,945 voter were missing (the precinct rolls reported 12,867 where the GEMS system only showed 922).
Apparently, 15 early voting memory packs didn't work in the standard (AccuVote-TS) process of accumulating the memory packs onto one unit and then "downloading" them into the GEMS server.
That didn't work in a reproducible fashion.
These memory packs had to be directly loaded onto the GEMS server by a DESI technician.
This seems to have been done in a controlled manner.

Hand entering of ballots: finally, the document describes hand-entering of "three to five" ballots.

Here is the legal correspondence
http://www.josephhall.org/nqb2/media/GastonDiebold2004....

The full article describing the 2004 Gaston County Diebold meltdown here
http://josephhall.org/nqb2/index.php/2005/10/03/p691

1998 "Bugs" in Diebold TS in Gaston County Voting Machines and Central Tabulators -
I turned up similar problems with the memory packs and the central tabulator
capacity issue in the Diebold TS in Gaston County - dating back to 1998:

"In 1998, Gaston County was using direct-record electronic voting machines for the first time. Before the election, officials discovered some problems with the machines, and an upgrade was necessary. The machines were run through basic testing, but the upgrades cut into time for additional testing for known “bugs.”

Suddenly it was election night. The first sign of a bug appeared when the time came to report the totals. Nearly one-third of Gaston County’s precincts reported no votes because the cartridges that read the totals from the machines were not functioning. Another facet of this bug was evident for the county at large: when the vote totals reached 32,000, the tabulators would not tally any higher.

The chair of the county board of elections opened the process up to all who wished to observe, as the computer experts opened the machines and retrieved the source code that showed the proper vote totals.
The internal drives that showed how each voter had voted were intact and retrievable. Two statewide court of appeals races, one state House of Representatives race, and one local race were close enough that they were determined by the data retrieved from the machines. Because of the safeguard of public scrutiny, confidence in the way the matter was handled was high, and there were no election protests." (page 8)

http://ncinfo.iog.unc.edu/pubs/electronicversions/pg/pg...

How many times were voting machines re-booted in Maryland in the primary,
and what is to prevent that from being done on purpose?

[mom cat]

1. Paper only with a million eyes watching is the only solution.

[InsultComicDog]

14. Those big old mechanical machines

worked very well, I thought.

Plus, they were cool. When you opened that curtain with the big handle that made the votes register, it really felt like you made a statement. Like an exclamation point on your vote!

[OnTheOtherHand]

17. hey, don't talk about those machines in the past tense

We use them in New York. It's not that New Yorkers are afraid of technological innovation -- we just don't like to buy (expletive deleted).

And yes, they are very cool. All other considerations aside, what is the point of bringing one's children to watch one vote on a touchscreen? Sigh.

[texpatriot2004]

2. K & R for Transparent Democracy nm

[skids]

3. Don't forget the convenient early-in-the-day power surges. n/t

[GuvWurld]

4. Today's election is being brought to you by Banana Republic

[yellerpup]

5. Very fishy, that election. Rove's favorite...

Congressman Patrick McHenry became only the third person in forty years to represent the people of North Carolina’s 10th Congressional District with his victory in the 2004 election. The congressman won a competitive four-candidate primary and then an extremely close runoff election by 85 votes. The district stretches from the Blue Ridge Mountains to the South Carolina border. McHenry attended high school in Gaston, NC.

[htuttle]

6. "when...vote totals reached 32,000, ...would not tally any higher"

Somebody is using shorts when they should be using integers, I think (explanation for laypersons: some computer data types, short integers, overflow at 32,767 and go negative if you add any more to them. Apparently, their work around to avoid this was to simply stop counting at 32,000).

That's a SERIOUSLY rookie programming mistake! Like a first-year programming class type mistake!

[yardwork]

13. It's not a mistake. It's a feature!

Diebold makes the ATMs that my bank uses. Gee, isn't it interesting how they always manage to keep track of all the deposits and withdrawals. The only time they ever make a mistake it's in their favor.

[troubleinwinter]

7. K & R! "Nearly one-third of...precincts reported no votes"

[Kurovski]

11. troubleinwinter, good to see you!

[rumpel]

8. Has anyone done a comprehensive count of how many votes

were definitely lost and not counted in 2004 throughout the nation?

[diva77]

9. I wonder if this relates to when the tabulation goes into

"manual mode" in the wee hrs of the morning at hq on election nite as well. Anyone know if there's an explanation for "manual mode" ??

[Kurovski]

10. aside from everything else, and just on the most basic and understandable

level, these machines are absolute pieces of junk.

[cyberpj]

12. "and the memory card had been removed and reinserted into the machine."

Isn't this exactly what the Princeton film showed as THE WAY to load your falsify the count programming?

[cui bono]

15. if memory serves, the Princeton film said that the virus was spread with

memory cards. So if you put the card in another computer it would install the virus. If the card was simply taken out and put back in it would have no effect that hadn't already taken place with the initial insertion.

[WillYourVoteBCounted]

18. if you reboot, card will show 0 votes - how many states losing votes?

I found out that our State BOE knew about this "feature" since 1998.
I believe they found out the hard way.

However, the BOE in Montgomery County didn't know about the feature
untill a few weeks ago when they asked Diebold what happened to the votes on
some of the machines.

How many states, how many counties - are losing tons of votes on these
machines, either by accident or on purpose?

The following is an excerpt from an email I received from a gentleman named Walter Mancuso who was a Republican Chief Judge in the September 12 primary in Montgomery County, Maryland.

"Approximately a week after the primary I received a telephone call from the Montgomery County Board of Elections inquiring as to why one of the eight touch screen voting machines that we used on election day had recorded no votes, even though 55 voters were logged onto the machine. Neither I nor the Democratic chief judge had any explanation. The person at the Board of Elections (B of E) told me that they would investigate, talk to Diebold, and get back to me. After a week I called the Board and asked what they had discovered. I was told that at 6:50 AM (prior to opening the polls) that particular touch screen machine had been rebooted, and the memory card had been removed and reinserted into the machine. I was told that removing the memory card activated a security feature of the touch screen unit, and thus nothing was recorded on the memory card."

http://avi-rubin.blogspot.com/2006/10/following-is-exce...

[cui bono]

16. so if the power goes out there's the same problem since I"m sure they

aren't using a back up power supply.

The other issue is how early voting is handled. In Los Angeles you can vote early from Oct 25-Nov 3. So this would mean for there not to be this memory card problem the machines have to be on continuously for 10 days with no power interuption. Possible, but pretty risky imho.

[Kurovski]

19. Kick.(nt)