(DIEBOLD BBV)Firm Blames Equipment for E-Vote Glitch

[papau]

http://www.guardian.co.uk/worldlatest/story/0,1280,-3981547,00.html

Firm Blames Equipment for E-Vote Glitch

Thursday April 15, 2004 6:46 PM


SAN DIEGO (AP) - The manufacturer of San Diego County's electronic voting equipment blamed a faulty power switch for problems that delayed the opening of 40 percent of polling stations, according to a report released by Diebold Election Systems Inc.

Switches on computer encoders malfunctioned, the company said Tuesday.

During the March 2 primary, poll workers were not able to activate electronic devices that encode the magnetic-striped cards used to activate touch-screen voting machines because the computer displayed the wrong screen.

The problem occurred because the faulty switches did not shut off the units, and the battery power drained. When poll workers then turned on the machines, a Windows operating system screen appeared, causing confusion for workers trained to see another screen. Voters were delayed, or forced to leave without casting ballots, while poll workers awaited instructions on how to find the proper screen. <snip>

[SeattleDem]

1. glad the UK is on top of this. n/t

[qb]

2. True Republicans - Pass the buck

Diebold did, in fact, select these switches for the encoders, did they not? Do they have anything resembling Quality Control in their organization?

[RoyGBiv]

4. Not a Bug

It's not a bug, it's a feature. :think:

[TrogL]

3. Unbelievable

What part of on-off can a switch screw up?

I'm suspecting the machines had a vampire circuit and were on all the time. The "go active" switch was probably busted.

[Mithras61]

5. Are you telling me that...

battery powered units didn't have a hibernate feature, so that when they drained the battery they powered off completely and then on reboot, not only didn't load the only software they are designed to run, but also that the poll workers were never taught that this could happen and never shown how to manually start the program if it didn't load at boot?

Sounds pretty damn weak to me...

[klook]

6. Were these voters just SOL?

Or did they go back to the polls later and cast their votes? (Or use paper absentee ballots.) And was there any pattern to what precincts this happened in?

If these people never got to cast their votes, I hope some of them are filing a class-action lawsuit.

See "How to Hack the Vote" by Brendan I. Koerner in the April 2004 Harper's Magazine for some tongue-in-cheek (?) suggestions for playing havoc with BBV systems.

[JohnOneillsMemory]

7. My MS Windows 98 software sometimes won't shut down.

I have to pull ac and battery and go through the whole scansearch routine regularly.

Sure don't wanna base democracy and the choice to go to war on this system.

[hunter]

9. Windows 98 shutdown...

On some computers, when all else fails, you can simply hold the "power" switch down for five to ten seconds and force a shutdown.

Meanwhile, we all have just learned yet another back door into these machines that was left wide open.

[Mithras61]

10. Well, ya see...

The story implies that the systems crashed because the power switch didn't work, and that when they were booted up again, the system didn't load the card-enabling software correctly. I may have misunderstood what they meant. I work closely with computers and have made bad assumptions about what people know about them before.

I suppose it is conceivable that the systems hibernated when the power switch was pushed, and when it was pushed again, the system came back up to where it was when the initial push happened. I can't imagine a stand-by causing the application to "disappear" and not load, but to still be able to recover the Windows state, and the article also implies that the systems performed an actual boot sequence, both of which imply a hibernation or off state. As far as it goes, these could be assumed to have been the root of the problem, but when a system comes out of hibernation, it starts up right where it was when it was put into hibernation, which means that the users would have had to close the program, which implies that they knew it COULD be closed, and therefore they should have known it COULD be launched.

The alternative is that which I outlined above... on startup, the app didn't load as it should have (again, these are single purpose machines).

Based on what I know about Diebold election systems (not as much as I'd like, I'm afraid), I'd guess that we're talking about a WinCE kernel, or a WinNT kernel, not a Win98 kernel (such as it is), simply because the system would need to have SOME form of security, and Win98 doesn't provide any. WinCE is a little better, but WinNT (as in Windows 2000, Windows XP and Windows Server 2003) actually has security built into it. These two kernels are also much less likely to hang at any point than Win98.

But even assuming a Win98 kernel (unlikely based on how they "update" their software), the systems would have had to have been in an unacceptable state at standby/hibernation or have failed to start properly, and that isn't a hardware problem, it's a software design flaw.

[hunter]

13. It is WindowsCE, and it is very flawed...

It is an OMG, poke a stick in your eye, kind of flawed.

I first jumped into this issue as you did here, by examining a single flaw in the system. It is probably far, far, worse than you imagine.

If you really want to look into electronic voting machines and see all the slimey wiggley sucky things that live in there please go to:

http://www.verifiedvoting.org

or

http://www.blackboxvoting.org

Verifiedvoting was spearheaded by Stanford University professor David Dill, and has an academic feel to it.

Blackboxvoting was started by activist Bev Harris. Bev's "pitchforks and torches" atitude is invigorating, if you like your politics rough.

If you want to see actual source code and specifications for these machines, you can find it. It may not be legal. Quite a bit of this information has been "stolen" by people who believe that citizens have every right to see EXACTLY how our votes are counted.

Good hunting!

[TroglodyteScholar]

15. And we all thought the abbreviation "WinCE" was just funny (eom)

[Mithras61]

16. Do you happen to know WHICH WinCE kernel it is?

I'm curious, because different WinCE kernels behave differently. My PDA (a PocketPC 2002-based system, kernel build 3.00.11171) acts in a pretty predictable manner when my battery dies - it shuts off & does a warm reset when I plug it back in... no programs are lost, but my active data is gone, as are any applications that may have been running when it lost power.

Which, by the way, tends to reinforce my belief that this is a bushit claim. If the system is built even HALFWAY like it should have been, dead batteries wouldn't have prevented the OS from loading the card enabling app at post, just like my system detects expansion cards and disables any apps that aren't found because of a Startup application that I've loaded.

I love blackboxvoting.org & Bev Harris' work. I haven't spent as much time on Professor Dill's site, but know of his work on the Diebold code, as well as the work of several other groups.

I am NOT in favor of electornic voting with a receipt. I want electronic voting that generates a human readable paper ballot, which is different than a receipt because a receipt isn't good as the "Official" vote. It is critical that these machines generate human-readable ballots, or there can be no meaningful recounts.

I'm not so sure that I think that even electronic vote counting is such a good thing. I like what I've heard about the French method... local heads of all parties gather at the election central location, count the ballots into lots of 100 ballots each, and seal them in canvas bags. After that, each bag is opened in turn, and the ballots are counted out by hand with all parties agreeing on the count for each lot, and then having the totals posted on a large bulletin board for all to see. After the count is completed, they create a conference call to the district/regional office with the totals. This happens again at the district/regional level (the conference call, I mean) and on up the line to the national level. All of this happens under the supervision of as many people as can pack themselves into the rooms to see/hear what's going on, and all the news media besides.

Of course, this requires that the media and the people actually care about the election and don't try to fix/rig the outcome, and that the Politicos actually campaign to the people and not the corporations...

[JohnyCanuck]

8. E voting problems continue to receive world wide attention

A columnist from Today's Daily Nation of Barbados quotes excerpts from another Guardian article on electronic voting and derides the US government's hypocrisy in proclaiming support for democracy abroad while apparently not giving a whit that it is being snuffed out at home.

US Impacts
by Peter Morgan

<snip>


Their (i.e. hanging chads /jc)importance was amply illustrated by the fact that they caused the election of the President of the United States to be left in the hands of a politically biased Supreme Court which handed the victory to Mr George Bush by a 5-4 majority. Since that time, Dubya and his pals have never ceased to preach the gospel that free and fair elections are the very bedrock of a democracy. Since no Weapons of Mass Destruction have been found in Iraq, the official American justification for the pre-emptive attack, killing thousands and costing the American people billions, has now been changed to have been the imposition of such a democracy. Guess by whom!

There is another election coming up in November and no doubt the world is hoping that they might use rather more conventional methods of determining the wishes of the voters on this occasion. It seems this is not so but, in view of the fact that nothing has been mentioned on the subject by our local news media, I believe it worthwhile to use this week’s column to reproduce relevant extracts of an article published in the British newspaper the Guardian -

“United States voters will go to the polls in November using electronic machines that cannot be verified, a computer scientist has warned”. David Gill of Stanford University told the American Association for the Advancement of Science meeting in Seattle that 1 600 technologists and 53 elected officials had joined his crusade for a ‘paper trail’ so that voting machines could be checked.

<snip>

At the time when there was utter confusion regarding the Florida votes in the 2000 election, President Fidel Castro of Cuba graciously offered to send electoral experts to solve their problem. I hope someone has alerted him that his offer may yet be needed.

US Impacts

[yardwork]

11. People who vote Democrat are those most likely to have low-wage jobs

that require the worker to punch a time-card. Many Democrats don't have the luxury to hang around for hours waiting for the polling place to get its act together. They don't have the luxury of coming back later that day.

Soccer moms who don't work, executives, wealthy retirees - they have the luxury of time. They are also more likely to vote Republican.

What a coincidence!!!!

[0rganism]

12. I smell a class-action lawsuit...

"Voters were delayed, or forced to leave without casting ballots, while poll workers awaited instructions on how to find the proper screen."

Something about that second part seems actionable to me. Maybe one of the frustrated voters can get the ball rolling; if not, the SoS could file to cover his own ass.

[Endangered Specie]

14. Least they didnt blame it on Clinton

[mulethree]

17. hmm democracy in the hands of a company that don't test?