GEMS voting system. Can a programmer read this please?

I'm not a programmer, but this looks very bad. The electronic voting system that covers about a quarter of the country - including swing states - is using a vote counting system that appears to have been built with fraud in mind.

This is a series that isn't completed yet, but take a look at what's here and explain it to us morons. The piece is at Election Watch.

Source

a reply to: bbarkow

This is from a study into this system:

The GEMS DB has not been designed to adhere to fundamental normalization principles that can permit a DB to reach high standards of accuracy and reliability. But the federal regulatory apparatus has also failed, for it imposes differential documentation requirements and financial burdens on vendors seeking certification for their election DB software. Those vendors who attempt to achieve higher design standards face far greater burdens and costs, including possibly more delays, than vendors who settle for DB designs with lower horizons. None of the 2005 VSS standards constitute a mandatory federal floor for voting systems to be deployed in federal elections. The certification of the GEMS software notwithstanding the significant demonstrable design flaws, offer a clear demonstration of the inadequacy of the current certification regime.

GEMS Tabulation Database Design Issues in Relation to

Voting Systems Certification Standards

Be back with more in a bit.

a reply to: jadedANDcynical

They aren't bugs, they're features.

a reply to: bbarkow

What it means is simple, the vote can be rigged based on who is voting. All votes are weighted by a selected multiplication factor based on political alignment, race and gender, from 0.0001 to 25. As an example, 1000 black female Democrats could be counted as 25000 votes. However, 1000 white male republicans could be counted as 1 single vote.

Dirty politics and election fraud at its finest!

Cheers - Dave

a reply to: bbarkow

I just read the source. I am a computer analyst and have been programming large computer systems for the past 25 years.

Here is the main kicker in the article:
"Source code: Instructions to treat votes as decimal values instead of whole numbers are inserted multiple times in the GEMS source code itself; thus, this feature cannot have been created by accident."

This is very true. There is no reason to do this in the code for this type of application.

a reply to: Starbuck799

In your estimation, would that be because of poor coding or purposely leaving a lot of doors open?

Look - I don't know how much I don't know. The reason I'm asking is because I got engrossed in the series because it was utterly unbelievable that a system like that would be in place.

Here's the best explanation I can come up with...

The core processing of most software (the 'engine') is usually very simple and straightforward. Complexity comes in for error-checking and 'bells and whistles.' For instance, the engine that makes Windows run is small and efficient, but the error-checking to make sure the user can't screw up something important and the nice visuals/wizards/automatic features are huge and complex.

Also, the more complex software becomes, the easier it is to have unwanted effects. Using Windows again as an example, this complexity is why they have continual updates; they have to continually fix problems they didn't know were there.

But anything can be manipulated. It is completely possible to purposely add complexity to hide designed weaknesses.

A voting verification engine should be the simplest thing in the world... once the veracity of the vote is verified (citizen, no duplicates, clear results) the various totals in each race should be incremented by one. Once voting is complete, the results should be made available to the proper authorities. There isn't much need in my mind for bells and whistles since it is so simple and only designed to be operated by election officials.

In this case, it sounds like the increment isn't necessarily by 1. There are many ways to code a number into software: integer values are the simplest and hold only whole numbers, while doubles, floats, etc. rely on a coding scheme and specialized mathematical modules to handle decimal numbers. Obviously vote totals should be integers, since there should never be a vote total that isn't a whole number. But it appears float values are being used, which is either an extreme waste of memory/processor time that indicates a thorough lack of expertise on the part of the programmers, or an attempt to allow fractional or weighted voting.

Either way, this disturbs me.

TheRedneck

The real travesty is that almost no one in the electorate will know about this or care enough to do anything about it.

Sorry, Ben, I guess we couldn't keep it. Link.

a reply to: TheRedneck

But it appears float values are being used, which is either an extreme waste of memory/processor time that indicates a thorough lack of expertise on the part of the programmers, or an attempt to allow fractional or weighted voting.

Great post (everyone else too). That's what I thought. One vote should be 1, no vote should be 0. The only reason to introduce complexity is to be able to massage those ones and zeros into something else.

I also found Part 6 available Here. Part 6...

When you combine the information in this thread along with this post:

Ahead of General Election, Democrat’s Most Powerful Campaign Software Gets Makeover

Chris Bing // DCInno

D.C.-based NGP VAN, a software-as-a-service (SaaS) tech company that specializes in providing campaign software to progressive campaign organizations, has been working to add a new and important feature to it's platform: a sidebar navigation bar feature. The update, however, is more significant than perhaps understood at face value because it represents a larger development trend for NGP VAN ahead of a complex general election. The move effectively comes at a time when NGP VAN's internal design team, led by Justin Lewis, is following through on multiple projects that aim to make the platform—which is used by volunteers, canvassers and campaign staff to run on-ground operations—more user-friendly.

https://________.___/dnc-emails/emailid/3658

And this post:

Votebuilder Basics
Creating a list of strong Democrats by precinct
(+ MiniVAN) v1.2
Votebuilder is our Democratic voter database that we use to keep track of our volunteers and voters so we can communicate with them and get them out to vote on Election Day. Votebuilder is supported by NGP/VAN. The precinct is the basic building block for all of our elections and getting volunteers connected and talking to their neighbors is the most effective way to increase turn out and win elections. The first step of organizing your precinct is recruiting volunteers and we use Votebuilder to create lists of likely volunteers (Strong Democrats) and then call or knock on their door to recruit them to attend events and help contact voters by phone banking or knocking on doors.

Votebuilder Basics(direct .pdf link)

And here is another:



VAN Basics(direct .pdf link)

Plus throw in Greenberg Quinlan Rosner, a polling firm with close ties to the Clintons, and you have a lot of electronic weight thrown behind a candidate.

And there is the "massive" support Hillary has when nary a real person can be found who actually supports her rather than the party choice.




Edited to add:

Here is an interesting and information rich comment:

"The American vote count is controlled by three major corporate players, Diebold, ESS, Sequoia, and a fourth, SAIC, Science Applications International. All four are hard-wired into the Bush power structure, the Bush crime family.

"They have been given millions of dollars by the Bush regime to complete a sweeping computerization of voting machines that were used in the 2004 election. The technology involved had a trial run during the 2002 mid-term elections. Georgia had Diebold machines in every precinct. As a result, a popular Democratic governor and senator were both unseated in what the media called an "amazing" 16 percent swing.

"Diebold's Walden O'Dell, a top Bush fundraiser, publicly committed himself to delivering his home state Ohio's votes to Bush. At Diebold, the election division is run by Bob Urosevich. Bob's brother, Todd, is a top executive at "rival" ES&S. The brothers were originally staked by Howard Ahmanson, a member of the Council For National Policy, a right-wing steering group stacked with Bush true believers. Ahmanson is also one of the bagmen behind the extremist Christian Reconstruction Movement, which advocates the theocratic takeover of American democracy.

"The four companies are interconnected; they are not four "competitors". Ahmanson has large stakes in ES&S, whose former CEO was Republican Senator Chuck Hagel of Nebraska. When Hagel ran for office, his own company counted the votes, and his victory was considered "an amazing upset". Hagel still has a million dollar stake in ES&S.

"Sequoia is the corporate parent of a private equity firm, Madison Dearborn, which is partner in the Carlyle Group.

"Meanwhile, SAIC is referred to a "shadowy defense contractor". They have gotten into the vote count game both directly and through spinoffs by its top brass, including Admiral Bill Owens, former military aide to Dick Cheney, and Carlyle Group honcho Frank Carlucci and ex-CIA chief Robert Gates. SAIC's history of fraud charges and security "lapses" haven't prevented it from becoming one of the largest Pentagon and CIA contractors, and will doubtless encounter few obstacles in its entrance into the vote counting business. "

Original Townhall article: Obama Likely Won Re-Election Through Election Fraud

If there are shenanigans occurring behind the scenes, I would not be at all surprised to find the Bushes involved right alongside the Clintons

Why would a vote in the programming even be designed as something other than an integer unless some bias was planned to be introduced.

a reply to: roadgravel

Read part 6 - Here's the source again.

Surveillance video in Saguache County, Colorado, taken with an ES&S M650 tabulator, shows an unusual sequence of events. First, it is reported that the election administrator, who was running for her own position in the election, lost the election. Then, she sits down with the voting system technician and berates him so aggressively that he reaches for a box of tissue and cries. Then, the technician surreptitiously locks the door and inserts a zip disk into the ES&S M650 tabulator. Next, the elections administrator announced that she has, after all, won the election. 22

a reply to: bbarkow

While I have no actual links to support me, my observations over the past several cycles are that both parties have used the 'civic duty to vote' argument to introduce collaboration between the voting process itself and the various voting advocacy groups. I guess that sounds like a good thing, but when the advocacy groups are partisan or agenda-driven (as they invariably are) it introduces a bias into the voting system itself.

That's just plain wrong, and there's no argument to make it right.

It's not a wonder then that different groups want and get access to the voting software, and also no wonder that they'd like to be able to manipulate the vote. Not hard to do if you have access to software that has a built-in back door.

TheRedneck

a reply to: bbarkow

Hory shet!

I've been doing a little digging around (plus I'm at work) and hadn't gotten to read this yet, but this says a lot.

This sums up part 6 quite nicely:

To sum it up, direct access to central tabulation data appears to be accessible to multiple persons from each jurisdiction, and there are over 3,000 jurisdictions. Access is also granted to voting system vendors for support services, and to a number of middlemen who have access to multiple jurisdictions, and whose role is often unclear and also undisclosed to the public.

There has been demonstrable fraud that has occurred. No wonder Trump says the elections are rigged.

I don't understand how government can use software that the owners refuse to allow auditing due to their proprietary claim. Given election rigging is a long tradition, this is not acceptable.

Election software shouldn't be a competitive, money making venture.

originally posted by: bbarkow
a reply to: Starbuck799

In your estimation, would that be because of poor coding or purposely leaving a lot of doors open?

Look - I don't know how much I don't know. The reason I'm asking is because I got engrossed in the series because it was utterly unbelievable that a system like that would be in place.

Programming can be very complex. But it is also very simple most of the time, logically I mean. When you write code, this type of thing is Not from sloppy coding. It would be purposeful. Sloppy coding would be something like not anticipating the user's actions when you know the users may try to circumvent the computer.

eg:
My first big programming job was at a bank. 99% of the people who would be using my programs were just normal folks who just wanted to do their jobs and then go home to their families at night. My 2nd big job was at a Tier 1 automotive assembly plant. Where we were supplying Ford for their car builds. I live in Toronto, so the people who worked in the plant were CAW union, (Canadian Auto Workers.) What I failed to realize was that the workers on the floor were not at all like the workers at the bank. I never coded into my programs to ensure that the workers could not circumvent the programs. I just never thought the people would want to do that. But I learned that a lot of the floor CAW workers would do that to either slow the line down or to stop the line completely so that they would not have to work for a while while the computer was fixed. I learned that I had to change the way I programmed to make my programs Bullet-proof.

So... what I am saying here is that it is Not just sloppy programming to code the 1 to 1 vote as a decimal. For this type of application, as a programmer, you have a very clear choice as to whether or not you code it to be 1 to 1, or a decimal. To code it as a decimal is not just a mistake, it is for sure on purpose.

a reply to: bbarkow

Oh crap. I just read TheRedneck's response. He is correct with regard to whole numbers, doubles, floats, etc. for data types. It could be sloppy programming where the programmer used the incorrect data type, which would result in wrong tabulation of the votes.

TheRedneck must be a programmer to know that.

But this would also be a way for them to say that it was a programming error, when in fact it was on purpose. There would be no way to tell.

Looking a bit further back regarding the GEMS software:

The GEMS central tabulator program is incorrectly designed and highly vulnerable to fraud. Election results can be changed in a matter of seconds. Part of the program we examined appears to be designed with election tampering in mind. We have also learned that election officials maintain inadequate controls over access to the central tabulator. We need to beef up procedures to mitigate risks.

Diebold GEMS Central Tabulator Contains Stunning Security Hole by Bev Harris

Just because a floating point is used to store the value doesn't mean it will be wrong. But it makes it easier to be wrong and could be misused.

Another point brought up in the article is the access to the boxes. Seems as though quite a few people have access. What about backdoors or access disguised as another process even if there is logging..

Seems like a fubar situation.

I'm a Dev - c# mainly but from a C/Java background.

I suspect they're not using ints as they only go up to a certain value (can't be arsed to check). If its unsigned (so you ignore whether it's negative) then that doubles the range, but even so.

Given its a voting system, and there'd be a shedload of votes - I guess they chose the data type they did to cope with the range. Just a theory.

Floats are also notoriously inaccurate when rounding, as there's no such thing as a floating point number in coding, strictly speaking. It's all binary, don't you know (for example, read this link