Last week we told you about the hard fought victory by the Pima County (Tucson), Arizona, Democrats to finally gain access to the Diebold databases which include the tallies of how folks voted in elections going all the way back to 1998.
The county had argued (on Diebold’s behalf, natch) that the database files were both proprietary, and a security risk should they be released. The judge, finally, found that to be nonsense, and ordered the release of the files which the local Election Integrity advocates now hope to comb over for evidence of fraud and/or other malfeasance.
Now that they’ve got all the info, they’re seeking a geek or two who may be able to help build a tool to make it a bit easier to go through the mountains of data they now finally have access to.
AZ’s Libertarian Election Integrity champ Jim March, whose been working closely with the Dems, sends us the following “Help Wanted” ad, seeking a programmer to help ’em make sense of “the world’s largest private collection of Diebold election data files anywhere.” (Note: If someone could kindly Slashdot this item to maximize the number of coder geek eyeballs it gets in front of, it’d be much appreciated!)…
To help process this unruly mess, Pima Dems consultant Jim March (footnote: who is actually a Libertarian but they forgive him) has designed the “feature specs” for a test tool. Now they need somebody to code it. The spec sheet for the tool may be downloaded here [PDF]. The “mark one” version can be very basic, command line driven and workable only with Diebold databases for now.
HELP WANTED: Computer programmer on this contract (paid) project with a donor already lined up. Must be very competent in database programming in general, the Microsoft JET database engine in particular.The project can be coordinated by EMail.
Note that the tool will be “open source” – free for download by ANYBODY, both running program and source code so anybody can see how it works and modify it as needed if we get election data files from any other vendor for instance.
I *know* there’s geeks reading Bradblog, folks. Speak up! And email me here if you’re interested, or have any questions.
Why don’t those Fair light guys help out? Aren’t they are really good with election data?
Submitted to Slashdot…lets wait and see.
We’ve worked hard getting these databases. There are some 1,300 databases going back as far as May of 1998. Most are backups that are time dated snapshot before and after election to closing called .GBF (Global Backup Files.) At least one per election is in .MDB (Microsoft DataBase) Files that open in MS-Access. We will have all passwords to open all the .GBF files and convert them to MDB.
The largest public release before these backup files that we know of was from the state of Alaska and that was 3 files from Election Day of November of 2006 only.
Any questions, feel free to contact me.
John R Brakey
AUDITAZ@cox.net
Phone 520-578-5678
Cell 520-250-2360
Tucson, Arizona
Subject: Pima County, the RTA and the 2 billion dollar heist.
4/21/08 Pima Co: KOLD TV Tucson: “Who Checks The Vote Counters?†Facts learned in the Pima Co Democratic Party’s Election integrity lawsuit against Pima County: From deposition of State Election Director Joseph Kanefield: “Who Checks the Vote Counters?†NOT the Secretary of State! Not the Attorney General! Not the County!
Must see TV! By Bud Foster KOLD TV TUCSON:
2 minutes long:
http://ca.youtube.com/watch?v=fqlefIQVkrk
Sourceforge it?
Where is Clint Curtus of Florida on this?
Curtis is currently busy running for Congress, trying to win the Democratic nomination in order to run against Tom Feeney in the fall.
I don’t understand the problem. It is in a database — a MS Access db as I understand it. If what is needed exceeds MS Access’s query and report writing then import it to MySQL. Use SQL and stored procedures.
You will most likely get a lot more ‘sourceforge’ help if it is in a MySQL database.
Why not just put the data up for download?
You’ll find a ton of people who’re interested in digging through that kind of data, but don’t want to commit to a contract.
Hey John 3#, now that’s what I call local home town news reporting! Thanks!!!!!
We need the automated review tool for a number of reasons:
1) Judge Miller has ordered these files be made public in future elections on the day of the official canvass. This isn’t a bad choice, if a bit conservative – I would have argued for “late on election night” but it’s not something worth squabbling over further in court. Problem is, ALL official challenges in AZ have to be filed within 5 days of the canvass. Without an automated review tool, challenges within 5 days of the canvass based on findings of database problems will be forever impossible.
2) We also want to export this whole data analysis concept nationwide – or even worldwide, wherever Diebold lurks. We could get enough volunteers together to check out PIMA elections, yeah (not in five days), but nationwide? Impossible in any timescale.
Jim
I agree all the databases should be converted to MySQL. This would take time but getting them into one or multiple database files (depending if all the files have similiar structure) that can be cross referrenced would be crutial.
Access can be slow and is not the best interface for a large number of records.
I have used MySQL for large country wide voter registration records without a problem.
The program analysis tools and reports could be written in PHP (open source) or JSP and provide web or local computer based access.
Let me know if you need help – I am very experience in all the above.
For all those “they should be in MySQL” – that’s more or less not possible. The goal of this is to analyze the *original* Diebold databases. Any analysis on a converted format will allow them to throw up a smoke screen, implicating the conversion in any errors found.
The less you need to modify the original data before evaluating it, the less of an attack opportunity you offer to criticize the results.
Robert #12:
You are right of cause, but they will try to obfuscate and belittle the results _anyway_. And the public can’t verify anything because they don’t know Access or MySQL.
What should be done is: 1) analyse the raw data with the best tools available (that is MySQL in my optinion) 2) if mischief is found get the FBI, Congress, whoever involved and have _them_ re-analyse the data (using _their_ methods and tools) and verify the results found by the party.
And then hit Diebold like a ton of bricks…
As long as the conversion is correct, any problem can then be tracked back to a corresponding issue in the Access (or whatever) Database.
For example, the scenario in the PDF can be detected by putting all of the precinct totals into a spreadsheet, and taking sums for the county.
Unless the data is very large the best approach may be to build a tool that dumps all of the data (including timestamps) into an easily read format and then generic tools can be applied to it.