BLOGGED BY Brad Friedman ON 4/8/2008 3:00PM
SNIP...Secretary of State Jennifer Brunner, who completed a massive study last year of Ohio's e-voting systems, finding numerous programming errors and "critical security concerns," has recommended that the state permanently ban the use of all touch-screen systems. Following the findings of her landmark EVEREST tests, last December she urged counties in the Buckeye State to move from touch-screen voting to paper-based systems. Almost all of the state's Election Directors have ignored her recommendations.
Diebold's voting systems have become legion for their persistent failures during elections, and the ease in which study after study across the country have found that they can be hacked in a moments time by insiders who gain access to either a single voting machine or their central tabulation computer. Vote-flipping viruses can be added via their memory cards, the machines are all opened with a single matching key, and the database which stores vote totals on Microsoft Access easily corrupted to change vote totals without detection.
One other rather notable revelation is also made in the letter sent to Diebold/Premier by Butler's officials:
Holeman speculated that the problem could stem from the size of a database being larger than 2 gigs. Mr. Holeman explained that the central count server used for scanning and uploading Absentee and Optical Scan Ballots was taking up too much space in the database and outlined a procedure for backing up, deleting and compressing data before continuing with the uploading of memory cards. He went on to say that Premier was aware of this problem and that they would be issuing a product advisory. To date, we have not received such notification. Mr. Bunch then reviewed the size of our database and found it to be an acceptable size (245mb), causing our office to question Mr. Holemans conclusion of too large of a database causing the problem.
So with just a few months left before what is likely to be the largest election in the history of the United States, in addition to the possibility of memory card totals showing as having been uploaded --- when they haven't been --- we learn that there is also a size limit that could effect the proper tabulation of election results on Diebold's widely used GEMS central tabulator.
Not good.
http://www.bradblog.com/?p=5879