We use a "scan-tron" type voting apparatus in our King George precinct.  The voter marks a paper ballot with a marker and feeds the marked ballot under witness to the tally machine.  It provides an electronic tally of votes cast but also (most importantly) retains the paper ballot used.  I find the ballot set-up easy to understand and use as well as being unambiguous.  If a voter commits an error, he/she can request a replacement ballot prior to casting their final official ballot (the ballot in error is destroyed).  Votes can still be cast in the event of a power failure. 

Yes, all electronic machines can provide greater efficiencies but having a physical paper ballot as a last resort for accountability is an excellent hedge against any fault that may come with a software based apparatus.  I feel having that paper ballot in a secure container provides a sense that my vote is there and ready to be counted in case the apparatus itself fails or comes into question.  The beautiful thing about paper is that it requires physical access and alteration or destruction to become compromised, regardless if the apparatus itself fell into an error state or was compromised electronically.

Not so with electronic machines not hardened against cyber-attack and come with a built-in vulnerability.  It defies basic security precepts having a voting apparatus with wireless capability.  The function of these machines is comparable to that of any critical infrastructure element in our society.  They must have the ability to record and tally votes in a means above reproach.  Having the ability to be remotely accessed automatically renders them suspect given the level of sophistication demonstrated by various cyber criminals to date.  This vulnerability is a glaring red flag and a cause for grave concern.  Just because something has not happened so far does not mean that something cannot happen.  Absence of evidence is not evidence of absence of a vulnerability.  How any community overlooked this aspect of their voting system is baffling.  To me this represents a major failure of due diligence of the governing agency(ies) to properly assess and address all risk associated with a particular apparatus proposed for use.  This is in addition to the other aspects of the system which are subject to compromise, disruption, or failure, such as the USB based flash memory.

The question put before the SEB is whether or not we should be using this particular voting system, which is open to attack and compromise by virtue of its design.  I think due diligence on behalf of the SEB requires them to assess the wireless capability as a grave vulnerability and decertify the system.  I understand this would upset a good many folks and agencies, but as a voter I would want my state and local governments taking appropriate measures to ensure that the voting process is secure from deliberate or inadvertent compromise.  Sunk cost is nothing compared to the integrity of the voting process.