Reply #354: 'Very interesting,' said the computer programmer [View All]

[Jack Rabbit]

Speaking as a professional computer programmer, if I were to inspect the source to determine if it is intended to steal elections, I would first have an idea of what a bona fide vote counting program would look like. It would have a straightforward method of counting votes and some modules to correct a mis-vote before the ballot is sent and secure the count afterwords. That isn't a lot of code.

I would question anything more than that. There is no such things a undetectable electronic voter fraud provided that one is inspecting the actual source code. As Mr. Curtis states in this affidavit (paragraph six), it would be impossible to detect such fraud if the code were compiled first; in other words, write one program and show the authorities another.

The basic algorithm to rig an election would simply be to count all of the first n-number of votes cast for candidate X, regardless of whether the voter actually selected some other candidate. In this hypothetical program, written in psuedocode, tallies votes in a two-way race. An integer value is passed to the procedure; a value of 1 indicates a vote for Candidate1 and of 2 for Candidate2. The values of the total number of votes cast and the total number of votes for each candidate are passed to the program and returned to the parent program. The code written in red doesn't belong there.

PROCEDURE CountVote:

INPUT PARAMETER iSelectCandidate AS INTEGER.
INPUT PARAMETER iVotesCastIn AS INTEGER.
INPUT PARAMETER iVoteCandidate1In AS INTEGER.
INPUT PARAMETER iVoteCandidate2In AS INTEGER.

OUTPUT PARAMETER iVotesCast AS INTEGER.
OUTPUT PARAMETER iVoteCandidate1 AS INTEGER.
OUTPUT PARAMETER iVoteCandidate2 AS INTEGER.

DECLARE iRigLevel AS INTEGER CONSTANT VALUE 100.

ASSIGN iVotesCast = iVotesCastIn.
ASSIGN iVoteCandidate1 = iVoteCandidate1In.
ASSIGN iVoteCandidate2 = iVoteCandidate2In.

If iSelectCandidate = 0 THEN RETURN. /* No vote Cast */

ASSIGN iVotesCast = iVotesCast + 1.
IF iVotesCast <= iRigLevel THEN DO:
ASSIGN iVoteCandidate1 = iVoteCandidate1 + 1.
END. /* IF iVotesCast < iRigLevel */
ELSE DO:
CASE iSelectCandidate:
WHEN 1 THEN ASSIGN iVoteCandidate1 = iVoteCandidate1 + 1.
WHEN 2 THEN ASSIGN iVoteCandidate2 = iVoteCandidate2 + 1.
END CASE.
END. /* ELSE */

RETURN.

END PROCEDURE.

In this program, the event that triggers the code in red, which rigs the election in favor of Candidate1, is that 100 votes or less have been cast; in that case, any vote cast is counted for Candidate1; after 100 votes have been cast, the votes are counted as they should be. Notice that the total of the number of votes cast for each candidate balances with the overall total number of votes cast, which avoids such red flags as Candidate1 getting more votes than the total number of registered voters in the district.

Mr. Curtis' vote-rigging algorithm would require a great deal of code and could be detected by examining the source code. It is really little more than an elaboration of the above algorithm, with a more complicated set of events triggering the code in red.

What needs to be emphasized is that in an environment where there is no paper audit trail of the ballot and where the software is the property of the voting machine manufacturer and thus beyond public scrutiny, anybody who has ever received a passing grade in a rudimentary programming course would have the skills to rig an election. Mr. Curtis has more skills than that.