John,

Speaking of printing out paper ballots ... I wonder if The Roman Empire had paper ballots?

The Inspector General of the Government Accounting Office, United States of America, David Walker, reports:

Drawing parallels with the end of the Roman empire, Mr Walker warned there were “striking similarities” between America’s current situation and the factors that brought down Rome, including “declining moral values and political civility at home, an over-confident and over-extended military in foreign lands and fiscal irresponsibility by the central government”.

(London Financial Times, emphasis added). John, I suppose that the prescient among us would say that if Brad is correct that the paper ballot wars is "just short of a shooting war", then political civility has gone by the wayside in the EVM / EI movement?

If so, then we must be very close to paper ballots only in our election systems?

If so, have we have attained the glory of Stalinist Russia and The Roman Empire (they were big paper ballot afficionados, since they did not have EVM's).

But, what is even more important is answering the question "how did they all forget that the integrity of anything is first and formost ultimately dependent on the humanity involved (or the lack thereof)", of any political system?

I think the answer to the latter question is that it was probably superstition that made them forget.

It's about time the "Black Box" through which our votes pass was removed! I am convinced the rush to replace punch card ballots with all electronic machines was really designed to hide serious election rigging. I am convinced the result of every major election since 2000 has been "modified" by these "black box" machines.

The "smoking gun" for me was learning Diebold (and others) used "Signed Integers" for counting votes in their machines. Bev Harris found this several years ago. I believe the use of signed integers has also been found in several other electronic voting machines. How else can you explain negative vote totals? REcall that negative vote totals have been reported in many places including Ohio. These negative totals have always been explained away as "computer glitches".

Think for a moment, why would a programmer decide to use signed integers instead of ordinary integers, when ordinary integers will do the job of simple counting? Obviously, there must be some reason for including negative numbers in those machines. But why would anyone need to use negative numbers when all the machines do is to add votes to running totals as the votes are cast? If using signed integers in these machines isn't a "smoking gun" pointing to rigged election results, I don't know what would be.

Why are we using machines with proprietary hardware designs and proprietary code inside? What is so secret about counting votes that the inner workings of these machines cannot be made available for inspection by anyone at any time?

LinearBob #2

The use of signed integers is not in and of itself conclusive of anything other than a style of design flaw.

One must also consider the number of bytes in the integer as well to detect a scenario where count progression will jump backwards into negative territory. Is the integer a long or a short?

For instance if a long integer (4 bytes) is used instead of a short integer (2 bytes), there will be no roll over into negative figures once 65,535 is reached, because of the vastness of the quantity ability even a signed long integer has.

But I do agree with you never-the-less that it is a style of design flaw, because negative numbers are not a part of the reality the software is designed to work in. There is no such thing as a negative vote.

The architechs of the EVM software, according to the reviews done in California recently, was very poor and slipshod.

As you know, any successful design considers all functionality, security, and exception handling from the very beginning, and before coding begins.

The software for these machines was indicative of waves of afterthought, which injected performance, as well as, security flaws into the system.

Experience tells us that there is rarely, if ever, a robust recovery from such fundamental design flaws.

LinearBob #2

In my post #3, I said "once 65,535 is reached", but I should have said "once 32,767 is reached" instead.

The signed 4 byte (32 bit) integer will reach 2,147,483,647 before it switches to a negative number on the next iteration (which would set the sign bit).

The 2,147,483,647 figure is beyond the scope of any vote tally contemplated in elections on this planet, so it is not a performance flaw.

But like I said in post #3, I think it is a style flaw because in elections there is never a negative vote and so the integers should be unsigned long integers.

It would be a performance flaw if a two byte (16 bit) integer was used whether that integer was signed or unsigned.