I disagree flatly. For starters, our campaign to demand a voter-verifiable audit trail is not about voter fraud. It is based on the recognition that computers and their programs are subject to failure, and these systems DO NOT have the necessary safeguards established in them.
However, the resolution we propose, by its very nature, helps eliminate fraud by requiring random, mandatory audits.
Dave Dill is professor of computer science. I'm a 30-year career programmer. Here are some talking points that I would ask you to consider, which I present strictly from the perspective of a geek that has written more programs, and introduced more errors into said programs, than I care to acknowledge:
What is not currently evident is whether other manufacturers have done the same. We may know more about that on Thursday after California reveals the next phase of the statewide audit.
The failures that have been detected in VA, MD and FL provide ample evidence that simply having properly trained election workers simply does not bring us a solution.
I disagree flatly. For starters, our campaign to demand a voter-verifiable audit trail is not about voter fraud. It is based on the recognition that computers and their programs are subject to failure, and these systems DO NOT have the necessary safeguards established in them.
My concern is voter fraud. As we saw in 2000 (and other elections too numerous) paper systems are subject to failure (and not necessarily accidental).
However, the resolution we propose, by its very nature, helps eliminate fraud by requiring random, mandatory audits.
Here we agree, just don't be fooled into thinking paper trails will solve the problem by itself.
Dave Dill is professor ... than I care to acknowledge:
I have a M.S. in Engineering and Registered P.E. Also have three years experience as an election judge (paper and e-slate computer systems). Just so you know where my expertise is.
We have not invented the means by which an error-free program can be written. At least not one of the sophistication needed to configure and run elections.
Do you live in TX? We used e-slates in Houston during the 2002 and 2003 elections, and except in precincts run by incompetant people (more about that some other time), we had no problems. All of the judges and clerks were required to attend training regarding poll operation and safeguards, so the problem precincts had no excuses. Your error free comment is handled better in the next statement.
Current law does not REQUIRE that ...
The rest of your bullet items touch on QA, QC, ISO-9000, and related issues where in general, I am in agreement. The only certain method to verify code is to obtain a copy of said code, not check paper trails. There will be proprietary issues, but they can be handled. I touched on this in a recent previous post above.
To recap, the certification process must be looked at and refined, the ethics of the manufactures have to be scrutinized, controls on all of the above need to be strengthened, and YES - we need a paper trail.
Agreement regarding certification process (the machines, equipment, software, etc) and if you want a paper trail for it, fine. But that process must be completed BEFORE election day. Election day is too late. Paper trails are fine, but too many people think that will solve the problem alone. It will not.
The failures that have been detected in VA, MD and FL provide ample evidence that simply having properly trained election workers simply does not bring us a solution.
You forgot something. The word honest. What happened in Florida was not incompetance; it was dis-honesty. Paper can be manipulated more easily than computers. I have read post from people who want the old punch card system, and with that, you don't even need to go to the booth to stuff the totals.
My Summary: Put competant and honest people in charge of the election process, and add you "quality control" checks toward system development, and you won't need a paper trail. See, no name calling, no insults, no profanity.