Erik Reuter wrote:
>
> On Tue, Nov 05, 2002 at 06:07:31PM -0600, Julia Thompson wrote:
>
> > (I don't trust the programmers not to stick in something that they
> > could be bribed to fix. I might trust specific programmers once I got
> > to know them, but I figure there will be a couple of bad apples in the
> > bunch and so I don't trust *all* of them as a *class*.)
> >
> > I'm happiest with the system under which I voted -- paper ballot
> > that you mark to be scanned electronically and votes tallied
> > electronically. Not paper ballot to be tallied by human, not paper
> > ballot to have holes punched out, but paper ballot to be marked by
> > writing instrument and scanned by machine.
>
> I seriously doubt the machine uses gears and levers. It had to be coded
> on some level, micro-code, machine-language, or higher level language.
> So, it depends on how it was bought. Was it bought on the used market,
> from, for example, ETS, or was it bought from the manufacturer who knew
> what it was to be used for?
>
> Actually, I think worrying about that OR programmers is rather over the
> top. Just wanted to point out that it is inconsistent to worry about one
> and not the other.
More that the paper ballot gives a backup of the data a lot better, and
you're not depending on machine programming at step 1. You can start
scanning the paper ballots 10 at a time and have people eyeball them to
make sure the machine is recording things accurately, or run a few test
ballots through to make sure things are running correctly. If someone
goofs in the programming where the initial ballot is entered (as is the
rumored case in Dallas), you've got nowhere to turn.
(If it weren't for the rumors about Dallas, I would have less of a
problem with the whole thing right now.)
Now, as my friend who voted early in Travis County pointed out, you can
lose boxes of ballots or stuff the boxes, so even that's not ideal, but
it's still the system I'm most comfortable with.
Julia
_______________________________________________
http://www.mccmedia.com/mailman/listinfo/brin-l
