RE: DC Article: "Larimer Shows Up Boulder"

["Paul Tiger" <paul.tiger@xxxxxxxxxxxxx>]

Quick is not the issue. Accuracy is. Boulder may have had some of the most
accurate elections of any county in this state (maybe other places too), and
also the slowest. There have absolutely been problems, and most of them have
been noted by the humans. Some haven't been fixed yet, but they've been
acknowledged. Things go slowly as the electronic systems actually force them
to be slower. Procedurally most things are being dealt with, except
catastrophic failures. And there have been those and probably we will see
more of those.

We are forced to use a modicum of computer equipment. I believe that we
should choose wisely and not invest in more systems that we might rely on,
for they will fail. I repeat - they will fail. We just don't know when they
will fail, but in a true Murphyism we can be certain that computer systems
(all systems) will fail when the heaviest load is placed on them. AKA when
we need them most is when they will go to hell in a handbasket.

What we lack is failure analysis, largely because it is difficult to test
election systems outside of an actual election. Mock elections provide mock
results.
There are ways of testing Mean Time Between Failure, only Boulder is not
doing that. Not just Boulder, but no one is doing that. If elections
equipment vendors have MTBF specs, it is on hardware that we couldn't care
about or could easily be replaced in a matter of minutes. Component parts
have MTBF specs, but not the integrated systems.

Paul Tiger

-----Original Message-----
From: Joe Pezzillo [mailto:jpezzillo@xxxxxxxxx]
Sent: Thursday, November 09, 2006 1:50 PM
To: Liss, Josh
Cc: CVV Voting; Hillary Hall
Subject: Re: DC Article: "Larimer Shows Up Boulder"


What do you mean? I live here and I've been involved in this issue
for several years. I might even be willing to help out if I didn't
have to fear getting investigated by the Sheriff should I happen to
ask any tough questions. -Joe