RE: commissioners presentation this thurs

["Paul Tiger" <tigerp@xxxxxxxxxxxx>]

I'm with Neal, and everyone else on this topic of full disclosure of the
software. One might remind the commissioners that the software that was used
for the DataVote system was open. The county itself writes almost all of its
own software. While it may be complied and linked, the source sure as hell
is available to them.

In every case that I can think of were the county has purchased closed
systems, there's been trouble, and continues to be trouble. Ask them if they
want another repeat of the Jail Management System or the Computer Aided
Dispatch system. Both boondoggles that the commissioners have been paying
through the nose for for years.

Make a comparison that makes sense for them.

Paul Tiger

-----Original Message-----
From: Neal McBurnett [mailto:neal@xxxxxxxxxxxxxxxxx]
Sent: Monday, December 01, 2003 4:43 PM
To: bcv
Subject: Re: commissioners presentation this thurs

On Mon, Dec 01, 2003 at 03:02:16PM -0700, Neal McBurnett wrote:
> ...we don't have to buy DRE's until 2006, based on the extension
> that the secy of state will get (in all likelihood).
>
> Ideally we would, in 2006, get
> precinct scanners that would have fully-disclosed software.  And they
> could scan DRE-generated paper ballots and read them back so even
> blind voters could verify them.

Al gently pointed out to me that "DRE-generated paper ballots" is a
misuse of terms, since "DRE" is generally understood to imply that the
vote is Directly Recorded Electronically right then and there.

I'll go with his terminology.  In 2006 we want hand-marked paper
ballots for most people and "Vote Marking" machines -- such as the
Vogue Elections system -- to make the polling places accessible to
all.  We never want to see a traditional DRE in Boulder County since
it does not generate voter verifiable paper ballots that could
actually be used in a normal recount.

In order to have some protection against flawed, hacked or fraudulent
software, I also think that Boulder County should institute, as a
standard part of its vote counting procedure, some sort of
confirmation hand-count of randomly selected ballots.  Perhaps via
random selection of a suitable number of stacks of ballots.  Linda
Salas seemed to think this would be workable in the meeting we had
with her a few weeks ago.

But that is no substitute for fully disclosed software, which would
better protect against fraud and lead to improved software quality,
reliability and security.

-Neal