Possible OVC "sub project"?

From: Jim March <jmarch_at_prodigy_dot_net>
Date: Tue Apr 26 2005 - 21:23:39 CDT

Ed Kennedy wrote:

> Hello JamBoi and others:
> Doug Jones, the election scholar at the University of Iowa told me
> that you can't get anything certified unless it also includes a
> registration and tabulation system. If I misunderstood him, perhaps
> he'll let us all know. If we can't get the EVM (Electronic Voting
> Machine) certified without these other items then they are part of the
> minimal standards. I would love to focus like a laser on getting the
> EVM ready for certification as it's real hard to go around and talk to
> election officials and tell them that they shouldn't use a DRE when
> you haven't got an attractive alternative.

OK, but this isn't necessarily a bad thing.

Right. The situation is that yet MORE dirt on Diebold is turning up.
I'll have it all on paper soon but the upshot is that ethical and legal
violations Shelley's office caught them at back in April 2004 have not
only not been fixed, they've been breeding.

BBV is going to call for statewide de-cert of Diebold on May 19th. It's
going to be *brutal* in that hearing, WW3 level.

We need to be able to propose an alternative for the existing Diebold
client counties. A *FAST* alternative that can be working by June '06.

Here's what I'm proposing: OVC (and the UC system funded with HAVA R&D
cash) crash-course develops a new tabulator, one that has a bank of 2 to
6 honkin' big autofeed scanners that can read Diebold paper ballots.
For this "first stage", we let the voters vote on paper and use Diebold
small precinct optical scanners to do the HAVA-required under/over vote
scan at the precinct - but *ignore* the "electronic ballot box" from
those Diebold terminals and scan the paper at elections HQ.

This means a single Linux PC per county hooked up to two or more big
gnarly scanners, and custom tabulator/scanner integrated open-source

Before you get uptight about costs, remember that the GEMS software
alone is $44k. A couple of $10k scanners, a $5k PC, another $5k for
misc. and we're still in under the costs of the Diebold software alone.

It should be possible to read the contents of a GEMS database file
(MS-Access, a well understood file format) to get the required ballot
layout data and import that into the Linux/SQL box. Under this model,
the Diebold GEMS and precinct scanners would be used to develop the
ballot layout, while the OVC box would do ALL the tabulation functions.

Now here's the cool part: the Linux-based tabulator you build would also
serve as the core tabulator for the next phase: an all-OVC setup.
Remember, you'll still need to do optical scan for absentee ballots so
the R&D you spend on this "stage one, prop up a crippled Diebold
critter" stage will be almost completely reusable for the "total OVC
package" down the road.

I would submit that this "stage one propup" project would be less
intensive than the complete solution...and would allow BBV to propose an
alternate trustworthy tabulation process in time for the '06 primaries
(so long as co-development via the UC system was funded via Bruce's HAVA
R&D funds).

If Doug is right and registration needs to be in there too, fine, add it
to the stage one Linux-based tabulator. As with the tabulator, the work
done will be applicable to the total solution later.

Jim March
OVC discuss mailing lists
Send requests to subscribe or unsubscribe to arthur@openvotingconsortium.org
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
Received on Sat Apr 30 23:17:16 2005

This archive was generated by hypermail 2.1.8 : Sat Apr 30 2005 - 23:17:22 CDT