Re: Fwd: Joe's Notes from 7/12/2004 meeting with CA asst. SoS Carrel

From: Alan Dechert <alan_at_openvotingconsortium_dot_org>
Date: Sat Jul 17 2004 - 13:30:16 CDT

> ---------- Forwarded message ----------
> From: Bruce Perens <bruce@perens.com>
> Date: Wed, 14 Jul 2004 21:24:02 -0700
> Subject: Re: Joe's Notes from 7/12/2004 meeting with CA asst. SoS Carrel
> To: joehall@pobox.com
> Cc: voting-project@lists.sonic.net
>
> Joseph Lorenzo Hall wrote:
>
> >-They have issues with off-the-shelf (COTS) hardware... 1) it is not
> >tamper-resistant (Alan mentioned plans for a cage) and 2) it doesn't
> >look good and isn't light or easily packaged (Amy and Karl mentioned
> >that they have experience (or contact with people who do) with
> >delivering solid products).
> >
> >
> The biggest problem I saw in the presentation was that the group did not
> come in with a system integrator as partner, nor did we describe a model
> for coupling the non-profit software research project with a for-profit
> system integrator.
>
This is a reasonable criticism. We're just not that far along yet. I have
been in touch with a number of potential industrial partners but no
relationship has sufficiently solidified to bring them in at this point.
Maybe next time.

I had hoped to meet with someone at Novell while in Utah but was unable to
get this organized. I made some progress establishing some contacts there
and may have something to announce about that in the not-too-distant future.

Sun and IBM are also high on the list of desirable industrial partners.
I've heard a few encouraging words from people at these companies but no
real movement.

> The model I think would work is for the academic project to create the
> software, and then make it available to for-profit system integrators
> who will combine the software with hardware and service. Our
> presentation said very little about service, and it's clear that they
> expect a lot of it - Diebold was said to have provided one
> troubleshooter per three precincts, although this level of service would
> probably not be preserved over time if they were able to get the bugs
> out of their system. We can establish a competitive market for the
> system integrators. .....
>
This is pretty much the plan we've been talking about for over a year now.
There are lots of things we didn't have time to cover in the meeting.
Deirdre's presentation and my presentation were both severely abbreviated.

> It's not clear to me yet whether the software should
> be certified as part of the academic project, or if the system
> integrators should certify their particular hardit ware-software
> combination, or whether we can do some combination of both.
>
The plan is for the OVC to handle certification of specific hardware
platforms. We'll start with one specific set up and get that certified.
Over time, we'll get other set ups certified. Ideally, we'll reach a point
where we'll be able to get a system certified where the PC will meet a
general specification rather than a specific make/model.

> Alan came in with a little cheeze-box PC which is great for a prototype
> but not, IMO, sturdy enough to deliver as a real product. For an example
> of what would work in a delivered product, please see
> http://www.citadelcomputer.com/ . These things will stand being banged
> and dropped and having coffee spilled on them, and have an optical
> touch-screen. The detection mechanism is LEDs and photodetectors in the
> screen frame, and the screen has a polycarbonate sheet in front of it
> for protection. You can scratch or replace the polycarbonate without
> disrupting the touch sensor.
>
> The bad news is that they probably 10 times as much as Alan's device. I
> haven't done comparison shopping, but I suspect that we would be talking
> several times the price of consumer equipment to gain the reliability.
>
Right. However, the OVC will not make cost/performance decisions for
jurisdictions. Over time, we will offer a variety of solutions from which
they can choose. I have spoken with enough elections people to know that
one solution will not fit all. Some are very interested in multi-use, some
are not. It's also possible that a county might buy a system using OVC
software that they intend to be dedicated and warehoused much like DREs are
handled today... and when new decision-makers come in and say, "that was
dumb" (pointing to a successful multi-use model in some other county) they
might decide to change to multi-use. Or it could happen the other way
around (start with multi-use model and decide to go dedicated/warehoused).

Alan D.

==================================================================
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
==================================================================
Received on Sat Jul 31 23:17:05 2004

This archive was generated by hypermail 2.1.8 : Sat Jul 31 2004 - 23:17:15 CDT