Re: Omidyar group

From: Arthur Keller <arthur_at_kellers_dot_org>
Date: Mon Dec 06 2004 - 12:03:22 CST

At 4:39 AM -0800 12/6/04, Fred McLain wrote:
>On Sun, 2004-12-05 at 23:41 -0800, Arthur Keller wrote:
>> The following is my personal opinion and not official OVC policy.
>> The best approach for satisfying HAVA deadlines is precinct-based
>> optical scan systems with Electronic Ballot Markers for disabled
>> access.
>> The OVC will likely NOT be able to produce a certified system in time
>> for the expected RFP's in the spring and summer of 2004. We don't
>> have a delivery infrastructure either.
>Frankly Aurthur, I believe you are mistaken here. I presume you're
>talking about 2005, not 2004. If, and only if, a state can make their
>election requirements visible, we can produce a voting system for Utah
>or North Carolina in less then 7 months. The developers here can do
>that, but we have to have concise requirements from the responsible
>parties. We're clearly the best choice, we need to demonstrate that

Yes, you are right. I meant 2005, not 2004. The key word is
certified. It takes several months and major $ to get a system
certified. A large part of the reason we didn't get funding from the
California Secretary of State was that even in September 2004 they
didn't think we'd have a system certified by the ITAs and the state
of California. Here we are 3 months later.

I would expect the RFP requirements to reference many other
requirements, such as FEC 2002, HAVA, etc. Since I've been serving
on two IEEE standards committees, I'm exposed to how complex these
standards can be, and how much testing is required. And the ITA
testing, although time consuming, is still insufficient for the job.

>I want to correct a misconception. The people developing the software
>are not the bottleneck here. It's the plan that is missing. Give us a
>goal and you will see the software. What motivates us is the idea of
>creating an open voting system. Think of the folks who are still
>working on the X-Prise project, they don't care that they didn't get the
>$10M prise, they are still working toward the goal. So should we.
>Missing an election is harsh, but we can produce the system regardless
>of these deadlines.

I think the volunteers should continue to work on a system. I've
written up some architecture documents. I've also written up some
next steps. Feel free to write more. Go ahead and build it. I'm
not stopping you. But look at the larger picture: if the OVC system
is NOT ready, would we rather have precinct-based opscan or DRE's
with AVVPAT (accessible voter-verified paper audit trail)?

> > I know Alan doesn't agree with my opinions here (which I have
>> discussed here earlier), but I know that others within the
>> voter-verified paper community do agree with me.
>I am a member of that community. Alan is a powerful spokesman for what
>we are doing, I respect his opinions on matters he tracks. If you need
>to know what we are able to do with the software he's certainly the
>wrong person to ask.
>Yes, the team can certainly be rebuild, reorganized and complete a demo
>for what Kathy is wanting to do in Utah. I discussed this with her this
>evening. It will be a bit of a push, but the SW team can accomplish
>these goals. With funding a 2006 election is well within our reach.

Give me a timeline that will get certified software ready by June
2005. And then tell me what the delivery vehicle will be, so that we
can support the needs for vendor-provided election administration.

>After having been involved with this group through more then one
>election, I believe the "lack of time" argument is moot. Our developer
>community can, and will, produce the software we need in a reasonable
>amount of time given a correct specification.

The federal specifications are published. What specifications exist
for the "Dechert architecture" are published. Feel free to flesh it

>This is IMHO, not the official position of OVC.
> -Fred-
>It is good to have an end to journey towards, but it is the journey that
>matters in the end - UKL.

I'm more interested in increasing the probability that the best
then-available stuff is adopted for 2006 (meaning selected by the
RFPs in spring/summer), whether or not it is our architecture, or
even open source. For me, a fun, enjoyable journey with a bad
outcome is not desirable.

Best regards,

> >
>> "The best is the enemy of the good." --Voltaire
>> Best regards,
>> Arthur
>> At 10:16 PM -0800 12/5/04, Ed Kennedy wrote:
>> > Hello Fellow Omidyar Correspondents:
>> >
>> > Can anyone add to this? This is under the smoke alarm subject
>> > matter.
>> >
>> > Hello David,
>> >
>> > Several states are involved in this process. Kathy Dopp is working
>> > on the same issue in Utah. Her web site is,
>> > <>. I will once again plug, OVC,
>> > <> as they are work on this from the
>> > software, hardware and legal end. They have a good web site.
>> >
>> > Even without all these people in motion there are things you can do
>> > on your own. Some the useful things you can do yourself are:
>> >
>> > 1. Educate your self. Both these sites above have useful links to
>> > sites like Black Box voting. Some of the laws and issues regarding
>> > voting can seem a little arcane. You need to be up on these so you
>> > can't be snowed by some vendor or some 'expert'.
>> > 2. Write letter(so) to the editor explaining the problem as you
>> > understand it.
>> > 3. Get acquainted with your County Registrar of Voters and the laws
>> > they work under. I live in California and the State's law are on
>> > the web and are searchable. I hope many other states have similar
>> > arrangements.
>> > 4. Get the laws and/or specifications for voting equipment in your
>> > state written to require VVPAT (voter verified paper audit trail)
>> > and preferably that the paper document is the actual legal ballot.
>> > 5. Also get the laws and/or specifications for voting equipment to
>> > be Open Source. Just putting the proprietary code into Escrow isn't
>> > enough. The code has to be available on paper or on the web for
>> > anyone to look at.
>> > 6. Insist that voting equipment not, repeat NOT, be networked
>> > together and that it be based on commodity PC's.
>> > 6. Stick with it. Commercial computing equipment does not last
>> > twenty years. Computers are usually technologically obsolete in two
>> > years. Before you say it, "The sooner the better," is a sentiment I
>> > agree with.
>> > 7. Is this all? Heck no, but that's enough for tonight.
>> >
>> > Thanks, Ed Kennedy
>> >
>> >
>> >
>> > Context:
>> >
>> > By David Goldbrenner (2) +|-, Sun, 05 Dec 2004 13:45:46 PST
>> > Comment feedback score: 0 +|- (net 0 from me)
>> > Joyce Mc Cloy said:
>> > North Carolina activist here again, we need to head off the
>> > private vendors at the pass. Our Chief of Elections wants to
>> > make North Carolina voting systems more universal like
>> > Maryland and Georgia. (both have Diebold TS!)
>> > That is my worst nightmare.
>> > Must have something to offer that makes sense and is fully
>> > developed.
>> > This is critical. Once a state spends $100 million on new systems it
>> > will be incredibly hard to dislodge. Do we really want to wait 20
>> > years for the next technology cycle? We need a viable open
>> > alternative NOW. The thought of more entire states using Diebold TS
>> > terrifies me. Joyce, is there any way we out of staters can help in
>> > the lobbying? Can we work to identify what states around the country
>> > are in the same position (Facing a big adoption?)
>> >
>> >
>> > --
>> >
>> >
>> > 10777 Bendigo Cove
>> > San Diego, CA 92126-2510
>> > USA
>> >
>> > "Let us all tend to our gardens." Candide - Voltaire
>> >
>> >
>> >
>> > _______________________________________________
>> > OVC discuss mailing lists
>> > Send requests to subscribe or unsubscribe to
> > >
>> --
>> Arthur M. Keller, Ph.D., 3881 Corina Way, Palo Alto, CA 94303-4507
>> tel +1(650)424-0202, fax +1(650)424-0424
>> _______________________________________________
>> OVC discuss mailing lists
>> Send requests to subscribe or unsubscribe to
>OVC discuss mailing lists
>Send requests to subscribe or unsubscribe to

Arthur M. Keller, Ph.D., 3881 Corina Way, Palo Alto, CA  94303-4507
tel +1(650)424-0202, fax +1(650)424-0424
OVC discuss mailing lists
Send requests to subscribe or unsubscribe to
= The content of this message, with the exception of any external 
= quotations under fair use, are released to the Public Domain    
Received on Fri Dec 31 23:17:05 2004

This archive was generated by hypermail 2.1.8 : Fri Dec 31 2004 - 23:17:22 CST