Re: OVC Voting

From: Keith Copenhagen <K_at_copetech_dot_com>
Date: Wed Dec 08 2004 - 23:20:27 CST

Hi Ken,

On Wed, 08 Dec 2004 21:45:18 -0500, Ken Pugh <> wrote:

> At 08:12 PM 12/8/2004, you wrote:
>> I'd like to advocate the primary vision as "One person - One vote"
> Could you be more specific as to the meaning? Does this mean every
> eligible person should be able to vote? So would one of the possible
> goals be reform of the registration process to insure that voters are
> not denied the right to vote?
> I could take interpret this as being the same as my vision, for example
> "One person gets their one vote counted"

I may be being more subtle or breif than I meant to be,

A Vision of "Insure ballot accuracy" tells me that the OVC's mission
focuses on
helping people vote, probably with technology.

FWIW, My opinion is that is too narrow.

OVC needs to frame and advocate a complete voting system from voter rolls
canvass and tabulation. The value added of the OVC is the open design of a
compelete system of checks and balances.

Using a vision of "one person - one vote", opens the mission to address a
of security issues and procedures.

>> While one of the most attractive promises of voting technology is to
>> provide ballot
>> access to more people.
>> I claim the key contribution that the OVC can provide is to
>> build confidence and trust by building and deploying good technology.
> How would the technology that OVC develops be different than optical
> scan technology or DRE's that provide printed ballots (both of which are

The open development is a principle that needs to span the entire voting
Personally I want to see a clear framework with a best of breed modules.
The voteing booth is only one module, and for that matter untrusted, so I
like OVC to advocate that ALL information in and out of the booth be human
(in = voter precinct, voter id, etc; out = completed ballot, provisional
voter status, ...)
Also that regardless of voting technology, the VVPAT is the ballot, and
the ballot
is cast as placed into the ballot box.

>> "One person - One vote" is the foundation for
>> (1) VVPAT, audit and procedural protections from vote fraud.
> With VVPAT, what types of audits (other than manual recount) might be
> involved?
I'm advocating an worm audit log in every computer, and also associate
with each election
official (for example their audit trail could travel on their USB Drive
with their private key.)
So ballot counts, spoiled ballot counts, voter registrations, # ballots
printed, ...

> Do you mean audits of the registration process?
> Do procedural protections involve the registration system itself or the
> procedures at the voting place ?

Soup to nuts.
The framework needs to provide for testing and auditing each module and
the overall

>> (2) Political sunshine and it's cousin Open source software.
Thats why we're open

>> (3) Anonymous voting to avoid coercion.
> I agree with that.
>> While I heartily agree with Ken's "count every vote", in practice this
>> country has had over 1%
>> loss in virtually all elections,
> I think it depends on how these votes are lost. If they are due to
> technological errors, I feel they are preventable. That should be the
> purpose of an OVC system.
> If they are procedural errors (lost absentee ballots, etc.), then they
> are much less avoidable.

I want to protect all the votes, I think if we can define the interfaces
any software must match to, we can pick where we get the biggest bang for
the buck.

>> I think the limited resources of the OVC should focus on
>> 1, 2 & 3 above.
> What measure objectives are you suggesting? Are they something like:
> a. By the election of 2006, all [or some percentage of ] electronic
> voting systems will have VVPAT.
> b. By the election of 2006, all [ or some percentage of ] electronic
> voting systems will have independently verified voting software
> c. By the election of 2006, all [ or some percentage of ] electronic
> voting systems will have been tested to standards set by OVC.
My crystal ball is pretty hazy, I think we as OVC need to complete the
(thanks for pitching in) prioritize and create a couple of scenarios.
(1) what can we do for free
        (Maybe define interfaces and work on their acceptance)
        (Prioritize and make progress on open versions of each module)

(2) What can we do as a for-profit consortium, leveraging future revenue
        providing turnkey voting systems.

(3) What we can do as a not for profit group, working from grants

I would claim the most critical thing we can do is to educate ourselves
and then the
      election community about open design and open standards based
security. I'd go
      for (a) above with an added thread of voter/ballot management during
the voting
      process. And performance standards that the module must live up to.

>> Thanks for reading
>> On Wed, 08 Dec 2004 17:15:03 -0500, Ken Pugh
>> <> wrote:
>>> In parallel with Keith Copenhagen’s Strawman Proposal, I suggest the
>>> following draft charter for OVC. If such a charter already exists, I
>>> missed it and I apologize for the duplication. Without having a
>>> vision and objectives, it’s hard to know where one is going and when
>>> one gets there. The objectives are measurable tests that should be
>>> achieved within a given amount of time.
>>> -----------------------------------------------------------------
>>> * Vision of OVC
>>> To insure that every registered voter who casts a ballot has their
>>> vote accurately counted.
>>> [ Alternative - To insure that every eligible voter has the
>>> opportunity to cast a vote which is accurately counted. This
>>> alternative would expand the scope to the registration process itself]
>>> * Goals
>>> 1. No ballot shall be disqualified due to any problems with the ballot
>>> (illegibility, overvote, etc.)
>>> 2. All votes shall be capable of being recounted using paper ballots
>>> that have been verified by the voter.
>>> 3. The vote totals that are reported by any electronic means and the
>>> totals reported by hand counting shall agree exactly.
>>> [Alternative ­ If the vision was expanded, then another goal might be:
>>> Every eligible voter shall be registered to vote]
>>> * Objectives
>>> 1. Ballots with overvotes, other than ballots that have been submitted
>>> by absentee ballot, shall be zero [or less than .1% of the ballots
>>> cast in a precinct] by the general election of 2006.
>>> 2. The discrepancy between tabulated paper ballots and electronically
>>> tabulated ballots shall be less than .1% of the votes cast in any
>>> precinct and less than .01% of the votes cast in a state by the
>>> general election of 2006.
>>> [These percentages are arbitrary ­ they should be set as potentially
>>> acceptable goals. There will probably never be an exact match between
>>> the two numbers, but the match should be close enough].
>>> -----------------------------------------------------------------
>>> * Courses of Action
>>> The Strawman proposal suggests some tracks ways for achieving these
>>> objectives. I would realign the tracks into the following
>>> possibilities:
>>> a. OVC could produce a software/hardware system that met the
>>> objectives.
>>> b. OVC could produce a framework for systems to meet these objectives.
>>> c. OVC could create a set of tests that commercial systems had to pass
>>> in order to give reasonable certainty of meeting these objects
>>> d. OVC could suggest legislative wording that would require these
>>> objectives to be met.
>>> With the exception of legislative action, I expect that any of these
>>> tracks are possible within a reasonable period of time and I’d be
>>> happy to work on any one of them. After the 2000 election, I proposed
>>> a system to some NC legislators, but nothing came to pass at that time.
>>> Ken Pugh
>>> At 08:30 PM 12/7/2004, Keith Copenhage wrote:
>>>> Strawaman Proposal & Plan for the OVC community.
>>>> Overview : A two pronged approach,
>>>> Track 1 :
>>>> Specify and publish a voting system component framework
>>>> (Stepping Stone), and then working with the existing election &
>>>> political structure, building expectations & evaluation criteria, and
>>>> work toward adoption.
>>>> Refine and enhance the data models, data validation models,
>>>> audit and self-checking.
>>>> Track 2 :
>>>> Focus on the value added that the OVC can provide, prioritize
>>>> and build components within the Stepping Stone Framework, as well as
>>>> general supporting tools and methodologies.
>>>> Deploy and enhance systems, create redundant mechanisms and
>>>> enhanced security modes.
>>>> Track 1 : Stepping Stone Voting System Vision
>>>> Phase 1 :
>>>> a. Adopt and publish a component (functional) framework reflecting
>>>> critical interfaces and data flow in the voting process.
>>>> b. Identify the evaluation criteria for each component & interface.
>>>> Develop proposals for test & audit of each component.
>>>> c. Anticipate issues with the Framework, and provide for remedies.
>>>> Phase 2 :
>>>> a. Support legislative & election adoption of the framework &
>>>> evaluation criteria. Including providing draft language for
>>>> legislation & criteria.
>>>> Track 2 : Voting System Component Development
>>>> Phase 1 :
>>>> a. Clarify the OVC value added and publish the OVC design criteria &
>>>> methodology.
>>>> Phase 2:
>>>> a. Harden the data models.
>>>> b. Build structural methodologies and components.
>>>> c. Prioritize and implement Stepping Stone components.
>>>> ----
>>>> _______________________________________________
>>>> OVC discuss mailing lists
>>>> Send requests to subscribe or unsubscribe to
>> --
>> Keith Copenhagen
>> _______________________________________________
>> OVC discuss mailing lists
>> Send requests to subscribe or unsubscribe to

Keith Copenhagen
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 23:17:08 2004

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