Re: Voting System Standards

From: Douglas W_dot_ Jones <jones_at_cs_dot_uiowa_dot_edu>
Date: Sun Jul 27 2003 - 17:56:52 CDT

> I certainly didn't want to imply that the Federal standards are good
> ones! There are lots of holes in them. On the other hand, they
> include lots of useful stuff -- handicapped accessibility rules,
> etc.
>
It might be nice if someone would sort through this and pick out the stuff
that might be relevant to the demo project. Any volunteers?

> > For example, strictly speaking, this voting machine is NOT a DRE.
>
> Indeed, your system, and for that matter, the Populex system and the
> Avante system, both of which print "receipts" that serve, legally,
> as ballots, are not, strictly speaking, DRE systems.
>
> But, they will be evaluated as such by all who see them, including
> the eventual source code auditors at the independent testing
> authority. ...
>
Okay, but we're not going for certification of the demo software. It's
throw away software.

> Legalistic wrangling won't convince anyone that it
> shouldn't be evaluated that way! Certainly, all the user interface
> components will be seen that way, and your machine must retain
> electronic images of the ballots as well as printing them out --
> so in effect, it will meet the redundant storage requirement for DRE
> machines by producing one paper copy and one electronic copy of
> each ballot.
>
Doesn't sound too difficult.

> They have their good elements, and as you've noticed,
> they have laughable ones. Sadly, some of these bad elements are
> being preserved in the IEEE draft standard I just read.
>
It's not sad, really. It means we have something to do. The larger full
blown [funded] study will address these things. We'll write the
standards -- "Regulatory Capture," as they say. If we do it right, our
system will fully comply with all the rigorous standards that will be
developed.

For one thing, we should define some minimum amount of data that the voting
machine should be able to present on one page. After all, in the document
on Usability

http://www.fec.gov/pages/vssfinal/v1/v1ac.doc

it says,

      - The design should minimize voter inputs, e.g., don't
         add unnecessary steps, minimize need to turn pages,
         and to navigate displays;

see that? MINIMIZE NEED TO TURN PAGES. With a large screen and high
resolution, we can get more than any existing system.

Add to this other requirements:

- - Must easily accommodate multiple languages
- - Must enable blind voters to have a secret ballot.
- - Poll site system must follow the absentee system in the following ways
(same in both systems):
        Electronic ballot image
        Paper ballot format
- - Source code must be open to public audit

etc. etc.

Alan
Received on Sun, 27 Jul 2003 15:56:52 -0700

This archive was generated by hypermail 2.1.8 : Wed Aug 06 2003 - 12:50:26 CDT