Re: Brand new concept in audit trails

From: Jim March <jmarch_at_prodigy_dot_net>
Date: Wed May 04 2005 - 00:07:24 CDT

Ed Kennedy wrote:

> Hello Jim:
> Actually, I like Joe Lunchbucket but Aunt Millie will do fine. We had
> been talking about ink jet or laser printers. I think this scheme
> would work better with two ink jet printers. The ballots can be
> essentially a duplicate without any changes in the design except for a
> screened, "Audit Only Copy," printed on the ballot vertically while
> they are being produced but in a way to not obscure the actual votes.
> I'm a little confused on where we would put the second printer for
> each voting machine or who would go around and pick up the second
> ballot. The model we had been looking at was a cpu and a printer
> attached under a table, say in some sort of thick plastic basket. I
> guess we could find room for another printer and perhaps some sort of
> bin that the second printer's output would drop into. Then these
> audit only ballots would be picked up after the closing of the poll.
> Still, considering the ability level of a typical poll worker there's
> going to have to be a lot of training. Let's keep working on this.
> One thing for sure though, you have to be clear to express that this
> isn't a voter receipt.

Joseph, the "complexity for the pollworkers" issue is why I'm leaning
towards single printer, two-part forms that have to be split back at
county elections HQ. This way, the voter is handling just "one sheet"
and the paper is going into "one box" exactly in traditional fashion.

We agree this ain't a "reciept". The only "gotcha" in this new proposal
is you need warnings at the voting booth saying "DON'T SEPARATE THE
PRINTED SHEETS". This could be on the screen in the voter's language or
on the audio track. To aid this, we can do the sort of multi-part forms
that are perforated both top and bottom and have to be split in two
places to separate.

Now...somebody else pointed out a possible voter secrecy error: if a
particular vote station had a screwed-up printer or was running out of
ink, those ballots could possibly be identified in the binders later. solution if we're going to do the "single printer per
votestation" would be to use DOT MATRIX printers. Yup. They still
exist. In fact they're still the industry standard for any situation
where you're doing multi-layer forms for things like renta-car
paperwork, etc. Prices are still quite reasonable.

So here's the cute part: you order two-level forms which have tractor
feed perforations up the sides. These perforations add extra width to
the 8.5" width...once pulled off the two sheets are separable and become

Now, you'd maybe have to do the bar-code strips for the visual/reading
impaired thing on the top and bottom instead of the sides? Not a big deal.

The advantage is, even if the ink starts to go soft towards the end of
the day, the content of the second sheet will be identical because it's
impact-driven versus ink-based.

You also don't have two critical printers that can't fail at each voting
station, you've got just one, and some VERY tried and true printing tech.

If the ink does go soft on the top sheet, it's fine as long as it's
readable. If it's not, then it fails the voter's "visual scan" before
turning it in. But here's the cool part: say they completely brainfart
and turn in a sheet that doesn't have a molecule of ink on it. With dot
matrix, worst case the dots will leave an identifiable "voter intent
mark"! It'll be pretty damned light but still readable with care AND
the secondary sheet will look exactly like any other.

The perforations would have to be very fine and easy to tear off even by
a completely blind voter. But I would submit that that's possible to
get, it's a question of paper vendor source QA.


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 Tue May 31 23:17:15 2005

This archive was generated by hypermail 2.1.8 : Tue May 31 2005 - 23:17:52 CDT