Re: Barcode Vocalization Application

From: Jan Karrman <jan_at_it_dot_uu_dot_se>
Date: Tue Oct 07 2003 - 02:19:24 CDT

On Mon, 6 Oct 2003, Alan Dechert wrote:

> Now for the BUGS!!
>
> I am expecting a 39 digit number but I got a 40 digit number choosing the
> "Get bar code number" option. Then I got a 44 digit number scanning the
> barcode for the same ballot. Specifically, here's what I got for ballot
> number 1234 and selection the first option in each race (first 3 in cat
> catcher and ranking all the county commissions in order of listing, i.e.
> Packard 1st, Hewlett 2nd, etc.... write-in 8th:
>
> 1234164113698038190059864668966218971635
> 12342468164113698038190059864668966218971635 <-- scanned w/CueCat
>
> Unexpectedly, the first number gives the right answers fed into my bva app!
> The scanned number obviously doesn't work We should be getting the same
> number either way, no?
>

Arghh, a last minute change made when I no longer had access to
a printer... The ballot number was added twice (and the first
copy was obfuscated). I have fixed this now.

Talking about the obfuscation, I realized that it is not good
as it is. It will be very easy to recognize bar codes on ballots
with all votes blank. If the ballot number is abcd, then the
bar code number will be abcdabcdabcdabcdabcdabcdabcdabcdabcdabcd
(plus the additional start/stop/checksum) i.e. a repeating pattern.

Instead of obfuscating the digits with the ballot number repeated,
one can do it using the number:

(ballot_number + num1)*num2 modulo num3

where num1, num2 and num3 are some carefully chosen integers. There
is a theory on how this should be done.

/Jan
==================================================================
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
==================================================================
Received on Fri Oct 31 23:17:01 2003

This archive was generated by hypermail 2.1.8 : Fri Oct 31 2003 - 23:17:07 CST