Re: Scanner, write-ins, printer (was Re: Barcode -- printed ballot)

From: Alan Dechert <adechert_at_earthlink_dot_net>
Date: Tue Oct 07 2003 - 07:36:47 CDT

Jan,
>
> I have actually got the wrong code scanned by the Cue Cat once now.
> It got three symbols (pairs of digits) wrong. The checksum character
> for both the correct and the faulty code happened to be the same.
> The font size was probably a little bit too small, because it took
> several tries to get it to read at all.
>
I have never seen such an error. I have seen scanned output where I only
got a few spurious characters -- but never while scanning any of your
barcodes. Checking the length of the string should catch these errors. But
if you got the same string length with some wrong digits, then this is a
different error from anything I've seen.

> As I wrote earlier, I always add a leading zero to the decimal number
> to make it an even number of digits. To improve the reliability of
> the scan, this could instead be a checksum digit we calculate somehow.
> This will then not affect the length of the bar code.
>
Okay. For now, I would like to hear from other testers. Let's see if
anyone else sees such an error.

Hey Matt!

> The application that reads the input from the scanner should, beside
> from checking the checksum digit and length of the string, also issue
> some sound to indicate a successful scan.
>
Another checking routine could ensure that the number represents a valid
ballot -- i.e., only one (or none) selection in contests 1 - 11, no more
than 3 cat catchers, no more than one ranking possible for county
commissioners. I would guess that the probability is very high that any
mistake in reading the number would also cause an invalid ballot. I mean
there are 2^116 possible strings of zeros and ones 116 digits long but only
a very small percentage (tiny fraction of one percent) of those would
represent valid ballots.

This would help catch any incorrect barcode reads and would also help
validate the logic of the ballot data entry routine.

Alan D.
==================================================================
= 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 23:17:02 2003

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