Re: Stepping Stones -- Markamatic proposal. Secondround.

From: charlie strauss <cems_at_earthlink_dot_net>
Date: Mon Dec 13 2004 - 11:15:01 CST

One thing to consider in your requirements is handshaking and at least validation of data transfers. A real event that happened in both Nevada and New Mexico and maybe elsewhere was that the voting system produced its data in an MS access compatible file. Access read this in and Access was used to generate the final reports. The problem was that access had a silent read-error problem when reading the files and only read the first 32000 entries then stopped reading. No one knew this had happend. Since the final reports were aggregates by district not machines they could not be traced to specific machines. Thus they did not audit the output from MS access, just the input files.

So what would have fixed this would have been some real or virtual handshaking loop each time the data is transfered from one system to another. For example, even some simple checksums on the data sets would have been sufficient for access to validate that it had read the file with fidelity. A stronger form would have been for access to re-write the data it read in, and for an auxillirary (simpler) program to validate that the input and re-written data matched. The strongest from would have been for this re-written data to get passed all the way back to the original system the data was taken from so it would know the data had made it to the central tabulator intact.

  
_______________________________________________
OVC discuss mailing lists
Send requests to subscribe or unsubscribe to arthur@openvotingconsortium.org
==================================================================
= 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 23:17:11 2004

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