Editor's Note: In this open email, Commissioner Douglas A. Kellner, co-chair of New York State Board of Elections, publicly comments on the issues of certifying computerized voting technology in accord with a court enforced timeline. Vendor production schedule lags; incomplete systems provided; chain of custody documentation and instructions lacking. Kellner says:
I believe that there is still strong bi-partisan consensus within New York that we should stick to our policy that newly purchased voting equipment meet all of the currently applicable standards.
As most of you know, the New York State Board of Elections is doing certification testing on two precinct based digital optical scanning systems, the Dominion Voting System ImageCast (marketed in New York by Sequoia Pacific) and the ES&S DS-200. I believe that ours is the most comprehensive testing anywhere, including that done by the EAC (Election Assistance Commission).
The court ordered timetable provides for completion of certification testing by October 1, 2008, which appears to be increasingly unrealistic. The weekly report from SysTest, New York's independent testing authority, excerpted below shows just how difficult it would be to meet this target date.
What is particularly distressing is discussion that New York should consider overlooking what some are describing as hundreds of "de minimis" discrepancies from the VVSG or NY regulations. In my view, if they are really de minimis, I do not understand why the vendors cannot address them. I am trying to keep an open mind, but I believe that this may become a major issue within a couple of months as pressure mounts to have us overlook the shortcomings in the interest of replacing the lever machines in 2009. The industry and the Department of Justice will argue that if every other state is using equipment that does not comply with current federal standards, why should New York be the exception?
I believe that there is still strong bi-partisan consensus within New York that we should stick to our policy that newly purchased voting equipment meet all of the currently applicable standards.
This is an excerpt from the SysTest Weekly Report to the New York State Board of Elections:
SCHEDULE RISKS: Time and schedule continue to be the major constraint and risk to the successful completion of this project. SysTest Labs remains committed to providing thorough and complete functional testing, as well as initiating the run-for-the-record test pass in accordance with the current schedule and timeline. However; NYSBOE's ability to meet its court-mandated timeline for complete and thorough testing is at significant risk for the reasons described below.
Issue 1: Test Case Execution. With only 16 working days until the commencement of the run-for-the-record test pass, there remains insufficient time for a complete test pass through the entire test case suite for both initial and regression test passes. Our risk mitigation strategy is to configure five test labs for each vendor, allowing us to execute test cases in parallel. The likely outcome, however, is that not all of the test cases will be completed within the remaining allotted time, and the Vendor's will be unable to address all of the discrepancies discovered.
Issue 2: Documentation Discrepancies. Numerous documentation discrepancies, caused in large part by missing information from both Vendors' documentation, prevent us from finalizing the test procedures in all of the test cases. As of close of business on Tuesday, July 22, Sequoia/Dominion has 148 open documentation discrepancies and ES&S has 414 open documentation discrepancies. Without timely receipt of this missing information and documentation discrepancy fixes, thorough and complete testing cannot be accomplished per the schedule.
Issue 3: Functional Discrepancies. Based on our experience, both as an ITA/VSTL and with the current NYSBOE test project, we anticipate that a large number of discrepancies will be discovered as part of the functional test effort. There is a significant risk, given the remaining allotted time, that the Vendors will be able to address all of the expected discrepancies and SysTest Labs will have sufficient time to regression test the fixes. [sic] If a significant number of discrepancies are uncovered, the run-for-the-record test pass will have to proceed with these remaining open.
[Given the last sentence, it seems SysTest meant to say "There is a significant risk ... that the Vendors will not be able to address..." and that "SysTest Labs will not have sufficient time..." Ed.]
Issue 4: Hardware/Software Deliveries. SysTest Labs does not yet have all of the necessary hardware and software, which is still being delivered by the Vendors. Moreover, as we configure the voting systems, we are discovering additional documentation and procedural discrepancies.
Issue 5: NYSBOE Approval of Deliverable 7. Based on direction from the NYSBOE, SysTest Labs cannot begin testing without formal approval of Deliverable 7, Voting System Specific Test Plans. Therefore, without timely approval, the start of functional testing will be delayed.
Issue 6: Required NYSBOE Policy Guidance. SysTest Labs requested from NYSBOE policy guidance with respect to the configuration files for both Sequoia/Dominion and ES&S. At issue here is whether the State will allow changes to configurations in set-up parameters for individual counties or whether a single configuration will be mandated. This decision impacts vendor-specific test case development and thus will effect test schedules.
The following is a sample of the Vendor-specific delays that are keeping us from beginning functional testing:
In 2004, Rady Ananda began contributing to the Web, as part of the growing community of citizen journalists. Focusing mainly on elections, her blogs also address religious, gender, sexual and racial equality, as well as environmental issues; and are sprinkled with book and film reviews on various topics. She currently serves as a senior editor at OpEdNews.
All material offered here is the property of Rady Ananda, copyright 2006, 2007, 2008. Permission is granted to repost, with proper attribution including the original link.
In a time of universal deceit, telling the truth is a revolutionary act. Tell the truth anyway. Sign this petition: http://www.electiondefensealliance.org/ny_levers_petition
I'm in the software business, and I could write the code in 1 week to execute an election. Second, all software should be open source. Any company that claims to have "proprietary" software should be handled with extreme scrutiny. The state should not do business with any of these companies. The transaction to a DB/log is extremely simple and there is nothing "proprietary" with counters.
by
Son of our Father's (0 articles, 0 quicklinks, 0 diaries, 12 comments)
on Friday, July 25, 2008 at 11:47:34 AM
I have come across a system that meets all requirements of a reliable system. It is open source. It verifies the voter's intent and gives a paper receipt plus two paper ballots of that vote. It performs an internal audit by comparing the "official" ballot with the "verifying" ballot, and it further allows the voter to verify, after the vote, that his/her vote was counted as they intended.
Here are just 12 quotes from computer security experts regarding software:
Debunking Pre-Election Testing MythsJuly 19, 2008Debunking myths can be a full time job in the election integrity world. Here are 12 expert quotes on testing election software. Malware (malicious software code) can easily defeat pre-election testing and certification processes: logic and accuracy tests cannot "prove" that software is free of malware.
A thousand hats tipped to Doug kellner. It is refreshing to see an official fight for transparency- The open source community salutes New York's decision to waive fees for OS systems-
One problem with the " activist" community is the industry creates by fund raising- When " activists" start earning a living off the debacle- Solution might not be paramount. I find it interesting that the mostly Libertarian / Republican dominated election reform " groups" are hesitant to endorse open source. Even the Jona Goldman / David Dill advised larger groups ( PFAW, MoveOn etc. ) often fail to embrace better solution practices. This may be the work of Microsoft, or just innocent lack of understanding. We'll never know.
Please note the OVC will be demonstrating open source paper ballot printing systems at LinuxWorld August 5-7th in SF-
Brent Turner
by
Brent Turner (1 articles, 0 quicklinks, 0 diaries, 76 comments)
on Saturday, July 26, 2008 at 11:19:58 AM
I wonder if it is possible to have voting machines that will protect our votes in the November election? After the 2000-2004 cheating that went on, I am not so sure my vote will be counted correctly, as many other millions. We had that problem in those last two elections. It has been proven, so we must have a secure voting machine for this most important election.
by
Margaret Denson (0 articles, 0 quicklinks, 0 diaries, 11 comments)
on Friday, July 25, 2008 at 3:44:36 PM
It is obvious, to anyone, that the new voting machines are wide open to manipulation and fraud.
Unlike the voting machines that have been in place that demonstrate a transparent counting method.
I am really trying hard to understand what it is about the NEW machines that is motivating the politicians to put them in place as soon as possible, because the evidence indicates that we should be running away from the technology as quickly as possible!
Ciao, CZ
by
steve scheetz (1 articles, 0 quicklinks, 1 diaries, 482 comments)
on Friday, July 25, 2008 at 4:27:11 PM
"I believe that there is still strong bi-partisan consensus"
There is also a strong bi-partisan and non-partisan consensus for functioning, secure, transparent, accurate, reliable, trustworthy, theft-deterring voting machines which software-driven machines can never be.
We New Yorkers are fortunate- we don't vote on theft-inviting software. Yet!
New Yorkers know they can trust their lever voting system because its mechanical design deters fraud and our electoral safeguards are designed to detect that fraud and expose it- it's visible to ordinary people, all of which serves to prevent fraud. Those are the kinds of checks required in a voting system. It's insider tampering that has always been the greatest threat to our freedom. And it's insider tampering that optical scanners and DREs open the door wide open to.
Why would NY choose a theft-inviting software driven system when we have a theft-deterring lever voting system? Especially since there's strong consensus in favor of our lever voting system.
STOP SECRET VOTE COUNTING IN NEW YORK SIGN THIS PETITION
