Mark Crispin Miller: No Opscan in New York!

[Wilms]

News From Underground

NO op-scans in New York!

July 09, 2008

As a friend of Bo Lipari, and longtime champion of New Yorkers for Verified Voting (NYVV),
I do regret that I must disagree profoundly with them on the matter of how New York casts
its votes. While they call for replacing this state's lever system with op-scans, I see op-scans
as unacceptable. As computerized machines, they don't permit the sort of open vote-count
that democracy requires, but only partial and belated audits--a lousy substitute for the
simplest and most honest counting method of them all: one that we can all observe together.

snip

As Andi Novick reminds us here, the hackability of op-scans was made crystal-clear in HBO's
great documentary Hacking Democracy (which, as she notes, will be screened this Sunday
in Poughkeepsie, for those concerned about the question of New York's voting system).

In the following open letter, Andi makes a powerful case against the use of op-scans in
New York.

MCM

snip


Dear friends and particularly the Voters of New York State,

New Yorkers are the only U.S. citizens who have a protected right to vote. Incredible but true. Almost every other American (except for those pockets of the country that hand count) are using software-driven DREs or Optical Scanners, both of which have been proven to vulnerable to undetectable tampering. So voting on them is a crap shoot- you have no idea if your vote will count. Speaking of crap- please go to http://re-mediaetc.blogspot.com/2008/07/eyes-wide-shut.html to see what your state election commissioners think of this crap (crap is a direct quote from one of the commissioners).

So why are we abandoning our secure lever voting system for crappy, theft-inviting software-driven optical scanners? To ponder this further see http://re-mediaetc.blogspot.com/2008/07/eyes-wide-shut.html

And the answer is-- there is no legitimate reason to give up our safe, reliable lever voting system for this unsafe crap. That's why we have to sue the State and the State Board of Elections to stop computerized vote counting software from coming into our state. That's what I've been working on (and that's why you don't get quite so many emails any longer). Here's a two page summary of the complaint LitigationSummary.pdf. we'll be commencing WITH YOUR HELP. Unlike all the other times I've said we could really use your help- this time it's more important than ever. If it's not everyone of you and everyone you know supporting this litigation effort than who are we doing this for?

Software-driven systems can steal your vote and you'd never know. That's a fact and here are the 2 dozen scientific studies Sci Studies in 6-21v.doc corroborating how unsafe and unreliable this crap is. Forcing us to vote on undetectably mutable software is unconstitutional! It is the essence of disenfranchisement! If we're going to fight to preserve our sovereignty we all have to mobilize behind this litigation now.

I had hoped to send around a Petition tonight asking you all to circulate widely and to sign up as plaintiffs to the litigation, but the Petition on Line people have yet to get back to me, so for now- read and learn and in a few days I will hopefully be able to send out the Petition and more about what you can do to help.

In addition to the litigation summary and Eyes Wide Shut, read NY Loves Its Levers as New Systems Fail and LET'S CLEAR THE AIR- clarifying the misinformation out there, distilling the facts from the myths. And check regularly into our Re-Media Election Transparency blog http://re-mediaetc.blogspot.com/ which we'll update to keep you apprised.

There's your homework assignment until the next email asking you to be a Plaintiff in this historic lawsuit and sending you the Petition to sign and circulate.

If you want to talk about this more, after you've read the articles linked above, I'll be at a showing of Hacking Democracy this Sunday at 2:00, July 13th, at the Muddy Cup 305 Main St, Poughkeepsie, NY.

For more information, http://www.activistresource.org/calendar/cal_event.php?id=11078. If you haven't seen how an optical scanner can be rigged without detection, come out and see Hacking Democracy this Sunday. PLEASE CIRCULATE TO YOUR LISTS

thanks
andi novick


http://markcrispinmiller.blogspot.com/2008/07/no-op-scans-in-new-york.html

[yowzayowzayowza]

1. Speaking of crap....

As computerized machines, they don't permit the sort of open vote-count that democracy requires, but only partial and belated audits--a lousy substitute for the simplest and most honest counting method of them all: one that we can all observe together.

OpScanz are not responsible for the fact they are implemented in insecure vendor-dominated systems. OpScans do not preclude inclusion of steps which "we can all observe together", rather the process as a whole does. Blaming OpScans for such issues is as ignorant as blaming ballot stuffing in a fully manual count system on the ink pen. :eyes:

[Wilms]

2. I don't defer to MCM.

Agreed it is not the fault of optical scan...or MCM...or Bo Lipari that optical scan isn't implemented anywhere with the necessary risk-based statistical audits.

Lever machines, however, obviate some of the ancillary requirements, and offer a process we can all observe together.

Problem solved.

[yowzayowzayowza]

3. If I'm not mechanically inclined...

how am I to trust a lever machine whose operation is beyond my comprehension? Problem far from solved.

[Wilms]

6. Understood.

Do you advocate hand-counting in that case?

[Bill Bored]

7. 1+1 = 2, 2+1=3, etc. That's how a lever machine counts votes. You can see it as it's being tested.

Can you now explain how this is done with software? Include a discussion of source code, object code, ballot programming, operating systems, databases, servers, clients, memory cards and networks -- just for starters.

Then we can move on to certification, security, buffer overflows and other bugs, malicious attacks, safe-use procedures, risk-based audits and the chain of custody issues with the paper ballots of record.

Not to mention the hand counting required to do those audits (not that there's anything wrong with that). So where's the transparency now?

By comparison, I think 1+1=2, 2+1=3, etc. is pretty transparent.

[yowzayowzayowza]

8. With software? My solution in two basic steps:

1. In order to wrest control of systems from vendors We establish a standardized ballot layout specification discernable by both humans & machines and to which all vendors must adhere. Based on this spec We define capability sets for various classes of machines from a basic model for smaller precincts to a large high speed model for county/state validation. With this in place no one vendor can dominate the system.

2. Establish an adversarial automated count protocol wherein We may utilize privately owned certified units to verify the count on election night and after the election. In democratizing the automated count many eyes equals many machines from many vendors deployed by candidates, the parties, the citizenry at large as well as the official elections apparatus.

...source code, object code, ballot programming, operating systems...

"Ballot programming" is not required as the spec includes all information for both the machines and humans to comprehend the ballot. Since We are not trying to secure a monolithic solution, the internals of the machine are much less important.

...security, buffer overflows and other bugs, malicious attacks, safe-use procedures,...

Obviously each machine must pass a L&A test before deployment. Anyone presenting a machine to utilize in counting official ballots must have documentation on the certification of the model and sign an affidavit that the unit has not been tampered with.

...databases, servers, clients, memory cards and networks .... risk-based audits and the chain of custody issues with the paper ballots of record.

Thatz moving beyond counting to tabulation which is a whole nuther kettle 'o fish that lever machines are not a part of.

1+1 = 2, 2+1=3, etc. That's how a lever machine counts votes. You can see it as it's being tested.

As one cannot 'see' *all* the machines being tested and given the known methods for tampering with 'em, I don't find lever machines transparent atall.

Napkin back two cents.

[Bill Bored]

12. Ballot layout?

Yowza,

You wrote, "Ballot programming is not required as the spec includes all information for both the machines and humans to comprehend the ballot."

Last time I checked, that's what ballot programming was! I'm sure you're smart enough to know that. Machines don't just "comprehend" ballots; they have to be programmed to interpret them. If you're saying this should be more standardized, I agree it would be nice, but it might have to be a different standard for each state. Certainly multiple standards.

I like the idea of adversarial machines counting votes for the opposing parties, sort of, but in the event of a discrepancy, which party's machine wins? And how do we know both machines haven't been rigged by either insiders, outsiders, DINOs, RINOs, spies or saboteurs? It seems that the solution you are proposing is as bad as what we have now, except that it would cost at least twice as much. If Ross Perot or the Greens get a lot of votes, then it would cost 3 times as much! The third parties need to know that their votes are counted too, don't they?

"Since we are not trying to secure a monolithic solution, the internals of the machine are much less important."

I'm not so sure for the reasons I stated above and about your other point:
L&A tests can be rigged. But we can talk about audits....

But before we do, people can see the lever machines tested if they show up. See:
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x496663

And what you see is what you get, unlike software-based testing which can produce one result one day, and another one tomorrow -- even with scanners. This has already been proven by the U. of Connecticut.

You may say, let them audit the system, and guess what; they do! I hope they are as honest as they look though, because they use the Hursti Hack to do it! They read the ballot programming right off the memory cards of those scanners! I'm sure it works well (Diebold has seen to it that the data on the those scanner cards is NOT secure), but how is this transparent to the public? Do they have to learn AccuBasic? At least they also hand count some ballots in Connecticut, but sometimes too many and sometimes not enough, like in most states.

No, I'm afraid if NY is going to dump its levers, we have a long way to go to get close to the level of security we now have.

[annabanana]

4. I believe that there will be opscans at the polls this year

only for the handicapped and those who, for whatever reason, can't use the lever machine. Those "good old" machines are useless for the handicapped.

I think that the machines, with non-proprietary software and a statistically valid sample for an automatic recount in all elections would be an acceptable substitution.

Bo has gotten a lot of grief from every direction, and he HAS managed to keep those accursed DRE's out of NY. I believe he is an incrementalist, an approach that might serve us better over the long run than the "all or nothing" crowd's methods.

[Wilms]

5. No one has suggested not using ballot markers for those with accessibility needs.

NY has already provided them, though not in every polling place as required by HAVA. That is one of the ways in which the state was non-compliant, though I think you are correct to say that will be straightened out for the next election.

BTW, those ballot marker ballots have been hand-counted.

And I applaud the call for risk-based audits. Once in place, you really don't even have to care if the software is public or not. It's still hackable, and the risk-based audits are designed to catch hacks.

Not to rehash, but I'm not sure what "the "all or nothing" crowd's methods" means. There are many who think it's a bad idea to have computers involved with elections, and therefore prefer to keep levers. That's not hard to understand. It merely conflicts with one of the ER movement "trademarks".

[WillYourVoteBCounted]

9. did someone forget their bi-focals?

or is that a slip of the key?

[EFerrari]

10. Kick

[Wilms]

11. I may not have realized what an issue this has been for quite some time now.

As far as I'm concerned, the main thing is to make it clear that what the DoJ is doing--trying to force this state (or any state) to toe the federal line, to the electoral advantage of the GOP--is wrong. It's my view that the people ought to be included in this argument, and that we'll all come to the right decision when the people are included.

- mark crispin miller

December 04, 2007

http://markcrispinmiller.blogspot.com/2007/12/nyvv-and-doj.html

[kster]

13. Risked based audits? Come on you can do better than that

just say, don't you chumps even think about hand counting all those paper ballots before the ballots leave the neighborhood/polling place, because we will do a "risked based audit" after we smuggle the ballots out of your neighborhood and out of your view and you will accept our "risked based audit" as proof that the machine counted ballots in your neighborhood/polling place was done correct.

I think "We the People" are catching on to your GARBAGE!

Carry on... :)