Name the top three Diebold experts you know of

[WilliamPitt]

I already know Andy, so three more besides.

[Wilms]

1. Bill Bored n/t

[HereSince1628]

2. Karl Rove for sure...

[New Earth]

20. lol

[Kelvin Mace]

3. Chuck Herrin

chuckherrin.com

Chuck tesified before my committee back in December (NC Joint Select Committee on E-Voting).

Roxanne Jekot
(contact me for her email)

And, uh, well, I know a bit about the files. <g>

David Allen
www.thoughtcrimes.org

[Andy_Stephenson]

7. Your too modest!

Kelvin is a wiz Diebold.

[greatauntoftriplets]

4. KelvinMace

n/t

[Andy_Stephenson]

5. Roxane Jekot

and if you call me I will give you her number.

check your pm's.

[elehhhhna]

6. She who shall not be named.

Oy.

[Melissa G]

23. LOL!

[Kelvin Mace]

8. Oh, I forgot

Justin Moore, a computer scientist who also testified before the NC committee.

[Andy_Stephenson]

9. Julie Anne Kempf

She ran 36 elections using Diebold GEMS and optiscan machines.

For practical everyday use...she is your gal.

Note she is a poster here at DU.

Check your PM's

[punpirate]

10. Still can't discount...

Avi Rubin. He's one of the people who's had a direct look at the software.

[Andy_Stephenson]

11. Avi has tested the DRE software.

While I agree with including him...a GEMS expert is also needed.

The sum of the parts count with Diebold software.

[Zan_of_Texas]

13. Yeah. Avi Rubin and Dan Wallach.

Both were on the first independent study of Diebold.

But Roxanne has looked at more of the stuff, I believe. Especially including GEMS. As you say, Andy. Plus she read the 20,000 Diebold memos.

Hi Andy! :hi:

[Andy_Stephenson]

16. Dan Wallach! Yes indeed.

Bev said he was working for the other side so he must be ok.

:hi: back at ya!

[Bill Bored]

12. Jim March, Avi Rubin and the guy who stole the election! nt

[Andy_Stephenson]

14. Jim March is a bit

too tied into Black Box Voting.

[Bill Bored]

19. I hear you. nt

[FogerRox]

15. Ummm--me ! !! ! ! NO --David ALLen

I did find the SEC filings that showed Wally Odell got 15,282 Million in stock options on Feb 14th 2004

then hmm Carmen Beck--nahhh not a diebold person though

And then there was the about 34 million he got on NOV 3 rd 2004

[Zan_of_Texas]

17. Stock options? No kidding. Interesting date too.

I missed that. PM me, will ya? With your source on that...or, if you have written it up somewhere.

[Bill Bored]

18. Who else has been looking at the "human factors?"

I've only been on it for a month or so, but it's obvious that GEMS is set up to facilitate all sorts of deliberate or accidental configuration problems on the Accuvote TS and OS.

I don't think most people know that the touch screens and op scans are actually programmed via GEMS. It's not your granddad's tabulator -- that's for sure.

[Amaryllis]

21. op scans programmed via GEMS?

[Bill Bored]

22. Diebold Accvote OS. The "OS" stands for Op Scan. nt

[Wilms]

24. Caught my attention when I first heard it.

[Andy_Stephenson]

27. Optical scans have "Firmware" installed.

in essence it is hard programmed and cannot be changed without quite a bit of effort.

The Diebold Tabulator is the big problem. However, that being said the opscans use accubasic a program designed by Guy Lancaster...who is still employed by Diebold. No one has ever really looked at it to my knowledge.

[Amaryllis]

28. What do you know about ES & S op scans? We have those in OR.

[Andy_Stephenson]

29. I know they are running on MS software.

My guess is it is a a dbase based on SQL. The ballots/elections for the ES&S system are done by the company. You also have some Diebold in Oregon...in fact the highspeed scanners.

[Bill Bored]

30. Andy, through GEMS, Op Scans can be configured to do the following:

First of all, by default, all the ballot error alerts that voters get are disabled. They have to be turned on in GEMS for undervotes and overvotes to be rejected by the Accuvote OS. Otherwise the voter gets no chance to correct such errors on the ballot.

Also, you can determine how straight party votes are counted. In particular, how a cross-party vote in a race will be handled if the straight party option has also been selected by the voter. For example, it's possible to set up the Op Scan to discard any cross party votes for President without the voter knowing it and it's also possible to exclude a candidate from a straight party ticket altogether, or even to assign a candidate from another party.

In fact, since it's possible to turn off the ballot rejections, you can probably set up Bush as the default choice for president in every straight party vote and the voter would never know it on an Op Scan. There were 17 states with Straight Party voting in 2004.

Since the Op Scan ballots are generated in GEMS, the names on the ballot can be switched independently from the candidates' database records so that votes for one candidate on the ballot actually go to the opponent in the database. This is pre-election configuration and does not require the direct access to the MS Access database demonstrated by Bev Harris.

All this is set up in GEMS and has nothing to do with hacking code, firmware, etc. They are just configuration options and GEMS can configure both Accuvote OS and TS remotely.

This is what I meant by human factors. Another way to say this is the user-level interface. And Op Scans are almost as vulnerable as touch screens because they are both configured through GEMS prior to the election.

[harmonyguy]

31. About Accubasic & Firmware

Both the TS and OS units use firmware.
In the TS units, the firmware can be easily updated since it's stored in flash memory. In the OS the firmware is more difficult to update since it is stored in UV-erasable ROM. (relatively old technology) and requires the physical removal of the (2)old ROM chips and their physical replacement with the new chips, which will have been 'burned' by Diebold.

All the ballot, race, precinct setup, of an election, which rightly or wrongly is often referred to as the programming, is handled in GEMS and then stored on a PCMCIA memory card which is then inserted into the OS unit - behind the locked door, if I recall correctly.

Accubasic - the Guy Lancaster piece - appears to be a simple reporting language written to run ON the OS unit in order to print out the poll tapes.
It's my understanding that the Accubasic language is 'burned-into' the ROMS and the poll-tape report programs are loaded FROM the GEMS computer INTO the PCMCIA card, along with the election programming info. I don't recall there being any facility in GEMS for actually creating or editing the poll-tape report programs, although I believe that a inquisitive individual with available documentation and a program like MS-Notepad COULD easily modify the various 'abo' files to suit their requirements. (just another area that needs to be 'certified')

Hope this helps.
HG

[Andy_Stephenson]

33. Yes...

"In the TS units, the firmware can be easily updated since it's stored in flash memory. In the OS the firmware is more difficult to update since it is stored in UV-erasable ROM. (relatively old technology) and requires the physical removal of the (2)old ROM chips and their physical replacement with the new chips, which will have been 'burned' by Diebold.

All the ballot, race, precinct setup, of an election, which rightly or wrongly is often referred to as the programming, is handled in GEMS and then stored on a PCMCIA memory card which is then inserted into the OS unit - behind the locked door, if I recall correctly."

I am not sure...but I don't believe accubasic was ever certified. I cannot find it in the NASED certifications. But as for testing and certification, I certainly don't trust CIBER and Shawn to do the job without a few advantageous "glitches." I am beginning to think federal elections should be seperate from local elections and they should be all paper, all handcounted in the precinct.

[Bill Bored]

36. Thanks HG. This was my main point:

You wrote:

"All the ballot, race, precinct setup, of an election, which rightly or wrongly is often referred to as the programming, is handled in GEMS and then stored on a PCMCIA memory card which is then inserted into the OS unit - behind the locked door, if I recall correctly."

The ballot and other options you mention include straight party voting options, tally options for SP voting, and also the ballot reject options that tell voters they may have made a mistake (which as I said, are disabled by default!). Also ballot rotations, the order of candidates' names on the ballot (which do NOT have to agree with the candidates' actual database records where the votes are tallied), etc.

Just to be clear, I refer to this as "configuration." In my book, programming is writing code. Configuration is point and click or simple text editing. User-level stuff. But I digress.

Are you saying that there's no networking of the OS to the GEMS server via modem? If so, this is a little more secure than the TS, but it also means the memory cards could be swapped and put in the wrong scanners, unless there's some authentication process for the cards, which I doubt.

So can the Op Scans make phone calls or not?

[Andy_Stephenson]

37. The OS machines

have a built in modem. If a county is xmitting votes to GEMS the os is plugged in.

King County uses a Depot method to run the election. Whereby results are transmitted via modem, from far flung parts of the county.

Using unencrypted straight ASCII text to an unpached windows machine... using RAS. Very secure IMHO (sarcasm).

[harmonyguy]

40. And also can be networked....

I don't recall if this was standard or an option, but the communication from the OS to GEMS can also be handled over an ethernet connection. I believe that this is/was mainly used for those 'Central Count' situations where they wanted to have several scanners hooked up to a GEMS system.

Here's a sample portion of one of the Accu-basic abo files. Although it may look like gibberish to some, look closer and you'll start seeing a pattern - rinse and repeat.......Your mileage may vary etc.

HG

F "^AA^PA{ZERO TOTAL REPORT}^OAA^BH^C^AD^PA{TEST ZERO REPORT}^OA"
F "A^BH^C^AE^PA{TEST RESULTS REPORT}^OAA^BH^C^AB^PA{ELECTION ZER"
F "O REPORT}^OAB^BH^C^AC^PA{ELECTION RESULTS REPORT}^E(^SUAC==E)"
F "^OAC^G^OAA^L^BH^C^AF^C^AG^OXA^OZB^OYY^OWD^PZ{4.4d}^OBB^NA{ PR"
F "INTING LABEL}A^DY*\n^BI^DY*\n^C^AI^PB^SVC^BJ^E^T^SVD^PB^SVD^B"
F "J^LDATE: %8.8s^SVB   TYPE: %1.1s^SVA\nPOLL CTR:%15.15s^SVE\n^"
F "E^QB\nVERSION:%3u^SUBA^DG COPY:%2u^SUBB\nCOUNT:  %3u^SVP   SI"
F "ZE:%5u^SUBD\nACCU-VOTE RELEASE:%6.6s^SUAA\nPRECINCT CHECK:   "
F "%6u(^SUCA+^SUCC)\nCOUNTER CHECK:^DE %6u^SUCB\n^L^E^SUDAPCT DA"
F "TA ERROR DETECTED\n^L^E^SUDBCOUNTER ERROR DETECTED\n^L^E^SUDC"
F "TEXT ERROR DETECTED\n^L^C^AH^OXA^OZB^OYY^OWD^PZ{4.4d}^E(^SUBL"
F "==B)^OCA^ODB^F(^QA==B)^OCB^ODA^F(^QA==C)^OCA^ODB^F^NA{   PRIN"
F "T FULL\n TOTALS REPORT?}L^OCB^ODB^F^NA{ PRINT SUMMARY\n  TOTA"
F "LS ONLY?}L^OCB^ODA^F^NA{   PRINT ALL\nPRECINCT TOTALS?}L^OCA^"
F "ODB^G^OCA^ODC^L^E(^QC==B||^QD==B)^BK^E(^QA==C&&^NA{ PRINT SUM"
F "MARY\n^DF TOTALS?}L)^OCB^ODA^BK^L^F(^QD==C)^HW^E^NA{ PRINT PR"
F "ECINCT\n NUMBER %-6.6s^SWA}L^BK^L^L^L^C^AK^OBA^E((^SUAC<E||^S"
F "UAC>F)&&^NA{   PRINT LONG\n^DE REPORT?}L)^OBB^L^OEA^I(^QE==A)"
F "^NA{PRINTING REPORT}A^DY*\n^MC^SUEEM^RA\n^DY*\n^BI^E^SUAB\nTI"
F "ME: ^WB  ^WA\n^L^E^QB\n=== OPTIONS SETTINGS ===\nREJECT BALLO"
F "TS WITH:\n^OF(^SVM&B)^BLOVERVOTED RACES^DG %3.3s^RB\n^OF(^SVM"
F "&C)^BLBLANK VOTED RACES^DE %3.3s^RB\n^OF(^SVM&E)^BLUNDERVOTED"
F " RACES^DF %3.3s^RB\n^OF(^SVM&I)^BLALL RACES BLANK^DG %3.3s^RB"
F "\n^OF(^SVM&Q)^BLSTR PARTY OVERVOTES  %3.3s^RB\n^OF(^SVM&32)^B"
F "LMULTI-PARTY VOTES^DE %3.3s^RB\n^OF(^SVM&64)^BLDUPLICATE VOTE"
F "S^DG %3.3s^RB\n\nPRINT OVERRIDE LOG FOR:\n^OF(^SVN&B)^BLOVERV"
F "OTED RACES^DG %3.3s^RB\n^OF(^SVN&C)^BLBLANK VOTED RACES^DE %3"
F ".3s^RB\n^OF(^SVN&E)^BLUNDERVOTED RACES^DF %3.3s^RB\n^OF(^SVN&"
F "I)^BLALL RACES BLANK^DG %3.3s^RB\n^OF(^SVN&Q)^BLSTR PARTY OVE"
F "RVOTES  %3.3s^RB\n^OF(^SVN&32)^BLMULTI-PARTY VOTES^DE %3.3s^R"
F "B\n^OF(^SVN&64)^BLDUPLICATE VOTES^DG %3.3s^RB\n\n^E(^SUBB==A)"
F "COPIES MADE^DK %3u^SUBC\n^LABS TALLY OPTION^DH %1u^SVI\nMESSA"
F "GE DELAY^DI %3u^SUBF\n^OF(^SVL&B)^BLSORT WRITE-IN VOTES  %3.3"
F "s^RB\n^OF(^SVL&C)^BLSORT BLANK BALLOTS   %3.3s^RB\n^OF(^SVL&I"
F ")^BLCOUNT DUPLICATES^DF %3.3s^RB\n^OF(^SVL&E)^BLCOUNT MULTI-V"
F "OTES^DE %3.3s^RB\n^OF(^SVL&Q)^BLSNGL BALLOT PRIMARY  %3.3s^RB"
F "\n^OF(^SVL&32)^BLADDITIVE STR PARTY   %3.3s^RB\n^OF(^SVL&64)^"
F "BLCOMBINED STR PARTY   %3.3s^RB\n^L^E(^QC==B)^BM^L^E(^QD==B)^"
F "HW^BN^L^F(^QD==C)^BN^L^BO^E(!^NA{  NEED ANOTHER\n^DF COPY?}L)"

[Bill Bored]

44. I see those options settings!

They are set through GEMS with mouse clicks but if this is a configuration file, as opposed to a report file, they could be changed there too. This is what determines which ballots are rejected or counted, or whether the voter is notified of the rejection and given a chance to fix the ballot, etc.

COMBINED STR PARTY means that a vote for a race will be discarded if there's straight party selection along with a cross-party vote in the same race. I.e., Republican for Kerry who tries to vote SP for the rest of the ticket. If there is no warning set, the vote for Pres is gone!

[Bill Bored]

42. Well, at least it's not on the Internet unencrypted, or is it?

What amazes me is all the ways there are to screw up the counting without hacking anything. It can just be done through GEMS configuration options.

[Andy_Stephenson]

43. Oh the opscan machones can be

hooked via internet if need be.

King county in a public records request gave out the number to the central tabulator. Anyone...with a modicum of computer skills could rig an election and no one would ever be the wiser.

[Andy_Stephenson]

26. Ballots are also designed in GEMS

and GEMS interfaces with Diebold's DIMS voter registration system.

[Melissa G]

25. I LOVE all you people! Your dry humor makes this Horrible

place we are in with this Hideous stolen election almost fun. Except for your DU corporate high caliber company I would be utterly despondent.:loveya: :loveya: :loveya: :toast:

[Cocoa]

32. Bev Harris

[Andy_Stephenson]

34. Darn cocoa...

why did you have to go and spoil a perfectly nice thread?

:D

[keepthemhonest]

35. Will are you going to the Nashville conference?

I did not see you on the list of speakers.

[Peace Patriot]

38. Kevin Shelley, the California Secretary of State who sued Diebold,

decertified their machines, and gave Californians a paper ballot option prior to the 2004 election, now gone--resigned last month after a vicious smear campaign by the SF Chronicle, joined by corrupt and collusive CA Democratic leaders, in fear of, and/or in the service of, the Bush Cartel/Schwarzenegger.

You said, "Name of the top three Diebold experts you know of." Shelley would have to be on that list.

He was one of the few Secretaries of States who made it his duty to become knowledegeable about Diebold and other electronic voting systems, and established the highest standards in the nation on electronic voting. He was also in a key position to find and expose the "smoking gun." I think that's why he was destroyed and removed.

(I think there was a 5% Bush popular vote padding in Calif., and, given Shelley's knowledge and vigilant sense of duty, he would have eventually found out how it was done, and blown the whistle on the election. There are other reasons why he had to be gotten rid of--Schwarz and the Bush Cartel have big plans for Calif.--for instance, a special election to redistrict the state, a la Tom DeLay in Texas, which they want to have happen prior to the kickin of the paper trail requirement in all counties, that Shelley had insisted upon. But I think there was something more--that Shelley was hounding Diebold and the others for their source code--and was courageous enough, and savvy enough, to figure out what they did.)

The ruination of Kevin Shelley is the ugliest thing I have ever seen happen in California politics, bar none.

We should honor the fallen. (And I hope he writes a book, and joins our movement. I've written him and encouraged him to do so.)

[Andy_Stephenson]

39. I agree!

Kevin is a hero IMHO.

[Helga Scow Stern]

41. Right On!

n/t