John ryan deposition sept 23 2005

pg 97-98: As late as december 13 2004 canvassing results were being sent back to Seewolf and they were putting the information into the database and producing reports.

pg 98-99:
Q. Earlier I was asking you whether you remembered any changes that you were involved in making in the results data following the November 2nd, 2004 election. This exchange appears to concern a change both in the numbers for Bernalillo County, but also the addition of two fields to the official results database. Does that appear to be correct?
A. What date was that.
Q. December 13th, 2004?
A. That appears to be correct.
Q. Did your wife, Sandy, have the ability to
add fields to the election application?
A. Technically?
Q. Yes.
A. Yes.
Q. And would she also have had the ability to update or correct numbers for a particular county?
A. Yes.

pg 69:
Q. Did you maintain the servers on your own facilities that were used for the public and private websites for the elections?
A. Yes.

pg 71:
Q. This morning we talked about how Mr. Rainey originally came to you with the request for development of this set of applications. Did he at any time provide you with requirements in written form?
A. Not that I recall.
Q. And just to be clear, do you recall anyone else providing you with requirements in written form?
A. No.

pg 45:
Q. In the case of the primary and general election applications, did you use version numbers?
A. I think by default there is a version number, but did we do formal version control, no.

Explanation of why this is significant: formal version control is a method by which every change in the software is recorded. (ref: en.wikipedia.org/wiki/Revision_control ). By not using that, Seewolf could have changed the code at any point easily and there would be no record of that having happened. Of course, even with version control, copies could be created and then modified, so it isn’t a foolproof method but at the very least it would allow the process of development to be tracked. In addition, it is very clearly industry practice and the lack of using it demonstrates at best incompetency on the part of Seewolf Auburn (correlated by the fact that less than a year later, at the deposition, the person who single handedly programmed the whole system seems not to be able to remember any details about the system).

Notes from public comment

pat leahan: codirector of peace and justice center. was dissapointed to hear the issue was removed from agenda. did want to make a comment on the record about that. when there was a discription of what the candidates should and should not do. voters were the losers. so many voters affected by it they had lawsuit which caused paper ballots. votes switching. asking for injunction against machines. instead a settlement and paper ballots. we do not like ES&S but were grateful for a paper trail. so somewhat of an audit can be done. briefly make comments about auburn seewolf. if you read the thing you would think there was no problem with results. they only touch unofficial results. would like to send 111 page deposition by john ryan. you will be stunned. more than they receive data. they are involved in our elections from the programming to develop our web sites. ‘our programs needed to allow for the county clerks to enter data.’ ‘election night functionality’. no contact with SOS? “yes, if they needed help we would answer problems”. there were problems with PCMCIA. instead of AES it was seewolf. we do in fact have concerns with AES and seewolf. we tried to obtain the contract between them. towards the end of the dep. see pg 71 comment (read out). seewolf has much more control in our elections than we are aware of. when we learned of it, we asked SOS who is seewolf, she did not know who they are. wanted to make these comments and get you that deposition. We need oversight of the contracts, any contracts that affect our elections. thank you very much.