response 47 of 54:

Dec 19 19:02 UTC 2003

  re #42:  
  > The BIN will be such that only the voter can identify his ballot later.

  Any security scheme which relies on the voter to "identify his
  ballot later" sounds to me to be both (a) unworkably cumbersome,
  and (b) too susceptible to post-election voter manipulation.
  Imagine this kind of validation were used in the 2000 presidential
  election.  How much do you think it would have been worth to either
  party to change or invalidate a few hundred votes in Florida?  

  re #43:
  > Diebold's president, Bob Urosevich, said the changes were "cosmetic"
  > and blamed the counties for not tracking the software more closely.

  So Bob Urosevich is blaming California for not repeating what one
  would hope is a lengthy and thorough certification process for the
  sake of "cosmetic" changes to the software.  (Note:  one *hopes*
  that it's a lengthy and thorough certification process, but one fears
  that it is not..)

response 49 of 54:

Dec 19 20:25 UTC 2003

  I understand that you could do that but if the whole vote relies on
  voter-held information for verification purposes, then a suborned voter
  could essentially withdraw their ballot by claiming it invalid.  In a 
  close election, that could be enough to throw the race to the other side.
  And if the information supplied to the voter isn't necessary to validate
  her ballot, then what's it for?

response 51 of 54:

Dec 19 23:05 UTC 2003

  If the presumption is that the counted ballot is valid unless
  the voter can prove otherwise, how does that differ from the
  current system?

response 53 of 54:

Dec 19 23:36 UTC 2003

Are these paper ballots or just electronic ballots?  If the latter, how is
a voter to know that the counting of her ballot _should_ be challenged?