[clug] e-Voting: What would you want in a smartphone App?

[Kevin Pulo]

On Thu, Apr 01, 2010 at 08:57:52AM +1100, Sam Couter wrote:

> Martijn van Oosterhout <kleptog at svana.org> wrote:
> > I was actually explained by someone once that it was theoretically
> > possible to build a voting system such that you could after the fact
> > prove that your vote was correctly counted in the end result, but in
> > such a way they you could not prove to anyone else how you voted.
> 
> If I can prove to myself how I voted after I left the polling place, I
> can also prove it to the guy watching over my shoulder.
> 
> Kevin's TL;DR scheme has the same problem.

No, it certainly does not.  It only proves that your vote has been
included unmodified in the final tally.  It does not reveal how you
voted to anyone: not to you, election officials, politicians, police,
the guy threatening your family, your boss, etc.  Proving that your
vote hasn't been fiddled with is not the same thing as proving what
your vote was[1].

This is all given the assumption that the smartcard is uncrackable,
that is, that it's not possible to get the private key out of the
smartcard[2].  I assumed that the card is designed to make this
sufficiently difficult as to effectively eliminate the risk of
large-scale organised coersion[3].  I don't know much about
smartcards, but I figure this is a pretty common goal - store private
key material, never let it leave the card, use it to encrypt/decrypt
and sign/verify stuff.

Obviously this makes the card the weak point in the system (apart from
enrolment and other social engineering type issues), and so it would
best be designed and tested openly and extremely thoroughly, and well
in advance of actually being used.

Kev.

[1] Knowing how you voted provides an easy way to check that your vote
wasn't fiddled with.  But that is not the method used here, and that's
the whole point of using cryptography in this setting.

[2] Well, it assumes crypto is secure as well, so if someone comes up
with fast factorisation then this is hosed just like the rest of
cryptography.

[3] Which is nearly enough.  If it was commonly accepted that the
cards can't be used to find out how someone voted, then who would
bother trying to coerce people who can't be checked up on?

Much the same as how the current system prevents checking up on people
by forcing one person per booth, with no records afterwards.  Sure,
you could check on the person you're coercing by forcing them into the
booth next to you (somehow) when you vote, and then carefully peering
over at their ballot, but this is too much effort and too risky to be
able to do it enough to make any real difference - so people don't
bother even trying.

-- 
.----------------------------------------------------------------------.
| Kevin Pulo                Quidquid latine dictum sit, altum viditur. |
| kev at pulo.com.au               _ll l_ng__g_e_ _r_ hi__ly p__d_ct__le. |
| http://www.kev.pulo.com.au/         God casts the die, not the dice. |
`--------------- Linux: The choice of a GNU generation. ---------------'
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://lists.samba.org/pipermail/linux/attachments/20100401/be215d82/attachment.pgp>