[ale] Thoughts on Electronic Voting

[Joseph A Knapka]

Jeff Hubbs wrote:

[a lot of thought-proviking stuff, which I snipped in order to get
to:]

> Inasmuch as I know that it would be possible to design and deploy
> electronic voting systems with a high level of integrity, their inherent
> complexity makes auditability by reasonable means impossible. 

Difficult, perhaps. Not impossible. It would be up to the implementor
to produce auditable code, not a mess of spaghetti. Of course,
crypto code should be obviously correct, not obfuscated (the
adage about security through obscurity applies).

> You can
> write in all the crypto you want, and it can be "walked around" by
> simple trickery [...].

Actually, it can't. The point of cryptographic voting protocols is
that the means of verifying that the process is working as advertised
is embodied in the relationships amongst the information possessed by
each of the parties to the protocol at each stage of the game. It is
implossible (or at least computationally impractical) to meddle with
that information due to the cryptographic techniqies employed, and
that safety is implicit in the protocol itself, independent
of any particular implementation. An incorrect implementation, whether
accidental or intentional, can easily be detected by any individual
voter with the motivation to do so, simply by examining the output
of the system.

> I therefore call for an end to "black box" voting systems of all sorts. 
> I want it to be possible for independent parties to audit voting results
> incontrovertably.  I don't care if the systems use nothing more advanced
> than brass cogs; I want something that's examinable in full.

I agree nearly 100%, the above notwithstanding; except that I don't
think "in full" should include the ability to associate a vote with
a voter.

Cheers,

-- Joe



---
This message has been sent through the ALE general discussion list.
See http://www.ale.org/mailing-lists.shtml for more info. Problems should be 
sent to listmaster at ale dot org.