Re: [atlarge-discuss] Internet Voting...

[Walter Schmidt <walts@dorsai.org>]

On Thu, 5 Jun 2003, J-F C. (Jefsey)  Morfin wrote:

> I dont think there is any problem in that. The basic problem is to make
> sure that the ballot cannot be traced to a voter (like real ballots) while
> making sure everything is reported as it happened.

   ...yes it is.

> A non reversible algorithm like MP5 permits the fist thing. Synchronous
> reporting registered by trusted third parties permit the second one.

   ...I do not know of "Synchronous reporting registered by trusted third
parties" so I do not know if that would work. What I know would work is:
"(T)he voting process should leave a verifiable audit trail not only to
guard against election fraud and allow for recounts, but also to ensure
that every vote cast is counted." Doug Chapin, director of
ElectionLine.org, a nonpartisan research center in Washington
(http://www.msnbc.com/news/826193.asp?0si=-).


> The problem is not the voting mechanism but the trust.

    ...trust is not what is in question - reliability, understandability,
verifiability, reproducibility, accuracy and completeness, are.


> Anyway, you challenged me on the last vote on a precise point you and I
> make the core of the solution and you did not commented my response. Be
> sure this is not dispute: I do believe we can very easily find a good
> voting solution which scales - different from the physical one - but
> that people may trust the same on the long range. And I am interested in
> building it as I think it is rather simple ... but free.

   ...they will trust it, if its functioning can be attested to - and yes,
many would like to see it built.

> I thing e-vote may be _less_ tampered, only it permits because a larger
> number of different concerned people to watch them through different
> synchronous and asynchronous processes.

   ...I'm not sure what that would add or not add.


> This does not solve the ID problem. But IMHO the ID problem can only be
> solved the day e-Human Rights are first written and the e-Naming Rights
> defined. Joey's and my own attacks in here plainly shown the problem and
> how the demands of some (who are genuinely cocnerned) conflict wiht many
> existing laws and moral duties.

   ...as to what problems exist, today, when asked "Would Internet voting
add to the potential for confusion and fraud?" Rebecca Mercuri, a computer
science professor at Bryn Mawr College and founder of Notable Software,
was certain it would. "Were taking an inherently insecure medium, the
Internet, and layering security on top of it, she said. It doesnt work."

Mercuri, is skeptical that Internet voting could ever be made secure. "All
of that is completely susceptible to the latest virus attack, the latest
denial-of-service attack, sniffers and snoopers," she said.  "There are
vendors out there who are trying to mislead the public and election
officials into thinking that they have secure cryptography."

Further, Mercuri sees nothing that would stand in the way of a voter
selling or transferring his or her voting codes to someone else unless
election officials employed an intrusive biometric ID system. She and
other experts say the incentive for fraud or just plain mischief will
increase as electronic voting becomes more widespread.
(http://www.msnbc.com/news/826193.asp?0si=-).

I think she is wrong when she says that Internet voting could (n)ever be
made secure. Difficult, yes - impossible, no. It all boils down to
implementing the right controls. As to what those are - I do not think we
yet know...

--

 ---  REgards, walts@dorsai.org  Walter C. Schmidt, IT CPA  Blue(^) ---
 - -   Microsoft MVP - Windows XP Media Center Edition - HPMC 873n  ---
 - -                 Associate Expert - Expert Zone                 - -
 ---         http://www.microsoft.com/windowsxp/expertzone/         ---
 - - http://www.microsoft.com/windowsxp/expertzone/columns/schmidt/ ---
 - - 52 Ken           http://www.dorsai.org/~walts/          Sun 57 - -


---------------------------------------------------------------------
To unsubscribe, e-mail: atlarge-discuss-unsubscribe@lists.fitug.de
For additional commands, e-mail: atlarge-discuss-help@lists.fitug.de