Re: Hand count or open source

[Nicholas Bernstein <nicholas.bernstein@xxxxxxxxxxxx>]

I encourage you to read the spectrum article for Chaum's take on this. 
Here's mine:

Yes, each ballot needs a unique ID. But that ID can be generated on the 
fly, so it doesn't sacrifice anonymity.

*Voter goes into electronic voting booth and makes selections.
*At the end of selection process, the voter is prompted for a password.
*A unique field  is created by merging information like polling place ID 
and the millisecond the selection process ended.
*This unique field is encrypted using the voter's password to create a 
unique ID
*The voters selections are also encrypted with the voter's password.
*The unique ID and encrypted selections are printed to a receipt that 
the users can take.
*The unique ID and encrypted selections are printed to a second 
"recount" receipt that held at the polling place.
*The unique ID and encrypted selections are posted on the web.

*A voter can look on the web to make sure that his/her vote was received 
properly.
*Any discrepancies can be reported to a voting authority .

This processes doesn't sacrifice anonymity. (Unless you give your ballot 
away, no one can link you to it. And even if you do give your ballot 
away, they still don't know how you voted.) And it allows users to 
verify their own ballot publicly. If the code that saves the electronic 
ballots is open source, then the tallying of votes is also protected 
from fraud. I don't know about other people on this list, but I believe 
this sort of end-to-end verifiability is the only way to conduct an 
election.


Nick


Paul E Condon wrote:

> On Thu, Nov 11, 2004 at 07:40:14PM -0700, Nicholas Bernstein wrote:
>  
>
> > Computers have the potential to be faster and more accurate. But we need 
> > a way of verifying their accuracy.
> >
> > I have said it before and I will say it again...
> >
> > What we need is a way for people to check whether or not THEIR INDIVUAL 
> > vote counted. Statistical sampling is a poor way to verify the accuracy 
> >    
>
> Checking an individual vote seems to me to require that each ballot
> have a serial number or ID number that the voter can write down and
> use to check the his vote record in a database of counted votes. I can
> contemplate this idea, but I have a hard time believing that it will ever
> receive the blessing of this group, or of any other in my lifetime.
>
> Is something else meant by "check"? 
>
>
>
>  
>
> > of a tally. Letting individuals check their own ballot is more accurate 
> > with less work. David Chaum's ideas are a step towards this. I'm not 
> > crazy about his specifics but I like the idea: everyone gets an 
> > encrypted paper receipt that they can check against a public list. Their 
> > are, of course, many variations on this theme. All of which make it even 
> > more incredible that their are machines that give no receipt at all.
> >
> > BTW, IEEE Spectrum had a good article on e-voting in their October issue 
> > with a sidebar explaining Chaum's idea. I have downloaded the pdf to my 
> > webpage. You can get it at osl-www.colorado.edu/~bernsten/chaum.pdf
> >
> >
> > Nick
> >
> >
> >
> >
> > Mary Eberle wrote:
> >
> >    
> >
> > > I support computer counts backed up by statistical sampling and full 
> > > hand counts if necessary, which is to say that I think we need to make 
> > > our first goal to change the Colorado law that stands in the way of 
> > > such a rational, modern approach.
> > >
> > > Mary
> > >
> > > Some Guy wrote:
> > >
> > >      
> > --
> > My razor-sharp wit was confiscated at airport security.
> >
> >
> >
> >    
>
>  


--
My razor-sharp wit was confiscated at airport security.