[EM] untraceable receipts

[David GLAUDE]

Ernest Prabhakar wrote:

> There's also something to be said for not over-optimizing for the corner 
> cases.  At least in the vast majority of precincts in the U.S., active 
> coercion is quite rare. We should design to avoid the most common 
> problems, as long as there was some way to address the edge cases with 
> more effort.

Belgium is not a country where we think the government or a specific 
party will try to modify the election result to get more power or stay 
for ever in power. At least not here and now.

However if electronic voting become common place, in many years, no-one 
will care to verify or distrust the system. And at that time, maybe the 
less democratic party of all will be in power and try to stay there by 
all mean.

Also there is another risk. If Cuba start doing electronic election 
taking Belgium or the US as an example and mimic it. Then we (democratic 
country) will have no way to say that the system Cuba use is not democratic.

Actually Belgium frequently send observer all arround the world (in 
country where the democracy is weak) in order to verify if everything 
goes well. But WHO ARE WE to judge the level of democratic election in 
other country where in our own country it is not possible to recount and 
only few the happy few expert or worker from the vendor of the e-voting 
system could control the election result.

> Of course, in other countries I imagine it would be quite different.  
> But I think the more important question is whether we can reach 
> consensus about -what- information should be available to -whom-. 

My vote should only be known by me (that is the principle of a secret). 
I don't even trust the computer in front of me. ;-) (many ATM machine 
have a "web"-cam for security reason. How long until voting machine have 
one?

Then my vote exactly as I made it should be available (mixed and shaked 
with other) to the counting team... counted with whitness from each 
party. Then after the count, keeped securely for potential recount.

I think that before any electronic system get introduced, you need to 
carefull set the goal and define what democratic election are.

My definition is:
1) The vote should be secret.
2) The election result should be under citizen control (not expert 
control) [Where expert are the computer scientist for the state or from 
the vendor or ...]

If a system exclude 99.9% of the population that must trust the other 
0.1% able or capable to understand the system, then we have a problem.

I had the plan to present a list of computer scientist durring the 
election. Then if our list get absolute majority... who will trust the 
result of the election.

 > If we can agree on that social question, the technical problems 
should > be simple to resolve.

Trust me it is NOT SIMPLE at all.

Securing a system from the user might be easy.
Securing a system from the creator/operator of it (and hacker) is way 
much more difficult.
Making such a system transparent as election should be without 
compromising the secrecy of the vote make it even more difficult.

David GLAUDE