[Fwd: Re: [E-voting] About Estonian e-voting]

Craig Burton caburt at alphalink.com.au
Thu Oct 27 09:42:33 IST 2005


In case my next mail bounces, here is a non-technical article which I 
believe touches on Colm's concerns and may round off the argument.
http://www.csoonline.com/read/090103/shop.html

You refer to the input/output layers of a computer programme and I 
assume you refer to the fact it is impossible to construct any kind of 
query for a "black box" that can reveal the box has been compromised.  
Is that right?

The above article advises that there is no way to examine a program and 
determine what it is going to do, even if it is a open box, let along a 
black box we can only query.  This seems to up the game a bit. 

The solution the article suggests is to limit what programs can do and 
reduce the general functionality of computers.  The partial solution I 
offered was a program on a limited virtual computer, the JVM, then 
reducing the voting application complexity, improving its transparency 
and using known protocols for transport of data.  Does this mean we can 
know absolutely that the voting application won't go in to an infinite 
loop?  Of course not.  Does knowing how it works improve our 
confidence?  Does parallel testing improve our confidence?  Does 
limiting the execution environment of the program improve things?  
Keeping the system as simple as possible?

I think these are good things for trustworthy e-voting, I'm sure it 
should go further, but there are no absolutes.

If you want to throw me off the this list for thinking and writing like 
this, you're doing me a favour.

Colm MacCarthaigh wrote:

>On Wed, Oct 26, 2005 at 06:26:21PM +1000, Craig Burton wrote:
>  
>
>>Thank you Noirin,
>>
>>My Google search was an example of the fact that the poster had offered 
>>up something that was perhaps expressed in an obscure manner.   If 
>>providing your (useful) query was that easy he might have done that 
>>instead. 
>>    
>>
>
>Craig, if you continue to troll this list: you will be banned. You have
>been warned. Once again: please stop wasting our time. 
>
>We have never banned anyone, and I *really* don't want to. But if you
>don't stop, it's going to happen.
>
>  
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.stdlib.net/pipermail/e-voting/attachments/20051027/edd2902b/attachment-0001.htm


More information about the E-voting mailing list