[E-voting] Source Code

Mark Dennehy Mark.Dennehy at cs.tcd.ie
Sat Mar 20 17:18:37 GMT 2004

Ref:[Aengus, Sat, 20/03/04 @ 16:53 -0000]
> No, it doesn't. The majority are satisfied that they don't need to
> individually exercise their right to oversee the process. They aren't
> being discriminated against. They aren't being excluded from the process
> based on their lack of training or ability.

Hmmm. So you're saying that the fact that not everyone
*could* watch *all* stages of the process doesn't count as
discrimination or a practical limitation? 

> And frankly, the argument that it's okay to rely on a system that is not
> transparent to the vast majority simply because "the system already
> discriminates against the majority".

...is just as valid as the argument that we shouldn't release
the source code because it's a commercial piece of software;
or to allow us to rely on security through obfuscation; or
because you don't agree with "Open Source" (even though
that's not what we're talking about); or because VVAT is more

> It's watched by ordinary citizens who are members of the parties that
> the voters have elected, not by some technical elite. 

So a political elite is fine, but a technical one is not?
Where's the logic in that?

> There's nothing
> happening in a PT-STV count that can't be explained in a 15 minute
> training session.

And there's nothing in the source code that can't be
explained either - it just takes longer for some people. 
Believe it or not, many people have an equal amount of
difficulty in grasping the maths and nuances behind the
PR-STV system as well.

> We don't need the source code if Nedap go belly up tomorrow.

No, we don't. We could just go buy another turnkey system.
It's just that I don't happen to believe that spending
*another* forty million euro is justifiable when it could be
prevented now.

Or are you saying that we could continue to use the Powervote
system without the hardware designs, or source code, or any
way to maintain the system or correct errors or update it to
handle different elections?

> > So what's the problem?
> The problem is the only argument that has been presented in favour of
> an open source versus a (insert propaganda term of the week for non open
> source) system with VVAT is the purely ideological argument "open source
> good, non open source bad".

Excuse me, but that's not true. I've already given two in
both this email and a prior one:
1) The system should be something that anyone can examine
because it directly affects everyone's lives;
2) If Nedap goes belly-up tomorrow, we'll need the source
code and the hardware designs to continue maintaining and
updating the system we've just bought. The electoral process
of a country should *not* be the property of a private firm
in another country. 

> There's nothing wrong with open source
> software. But there's nothing about the open source model that adds the
> tiniest value to the problem we are trying to address. An open source
> voting system with VVAT is no more worth of trust by the average voter
> than an  system with VVAT is.

The thing is, I'm not talking about "Open Source" because
"Open Source" means that I can download the code, make a
change for a bugfix or what have you, and then *upload the
code* and have it become part of the build. 
That's not what I'm saying should happen here. I'm saying
that the code in the machines on election day should be up on
the website or available on CD-ROM for any member of the
electorate to see. There's a *very* big difference between
that and "Open Source"/"Free Software"/whatever.

> Source code is a distraction from the issue at hand.

I agree that it is not by any means as important as having a
VVAT in the system. That is the number one priority. But
having access to the source code and hardware designs is not
a triviality. It's an important point, and ignoring it is
similar to demanding a VVAT and not demanding procedures for
checking the VVAT that equal our current paper ballot; or
ignoring the primary legislation proposed to prevent a
challenge of the election results as delivered by the
Powervote system.

Mark Dennehy,
Computer Vision and Robotics Research Group,
Computer Science Dept., Trinity College Dublin		
email:	Mark.Dennehy at cs.tcd.ie
www:	http://www.cs.tcd.ie/Mark.Dennehy

More information about the E-voting mailing list