Re: [council] RE: motion to waive rules

[Elisabeth Porteneuve <Elisabeth.Porteneuve@cetp.ipsl.fr>]

Thanks Milton for reminding this fundamental characteristic 
of the NC vote, and since its origin. The vote is under NC control
and visible.

Because the NC are elected reps, and as such representing
various groups, it was considered since the origin of the 
NC in June 1999 that the voting system be e-mail must provide 
for 2 stages of verification:
  (1) by code (which is the personalized KxxxxZ, uniquely
      assigned to each voter-ballot)
  (2) by name once the vote is terminated

At the first stage each and every voter is able to verify
if what she voted for has been correctly recorded.
At the second stage everything is just explicit: who voted
for what is disclosed. Which imply that each group may verify
how their elected representatives vote.

The KxxxxZ random code is computer-generated. It is 4 hexadecimal
characters, which means that a probability to guess it in one try
is 1 of 65536 for each person. No need to say that the voting
software checks if there are multiple answers for each ballot
and should that situation arise, that case is double-checked
with full e-mail enveloppe by human.
If there would be any doubt about ballots received, the
Secretariat would be reporting immediately.

Elisabeth Porteneuve
(author of voting system and initial DNSO Secretariat for 28 months)
--

"Milton Mueller" <Mueller@syr.edu> wrote:
> 
> Agree with Bruce here. To add another argument: 
> 
> The Names Council is a group of only 20 people, it is not a large
> and anonymous group. All one has to do is make the results of 
> an email ballot public - after all, our votes in a teleconference are
> public and visible anyway - and the risk of fraud drops to zero.
> For I would be able to instantly see whether someone had voted
> for me in the wrong way.
> 
> >>> Bruce Tonkin <Bruce.Tonkin@melbourneit.com.au> 02/12/02 06:01PM >>>
> Hello All,
> 
> > 
> > I agree with Marilyn.  Given the potential for fraud, I think it is 
> > important that NC votes are ratified in person or via a telephone 
> > conference.  I am willing to share the burden by attending some calls 
> > outside of US business hours to accommosate our friends from other 
> > hemispheres. 
> > 
> 
> Well I am amazed at the concern for fraud.  But since you mention it:
> (1) how do you verify the voice on the other end of the line - no one
> checked it was me when I last called in - this was no checking of caller
> line ID, no scrambling of an encoded telephone line.
> (2) how do you know it is me when I turn up face-to-face - I have never been
> asked for passport ID when attending an ICANN meeting.
> 
> Technically it is possible to defeat the existing namescouncil security for
> both in person and phone calls.
> 
> In any case, it is about risk management.  From a personal point of view I
> believe that the risks for fraud via email in this instance are acceptable
> (but it might just be that I have been using email for about 20 years and I
> am perfectly comfortable with it for consensus building and decision
> making).  There are additional mechanisms for improving the security of
> email - digital certificates etc, just as there are for voice and
> face-to-face verification.
> 
> So I still stand by my earlier comments regarding email being acceptable.
> It is in many other business contexts and Boards I operate in.  Again we
> should be using the Internet constructively, and not relying on traditional
> means of communication.  If there is a concern about security of email -
> then lets use the readily available technologies to secure it.
> 
> This debate reminds me of those that are concerned about using their credit
> card over the Internet, yet will let a waiter take their credit card away at
> a restaurant to use from the kitchen while you wait for it to be returned.
> 
> In any case it is not a big issue for me.  If the majority wish to not use
> the Internet to conduct ICANN business that is fine by me - it does send a
> worrying message to the outside world though.
> 
> Regards,
> Bruce Tonkin
> 
>