ICANN/DNSO
DNSO Mailling lists archives

[council]


<<< Chronological Index >>>    <<< Thread Index >>>

Re: [council] e-mail votes v3




> 
> The first point probably provides a sufficient safety valve.  What is 
> the difference between a GA vote under the auspices of the DNSO 
> Secretariate v. a self-organized GA vote?
> 
> Harold Feld

Speaking as the one who has been conducting a number of votes
for the DNSO, I will limit my comment to describe you what
it really is.

To have a vote, one needs:
   1. a well defined electorate
   2. a time to start and to end, precisely respected
   3. a voting procedures to determine when a vote is valid 
      (such as conditions on minimal participation)
      and how to calculate it (simple majority, or single trasferable
      vote, or anything else which is mathematicaly defined)
   4. Watchdog Committe able to help solve problems, and to recalculate 
      results - to avoid a single point of failure
   5. public record in which each and every voting person may
      see his ballot, and still remain secret - to have who voted 
      for whom undisclosed
      (as someone recently pointed out: when elected people vote
      their vote must be completely disclosed; when an assembly vote
      each voter have rights to a secret vote)
and a committed, reliable person conducting the vote.

When the Secretariat is conducting a vote, all the above
are respected:
   1. the electorate list is used - by rule the email addresses 
      on the GA electorate list are kept undisclosed;
      each ballot get a persoalized key, valid for one vote;
      the received ballots are checked against posted ones
   2. the time is respected
   3. the voting procedure, which have been developped by
      the GA itself, is respected
   4. the GA Watchdog committee receive automatically replicated ballots,
      and is sollicited in case of claims
   5. the secret is preserved, and at the same time the whole
      public record provided; each vote have appropriate web pages
      providing for the full context

Please add that we are in the international environnment, which
oblige people to look on reference time in UTC and to deduce
their local time. They get lost sometimes.
That emails get bounced - for zillion reasons, temporary or 
permanent. That people may delete or lose their ballot,
and ask you one hour before the dealine to regenerate it (in some
votes the last 8 hours before the deadline are just rush hours).
That some will post in HTML format, sometimes even without knowing it
(thanks to some well known volume consuming softwares) 
not to mention that some will choose various international
character sets and personalized signatures, and then their 
ballot will arrive as encoded attachement. Etc.

In one sentence - a vote by nature is a complex and time
consuming. An electronic vote too.

Elisabeth Porteneuve

--
> 
> Philip Sheppard wrote:
> 
> > Harold, before we add in a possible override for the NC with respect 
> > to a GA vote we need to understand what this paragraph intends.
> >
> >  
> >
> > 1. The GA can vote whenever it wants, however it wants at any time if 
> > it organises itself to do so. The GA does not need the NC's permission.
> >
> > 2. If however the GA has a vote of sufficient importance that it wants 
> > to use the DNSO secretariat as returning officer and the DNSO software 
> > for the vote, then this rule requires it to check with the elected 
> > chair of those who pay for those resources. The rule says if its still 
> > in the budget, the answer is expected to be yes.
> >
> > 3. In the rare case that it is outside of the budget (eg the 50th vote 
> > the GA has asked for this year!!) the answer may be different. But if 
> > the answer is No, and the GA chair thinks that unreasonable, the GA 
> > chair can today choose to lobby any NC member or indeed propose it 
> > himself as an NC agenda item. If discussed then a simple majority of 
> > the NC could authorise the vote.
> >
> >  
> >
> > I hope that this is sufficient flexibility.
> >
> >  
> >
> > Philip
> >
> >  


<<< Chronological Index >>>    <<< Thread Index >>>