Repasso mensagem que recebi do prof, Michael Stanton (que tinha escrito um
timo artigo no estadao on line).
Contem uma outra mensagem de um americano, Brad Templeton, que, depois da
noticia de que a Microsoft, Unisys e Dell iriam desenvolver urnas
eletrnicas, comecou a debater como deveria ser esta urna.
A proposta tremendamente coincidente com a nossa do Voto-e:
1 - programas com cdigo 100% abertos
2 - garantia de carga do programa correto (eu no gostei do mtodo proposto
com o uso de hard-disks novos, acho melhor se fosse cd-rom)
3 - voto eletrnico + impresso do voto
4 - apurao imediata do voto eletrnico para a totalizao
5 - apurao por amostragem do voto impresso para conferncia.
Tem tambm uma segunda alternativa (o eleitor trazer o voto impresso de
casa) que no serviria para o caso brasileiro por permitir o voto-de-cabresto.
Tambm proposto um sistema em que o eleitor votaria em mais de um
candidato segundo uma ordem definida pelo prprio eleitor. Caso o seu
primeiro candidato fosse eliminado da disputa seu voto iria automaticamente
para o seguinte da lista. Assim se pode fazer a eleio em "dois turnos"
com uma votao s! (no sei se o eleitor comum entenderia o processo com
clareza)
A segue a mensagem...
[ ]s
Amilcar
--
At 13:42 13/01/2001 -0200, Stanton wrote:
Ol
Assino uma lista coordenada pelo prof Dave Farber da U. de Pennsylvania,
que incluiu umas mensagens a respeito da situao ali depois das recentes
eleies.
Abraos
Michael
=-==
Date: Fri, 12 Jan 2001 06:50:01 -0500
From: Dave Farber [EMAIL PROTECTED]
Subject: IP: Unisys, Microsoft, Dell to Create New Voting System
http://my.aol.com/news/news_story.psp?type=1cat=0700id=010252572125
Reuters
Jan 11 2001 12:52PM
NEW YORK (Reuters) - Three of the world's top computer companies
have teamed up to vaporize the paper chad by developing an
electronic voting system that would overcome the kind of ballot
confusion that wracked the U.S. presidential election.
Blue Bell, Pa.-based Unisys Corp. said on Thursday it will bring
together hardware from No. 2 PC maker Dell Computer Corp. and
software from Microsoft Corp. in the new voting system.
From: Brad Templeton [EMAIL PROTECTED]
For IP.
I think an interesting alternative to this would be to take advantage
of the huge resource of used, slow computers out there.
The open source movement would be called upon to make an open source
voting machine system. The source code, widely available, would be
highly scrutinized for fairness and elmination of protential fraud.
This software would then be placed on certified hard disks (which may have
to be bought new) and placed into any suitable PC. It would assume nothing
more than a low-memoried, slow pentium with VGA.
The public would then be asked to donate old PCs, and they would get a
tax deduction for it. I think there would be a flood of donations.
A quick check to assure the donated parts are standard, and you would
be able to get all the voting machines at close to zero cost (mostly just
the $50 new hard drive, since allowing a donated hard drive provides a
slight risk of fraud by very clever people who recode the firmware on the
drive.)
People would get a thrill out of donating their old PC to help an electoral
system in crisis.
The machines would not be internet connected. They would just have a
screen and printer. They would conduct the voter through their ballot
and then both record the ballot and print out a paper ballot which is
both human and machine readable -- with the machine reading what the human
reads, not some bar code or other non human readable info. When the
person confirms their paper ballot is correct, they end the process and
take the paper ballot to a ballot box.
At close of voting, the machines immediately report a tally of the votes
to the returning officers. However, randomly, or in case of a recount,
the paper ballots are counted by hand or with a scanner, as they are the
true record.
An alternative scheme would be to get donated scanners on a 2nd machine.
On the 2nd machine, the voter would take their ballot and insert it into
the sheetfed scanner. The machine would display their vote as read off
the paper ballot, and they would confirm it is correct, then place the
paper ballot in the ballot box. The machine with scanner would retain
the tally, and could rescan the ballots at any time. Again, the machine
would do some form of OCR on a ballot designed to make it reliable, it
would not rely on anything a human can't easily read.
In this system, voters could, in theory, prepare their ballots ahead of
time running the software on their own machines, and just bring them in
and present them to the scanner machine, confirm them and put them in the
box. If the scanner did not confirm what they want, they could go to
a machine at the polling station and re-vote.
Such machines would provide a quick accurate count, with