scottxyz
(1000+ posts)
Send PM |
Profile |
Ignore
|
Fri Nov-05-04 07:25 PM
Original message |
"I THINK I voted" - List of voting machine errors |
|
Machine Error Gives Bush Extra Ohio Votes http://story.news.yahoo.com/news?tmpl=story2&u=/ap/20041105/ap_on_el_pr/voting_problemsDear Federal Elections Commission - How many of the votes cast in the 2004 elections have been verified? Just wondering!
|
hedda_foil
(1000+ posts)
Send PM |
Profile |
Ignore
|
Fri Nov-05-04 07:42 PM
Response to Original message |
|
Somebody finally says it could be the tallying software.
<snip>
Kimball Brace, president of the consulting firm Election Data Services, said it's possible the fault lies with the software that tallies the votes from individual cartridges rather than the machines or the cartridges themselves.
|
scottxyz
(1000+ posts)
Send PM |
Profile |
Ignore
|
Fri Nov-05-04 08:21 PM
Response to Reply #1 |
2. The real fight is tallying |
|
Edited on Fri Nov-05-04 08:22 PM by scottxyz
Diebold has proven it cannot tally votes. (Strange, when they're so good at tallying your ATM withdrawals down to the last penny - and let's not forget the $1.50 fee!)
The challenge now is to design an "information architecture" which lets every citizen vote AND verify that their vote is included in the total. Results from systems where this can't be done will automatically be seen as less legitimate.
"Information architecture" could mean things like... duplicate or triplicate hard copies of a ballot, counted independently (by observers who can be openly partisan)... or some kind of billboard which publishes all the votes so each voter can (privately) verify that their vote was indeed included in the totals.
Interfaces (touchscreens versus punchcards) and implementations (Microsoft Access versus open-source) are totally separate from "architecture" issues. While computer scientists may be best qualified to address interfaces and implementations, non-programmers can address "architecture" - people with experience in auditing and security.
Diebold's very flashy but ultimately nondeterministic* black-box wide-area-network implementation of addition will go down as one of the most unverifiable totalling algorithms in the history of computer programming.
* Nondeterministic in computer science simply means a program that arbitrarily gives different outputs for the same input. This can be a "feature" in certain programs - such as coin-toss simulations - but it is considered a bug in voting or banking programs.
|
Eloriel
(1000+ posts)
Send PM |
Profile |
Ignore
|
Fri Nov-05-04 10:56 PM
Response to Reply #2 |
|
* Nondeterministic in computer science simply means a program that arbitrarily gives different outputs for the same input. This can be a "feature" in certain programs - such as coin-toss simulations - but it is considered a bug in voting or banking programs.
Sure it's not a feature in SOME voting machines? :evilgrin:
Hi, Scott. :hi:
|
TalkingDog
(1000+ posts)
Send PM |
Profile |
Ignore
|
Fri Nov-05-04 09:43 PM
Response to Original message |
|
Whip us up a "I think I voted?" logo? With the correct font and a checkbox with a question mark in it.
I'd buy that and put in in my car window.
|
DU
AdBot (1000+ posts) |
Fri Apr 26th 2024, 11:05 PM
Response to Original message |