On november 17th, the Online Policy Group will be involved in a legal action to cease the Diebold cease and desists against Indymedia, and the upstream provider the California Community Colo Project (where this machine is hosted). In the meantime, I've been requested to remove the diebold memos from this server until a ruling is granted next week.
On the topic of voting, I've gotten a number of responses on preferential voting. The first is the general complaint on Instant Runoff Voting. Mako correctly points out that IRV is really the weakest form of preferential voting there is, and a push for a voting system should be something more a long the lines of a condorcet counting method.
Also, Paul Harrison (developer of the RSPP protocol in linked to in an earlier post) writes an interesting perspective:
Here in Australia, we have preferential voting. In fact we have compulsory preferential voting -- you get a big fine if you don't vote.
Result: we still have the aussie equivalent of republicans in power, BUT both our "republicans" and "democrats" are way left of yours, and minor parties hold the balance of power in the senate and effectively stop the worst stuff getting through.
I would love to hear more about people's experiences with alternative tallies... And more to the point, any effective cooperative and consensus building tools that they may have used in their personal lives.
To implement a useable piece of voting software, you need security. Gnu.free is now defunct, but attempted to implement voting software on a government wide scale. My goals would be a litte less ambitious, as the problems free elections in governments seems to be a systemic cultural problem more than anything else. I think in small groups, cultural agreement simplifies trust models- but in a scenario with millions of people all of it goes out the window.
So perhaps the first place to start is to define a standard of what an internet voting system should minimally include security wise. The conditions for democratic voting are generally agreed upon as:
- Individuals are anonymous, ideally to the end tallyer
- the voting system needs to be auditable
- the voting system CANNOT issue verifiable receipts of an individual vote, otherwise it introduces the ability for a person to buy a vote
- the votes themselves should not be interceptable: while there is arguement whether it is good or bad to be able to forecast running elections, the client and the server connection should be secure.
- only legitimate partcipants can vote a set number of times
etc etc. As far as I am aware, there are no cryptographic standards in place for a non-governmental voting system....
Posted by Da Mystik Homeboy at November 8, 2003 12:00 AM