I'm a Wikinews editor, and was kind of interested in the LA Times experiment: over at the English Wikinews, we've been avoiding editorials since they're so personal in nature, and not NPOV.
I ended up on the Wikitorial wiki soon after it opened, and proceeded to help with the vandalism, and with providing some navigation, new user help, etc. Jimbo Wales (founder of Wikipedia) was also around from time to time, as were other Wikinews and Wikipedia people, trying to grow the wikitorial from a one page thing to something actually usable by a group of people.
I have both a Palm V and one of the first-generation PalmPC CE devices, and I find the CE grafitti-like recognition system (jot) to be _significantly_ easier to use.
At the same time, imagine booting your brain. All of those kernel messages... How're you going to do TCP/IP, anyways -- wireless? Using hair as your antennae?
I actually had one of those nifty Sidekick manuals with the Borland guy on it...
And used Turbo Pascal 3 in my CS class in high school. Of course our teacher, I remember, was a little bit of an old-timer and got amazed every time we added a little bit of color to our program (e.g. "ooh, color! that's wonderful!")
Well, actually if you read the appropriate chapter in APplied Cryptography you'll see what the right way to do it is.
There is a number of ways to make a system that does _not_ allow the verification service an easy way to find out how a certain person has voted, while making sure that the votes are actually unique and valid.
Yes, it can be made anonymous and at the same time secure, within certain bounds of both.
Other than the aforementioned Applied Cryptography's chapter, there's a couple of resources on the Net to go to see more secure election protocols.
In my experience, I was able to set up a fairly anonymous (except to people with real-time database access or access to transaction logs (which shouldn't be turned on)) system that also guaranteed the relative impossibility of voting twice, or of the people setting up the votes the ability to figure out what you've voted for. Of course, the validity of the votes was also guaranteed, and it was possible for the voters to double-check to see if their vote was counted.
This of course was for a very small set of people (about 30 people in a Computer Science department). But even there it is significantly more secure than the old paper-based system they were using.
Actually, I see this as the thing that will make me not want to use hushmail at all for any secure transactions. I can trust secure point-to-point communications, I can trust a good remailer chain. But to trust a small company, that's sketch. If this was being done by Microsoft or IBM or Sun I would trust it -- these are big guys who have enough people to make sure that stuff is secure, have enough expertise to do it right the first time (and not, like someone noticed, use the Date/Time as the seed for the random num generator), and someone who is going to watch the system very closely to make sure that they themselves never have the ability to read the messages or else they'd be involved in a major PR nightmare. But hushmail has nothing to be afraid about, and they are not big enough to guarantee that they'd even tell people if they start getting investigated by the feds in whichever country they're located in.
Slashdot Mirror
er, by "help with the vandalism" I of course mean "help clean up the vandalism"...
I'm a Wikinews editor, and was kind of interested in the LA Times experiment: over at the English Wikinews, we've been avoiding editorials since they're so personal in nature, and not NPOV.
I ended up on the Wikitorial wiki soon after it opened, and proceeded to help with the vandalism, and with providing some navigation, new user help, etc. Jimbo Wales (founder of Wikipedia) was also around from time to time, as were other Wikinews and Wikipedia people, trying to grow the wikitorial from a one page thing to something actually usable by a group of people.
I've written up about my personal view on the wikitorial experiment. Take a gander.
I have both a Palm V and one of the first-generation PalmPC CE devices, and I find the CE grafitti-like recognition system (jot) to be _significantly_ easier to use.
At the same time, imagine booting your brain. All of those kernel messages... How're you going to do TCP/IP, anyways -- wireless? Using hair as your antennae?
:-]
I actually had one of those nifty Sidekick manuals with the Borland guy on it...
And used Turbo Pascal 3 in my CS class in high school. Of course our teacher, I remember, was a little bit of an old-timer and got amazed every time we added a little bit of color to our program (e.g. "ooh, color! that's wonderful!")
Well, actually if you read the appropriate chapter in APplied Cryptography you'll see what the right way to do it is.
There is a number of ways to make a system that does _not_ allow the verification service an easy way to find out how a certain person has voted, while making sure that the votes are actually unique and valid.
I would say that Sequent merged with IBM, not vice versa. The larger company is, usually, listed second, I think.
Yes, it can be made anonymous and at the same time secure, within certain bounds of both.
Other than the aforementioned Applied Cryptography's chapter, there's a couple of resources on the Net to go to see more secure election protocols.
In my experience, I was able to set up a fairly anonymous (except to people with real-time database access or access to transaction logs (which shouldn't be turned on)) system that also guaranteed the relative impossibility of voting twice, or of the people setting up the votes the ability to figure out what you've voted for. Of course, the validity of the votes was also guaranteed, and it was possible for the voters to double-check to see if their vote was counted.
This of course was for a very small set of people (about 30 people in a Computer Science department). But even there it is significantly more secure than the old paper-based system they were using.
Yep, there we go. Geek is what I use mosly -- I even sometimes wear the CopyLeft 'geek' hat :-]
-Ilya Haykinson
Actually, I see this as the thing that will make me not want to use hushmail at all for any secure transactions. I can trust secure point-to-point communications, I can trust a good remailer chain. But to trust a small company, that's sketch. If this was being done by Microsoft or IBM or Sun I would trust it -- these are big guys who have enough people to make sure that stuff is secure, have enough expertise to do it right the first time (and not, like someone noticed, use the Date/Time as the seed for the random num generator), and someone who is going to watch the system very closely to make sure that they themselves never have the ability to read the messages or else they'd be involved in a major PR nightmare. But hushmail has nothing to be afraid about, and they are not big enough to guarantee that they'd even tell people if they start getting investigated by the feds in whichever country they're located in.