First he suggests Apple is tracking its users when it just neglected to encrypt location cache data on the devices and their backups (fixed about 2 years ago). And then he suggests you cannot delete cookies on mobile Safari when it actually has all the usual options (it has private mode also). Not nice at all. And quite irrelevant to his subject.
Even if the idea of a smartwatch turns out to be not very viable in the end, generating all kinds of ideas for it should be fun. Instead 'smart-ass' useless comments dominate here.
First, it could be beautiful, and adaptable to your own taste. There is the silent alert function, you feel something and look at your watch. Fast and natural. Then, there is the sensor platform, for sports and exercising, like the Nike FuelBand, but all kinds of sensors can be imagined. With Bluetooth, it might even download/upload data (even time/location/internet) from/to a nearby device such as a phone or camera, for direct display or as an input. It could be an app platform. Anyone could come up with useful or fun applications. It could act as a key/passcode or remote or wallet or NFC device It could trigger a SOS with the right touch-gesture.
There are endless possibilities, and only one needs to be a killer-feature to make it another must-have breakthrough device. Apple could throw a few hundred million at it, and it would be a good gamble. If only to develop the creative potential in the company. Probably, it already has done so. It's fun to work there, I guess.
You greatly exaggerate the lock-in effect. Why can people switch from a Nokia or other brand to an iPhone but not the reverse? It makes no sense. And what you call 'psychological lock-in' is just plain old brand recognition.
The app-store *does* have a lock-in effect, but Apple just has done a great job of creating this new market. It is hard to hold that against them, unless you hate to see others succeed. And with these low prices, few will have really spent so much that lock-in is a serious factor. You can buy apps for other phones too.
User satisfaction is the only important thing to look at. iPhone 4 users are evidently not prepared to return their phone, even with the flaw (plus all the publicity) and even though the full price is reimbursed. Only a third as many returns as the 3GS.
Still, Apple will improve the quality of the 'phone' part of their iPhone, and maybe also give their whole iOS/Mac OS network stack a good working-over.
The result will be an even better phone, which is good for me. So I'm glad of all this negative publicity, even if it is overblown. Keeps Apple on its toes.
There is quite a large gap between cost-sharing and for-profit operations. If you calculate the real cost of driving (car, fuel, maintenance, insurance, taxes) you get to at least CAD 0,30 per kilometer I would say (likely more). Splitting that cost with two people would get you $81.
Commercially, you could not get anybody to drive 541 km (and back?) for anything near that price.
The role of government is to serve the public and not restrict their freedoms arbitrarily to serve commercial interests.
Car sharing is not for profit. It would be wonderful to have more than 1.2 people per car on average! Good for everyone, except the redundant bus company. It would be easy to limit the amount of money involved so that unregulated commercial operations would not be viable.
Sm>> "Let's say there's 150,000 voters in a district. Each one needs to pick a unique, 6 digit number.... So how many tries do you think it would take for the last person to find one? If you think it's less than 5 you need to consider the IQ of the average voter... 111111, 123456, etc... what this "solution" does it make it REALLY inconvenient for the voter to verify their votes."
No! You *still* misunderstand! The number only has to be unique at one polling-station and over a time of one hour or so. That's less than 100 generally, I would think.
Sm>> "And since you know which votes have been verified (you'd have to, or else this would be meaningless... you can't claim confidence if you don't know how many votes were verified!) then you can tamper with the votes that HAVEN'T been verified, which will very likely be the majority of them."
Again, you misunderstand the way it is supposed to work. It is about Sherlock Holmes' "dog that didn't bark". No complaints means no fraud.
[...]
Sm>> "So you're saying that it would be easy enough for Joe Q Voter to select a random code that nobody else uses, but it wouldn't be easy for Jack D Fraud to select one and give it to Joe?"
Which numbers are available depends on the other voters, and is only known at the time and place of voting. Anyway, I think that the massive voter suppression in the US (due to legalities) is a far more significant problem. And like I said, if really needed there is a solution (dummy codes).
Sm>> "And even this doesn't help, because Joe could be coerced into giving up the code he DID use anyway. That's the problem: there is NO WAY to be certain that only the voter can see the vote and know it's his. Any system that allows a vote to be publicly viewed and still be correlated to an individual (via pass code, in this case) is already a thousand times more prone to corruption than the current system."
Now really, a person is generally able to keep a four digit code to himself, just like any bank assumes. Otherwise you must assume everybody is robbed blind routinely by people 'pressuring' him or her. Coercion is as difficult as robbery. And robbery is a lot more profitable. So I do not think it is a big problem (having read reports about it lately? It is perfectly possible in the current system.).
Sm>> "So what happens when the website is tampered with? You fall back on the paper ballots. What happens when the ballots are tampered with? For fall back on the website. How do you know which is correct?"
Simply, any tampering on the website will show up as complaints. Many people will check their vote, also because it is easy to pin a hardcopy version to the wall somewhere. Complaints will lead to a recount. A recount will lead to inspection of the paper bollots. If many people claim that their paper ballots are not the original ones, the fraud will be untenable.
Sm>> "You said only a small percentage of the votes need to be verified, that leaves a lot of room for screwups with no way of knowing which is right... maybe both of them are wrong!"
Well, if only 10% will check (and I am convinced that, especially in a state such as Florida, it would have been 90% rather than 10%), and very few complaints come forward, statistically this gives a high level of confidence in the result. Remember, the voter can at any time confirm with 100% certainty that his/her vote is still counted correctly.
Sm>> "And your idea does not secure this any more than the current system. Also, the "final tally" is not fool proof because you have no way of being 100% sure that the tallies were not altered before you counted them. It's only "fool proof" in the sense that you can't change them afterwards to any effect, which is true today anyway."
Sadly, it seems irrelevant what my proposal entails, as long as you're prepared or even determined to see fault where there is no fault, instead of trying to improve upon what still
Indeed, a good idea. But for important elections great care is needed to avoid linking voter and vote (maintain anonimity) and a long identifier cannot be remembered, so it must be in writing and is therefore open to inspection from others.
Thank you, Smidge, for your comments. I put my replys in between.
Sm>> "This still fails on several levels."
Sm>> "1) It assumes that everyone at least in the same voting district will select unique codes, unless you correlate each person's code to some other type of data to uniquely identify them which pretty much does away with the anonimity."
If you choose a code that is not available (anymore), you are asked to choose another code.
Sm>> "2) It assumes that the majority of people will bother to check it when they get home or wherever they may have internet access (if any at all.. yeah, those places still exist!) The chance of someone actually bothering to go to a library or somesuch to check their vote is exceedingly small."
Every (random) voter that checks his/her vote succesfully confirms the result, you need only a small percentage to reach a high level of confidence.
SM>> "3) It assumes that everyone who participates will actually remember their passcode correctly. If they type an invalid key, they will either get an error or someone else's vote (see #1). If they get an error can they be sure their vote was cast correctly? Since there is no way to link a passcode/vote result to an individual, password recovery/reset is impossible."
All results are displayed on the web pages for all to see. People navigate to their district to check their vote. Concerning the people that forget their code, see the reply to 2), not many voters are needed that actually check their vote.
SM>> "4) It does not eliminate the possibiliy of coercion or fraud. I can threaten/pay you to vote a certain way using a certain passcode, and then I can verify it."
That specific passcode may turn out not to be available. Coercion is always possible as long as voting by mail or by proxy is possible. If it really were a problem, dummy codes could be introduced at request and subtracted from the totals.
Sm>> "5) It assumes that the information online will be faithful to real world data. In short, you are creating another avenue for mistakes and/or fraud. The claims that false reports won't acheive anything is absurd: They would force a recount. Do you think it's easier to tamper with the original election or the paper records and try to invoke a recount?"
Not so. The information on the internet has to match the real-world voting, otherwise there will be many complaints. Tampering with the paper ballots is always possible, but the slips have to correspond to the internet (including the codes), and people are allowed to place an unobtrusive mark of any kind on their paper ballot (and check in case of a recount) if they suspect it might be replaced by a fraudulent copy.
Sm>> "In short, why not just have the vote checked by the citizen right then and there, before they leave the booth, and maintain a paper trail like everyone agrees we should. You are just adding another layer of complexity to the system and not really solving anything."
Checking something that is not clearly the final tally always leaves the possibility of fraud further along the track. The possibilities for fraud are endless (as history has shown and continues to show). Checking the final tally is fool-proof, and is able to give complete confidence among everybody involved.
Now that I think of it, a paper trail would simplify matters condiderably. A recount would be triggered if enough people complain about their vote being misrepresented, according to the list on the internet (and their secret code).
Complaining would neither require disclosing your vote(s) or code, nor would it require any corroboration, because false complaints cannot achieve anything (besides lost time and energy).
The list on the internet would show location and *approximate* time of the vote + the 4 (or even 3) digit code and of course the choice made. With (sub) totals.
It would be impossible to trace the vote so displayed to the person that voted. Only the voter can find his/her vote between all the other votes in the same general time-frame (and location).
So, the total number of votes is known, and every voter can verify his/her vote, thus the result MUST be correct Q.E.D.
Verification of complaints does not require that the relation between voter and vote is stored. The voter could mention the real time of the vote as displayed by the voting machine, or maybe an additional digit of the code. That would make te complaint credible.
- everyone who votes must see his/her vote back on paper and on a public website, plainly readable (on the web, anyway). - it must be seen to count towards the preferred candidate (correct tabulation, obviously). - no one else must be able to identify the connection between te voter and the vote.
Solution: display a 4-digit unique number (either generated or chosen) to the voter when s/he has voted, and together with place and (slightly randomized) time of vote this is published on the internet.
Complain when something is foul, and if there are too many (partially) validated complaints: take corrective action (maybe re-vote).
While internet voting is a very difficult subject, electronic voting is eminently possible now.
Many of the points Dr. Mercury raises are far-fetched and ignore the corresponding possibilities for election-fraud with paper ballots, as is amply demonstrated by the many unfair elections in developing countries.
With a few dependable officials (e.g. from the UN) in key-positions, and use of standardized software, electronic voting (and even internet voting) can be made to be much more tamper-resistant than existing methods (with the same number of independent officials).
Of course, Mercury's research is still useful, but her negative attitude towards the subject really prevents her from finding any solutions.
Others with a more positive mindset will make progress, and create a comprehensive and low-cost software-infrastructure for fair UN-certified elections in fledgeling democracies.
Buggy browser-included 1.1.x JVM's were a disaster for Java on the client-side.
Microsoft is doing us a favor by not including its own version of Java with its OS anymore.
Yes, there is a problem with preserving Intellectual Property once you let it onto the internet where copying is so easy.
But there is no justification for Bronfman's silly rant about buffalo's, Nazi Germany or the USSR, nor for his shortsighted greedy crackdown on anonymity, which is essential to protect us all from a future police-state government.
But there should be a technical solution to a large part of the problem. Just distribute copyrighted material only in a strictly identifiable digital form, every copy unique. With the sale/resale, transfer this unique 'ID' to the buyer. If you use/publish a pirated copy, with someone else owning it's 'ID', it will be very clear and can't be denied. The credit to this 'ID' should be on a bank account (this is new of course). The bank protects it's customers identity, and it protects the unique-ness of the 'ID' for the IP-owner (just like banks cannot make money just by 'duplication' of the accounts balance). Oops -- should I keep this idea of mine (hmm, surely, I'm not the first) a secret...? Nay, let's all profit from it.
Will some beneficent company do the same for streaming media?
Don't forget IBM, with it's Hotmedia player, they're pro Unix, pro Open standards, anti M$, seem competent enough to me...
Hotmedia is no-charge now (though not yet open and not available for Linux), and if it is (will be) good enough, it just may overcome the proprietary players.
It would be fun always to be able to one-click rate what you read. More rating-data allows for better and more dependable filtering. Karma is a very nice way to give users feedback about what they are doing.
Just give EVERYBODY (even AC's) two drop-down boxes:
Rate (interesting, informative, funny, original,...., troll, no content,....) Agree (very much, yes, no, not at all)
And maybe, just for fun, a drop-down box with a few fast statistics about the post.
The rest is some (very) smart programming (and maybe a heavy server).
Goals:
Privacy (AC, option to delete email for password, transient logs)
Flexible filtering (build up a collection of scripts to select from) Troll protect (try to discriminate valuable moderation from vandalism)
Troll protect criteria:
The many 'responsible' contributions will overcome the baddies.
Limit the effect someone (esp. with negative karma) can have in a short time Don't allow controversial posts (many pro/con) to drop below 2
Hey, all you kick-ass programmers, give slashdot the best open source rating/karma system and some good filter-scripts, maybe a bias-vector/formula for selection (and one for ordering) of posts (f.e. x*informative + y*long, funny, controversial, new,...).
Of course, Rob must do a lot of work first to make this happen. But yes, I think it's very worthwile for Slashdot (and all the sites that will adopt the system!).
Slashdot Mirror
First he suggests Apple is tracking its users when it just neglected to encrypt location cache data on the devices and their backups (fixed about 2 years ago). And then he suggests you cannot delete cookies on mobile Safari when it actually has all the usual options (it has private mode also). Not nice at all. And quite irrelevant to his subject.
What a depressing site this has become.
Even if the idea of a smartwatch turns out to be not very viable in the end, generating all kinds of ideas for it should be fun. Instead 'smart-ass' useless comments dominate here.
First, it could be beautiful, and adaptable to your own taste.
There is the silent alert function, you feel something and look at your watch. Fast and natural.
Then, there is the sensor platform, for sports and exercising, like the Nike FuelBand, but all kinds of sensors can be imagined. With Bluetooth, it might even download/upload data (even time/location/internet) from/to a nearby device such as a phone or camera, for direct display or as an input.
It could be an app platform. Anyone could come up with useful or fun applications.
It could act as a key/passcode or remote or wallet or NFC device
It could trigger a SOS with the right touch-gesture.
There are endless possibilities, and only one needs to be a killer-feature to make it another must-have breakthrough device. Apple could throw a few hundred million at it, and it would be a good gamble. If only to develop the creative potential in the company. Probably, it already has done so. It's fun to work there, I guess.
Haha, locked in, and I don't even own an iPhone... You're pathetic.
Apple-bashers have a certain psychological process going, it seems. Interesting, it is food for thought.
I'm thinking of getting an iPhone though, that's why, as I said earlier, I like this pressure on Apple to keep improving their stuff.
You greatly exaggerate the lock-in effect. Why can people switch from a Nokia or other brand to an iPhone but not the reverse? It makes no sense. And what you call 'psychological lock-in' is just plain old brand recognition.
The app-store *does* have a lock-in effect, but Apple just has done a great job of creating this new market. It is hard to hold that against them, unless you hate to see others succeed. And with these low prices, few will have really spent so much that lock-in is a serious factor. You can buy apps for other phones too.
User satisfaction is the only important thing to look at. iPhone 4 users are evidently not prepared to return their phone, even with the flaw (plus all the publicity) and even though the full price is reimbursed. Only a third as many returns as the 3GS.
Still, Apple will improve the quality of the 'phone' part of their iPhone, and maybe also give their whole iOS/Mac OS network stack a good working-over.
The result will be an even better phone, which is good for me. So I'm glad of all this negative publicity, even if it is overblown. Keeps Apple on its toes.
There is quite a large gap between cost-sharing and for-profit operations. If you calculate the real cost of driving (car, fuel, maintenance, insurance, taxes) you get to at least CAD 0,30 per kilometer I would say (likely more). Splitting that cost with two people would get you $81.
Commercially, you could not get anybody to drive 541 km (and back?) for anything near that price.
The role of government is to serve the public and not restrict their freedoms arbitrarily to serve commercial interests.
Car sharing is not for profit. It would be wonderful to have more than 1.2 people per car on average! Good for everyone, except the redundant bus company.
It would be easy to limit the amount of money involved so that unregulated commercial operations would not be viable.
Sm>> "Let's say there's 150,000 voters in a district. Each one needs to pick a unique, 6 digit number .... So how many tries do you think it would take for the last person to find one? If you think it's less than 5 you need to consider the IQ of the average voter... 111111, 123456, etc... what this "solution" does it make it REALLY inconvenient for the voter to verify their votes."
No! You *still* misunderstand! The number only has to be unique at one polling-station and over a time of one hour or so. That's less than 100 generally, I would think.
Sm>> "And since you know which votes have been verified (you'd have to, or else this would be meaningless... you can't claim confidence if you don't know how many votes were verified!) then you can tamper with the votes that HAVEN'T been verified, which will very likely be the majority of them."
Again, you misunderstand the way it is supposed to work. It is about Sherlock Holmes' "dog that didn't bark". No complaints means no fraud.
[...]
Sm>> "So you're saying that it would be easy enough for Joe Q Voter to select a random code that nobody else uses, but it wouldn't be easy for Jack D Fraud to select one and give it to Joe?"
Which numbers are available depends on the other voters, and is only known at the time and place of voting. Anyway, I think that the massive voter suppression in the US (due to legalities) is a far more significant problem. And like I said, if really needed there is a solution (dummy codes).
Sm>> "And even this doesn't help, because Joe could be coerced into giving up the code he DID use anyway. That's the problem: there is NO WAY to be certain that only the voter can see the vote and know it's his. Any system that allows a vote to be publicly viewed and still be correlated to an individual (via pass code, in this case) is already a thousand times more prone to corruption than the current system."
Now really, a person is generally able to keep a four digit code to himself, just like any bank assumes. Otherwise you must assume everybody is robbed blind routinely by people 'pressuring' him or her. Coercion is as difficult as robbery. And robbery is a lot more profitable. So I do not think it is a big problem (having read reports about it lately? It is perfectly possible in the current system.).
Sm>> "So what happens when the website is tampered with? You fall back on the paper ballots. What happens when the ballots are tampered with? For fall back on the website. How do you know which is correct?"
Simply, any tampering on the website will show up as complaints. Many people will check their vote, also because it is easy to pin a hardcopy version to the wall somewhere. Complaints will lead to a recount. A recount will lead to inspection of the paper bollots. If many people claim that their paper ballots are not the original ones, the fraud will be untenable.
Sm>> "You said only a small percentage of the votes need to be verified, that leaves a lot of room for screwups with no way of knowing which is right... maybe both of them are wrong!"
Well, if only 10% will check (and I am convinced that, especially in a state such as Florida, it would have been 90% rather than 10%), and very few complaints come forward, statistically this gives a high level of confidence in the result. Remember, the voter can at any time confirm with 100% certainty that his/her vote is still counted correctly.
Sm>> "And your idea does not secure this any more than the current system. Also, the "final tally" is not fool proof because you have no way of being 100% sure that the tallies were not altered before you counted them. It's only "fool proof" in the sense that you can't change them afterwards to any effect, which is true today anyway."
Sadly, it seems irrelevant what my proposal entails, as long as you're prepared or even determined to see fault where there is no fault, instead of trying to improve upon what still
Indeed, a good idea. But for important elections great care is needed to avoid linking voter and vote (maintain anonimity) and a long identifier cannot be remembered, so it must be in writing and is therefore open to inspection from others.
Thank you, Smidge, for your comments. I put my replys in between.
Sm>> "This still fails on several levels."
Sm>> "1) It assumes that everyone at least in the same voting district will select unique codes, unless you correlate each person's code to some other type of data to uniquely identify them which pretty much does away with the anonimity."
If you choose a code that is not available (anymore), you are asked to choose another code.
Sm>> "2) It assumes that the majority of people will bother to check it when they get home or wherever they may have internet access (if any at all.. yeah, those places still exist!) The chance of someone actually bothering to go to a library or somesuch to check their vote is exceedingly small."
Every (random) voter that checks his/her vote succesfully confirms the result, you need only a small percentage to reach a high level of confidence.
SM>> "3) It assumes that everyone who participates will actually remember their passcode correctly. If they type an invalid key, they will either get an error or someone else's vote (see #1). If they get an error can they be sure their vote was cast correctly? Since there is no way to link a passcode/vote result to an individual, password recovery/reset is impossible."
All results are displayed on the web pages for all to see. People navigate to their district to check their vote. Concerning the people that forget their code, see the reply to 2), not many voters are needed that actually check their vote.
SM>> "4) It does not eliminate the possibiliy of coercion or fraud. I can threaten/pay you to vote a certain way using a certain passcode, and then I can verify it."
That specific passcode may turn out not to be available. Coercion is always possible as long as voting by mail or by proxy is possible. If it really were a problem, dummy codes could be introduced at request and subtracted from the totals.
Sm>> "5) It assumes that the information online will be faithful to real world data. In short, you are creating another avenue for mistakes and/or fraud. The claims that false reports won't acheive anything is absurd: They would force a recount. Do you think it's easier to tamper with the original election or the paper records and try to invoke a recount?"
Not so. The information on the internet has to match the real-world voting, otherwise there will be many complaints. Tampering with the paper ballots is always possible, but the slips have to correspond to the internet (including the codes), and people are allowed to place an unobtrusive mark of any kind on their paper ballot (and check in case of a recount) if they suspect it might be replaced by a fraudulent copy.
Sm>> "In short, why not just have the vote checked by the citizen right then and there, before they leave the booth, and maintain a paper trail like everyone agrees we should. You are just adding another layer of complexity to the system and not really solving anything."
Checking something that is not clearly the final tally always leaves the possibility of fraud further along the track. The possibilities for fraud are endless (as history has shown and continues to show). Checking the final tally is fool-proof, and is able to give complete confidence among everybody involved.
Now that I think of it, a paper trail would simplify matters condiderably. A recount would be triggered if enough people complain about their vote being misrepresented, according to the list on the internet (and their secret code).
Complaining would neither require disclosing your vote(s) or code, nor would it require any corroboration, because false complaints cannot achieve anything (besides lost time and energy).
The list on the internet would show location and *approximate* time of the vote + the 4 (or even 3) digit code and of course the choice made. With (sub) totals.
It would be impossible to trace the vote so displayed to the person that voted. Only the voter can find his/her vote between all the other votes in the same general time-frame (and location).
So, the total number of votes is known, and every voter can verify his/her vote, thus the result MUST be correct Q.E.D.
Verification of complaints does not require that the relation between voter and vote is stored. The voter could mention the real time of the vote as displayed by the voting machine, or maybe an additional digit of the code. That would make te complaint credible.
in such a way that the voter can check his/her vote but nobody else can.
This is easy. Just let the voter enter a unique 4-digit number to go with the vote to make it traceable (for only that voter) on the list.
Complaints can be verified with help of the voter (and the printed, partially encrypted log).
Too many complaints (unlikely with this system), and corrective action follows (maybe re-vote).
With voting you want to have:
- everyone who votes must see his/her vote back on paper and on a public website, plainly readable (on the web, anyway).
- it must be seen to count towards the preferred candidate (correct tabulation, obviously).
- no one else must be able to identify the connection between te voter and the vote.
Solution: display a 4-digit unique number (either generated or chosen) to the voter when s/he has voted, and together with place and (slightly randomized) time of vote this is published on the internet.
Complain when something is foul, and if there are too many (partially) validated complaints: take corrective action (maybe re-vote).
While internet voting is a very difficult subject, electronic voting is eminently possible now.
Many of the points Dr. Mercury raises are far-fetched and ignore the corresponding possibilities for election-fraud with paper ballots, as is amply demonstrated by the many unfair elections in developing countries.
With a few dependable officials (e.g. from the UN) in key-positions, and use of standardized software, electronic voting (and even internet voting) can be made to be much more tamper-resistant than existing methods (with the same number of independent officials).
Of course, Mercury's research is still useful, but her negative attitude towards the subject really prevents her from finding any solutions.
Others with a more positive mindset will make progress, and create a comprehensive and low-cost software-infrastructure for fair UN-certified elections in fledgeling democracies.
While Windows wont be able to optimally take advantage of this processor, Linux will.
If only the GCC-developement team could make better progress...
Buggy browser-included 1.1.x JVM's were a disaster for Java on the client-side.
Microsoft is doing us a favor by not including its own version of Java with its OS anymore.
Yes, there is a problem with preserving Intellectual Property once you let it onto the internet where copying is so easy.
But there is no justification for Bronfman's silly rant about buffalo's, Nazi Germany or the USSR, nor for his shortsighted greedy crackdown on anonymity, which is essential to protect us all from a future police-state government.
But there should be a technical solution to a large part of the problem. Just distribute copyrighted material only in a strictly identifiable digital form, every copy unique. With the sale/resale, transfer this unique 'ID' to the buyer. If you use/publish a pirated copy, with someone else owning it's 'ID', it will be very clear and can't be denied. The credit to this 'ID' should be on a bank account (this is new of course). The bank protects it's customers identity, and it protects the unique-ness of the 'ID' for the IP-owner (just like banks cannot make money just by 'duplication' of the accounts balance). Oops -- should I keep this idea of mine (hmm, surely, I'm not the first) a secret...? Nay, let's all profit from it.
Bart Meerdink
Hotmedia is no-charge now (though not yet open and not available for Linux), and if it is (will be) good enough, it just may overcome the proprietary players.
More rating-data allows for better and more dependable filtering.
Karma is a very nice way to give users feedback about what they are doing.
Just give EVERYBODY (even AC's) two drop-down boxes:
And maybe, just for fun, a drop-down box with a few fast statistics about the post.The rest is some (very) smart programming (and maybe a heavy server).
Goals:
Troll protect criteria: Hey, all you kick-ass programmers, give slashdot the best open source rating/karma system and some good filter-scripts, maybe a bias-vector/formula for selection (and one for ordering) of posts (f.e. x*informative + y*long, funny, controversial, new,Of course, Rob must do a lot of work first to make this happen. But yes, I think it's very worthwile for Slashdot (and all the sites that will adopt the system!).