If it's just about taxes, should everyone below the income tax threshold be treated as a second-class citizen? Maybe people's votes should be weighted in proportion to their tax contribution? I'm sure that would lead to a utopian society of hard-working taxpayers and wouldn't quickly devolve into a corrupt oligarchy.
The collection and analysis of personal data shifts power from the individual to the corporation. The not-for-profit distribution of copyrighted data shifts power from the corporation to the individual. So it's not hypocritical to support both privacy rights and fair use rights.
It used to be that if you wanted a luxury you would work hard for it. Now the entitlement generation just bases their ethical code on the ever popular rallying cry of the two year old: "gimme! mine!"
People have been making that complaint since Plato ("they fill their bellies like the beasts"). I don't think it's anything new that people want to have luxuries without working for them.
I'm still trying to get my head round relational database concepts - if he disappeared while scattering his mother's ashes, would this be an example of a one-tomb-many-relation?
For your information they all look alike to me because I'm extremely short-sighted rather than because their similarities in age, income, skin pigmentation and gender somehow detract from their uniqueness as human beings. Thanks for drawing attention to my disability while implying that I'm a racist bigot, you self-confessed insensitive clod!
The ballots can be unique and anonymous. Each ballot is printed on two sheets of carbon paper, with a unique serial number that appears on both sheets. The candidates' names are printed on the top sheet only, in a random order, with a box beside each name. The voter marks a box, and the mark appears on both sheets. The top sheet is dropped in the ballot box and the voter keeps the bottom sheet.
The bottom sheet shows which box was marked, but not which name was written beside it. If a goon comes round to the voter's house and demands to see the bottom sheet, the goon can't find out who the candidate voted for without stealing all the ballot boxes and manually sifting through them for the matching top sheet.
Once the votes have been counted, 1% of the ballots are chosen at random for verification. (This can be done in public, using a lottery machine to select the last two digits, eg all ballots ending in 67.) For each of the selected ballots, the unique serial number and the position of the mark (but not the candidate's name) are published in the local newspaper. Any voter can easily check whether his or her serial number was published, and if so, whether the published mark corresponds to the mark on the bottom sheet. If there's a problem, the voter can prove it without revealing how he or she voted by showing the newspaper and the bottom sheet to anyone gives a shit about democracy, assuming such a person can be found.
You can't do a series of experiments on the same animal, because the effects of previous experiments could interfere with the current experiment. Animals are usually "sacrificed" after one experiment because they're no longer useful from a statistical point of view, regardless of whether it's physically possible to perform further experiments on them.
Using the network is a good way to monitor it: "The ability to route over the anonymous communication network, that anyone has, can be used to estimate the traffic load on specific Tor nodes accurately enough to perform traffic-analysis."
Re:How Many Nodes Do You Need to Own?
on
Tor Open To Attack
·
· Score: 2, Interesting
TOR has never claimed to provide strong anonymity, you need something like Herbivore for that.
Herbivore isn't vulnerable to traffic analysis but it's vulnerable to DoS: the attacker's nodes follow the secure entry protocol and get assigned to random cliques. Then they transmit in every round, jamming communication within their cliques. Jamming doesn't require any more bandwidth than normal participation in the protocol, and the source of the jamming can't be detected because communication within a clique is completely anonymous. With cliques of 128 nodes, an attacker who controls 1% of the nodes can jam 72% of the cliques at any given time. If the innocent nodes move to different cliques to escape the jamming, the attackers can move too.
Re:How Many Nodes Do You Need to Own?
on
Tor Open To Attack
·
· Score: 2, Informative
That's fine for small networks, but for a network with hundreds or thousands of nodes, controlling 5 to 10 percent may become infeasible.
Tor scales to a few hundred nodes, but it doesn't scale indefinitely - all the routers are listed in a central directory to ensure that all clients use the same set of routers and the same set of public keys.
It's something you can't do unless you have not just a pack hierarchy/dominance structure, but also a language/communications system capable of abstracting out concepts like "position" and "time".
Not necessarily true - the anvil group can be composed of risk-seeking individuals who are willing to fight an equally-matched enemy, while the hammer group is composed of risk-averse individuals who aren't willing to join the fight unless they have a good chance of winning. The anvil group attacks first and in most cases gets bogged down; the hammer group attacks later if it's clear that joining the fight will tip the balance. No need for communication or hierarchy (although I'm not denying that chimps have both).
If both sides have hammer groups waiting in the background, the situation is similar to an iterated game of chicken: each hammer group wants to delay joining the fight for as long as possible, allowing the anvil group to wear down the enemy, but neither hammer group wants the other to join the fight first, which would lead to a defeat. (Joining the fight corresponds to driving straight in the chicken game, and holding back corresponds to swerving; if both players swerve, the game is repeated.)
If a parrot can genuinely do what the article claims, then why not chimps?
One of the things that makes this kind of research very tricky is another amazing ability of Homo sapiens: the ability to read meaning into just about anything. Horst Hendriks-Jansen argues that this ability is crucial for bringing up children: we interact with them as if they understand far more than they actually do, which gives them the cues they need to start understanding what we thought they already understood.
So there's a strong selection pressure in favour of anthropomorphism, and historically there's been a weak selection pressure against it: there have been few situations where attributing human-like consciousness to natural phenonema would lead to faulty predictions (this has changed recently because our models of natural phenomena have become much more sophisticated, bringing them into conflict with institutionalised anthropomorphism in the form of religion).
It seems arrogant and almost pre-Darwinian to dismiss comparisons between humans and other animals on the grounds of anthropomorphism, but if we're going to ask questions such as "are parrots intelligent?" or "do chimps have culture?" we first need working definitions of intelligence and culture that aren't defined in strictly human terms. It's hard to imagine what those definitions would be, but until we have them the questions remain ill-posed.
The 439,000 wiretap requests resulted in 2,243 warrants - I don't know whether multiple requests can be granted in a single warrant. For human surveillance, which is covered by the second report, 2,177 authorisations were granted under the Police Act, of which roughly half involved drug offences, and 418 authorisations were granted under the Regulation of Investigatory Powers Act.
But for me, the most interesting part of the Chief Surveillance Commissioner's report was his opinion about automatic number plate recognition cameras:
...it is unlikely that the deployment could be authorised under RIPA or RIP(S)A. There may well be human rights issues arising in connection with any use of private information to build up pictures of the movements of particular persons or vehicles... The unanimous view of the Commissioners is that the existing legislation is not apt to deal with the fundamental problems to which the deployment of ANPR cameras gives rise.
After a bit of digging it appears there were 1,895,002 prosecutions and 1,484,424 convictions in 2005 (warning: large XLS file), of which roughly one fifth were for serious (indictable) offences. I'd be interested to know how many convictions involved wiretap evidence.
The figure seems particularly large when you consider that around 5,000,000 crimes were reported in England and Wales during the same period. Does one in twelve crimes require a wiretap? Or is it possible that at least some of the surveillance is politically motivated?
Skype contains encrypted code, self-modifying code, timing loops to detect whether it's running inside a debugger, and any number of other tricks to prevent reverse engineering. Which hasn't stopped people trying:
It's possible to classify 80-90% of traffic without looking at the payload or the port numbers, just by considering who connects to whom, for how long, with what distribution of packet sizes and inter-packet intervals. See this paper for details.
Mind you, BitTorrent makes up such a huge fraction of traffic these days that you could probably get 80-90% accuracy by just classifying everything as BitTorrent.;-)
Slashdot Mirror
Seriously though, would optical circuits be vulnerable to this kind of eavesdropping?
If it's just about taxes, should everyone below the income tax threshold be treated as a second-class citizen? Maybe people's votes should be weighted in proportion to their tax contribution? I'm sure that would lead to a utopian society of hard-working taxpayers and wouldn't quickly devolve into a corrupt oligarchy.
People have been making that complaint since Plato ("they fill their bellies like the beasts"). I don't think it's anything new that people want to have luxuries without working for them.
I'm still trying to get my head round relational database concepts - if he disappeared while scattering his mother's ashes, would this be an example of a one-tomb-many-relation?
Thanks for telling me not to worry, you sensitive clod!
For your information they all look alike to me because I'm extremely short-sighted rather than because their similarities in age, income, skin pigmentation and gender somehow detract from their uniqueness as human beings. Thanks for drawing attention to my disability while implying that I'm a racist bigot, you self-confessed insensitive clod!
The ballots can be unique and anonymous. Each ballot is printed on two sheets of carbon paper, with a unique serial number that appears on both sheets. The candidates' names are printed on the top sheet only, in a random order, with a box beside each name. The voter marks a box, and the mark appears on both sheets. The top sheet is dropped in the ballot box and the voter keeps the bottom sheet. The bottom sheet shows which box was marked, but not which name was written beside it. If a goon comes round to the voter's house and demands to see the bottom sheet, the goon can't find out who the candidate voted for without stealing all the ballot boxes and manually sifting through them for the matching top sheet. Once the votes have been counted, 1% of the ballots are chosen at random for verification. (This can be done in public, using a lottery machine to select the last two digits, eg all ballots ending in 67.) For each of the selected ballots, the unique serial number and the position of the mark (but not the candidate's name) are published in the local newspaper. Any voter can easily check whether his or her serial number was published, and if so, whether the published mark corresponds to the mark on the bottom sheet. If there's a problem, the voter can prove it without revealing how he or she voted by showing the newspaper and the bottom sheet to anyone gives a shit about democracy, assuming such a person can be found.
All middle-aged middle-class white men look alike to me, you insensitive clod!
You can't do a series of experiments on the same animal, because the effects of previous experiments could interfere with the current experiment. Animals are usually "sacrificed" after one experiment because they're no longer useful from a statistical point of view, regardless of whether it's physically possible to perform further experiments on them.
It might be possible to use retroviruses to insert the light-sensitive genes into the patients' cells.
Backus is also the B in BNF. Many will mourn his parsing.
Using the network is a good way to monitor it: "The ability to route over the anonymous communication network, that anyone has, can be used to estimate the traffic load on specific Tor nodes accurately enough to perform traffic-analysis."
Herbivore isn't vulnerable to traffic analysis but it's vulnerable to DoS: the attacker's nodes follow the secure entry protocol and get assigned to random cliques. Then they transmit in every round, jamming communication within their cliques. Jamming doesn't require any more bandwidth than normal participation in the protocol, and the source of the jamming can't be detected because communication within a clique is completely anonymous. With cliques of 128 nodes, an attacker who controls 1% of the nodes can jam 72% of the cliques at any given time. If the innocent nodes move to different cliques to escape the jamming, the attackers can move too.
Not necessarily true - the anvil group can be composed of risk-seeking individuals who are willing to fight an equally-matched enemy, while the hammer group is composed of risk-averse individuals who aren't willing to join the fight unless they have a good chance of winning. The anvil group attacks first and in most cases gets bogged down; the hammer group attacks later if it's clear that joining the fight will tip the balance. No need for communication or hierarchy (although I'm not denying that chimps have both).
If both sides have hammer groups waiting in the background, the situation is similar to an iterated game of chicken: each hammer group wants to delay joining the fight for as long as possible, allowing the anvil group to wear down the enemy, but neither hammer group wants the other to join the fight first, which would lead to a defeat. (Joining the fight corresponds to driving straight in the chicken game, and holding back corresponds to swerving; if both players swerve, the game is repeated.)
One of the things that makes this kind of research very tricky is another amazing ability of Homo sapiens: the ability to read meaning into just about anything. Horst Hendriks-Jansen argues that this ability is crucial for bringing up children: we interact with them as if they understand far more than they actually do, which gives them the cues they need to start understanding what we thought they already understood.
So there's a strong selection pressure in favour of anthropomorphism, and historically there's been a weak selection pressure against it: there have been few situations where attributing human-like consciousness to natural phenonema would lead to faulty predictions (this has changed recently because our models of natural phenomena have become much more sophisticated, bringing them into conflict with institutionalised anthropomorphism in the form of religion).
It seems arrogant and almost pre-Darwinian to dismiss comparisons between humans and other animals on the grounds of anthropomorphism, but if we're going to ask questions such as "are parrots intelligent?" or "do chimps have culture?" we first need working definitions of intelligence and culture that aren't defined in strictly human terms. It's hard to imagine what those definitions would be, but until we have them the questions remain ill-posed.
Why would the US government make 439,000 requests to tap people's phones when it has a perfectly good warrantless wiretap program?
Report of the Interception of Communications Commissioneer for 2005-6
Report of the Chief Surveillance Commissioner for 2005-6
The 439,000 wiretap requests resulted in 2,243 warrants - I don't know whether multiple requests can be granted in a single warrant. For human surveillance, which is covered by the second report, 2,177 authorisations were granted under the Police Act, of which roughly half involved drug offences, and 418 authorisations were granted under the Regulation of Investigatory Powers Act.
But for me, the most interesting part of the Chief Surveillance Commissioner's report was his opinion about automatic number plate recognition cameras:
After a bit of digging it appears there were 1,895,002 prosecutions and 1,484,424 convictions in 2005 (warning: large XLS file), of which roughly one fifth were for serious (indictable) offences. I'd be interested to know how many convictions involved wiretap evidence.
The figure seems particularly large when you consider that around 5,000,000 crimes were reported in England and Wales during the same period. Does one in twelve crimes require a wiretap? Or is it possible that at least some of the surveillance is politically motivated?
"Never underestimate the bandwidth of a covered wagon full of tapes..."
That's why I always write ransom notes by hand, using my own blood.
http://www.recon.cx/en/f/vskype-part1.pdf
http://www.recon.cx/en/f/vskype-part2.pdf
Mind you, BitTorrent makes up such a huge fraction of traffic these days that you could probably get 80-90% accuracy by just classifying everything as BitTorrent. ;-)
Democracy isn't the right to choose your rulers, it's the right to fire them.