But that's stupid. The data generated depends on thousands of tiny variables, and you have to admit that, even if you claim they are measurable, in contradiction to QM, and the math is doable, in contradiction to chaos theory.
These variables include external variables, like the air pressure in the room and the exact temperature of the device. You can't just 'calculate' them and somehow have future results, because they change.
Of course, at this point, I realize the people claiming there are no random numbers are just crackheads who somehow need to insist there is no such thing as 'random' for some philosophical reason, and don't actually have any logic to back them up.
Even in a completely predictable universe, there are still truely random numbers, because random is just the inverse of unpredictable, and prediction requires knowledge of the original state. I could drive to the moon, pick up a slice of moon rock, and generate random numbers based on the placement of the atoms, and then keep that chunk of rock hidden away forever. Good luck figuring out those numbers mathmatically.
A solution to the sun problem would be to built, say, three giant, one-shot transmitter, and hurl them out of the solar system at 120-degree angles from each other, as fast as we can. (By out, I mean north and south, so we know they're not going to hit anything.) They don't do anything for decade or so, until they get at least 1 light-year away from the earth, at which point they basically explode with a short burst message.
The nice thing about this plan is if anyone sees it, but misses the message, they'd be able to see the other messages a short time later. (We need three because it's possible that one would be directly in front or directly behind the sun, and thus no one could see them. Also one of them would probably fail.)
In fact, we might want to do four, or, hell, six. And we want to make them all slightly different, just to show it's not just some sort of weird echo. Possibly just include a diagram of the thing, and mark which probe is which in the message.
I'm not certain whether it would be best to do them all at once, or stagger them. If we stagger them, we could start with an 'about to transmit' message, that basically only explains that we are here, and there should be another message along shortly, making sure the first message is very odd and can't be mistaken for a natural one. And then, later, we can send bigger more complicated messages without also having to make them stand out, because we told the aliens what frequency to listen to and where to look.
We could also give the exact time we were going to broadcast, to make it easier, but that would be rather difficult to convey. We could try by telling them a time in divisions of the rotation of the galaxy, or in multiples of the length of time the first message took, though. I doubt we can fit complicated concepts like that in the first message though...we'll probably just have to put an index of pages at the start that's bigger than the amount of pages we send them, and hope they figure it out.
First of all, I seriously doubt any library does, in fact, want you to do that. Libraries want you to use their services. There's no reason for them to limit it. (Libraries especially hate spending money to limit things.)
However, there's an easy solution to the 'problem' here...turn off the goddamn AP when the liberary closes. Duh.
If they cared about people using them outside when the library is open, which I cannot imagine, a solution would be to change the WEP code every day, and only hand it out at the front desk, with the requirement that people only use it inside the library. Then get people outside the library for unauthorized access to a network/computer, which is, indeed, a crime.
However, I'd like to see that actually stated as a rule at any library, because it makes no sense. Why would they care where you use it from? And, no, libraries do not care about monitoring your connection. They just are sometiems required by law to pretend to care about it. (And, besides, if I carry my computer into the library, I can pull in as much of alt.binaries.erotica.lesbian.fishsticks as much as I want, in a tray icon, and no one would ever see. How me being in the library would stop that I have no idea.)
Public APs are public, period. I don't care if morons don't know they're running them. By the same measure, encrypted access points are private, period. I don't care if the password is 'password', you need permission to get on them.
A good example might be vending machines. If a store says 'closed', but has a vending machine outside that's turned on and lit up, it would be rather hard for that store to claim some sort of trespass or theft if you walked up and put in 75 cents and got a Coke.
If they didn't want you there, they shouldn't have had a machine advertising that it was selling something! If they didn't want you to take the Coke, they shouldn't have given it to you in return for money!
People in this thread are either idiots or know nothing about libraries. Libraries want you to use their services, for free.
If they could leave libraries unlocked 24 hours a day, they would. If they could do away with the overhead of checking out books, they would. But they can't, they'd loose too many books.
Libraries want you to use their stuff. If my library was providing an open-AP from outside the library, I'd certainly presume I could access it at any time, at least if I wasn't being abusive of their bandwidth by denying it to others. (Which I, by defination, couldn't be, if I was the only user.) And assuming they weren't being charged per-byte, but, if they were, they really shouldn't be letting it operate unsupervised.
And the porn argument is a red herring. The only reason libraries even vaguely care about porn (note most filtering is imposed by higher up government), is that patrons complain. They could give a flying fuck if you download porn in the parking lot, while they're closed, where people can't see you. They don't have some moral objection to porn.
Right, that's why I said you should transmit data constantly, a known pattern. (A known pattern before being encrypted, that is.)
This doesn't ensure that the line wasn't tapped before being purchased, but it does keep it from being so afterwards without people noticing.
There may be possibly a way to use photon entanglement to ensure the line is not tapped, now that I think of it. They may be able to retransmit the photon, but they couldn't re-entagle it.
Erm, actually, maybe they could just entangle another photon, and then read that one after you already read the other end. Ugh. There has to be a way to set it up securely, it's inane to have a link that's untappable without being noticed while in use, but no way to secure it beforehand.
Even if none of that was true (Which it is.), I'm always baffled by people who claim it matters.
Even if chaos theory and quantum mechanics didn't exist and we could predict future events based on measurements of the the initial systems...why are you letting strangers come in and install complicated measuring equipment inside your random noise generator? Even if it was possible to know the initial state enough to predict the random numbers, which it isn't, how the hell would attackers manage to do it?
Randomness isn't some mathmatical ideal, it's just the inverse of predictablity. It doesn't matter how the numbers come into existence, it just matters whether or not bad guys can predict them. It's completely irrelevent whether or not someone with microscopic vision magically located inside your random number generator could figure things out...because they aren't there.
And, yes, on top of that, there are things that literally are impossible to predict, like atomic decay.
It also allows you to convert two 'mostly secure' channels into one rather secure channel.
For example, let's say you're sending data into a hostile country that's trying to steal data from you, but they don't want to make it obvious. And you know your lines are tapped.
So you mail one CD, containing the OTP, to a trusted address, and you carry one CD containing the encrypted data in physically.
It's possibly they could intercept the mail in transit and copy that, and they could come up with some excuse and have to 'examine' the CD in customs, but they'd have to do both to get the stuff. If you stagger them, for example carry it in first, you can then choose to only mail the CD if they didn't have a chance to copy it during customs.
And I don't want to get into the math, but you can keep 'splitting' the data as much as you want. You can even make situtations where you need four of five CDs to get the data, or one of a set of two CDs and three of a set of seven. And this isn't some insecure password situtation, they are true OTPs, where you cannot mathmatically reconstruct the data until you get enough CDs.
NP!=P is not unprovable. The only options are the last three. Although proving that NP is not P is rather difficult compared to proving NP is P...it just takes one example for the latter, whereas the first requires a lot of math.
It's rather akin to claiming 'all people are under 50 feet tall'. It's trivial to prove that false if you can find a person over it. It's difficult, but possible, to map the human genome and prove that humans cannot possibly be that tall. (Assuming you can define 'human' correctly, but you get my point.) And it's currently not proven yet.
There are things that are provable unprovable, but NP!=P can't be one of them.
In fact, you don't even need to make the pad in advance, if you can write a quantum data stream that changes when intercepted. (Which has actually already been done.)
What you can do is just split the data in half. For each bit, come up with a random value, XOR it with the data, and transmit one of those now, and the other one later, when you're sure the first half got there. (For speed, you might want to do that in large blocks, 1024 bytes or whatever.)
At the other end, you need to, every X bytes, make a checksum and send it back the other way. (Of course, you need some way of making sure that the line hasn't been completely cut and the data retransmitted. Which you can do if you constantly keep the data line running, transmitting a known pattern.)
You can't go to jail for printing up a t-shirt with a patented formula on it...patents are public information. It may be illegal to build a device that play back a patented codec, but a t-shirt is not a computer, and cannot violate a patent, unless someone's patented 'Convaying infomation to people by means of text printed on clothing'.
That said, there's nothing like the DMCA for patents. You can build 90% of a patented device, hand it out expecting people to build the last 10%, and get away with it. As long as what you give out cannot do what is described in the patent, you're clear. (Erm, and don't give them instructions on how to build the rest. Let other people do that.)
I'm expecting some sort of 'modular' player that is missing the actual module to decode, but fully supports the format, and someone completely unrelated in Russia just happening to hand out a module that will plug right in. I'm also expecting mplayer-like hacks where they just grab the codec from Windows.
As another gag, almost all standards committies that allow patented standards require a 'fair' licensing arrangment. Aka, the company has to agree to sell a license to anyone for X dollars a player, and X dollars a disk, and whatever. That means the player software can tell you have to enable the patented things, and just has to explain you need to go purchase a patent license first...which, of course, no one will do.
People have tried to pull crap like that with copyrights, saying 'You have to get a license to play this illegally downloaded movie' to pretend to cover their back, but it wouldn't really work there, because copyright holders won't just sell random licenses without the content.
But it possibly could work with manditory licenses, where it's perfectly possible to walk up to microsoft, describe your Linux box, and they then be required, under their license with the standard organization, to give you a license for X amount of money. Having a explaination of how you needed to do this, and a checkbox saying 'I did this', migth be enough to fend off allegations of patent violation by the programmers.
Someone needs to figure out if these standards are required to be licensed to all comers, and then write some BSD code to impliment them. (You still can't do this under the GPL...if something a GPL program does is patented, it has to be your patent, and then you have to let people use it freely. Requiring them to get a patent from a third party to use the software is an illegal requirement under the GPL.)
If the application was filed by June 7, 1995 and issued before June 8, 1978, the expiration date is 17 years from issuance.
I think you could probably leave that first bit out. You can't have a patent that was issued before June 8, 1978, but filed after June 7, 1995, unless you just patented a time machine.
And, really, who cares about '17 years' anymore? Wouldn't it make more sense to say 'If the patent was issued before June 8, 1978, it's expired.'?
So, basically, in a way that's understandable: If it was filed after June 7, 1995, it's twenty years from that point. If it was filed before that point, it's seventeen years after it was issued or twenty years after it was filed, whichever is last.
Someone needs to be the fall guy here, like another post I read here said.
We need someone in a non-swing state to go in and punch in number_of_people_in_precinct * 2 votes for the guy who's going to win anyway, alert the media, and run like hell for Canada. And pick someone running locally in a unopposed slot, give them 1 vote instead of what they really got.
We need to have people tampering in obviously detectable ways, without changing the outcome of the election.
Actually, ATM misplace money all the time. But with an ATM, a customer with misplaced money can walk into a bank, and they pull up the log files, and the video camera, and see the ATM did, in fact, not give the person their money.
And you can't do that with voting. And banks have a billion backups of their database. And actually care about mistakes, or at least have to pretend to care.
Also shame on the people who decided to keep voting after they screwed up. Look, if you mark your ballot wrong, you take it back to the nice person at the counter, and they will happily give you another one. They'll keep giving your new ballots until you do one correctly.
You don't mark another box to try to fix it, or mark a box and write in a name.
Not that this in anyway okays what the Supreme Court decided to do, but honestly, people.
Quantum computers cannot copy quantum data, they can only move it, by particle entanglement, or, duh, just moving the quantum particles that contain it. Once you measure it, it stops being a qbit and is just a single measurement.
Once any quantum calculations have finished, of course, you can copy the final measurement as much as you want. But any measurement before that will cause the collapse of the state vector. Aka, drop you into one universe or another, or complete the transaction, or just collase the state vector, depending if you believe many worlds, transactional, or copenhagen interpetation, respectively.
I mean, the president, and even the Senate, can do very little about abortion, and in fact do very little about abortion. About the only change in abortion is going to be through a constitutional amendment.
It's something to wave around and scream and rant and rave about.
You twit, ignorance of commiting copyright infringement is an excuse. With ignorace, you're just liable for damages, without, you're liable for triple damages.
I read an interesting book recently, called 'What's the Matter with Kansas'. I came away from that book with the profound realization that the Republicans started, basically, 'outrage politics' in the early 90s,. (And, although the book doesn't go there, that the Democrats are following suit.)
Abortion is something to get people outraged about. It hasn't been a real political issue in 30 years. While people are wailing and gnashing their teeth over that, both sides are robbing us blind.
So, basically, during takeoff or landing, if a pilot wanted to crash the plane, he'd just have to wait till they were about 30 feet in the air, pull the stick to the right, and hold it there?
Slashdot Mirror
These variables include external variables, like the air pressure in the room and the exact temperature of the device. You can't just 'calculate' them and somehow have future results, because they change.
Of course, at this point, I realize the people claiming there are no random numbers are just crackheads who somehow need to insist there is no such thing as 'random' for some philosophical reason, and don't actually have any logic to back them up.
Even in a completely predictable universe, there are still truely random numbers, because random is just the inverse of unpredictable, and prediction requires knowledge of the original state. I could drive to the moon, pick up a slice of moon rock, and generate random numbers based on the placement of the atoms, and then keep that chunk of rock hidden away forever. Good luck figuring out those numbers mathmatically.
The nice thing about this plan is if anyone sees it, but misses the message, they'd be able to see the other messages a short time later. (We need three because it's possible that one would be directly in front or directly behind the sun, and thus no one could see them. Also one of them would probably fail.)
In fact, we might want to do four, or, hell, six. And we want to make them all slightly different, just to show it's not just some sort of weird echo. Possibly just include a diagram of the thing, and mark which probe is which in the message.
I'm not certain whether it would be best to do them all at once, or stagger them. If we stagger them, we could start with an 'about to transmit' message, that basically only explains that we are here, and there should be another message along shortly, making sure the first message is very odd and can't be mistaken for a natural one. And then, later, we can send bigger more complicated messages without also having to make them stand out, because we told the aliens what frequency to listen to and where to look.
We could also give the exact time we were going to broadcast, to make it easier, but that would be rather difficult to convey. We could try by telling them a time in divisions of the rotation of the galaxy, or in multiples of the length of time the first message took, though. I doubt we can fit complicated concepts like that in the first message though...we'll probably just have to put an index of pages at the start that's bigger than the amount of pages we send them, and hope they figure it out.
However, there's an easy solution to the 'problem' here...turn off the goddamn AP when the liberary closes. Duh.
If they cared about people using them outside when the library is open, which I cannot imagine, a solution would be to change the WEP code every day, and only hand it out at the front desk, with the requirement that people only use it inside the library. Then get people outside the library for unauthorized access to a network/computer, which is, indeed, a crime.
However, I'd like to see that actually stated as a rule at any library, because it makes no sense. Why would they care where you use it from? And, no, libraries do not care about monitoring your connection. They just are sometiems required by law to pretend to care about it. (And, besides, if I carry my computer into the library, I can pull in as much of alt.binaries.erotica.lesbian.fishsticks as much as I want, in a tray icon, and no one would ever see. How me being in the library would stop that I have no idea.)
Public APs are public, period. I don't care if morons don't know they're running them. By the same measure, encrypted access points are private, period. I don't care if the password is 'password', you need permission to get on them.
If they didn't want you there, they shouldn't have had a machine advertising that it was selling something! If they didn't want you to take the Coke, they shouldn't have given it to you in return for money!
People in this thread are either idiots or know nothing about libraries. Libraries want you to use their services, for free.
If they could leave libraries unlocked 24 hours a day, they would. If they could do away with the overhead of checking out books, they would. But they can't, they'd loose too many books.
Libraries want you to use their stuff. If my library was providing an open-AP from outside the library, I'd certainly presume I could access it at any time, at least if I wasn't being abusive of their bandwidth by denying it to others. (Which I, by defination, couldn't be, if I was the only user.) And assuming they weren't being charged per-byte, but, if they were, they really shouldn't be letting it operate unsupervised.
And the porn argument is a red herring. The only reason libraries even vaguely care about porn (note most filtering is imposed by higher up government), is that patrons complain. They could give a flying fuck if you download porn in the parking lot, while they're closed, where people can't see you. They don't have some moral objection to porn.
This doesn't ensure that the line wasn't tapped before being purchased, but it does keep it from being so afterwards without people noticing.
There may be possibly a way to use photon entanglement to ensure the line is not tapped, now that I think of it. They may be able to retransmit the photon, but they couldn't re-entagle it.
Erm, actually, maybe they could just entangle another photon, and then read that one after you already read the other end. Ugh. There has to be a way to set it up securely, it's inane to have a link that's untappable without being noticed while in use, but no way to secure it beforehand.
Even if chaos theory and quantum mechanics didn't exist and we could predict future events based on measurements of the the initial systems...why are you letting strangers come in and install complicated measuring equipment inside your random noise generator? Even if it was possible to know the initial state enough to predict the random numbers, which it isn't, how the hell would attackers manage to do it?
Randomness isn't some mathmatical ideal, it's just the inverse of predictablity. It doesn't matter how the numbers come into existence, it just matters whether or not bad guys can predict them. It's completely irrelevent whether or not someone with microscopic vision magically located inside your random number generator could figure things out...because they aren't there.
And, yes, on top of that, there are things that literally are impossible to predict, like atomic decay.
For example, let's say you're sending data into a hostile country that's trying to steal data from you, but they don't want to make it obvious. And you know your lines are tapped.
So you mail one CD, containing the OTP, to a trusted address, and you carry one CD containing the encrypted data in physically.
It's possibly they could intercept the mail in transit and copy that, and they could come up with some excuse and have to 'examine' the CD in customs, but they'd have to do both to get the stuff. If you stagger them, for example carry it in first, you can then choose to only mail the CD if they didn't have a chance to copy it during customs.
And I don't want to get into the math, but you can keep 'splitting' the data as much as you want. You can even make situtations where you need four of five CDs to get the data, or one of a set of two CDs and three of a set of seven. And this isn't some insecure password situtation, they are true OTPs, where you cannot mathmatically reconstruct the data until you get enough CDs.
It's rather akin to claiming 'all people are under 50 feet tall'. It's trivial to prove that false if you can find a person over it. It's difficult, but possible, to map the human genome and prove that humans cannot possibly be that tall. (Assuming you can define 'human' correctly, but you get my point.) And it's currently not proven yet.
There are things that are provable unprovable, but NP!=P can't be one of them.
What you can do is just split the data in half. For each bit, come up with a random value, XOR it with the data, and transmit one of those now, and the other one later, when you're sure the first half got there. (For speed, you might want to do that in large blocks, 1024 bytes or whatever.)
At the other end, you need to, every X bytes, make a checksum and send it back the other way. (Of course, you need some way of making sure that the line hasn't been completely cut and the data retransmitted. Which you can do if you constantly keep the data line running, transmitting a known pattern.)
That said, there's nothing like the DMCA for patents. You can build 90% of a patented device, hand it out expecting people to build the last 10%, and get away with it. As long as what you give out cannot do what is described in the patent, you're clear. (Erm, and don't give them instructions on how to build the rest. Let other people do that.)
I'm expecting some sort of 'modular' player that is missing the actual module to decode, but fully supports the format, and someone completely unrelated in Russia just happening to hand out a module that will plug right in. I'm also expecting mplayer-like hacks where they just grab the codec from Windows.
As another gag, almost all standards committies that allow patented standards require a 'fair' licensing arrangment. Aka, the company has to agree to sell a license to anyone for X dollars a player, and X dollars a disk, and whatever. That means the player software can tell you have to enable the patented things, and just has to explain you need to go purchase a patent license first...which, of course, no one will do.
People have tried to pull crap like that with copyrights, saying 'You have to get a license to play this illegally downloaded movie' to pretend to cover their back, but it wouldn't really work there, because copyright holders won't just sell random licenses without the content.
But it possibly could work with manditory licenses, where it's perfectly possible to walk up to microsoft, describe your Linux box, and they then be required, under their license with the standard organization, to give you a license for X amount of money. Having a explaination of how you needed to do this, and a checkbox saying 'I did this', migth be enough to fend off allegations of patent violation by the programmers.
Someone needs to figure out if these standards are required to be licensed to all comers, and then write some BSD code to impliment them. (You still can't do this under the GPL...if something a GPL program does is patented, it has to be your patent, and then you have to let people use it freely. Requiring them to get a patent from a third party to use the software is an illegal requirement under the GPL.)
I think you could probably leave that first bit out. You can't have a patent that was issued before June 8, 1978, but filed after June 7, 1995, unless you just patented a time machine.
And, really, who cares about '17 years' anymore? Wouldn't it make more sense to say 'If the patent was issued before June 8, 1978, it's expired.'?
So, basically, in a way that's understandable: If it was filed after June 7, 1995, it's twenty years from that point. If it was filed before that point, it's seventeen years after it was issued or twenty years after it was filed, whichever is last.
This is like the security on MS Bob, where if you mistyped a password three times it would assume you'd forgotten it and happily reset it for you.
The 'security feature' is completely insecure. Security sometimes means inconvinence.
We need someone in a non-swing state to go in and punch in number_of_people_in_precinct * 2 votes for the guy who's going to win anyway, alert the media, and run like hell for Canada. And pick someone running locally in a unopposed slot, give them 1 vote instead of what they really got.
We need to have people tampering in obviously detectable ways, without changing the outcome of the election.
And you can't do that with voting. And banks have a billion backups of their database. And actually care about mistakes, or at least have to pretend to care.
You don't mark another box to try to fix it, or mark a box and write in a name.
Not that this in anyway okays what the Supreme Court decided to do, but honestly, people.
What do you mean, why don't we impliment it? You think we're in charge of the government or something?
What's an "investigative journalist"? You mean like those guys who read things of the newswire?
Resorting to murder wouldn't accomplish anything at all, politically, so it's rather unlikely anyone would do it.
Quantum computers cannot copy quantum data, they can only move it, by particle entanglement, or, duh, just moving the quantum particles that contain it. Once you measure it, it stops being a qbit and is just a single measurement.
Once any quantum calculations have finished, of course, you can copy the final measurement as much as you want. But any measurement before that will cause the collapse of the state vector. Aka, drop you into one universe or another, or complete the transaction, or just collase the state vector, depending if you believe many worlds, transactional, or copenhagen interpetation, respectively.
I mean, the president, and even the Senate, can do very little about abortion, and in fact do very little about abortion. About the only change in abortion is going to be through a constitutional amendment.
It's something to wave around and scream and rant and rave about.
You twit, ignorance of commiting copyright infringement is an excuse. With ignorace, you're just liable for damages, without, you're liable for triple damages.
Abortion is something to get people outraged about. It hasn't been a real political issue in 30 years. While people are wailing and gnashing their teeth over that, both sides are robbing us blind.
The obvious solution is to kill the old people before the terrorists can use them!
So, basically, during takeoff or landing, if a pilot wanted to crash the plane, he'd just have to wait till they were about 30 feet in the air, pull the stick to the right, and hold it there?