At the moment, true. But as the article I linked speculates, they're certainly moving toward specific faces. I can't really imagine them going "well, that's neat, we're done."
When a site derives its content entirely from its users, that site ceases to be entirely under the control of its creators. Somehow it seems to be taking a while for some people to figure this out, but when the users want something badly enough, well... you better give it to them. You know how some people keep saying the internet will empower the people by giving them a voice? Well, it turns out they mean it -- especially when it's in relation to things on the internet.
Except that in many places in the US (and apparently the UK, too) we don't have that option. There's frequently *one* option for broadband, and it frequently sucks. Sometimes there are two options that both suck. When I only have one option and it sucks, how is it my fault? Don't blame people when they have no choice -- the current government regulatory system is largely to blame for the lack of competition, which lets incompetent ISPs advertise deceptively and create these problems (since when should it be legal for "unlimited" access to have usage caps?).
I imagine in practice you could make it difficult to tap by monitoring line resistance; however there's no theoretical basis that says Eve can't just be better at tapping than Alice and Bob are at detecting her.
Also note this isn't about line resistance -- it's about the resistors at the other end, which, incidentally, aren't being used very much like resistors. That is, they're not being used to convert between voltage and current in the normal fashion, they're being used as sources of noise with very specific characteristics.
They don't need to be all that identitical. 10% is probably good enough.
Also, here you don't *need* to detect eavesdroppers -- listening to the channel doesn't actually tell you anything. (It is vulnerable to a man in the middle attack, though -- I can sit in the middle with a pair of resistors and establish a key with alice, and a different one with bob, and then intercept the message traffic on the data channel and decrypt / read / reencrypt.)
The circuit looks like this: alice connects one end of her resistor to ground, and the other to the wire. Bob does the same. Note that neither of them applies a voltage or current. Resistors generate what's called "thermal noise," which is dependent on the resistor value. (Higher values generate more thermal voltage noise and less thermal current noise -- noise power is independent of resistance -- it's caused by Brownian motion of electrons.) It's so small that it's irrelevant in most applications (precision measurements and highly sensitive radio receivers come to mind as ones that notice). Alice and Bob (and Eve) can all measure the quantity of thermal noise generated by the pair of resistors, which is dependent on the value of the two resistors in parallel -- so if both hook up 100k resistors, they see 50k equivalent noise; if both hook up 1M resistors, they see 500k equivalent noise; one of each gets 91k equivalent noise. As long as you can distinguish between these noise levels, your resistors are precise enough.
The reason Eve can't get any information is that she's only reading combined resistance. Alice and Bob are reading the same thing, but they each know which resistor they hooked up. So, in the "medium" noise case (the only one where information is transmitted), all three know the two resistor values are different, but only Alice and Bob know which end had the high resistance and which end had the low resistance.
This is a perfect example of one of the big problems of patents as currently implemented. They're supposed to be there to reward inventors and promote innovation -- but here the patent was doing the exact opposite, it's preventing new grill designs. The headline shouldn't be "patent expiration enables new grills," but rather "patent expiration makes grills cheaper." In theory, the market should make this happen through patent royalties. But obviously this patent holder would be making more in grill royalties if the patent were being licensed at a reasonable rate, and the grill makers and grill users would be better off too. So why do we all too often see patents *not* being licensed? The more I see this occur the more I think a compulsory licensing scheme for patents would be helpful. Remember, patents aren't just supposed to reward inventors -- they're supposed to encourage inventors to share their ideas *so that society can use them*. Patents should be benefiting both the inventor and the rest of us!
No, it won't. The aluminum oxide is *hard* to convert into aluminum. That's the *reason* aluminum is expensive -- not because the oxide is expensive (it's dirt cheap), but because it takes *massive* amounts of energy to turn it back into aluminum. And, said energy has to come in the form of electricity. This is just an expensive way of storing and moving electrical energy -- and an inefficient one, too, when you remember that only some of the energy in the aluminum goes into cracking the water, and about half of it goes into heat.
Well, you can improve that reaction by running it at high temperatures -- around 800C, say. At that temperature, a significant fraction of the energy needed to crack the water comes from the heat. That is, in a normal (room temp) electrolysis cell, even if there are no resistive losses, some of the electricity is turned into heat. At high temp, with low or no resistive losses, you have to *add* heat to maintain a steady state, which can be the waste heat from your electric plant.
That said, it's not trivial to run the reaction that hot, but people are working on it -- and it looks interesting.
Supercaps are definitely worth watching. Commercially available (though pricey), they're about 1/10 the energy density of lead-acid batteries these days. They seem to run about $15/kJ for high density, high discharge rate caps. You can get 400F, 2.7V in a package the size of a D cell battery for $30 qty 1, much less in bulk. I give it 3 years till they're competitive with batteries for energy density, and another 3-5 after that till the prices look reasonable.
an energy density about 2.5 times less per pound than gasoline, and about the same density per volume
Yes, but what's the energy per dollar look like? I'm guessing it's poor in comparison to gasoline and especially electricity, seeing as aluminum is rather expensive -- $2-3/lb in medium quantities, before you count the gallium.
Mostly because it's really hard to get aluminum to burn in air, so you need something to react it with. Water gives you hydrogen, which is easy to work with. Of course, this whole idea seems rather hopelessly inefficient and expensive.
You could add sodium hydroxide (lye) or another base to the water, to dissolve the oxide layer. Their solution is probably safer, but mine you can buy at the drug store. And fill balloons with the H2. (Oblig warning: NaOH is nasty caustic, and H2 is ridiculously flammable with a *huge* explosive range in air. Don't do this without appropriate safety precautions.)
What I'm actually curious about is why they think this is useful. The energy released only partly goes into cracking the water; an awful lot of it comes out as heat, which is both wasteful and has to be removed from the system. And all that energy came from electricity to refine the aluminum from aluminum oxide ore. It seems to me you should just ship the electricity in the normal manner and use it to charge conventional batteries, which have really gotten rather efficient lately.
While I loved Homeworld, I thought the basic element they missed that Starcraft had was that the races should be *different*. There should be at least two, preferably 3 races that are unique yet balanced -- not the same units repackaged with a tweak or two. That and I have a pet peeve about "space" physics that include drag -- yeah, yeah, it made the game playable...
Homeworld was quite well done, especially from a UI / controls standpoint. I also felt it had less of the micromanagement requirement that Starcraft and the like had. I'd love to see something like that but with Starcraft-like variety in the races.
It's a competitive market. There are multiple battery manufacturers. If they all have basically the same product and modest market shares, then if one of them comes out with a slightly more expensive, much longer lasting battery, they make more money because they sell more batteries (not less), because customers switch from buying other manufacturers' batteries. The phenomenon the OP described only occurs in monopoly (or sufficiently similar) markets; I'm fairly sure batteries are not such a market.
And battery companies don't do it because they want your batteries to run out faster, so you'll buy more...
Do you have *any* evidence for this?
I'd say the opposite is true. Battery companies *do* come out with new, higher performance models, and they provide good data about how well they perform. For example, Energizer has their e2 line of batteries, which have a longer life under some discharge conditions -- and those conditions are thoroughly documented in the data sheet.
See also continued improvements in lithium ion rechargeable technology -- in the past few years both power and energy densities have improved dramatically.
I suggest you do some research into the current state of the art before claiming the battery companies just sit on technology so you'll buy more batteries.
If the fruit flies had no "free will" then their behavior would be completely determined by outside circumstances or be random. As the article says, "free will" must exist somewhere between complete randomness and complete determinism. The result of the study is that flies in sensory deprivation exhibit a non-uniform random distribution -- that is, their behavior shows structure, and is neither completely random nor completely predictable. Hence, a spark of "free will".
However, I don't know that we are the determining factor. We simply don't have enough information yet. There is a LOUD chorus of individuals who claim to be sure, and they drown our the scientists that say we need more study.
I agree completely; however I don't think that means it's ok to not do anything. There is a lot of evidence that we are an important factor. It's not obviously a closed case, and it does need more study, but we also need to avoid the trap of "paralysis through analysis." We can commission study after study and await results until it is either too late or the costs of fixing it have gone up. At this point, the evidence is strong enough that it should be clear we are better off starting to solve the problem *now*, while continuing to study it, than we are postponing a solution while the problem gets harder to solve in hopes that we've been wrong.
Put another way, "needs more study" vs "fix the problem" is a false dichotomy -- there is nothing to say we can't start solving the problem now, while it's still tractable, while *also* continuing to study it to make sure both that we're solving the problem in the best manner and that it actually exists / is solvable.
Slashdot Mirror
Rather, it will be posted on YouTube.
At the moment, true. But as the article I linked speculates, they're certainly moving toward specific faces. I can't really imagine them going "well, that's neat, we're done."
Could you perhaps provide us with a car analogy?
They may be small, but they would appear to be vocal enough to get themselves posted on Slashdot...
When a site derives its content entirely from its users, that site ceases to be entirely under the control of its creators. Somehow it seems to be taking a while for some people to figure this out, but when the users want something badly enough, well... you better give it to them. You know how some people keep saying the internet will empower the people by giving them a voice? Well, it turns out they mean it -- especially when it's in relation to things on the internet.
Google seems to be deploying something similar
.Apple has a two-button mouse, they just hid one of the buttons on the keyboard...
Except that in many places in the US (and apparently the UK, too) we don't have that option. There's frequently *one* option for broadband, and it frequently sucks. Sometimes there are two options that both suck. When I only have one option and it sucks, how is it my fault? Don't blame people when they have no choice -- the current government regulatory system is largely to blame for the lack of competition, which lets incompetent ISPs advertise deceptively and create these problems (since when should it be legal for "unlimited" access to have usage caps?).
I imagine in practice you could make it difficult to tap by monitoring line resistance; however there's no theoretical basis that says Eve can't just be better at tapping than Alice and Bob are at detecting her.
Also note this isn't about line resistance -- it's about the resistors at the other end, which, incidentally, aren't being used very much like resistors. That is, they're not being used to convert between voltage and current in the normal fashion, they're being used as sources of noise with very specific characteristics.
They don't need to be all that identitical. 10% is probably good enough.
Also, here you don't *need* to detect eavesdroppers -- listening to the channel doesn't actually tell you anything. (It is vulnerable to a man in the middle attack, though -- I can sit in the middle with a pair of resistors and establish a key with alice, and a different one with bob, and then intercept the message traffic on the data channel and decrypt / read / reencrypt.)
The circuit looks like this: alice connects one end of her resistor to ground, and the other to the wire. Bob does the same. Note that neither of them applies a voltage or current. Resistors generate what's called "thermal noise," which is dependent on the resistor value. (Higher values generate more thermal voltage noise and less thermal current noise -- noise power is independent of resistance -- it's caused by Brownian motion of electrons.) It's so small that it's irrelevant in most applications (precision measurements and highly sensitive radio receivers come to mind as ones that notice). Alice and Bob (and Eve) can all measure the quantity of thermal noise generated by the pair of resistors, which is dependent on the value of the two resistors in parallel -- so if both hook up 100k resistors, they see 50k equivalent noise; if both hook up 1M resistors, they see 500k equivalent noise; one of each gets 91k equivalent noise. As long as you can distinguish between these noise levels, your resistors are precise enough.
The reason Eve can't get any information is that she's only reading combined resistance. Alice and Bob are reading the same thing, but they each know which resistor they hooked up. So, in the "medium" noise case (the only one where information is transmitted), all three know the two resistor values are different, but only Alice and Bob know which end had the high resistance and which end had the low resistance.
This is a perfect example of one of the big problems of patents as currently implemented. They're supposed to be there to reward inventors and promote innovation -- but here the patent was doing the exact opposite, it's preventing new grill designs. The headline shouldn't be "patent expiration enables new grills," but rather "patent expiration makes grills cheaper." In theory, the market should make this happen through patent royalties. But obviously this patent holder would be making more in grill royalties if the patent were being licensed at a reasonable rate, and the grill makers and grill users would be better off too. So why do we all too often see patents *not* being licensed? The more I see this occur the more I think a compulsory licensing scheme for patents would be helpful. Remember, patents aren't just supposed to reward inventors -- they're supposed to encourage inventors to share their ideas *so that society can use them*. Patents should be benefiting both the inventor and the rest of us!
Actually, it does. It just doesn't make it Free Software.
No, it won't. The aluminum oxide is *hard* to convert into aluminum. That's the *reason* aluminum is expensive -- not because the oxide is expensive (it's dirt cheap), but because it takes *massive* amounts of energy to turn it back into aluminum. And, said energy has to come in the form of electricity. This is just an expensive way of storing and moving electrical energy -- and an inefficient one, too, when you remember that only some of the energy in the aluminum goes into cracking the water, and about half of it goes into heat.
Well, you can improve that reaction by running it at high temperatures -- around 800C, say. At that temperature, a significant fraction of the energy needed to crack the water comes from the heat. That is, in a normal (room temp) electrolysis cell, even if there are no resistive losses, some of the electricity is turned into heat. At high temp, with low or no resistive losses, you have to *add* heat to maintain a steady state, which can be the waste heat from your electric plant.
That said, it's not trivial to run the reaction that hot, but people are working on it -- and it looks interesting.
Supercaps are definitely worth watching. Commercially available (though pricey), they're about 1/10 the energy density of lead-acid batteries these days. They seem to run about $15/kJ for high density, high discharge rate caps. You can get 400F, 2.7V in a package the size of a D cell battery for $30 qty 1, much less in bulk. I give it 3 years till they're competitive with batteries for energy density, and another 3-5 after that till the prices look reasonable.
an energy density about 2.5 times less per pound than gasoline, and about the same density per volume
Yes, but what's the energy per dollar look like? I'm guessing it's poor in comparison to gasoline and especially electricity, seeing as aluminum is rather expensive -- $2-3/lb in medium quantities, before you count the gallium.
Mostly because it's really hard to get aluminum to burn in air, so you need something to react it with. Water gives you hydrogen, which is easy to work with. Of course, this whole idea seems rather hopelessly inefficient and expensive.
You could add sodium hydroxide (lye) or another base to the water, to dissolve the oxide layer. Their solution is probably safer, but mine you can buy at the drug store. And fill balloons with the H2. (Oblig warning: NaOH is nasty caustic, and H2 is ridiculously flammable with a *huge* explosive range in air. Don't do this without appropriate safety precautions.)
What I'm actually curious about is why they think this is useful. The energy released only partly goes into cracking the water; an awful lot of it comes out as heat, which is both wasteful and has to be removed from the system. And all that energy came from electricity to refine the aluminum from aluminum oxide ore. It seems to me you should just ship the electricity in the normal manner and use it to charge conventional batteries, which have really gotten rather efficient lately.
Release date?
While I loved Homeworld, I thought the basic element they missed that Starcraft had was that the races should be *different*. There should be at least two, preferably 3 races that are unique yet balanced -- not the same units repackaged with a tweak or two. That and I have a pet peeve about "space" physics that include drag -- yeah, yeah, it made the game playable...
Homeworld was quite well done, especially from a UI / controls standpoint. I also felt it had less of the micromanagement requirement that Starcraft and the like had. I'd love to see something like that but with Starcraft-like variety in the races.
Actually, what they're saying is that they observe a distribution of results that is specifically different from what your code would produce.
It's a competitive market. There are multiple battery manufacturers. If they all have basically the same product and modest market shares, then if one of them comes out with a slightly more expensive, much longer lasting battery, they make more money because they sell more batteries (not less), because customers switch from buying other manufacturers' batteries. The phenomenon the OP described only occurs in monopoly (or sufficiently similar) markets; I'm fairly sure batteries are not such a market.
And battery companies don't do it because they want your batteries to run out faster, so you'll buy more...
Do you have *any* evidence for this?
I'd say the opposite is true. Battery companies *do* come out with new, higher performance models, and they provide good data about how well they perform. For example, Energizer has their e2 line of batteries, which have a longer life under some discharge conditions -- and those conditions are thoroughly documented in the data sheet.
See also continued improvements in lithium ion rechargeable technology -- in the past few years both power and energy densities have improved dramatically.
I suggest you do some research into the current state of the art before claiming the battery companies just sit on technology so you'll buy more batteries.
If the fruit flies had no "free will" then their behavior would be completely determined by outside circumstances or be random. As the article says, "free will" must exist somewhere between complete randomness and complete determinism. The result of the study is that flies in sensory deprivation exhibit a non-uniform random distribution -- that is, their behavior shows structure, and is neither completely random nor completely predictable. Hence, a spark of "free will".
However, I don't know that we are the determining factor. We simply don't have enough information yet. There is a LOUD chorus of individuals who claim to be sure, and they drown our the scientists that say we need more study.
I agree completely; however I don't think that means it's ok to not do anything. There is a lot of evidence that we are an important factor. It's not obviously a closed case, and it does need more study, but we also need to avoid the trap of "paralysis through analysis." We can commission study after study and await results until it is either too late or the costs of fixing it have gone up. At this point, the evidence is strong enough that it should be clear we are better off starting to solve the problem *now*, while continuing to study it, than we are postponing a solution while the problem gets harder to solve in hopes that we've been wrong.
Put another way, "needs more study" vs "fix the problem" is a false dichotomy -- there is nothing to say we can't start solving the problem now, while it's still tractable, while *also* continuing to study it to make sure both that we're solving the problem in the best manner and that it actually exists / is solvable.