Given rockets spend tens of thousand dollars for reaching geosynchronous orbit the space elevators a few hundreds of dollars per kilogram sounds good...until you compare it with 10 cents per kilogram of Orion. P George Dyson's "Project Orion" is a great book on the subject. http://en.wikipedia.org/wiki/Orion_project is a good introduction.
Will they be able to produce each tube as long as the cable itself? If not, they will have to glue them together to withstand vertical tension too. Even if they could make thousands of miles long strands, since the cable thickness isn't constant, it will be required to glue strands at some regions.
They will have to take off their pink glasses if they want any more progress. Chemically binding tubes with each other should be possible. Tangling tubes and using glue to avoid untangling should be possible. Neither option would have the same weight per tensile strength ratio. They might work, perhaps they won't. In contrast, their proposal -as it stands- cannot possibly work.
It is absurd. If epoxy cannot be used alone, it means that attraction between its own molecules or bonds of ploymer chain cannot withstand tension elevator experiences. Which means that electrostatic force were an accurate analogy, the electrons would have fallen down from atoms, let alone lift anything else. If the electrostatic force can't overcome gravity even if full length of hairs overlap, how would it possibly hold two hairs together when they partially overlap?
I believe grandparent has a point. Spy satellites were sometime in the future during the early phase but reconnaissance planes and good old spy humans used to detect them just fine. IMHO, secret services' knowing some launch failed doesn't make much difference, it is keeping that information from public, especially from citizens of you country and your allies, that matters. At that, west had a disadvantage due to more free media.
You should have seen it before kernel 2.6, prelinking and KDE 3.2!
Linux distros are usually slower than windows for desktop use. That is especially true for big and heavy stuff like KDE, Gnome, Mozilla and openoffice.org. The nicest thing is most bottlenecks can be and are being eliminated.
You can use the personalizer (kpersonalizer) to quickly turn off eye candy. Or you can fire up the kde control center (kcontrol) and go over all appearance & themes stuff one by one. Don't worry, you can't break your install from kcontrol unless you try to do just that.
IMHO, the best theme for KDE is plastic but almost all themes are "low-impact". Only a few used to use fancy ways to render their widgets, I don't know if any survive.
In the case of 5.1, you can make a sound come out of exactly one speaker to convey point-source direction that you can't accomplish with 2.1.
Even if i comes out of only one speaker, you hear the sound with two ears. It is just the time of arrival, relative loudness etc. you can make out where it comes from. Which means you can give exactly the same impression using 2.0 headphones.
Doing the same with a speaker system is complicated but only because sound doesn't travel from speakers to your ears and just stop there, not because it has only two speakers. If the system could on-the-fly downmix n.1 to 2.1 with a perfect model and dynamic of your auditory environment, there would have been no difference.
That is actually configurable. If you have it enabled (which is the default), same instance of a program will get 3 to 1 more time when it is in foreground under 100% CPU utilization.
>> Do people really think that products are designed by an army of Mr. Magoo clones?
>Having dealt with commercial application software, by answer would be a resounding: "Yes!"
Do you realize that is a very sofware engineering thing? No other engineering discipline consistently market products of such low quality design and implementation.
Of course there are exceptions, there are always exceptions.
Have a look at this. Some of Radin's stuff is speculative, but this experiment has been reproduced a number of times by different groups.
I find that hard to swallow. fMRI is a relatively new technique but not really rare. The only other piece of equipment required to duplicate it is a desktop computer, loaded with stuff from rotten.org. So we should have seen papers published all around the world, yet, this is the first time I've heard of proof of clairvoyance. Can you find a similar paper that was published in a peer reviewed mag.?
If I can paraphrase, you conjecture that a small part of the universe (your head, but it could be any small part) bounded in space & time can be simulated *exactly*. Again, good luck. No one has ever succeeded in simulating anything exactly, not for reasons of computational inadequacy, but because of either ignorance of or unobservability of parameters / boundary conditions / undiscovered physical "laws". But, as you say, simulations teach us stuff. So I guess it depends whether you are happy with a simulation of limited accuracy - I would see little benefit, but you may see more (did I pick up earlier that you were in the cognitive sciences?).
Two glasses of water, at almost same temperature in very similar glasses behave exactly like each other for all thermodynamic purposes. However they would be in very different states and if we wanted to match quantum states of molecules in glass A to ones in glass B we would definetly fail. The matter is whether or not our "minds" are robust. Do minds change a lot with slight changes to brain, or do there exist several physical states that correspond to same mental state?
I think (but sure don't know) a procedurally exact simulation of my brain with slightly off data will behave (in the cognitive sense) exactly like my brain. Other poster in the thread (not the AC one) seems to think this is not so. I have two reasons to believe what I believe, but let me present some facts about color perception.
The color cones in our eyes are probabilistic. They fire with higher probability if the light intensity is higher or the color of incident light closely matches the color cone is responsive to, compared to lower intensity and off-color light cases. Of course, the signals they transmit to neurons in the visual cortex is different each time even if the scene stays the same. The neurons, in turn, are not 100% deterministic machines with respect to states of other neurons, so they may or may not fire during a given time interval even if it did fire when the states of other neurons connecting to it were in the exact same state previously. Given this two facts, even seeing a pure color create a different physical state of brain each time, yet, averaged over time and population of neurons, they are matching physical states.
This is an important property of perception in the brain. The matching perceptual states may be, and usually is, different physical states when analysed with too high resolution but matching physical states with sufficiently low resolution.
I have absouletly no proof that higher mental functions are like primitive perceptual functions. As such I cannot rule out important chaotic processes in the brain or possible role of superposition in cognition. My two reasons for believing that is not the case is simply:
A) The subjective experince of our own mental states are robust. We don't go around changing our beliefs, focus, course of action on time scales comparable to that of a single synapse or neuron's.
B) The brain is too large to make use of single quantum events and population of quantum events are very deterministic.
Anyway, this is somewhat irrelevant. Would a simulated brain act like a human brain (not necessarily act exactly like the human brain it was modeled after), given inaccurate initial state and boundary conditions but perfect pyhsical simulation? Is it in principle possible to run that simulation on a turing machine? If the answer is "yes" to both questions, we have Turing equivalent brains.
You are entirely missing my points. First of all, the bullshit is not that the universe may not be computable, but that some observed property of brain suggests it really is so. That is utter and complete bullshit. All supposed super-turing abilities I've seen so far come in two flavors:
a) Writer compare a human's conjecture (insights, feelings etc. also fall into this category) with valid and formal proof, decides computers can never ever do some stuff humans can. No wonder turing machines come with less abilities! Except that computers are equally well at guessing things without proving them.
b) Malfunction of a tur,ing machine can make grant it super-turing abilities. That is what you are basically saying, and if you can't see the problem with that by itself, I won't bother. The key question is "how exactly?"
You are also missing the point about term "computable." If a thing is "computable", there must be some procedure that computes it given infinite storage and finite, but possibly very long, time. It has nothing to do with computers.
Now, the significant thing about simulating part of a universe exactly is that, the simulated universe need not be closed and isolated like the universe must be. You can draw a boundary around your head, and if the simulated universe behaves exactly like your head inside the boundary, that says something about cognitive abilities and their roots. The partialness of that simulation is its unnatural (e.g. not emergent from the principles simulations runs on) space and time boundedness, unlike Kepler's which is not bounded in space or time, but explains only part of what is going on.
Finally, simulating universes is done every day. Just buy a strategy, FPS, racing etc. game. They don't qualify, not because they are deficient universes, but
a) no sentient being is simulated.
b) I unnecessarily reestricted myself to "same physics" rules when presenting the original argument.
But I feel as if my inability to pinpoint complexity value of consciousness hints about a serious problem with the whole idea. make that "But I don't think my inability to pinpoint complexity value of consciousness hints about a serious problem with the whole idea."
Computational emergence makes the most testable prediction of all: that we can build conscious machines out of ordinary computers given enough processing power. It is even possible to quantify that processing power to something not beyond reach of current computing technology given some assumptions about how exactly brain operates.
I'm not in a position to make an informed guess about whether or not the universe is computable. I'm not a physicist. But I'm actuall trained in cognitive science and there not a piece of single experimental data in the literature that suggests the universe is not computable because minds do funny things. It is as if, even if the universe is not computable, brains act as if it is.
The philosophical foundations are not that strong. For example, I can't answer your quick question. But I feel as if my inability to pinpoint complexity value of consciousness hints about a serious problem with the whole idea. Quick, when it is pink and when it is red?
Oh, but you see my statements weren't about whether we (as in "humans") can build such a machine but
a) whether it is in principle possible to perfectly simulate a universe which operate on the same physics laws as ours, or perfectly simulate part of our universe, allowing for as garbage in garbage out inperfectliness.
b) whether our minds always obey physics.
That is how I used the word "computable", a statement about minds, not our engineering ability. It may well be possible that we won't ever have computational power to simulate something size of a brain, but perhaps we will make do with much less instead (i.e. almost everything goes on in brain is unrelated to cognition.)
I'm sick of hearing such bullshit time and again. If our mental substrate is composed of computable medium, it can be programmed on a general turing machine and the seeming super-turing abilities of the mind cannot be real (not that I've seen any but most folks -like you- think they exist.) If our mind is a function of brain and if universe is computable, we are computable. No trans-logic, no quantum effects in synapses, no artistic inspirations, no high on drugs, no nothing will change that. Because, despite you can't imagine how seemingly incomputable behavior come from the strictly computing machine, there is no uncomputable component in the system
If universe isn't computable, or if we have "souls" in the very metaphysical sense of the word, we may not be computable. But that is not what you are claiming now, is it?
On top of my head, reiser, unlike NTFS (and xfs, among others) doesn't support streams. Also v3 has been a lot less stable for me than any other fs I tried, excluding -perhaps- JFS; even FAT16 handles hardware related crashes better. Journalling, b-trees, transparent compression and encryption is already in NTFS but is at proof-of-concept stage in reiserfs. So, I really don't see your point about reiser making NTFS look like a toy. Quite the opposite, in fact. If and when reiser4 gets the plugins it was designed for and really make use of its small file optimization, it will be superior to NTFS. It is not there yet, and v3 can never be there.
That said, I'm perfectly happy with ext3. Missing all the whiz-bang in NTFS doesn't make it any less useful.
And that would solve the problem how? They could have reserved any piece of the memory, as long as OS could access whole memory, there would have been no 640k limit. It is not like OS has to use a single continous block. The problem was that the 20bit scheme was with 16bit segments and 16bit offsets and could not be transparently (and in a sane way) extended to 32 bits. Even if they reserved no memory for IO and stuff, they could not address more than 1MB of memory and be compatible with older DOS programs without using awful kludges.
I'm pretty sure it was after introduction of the 56k modems. I don't remember hearing it before also 56k modems don't bonk when they connect other home dial-up modems either (since the connection is limited to 33.6 downstream too in that case.)
What if OS X and its bundled applications are every bit as bad as windows' and its market share is equal? What if we had 20 different and exploit-wise incompatible, nevertheless as insecure as each other , versions of Windows and each used as often as other? In both cases the exploits in the wild will be much less frequent. There will be less vectors to carry a particular worm/virus/exploit and they will be much easier to contain. So I believe preaching alternative, however insecure they may be, makes sense from a security POV.
you will have to do them *many* times until the result is the same.
That is only true if the computer can't interrupt and reliably store the state of calculation. If it can, the running the same computation only(!) twice gives the correct answer. Say the average rate of 1 bit errors is 1 per hour, if the computer checks its state every one minute, one has to do triple calculations only once every sixty minutes. There should be an optimum value in which the sum of overhead of storing and retrieving data and overhead of occasionally doing more than two calculations will be lowest.
300 teraflops is the upper bound of human brains' computational capacity based on neuron count and a very generous estimate of how much a neuron can represent and communicate, assuming no redundancy and no calculation outside neurons occur (both assumptions are known to be false, they are there for simplification.) 500 teraflops is the processing power of human brains' based on pretty accurate estimate of computational capacity of retina and extrapolating that to whole brain.
You are looking at the wrong level. Here, this device A is doing computation C and when doing so it creates k possible outputs from n possible inputs. If k is less than n it must be destroying data while operating (pigeon holes etc) on some or all of its inputs. It is the computation C that loses data, not the device A. But device A uses energy because it implements a data losing computation C.
Dude, make your mind: Are you for free trade or against free trade? You can't divide market into two and advocate free trade for one part (goods) but not the other (labor.) Because the goods you want to freely trade don't come out of blue. If you restrict outsourcing, you will make the poorer country much more competitive with the goods they sell, losing jobs all the same in the end.
The problem is, while the rich countries' labor force is much more efficient relative to poorer countries, they are not as efficient as their wealth suggests (again relative to poor countries.) In a world of completly free and fair trade, you Americans can't possibly ask half the wages you now get. That is doubly true for Europe. It goes without saying rich countries won't give up their relative wealth just because. Restriction of trade is one of the more humane ways of keeping it that way, all alternatives -short of actually making rich people as productive as they should have been, IMHO an impossible feat- involve some sort of destruction of competitiveness of others. Sabotage, terrorism and outright war are time proven ways of doing that.
Slashdot Mirror
Given rockets spend tens of thousand dollars for reaching geosynchronous orbit the space elevators a few hundreds of dollars per kilogram sounds good...until you compare it with 10 cents per kilogram of Orion. P George Dyson's "Project Orion" is a great book on the subject. http://en.wikipedia.org/wiki/Orion_project is a good introduction.
They will have to take off their pink glasses if they want any more progress. Chemically binding tubes with each other should be possible. Tangling tubes and using glue to avoid untangling should be possible. Neither option would have the same weight per tensile strength ratio. They might work, perhaps they won't. In contrast, their proposal -as it stands- cannot possibly work.
It is absurd. If epoxy cannot be used alone, it means that attraction between its own molecules or bonds of ploymer chain cannot withstand tension elevator experiences. Which means that electrostatic force were an accurate analogy, the electrons would have fallen down from atoms, let alone lift anything else. If the electrostatic force can't overcome gravity even if full length of hairs overlap, how would it possibly hold two hairs together when they partially overlap?
I believe grandparent has a point. Spy satellites were sometime in the future during the early phase but reconnaissance planes and good old spy humans used to detect them just fine. IMHO, secret services' knowing some launch failed doesn't make much difference, it is keeping that information from public, especially from citizens of you country and your allies, that matters. At that, west had a disadvantage due to more free media.
Linux distros are usually slower than windows for desktop use. That is especially true for big and heavy stuff like KDE, Gnome, Mozilla and openoffice.org. The nicest thing is most bottlenecks can be and are being eliminated.
IMHO, the best theme for KDE is plastic but almost all themes are "low-impact". Only a few used to use fancy ways to render their widgets, I don't know if any survive.
Even if i comes out of only one speaker, you hear the sound with two ears. It is just the time of arrival, relative loudness etc. you can make out where it comes from. Which means you can give exactly the same impression using 2.0 headphones.
Doing the same with a speaker system is complicated but only because sound doesn't travel from speakers to your ears and just stop there, not because it has only two speakers. If the system could on-the-fly downmix n.1 to 2.1 with a perfect model and dynamic of your auditory environment, there would have been no difference.
That is actually configurable. If you have it enabled (which is the default), same instance of a program will get 3 to 1 more time when it is in foreground under 100% CPU utilization.
>Having dealt with commercial application software, by answer would be a resounding: "Yes!"
Do you realize that is a very sofware engineering thing? No other engineering discipline consistently market products of such low quality design and implementation.
Of course there are exceptions, there are always exceptions.
I find that hard to swallow. fMRI is a relatively new technique but not really rare. The only other piece of equipment required to duplicate it is a desktop computer, loaded with stuff from rotten.org. So we should have seen papers published all around the world, yet, this is the first time I've heard of proof of clairvoyance. Can you find a similar paper that was published in a peer reviewed mag.?
Two glasses of water, at almost same temperature in very similar glasses behave exactly like each other for all thermodynamic purposes. However they would be in very different states and if we wanted to match quantum states of molecules in glass A to ones in glass B we would definetly fail. The matter is whether or not our "minds" are robust. Do minds change a lot with slight changes to brain, or do there exist several physical states that correspond to same mental state?
I think (but sure don't know) a procedurally exact simulation of my brain with slightly off data will behave (in the cognitive sense) exactly like my brain. Other poster in the thread (not the AC one) seems to think this is not so. I have two reasons to believe what I believe, but let me present some facts about color perception.
The color cones in our eyes are probabilistic. They fire with higher probability if the light intensity is higher or the color of incident light closely matches the color cone is responsive to, compared to lower intensity and off-color light cases. Of course, the signals they transmit to neurons in the visual cortex is different each time even if the scene stays the same. The neurons, in turn, are not 100% deterministic machines with respect to states of other neurons, so they may or may not fire during a given time interval even if it did fire when the states of other neurons connecting to it were in the exact same state previously. Given this two facts, even seeing a pure color create a different physical state of brain each time, yet, averaged over time and population of neurons, they are matching physical states.
This is an important property of perception in the brain. The matching perceptual states may be, and usually is, different physical states when analysed with too high resolution but matching physical states with sufficiently low resolution.
I have absouletly no proof that higher mental functions are like primitive perceptual functions. As such I cannot rule out important chaotic processes in the brain or possible role of superposition in cognition. My two reasons for believing that is not the case is simply:
A) The subjective experince of our own mental states are robust. We don't go around changing our beliefs, focus, course of action on time scales comparable to that of a single synapse or neuron's.
B) The brain is too large to make use of single quantum events and population of quantum events are very deterministic.
Anyway, this is somewhat irrelevant. Would a simulated brain act like a human brain (not necessarily act exactly like the human brain it was modeled after), given inaccurate initial state and boundary conditions but perfect pyhsical simulation? Is it in principle possible to run that simulation on a turing machine? If the answer is "yes" to both questions, we have Turing equivalent brains.
a) Writer compare a human's conjecture (insights, feelings etc. also fall into this category) with valid and formal proof, decides computers can never ever do some stuff humans can. No wonder turing machines come with less abilities! Except that computers are equally well at guessing things without proving them.
b) Malfunction of a tur,ing machine can make grant it super-turing abilities. That is what you are basically saying, and if you can't see the problem with that by itself, I won't bother. The key question is "how exactly?"
You are also missing the point about term "computable." If a thing is "computable", there must be some procedure that computes it given infinite storage and finite, but possibly very long, time. It has nothing to do with computers.
Now, the significant thing about simulating part of a universe exactly is that, the simulated universe need not be closed and isolated like the universe must be. You can draw a boundary around your head, and if the simulated universe behaves exactly like your head inside the boundary, that says something about cognitive abilities and their roots. The partialness of that simulation is its unnatural (e.g. not emergent from the principles simulations runs on) space and time boundedness, unlike Kepler's which is not bounded in space or time, but explains only part of what is going on.
Finally, simulating universes is done every day. Just buy a strategy, FPS, racing etc. game. They don't qualify, not because they are deficient universes, but
a) no sentient being is simulated.
b) I unnecessarily reestricted myself to "same physics" rules when presenting the original argument.
But I feel as if my inability to pinpoint complexity value of consciousness hints about a serious problem with the whole idea. make that "But I don't think my inability to pinpoint complexity value of consciousness hints about a serious problem with the whole idea."
I'm not in a position to make an informed guess about whether or not the universe is computable. I'm not a physicist. But I'm actuall trained in cognitive science and there not a piece of single experimental data in the literature that suggests the universe is not computable because minds do funny things. It is as if, even if the universe is not computable, brains act as if it is.
The philosophical foundations are not that strong. For example, I can't answer your quick question. But I feel as if my inability to pinpoint complexity value of consciousness hints about a serious problem with the whole idea. Quick, when it is pink and when it is red?
PS: I won't be around for some 12-16 hours.
a) whether it is in principle possible to perfectly simulate a universe which operate on the same physics laws as ours, or perfectly simulate part of our universe, allowing for as garbage in garbage out inperfectliness.
b) whether our minds always obey physics.
That is how I used the word "computable", a statement about minds, not our engineering ability. It may well be possible that we won't ever have computational power to simulate something size of a brain, but perhaps we will make do with much less instead (i.e. almost everything goes on in brain is unrelated to cognition.)
If universe isn't computable, or if we have "souls" in the very metaphysical sense of the word, we may not be computable. But that is not what you are claiming now, is it?
On top of my head, reiser, unlike NTFS (and xfs, among others) doesn't support streams. Also v3 has been a lot less stable for me than any other fs I tried, excluding -perhaps- JFS; even FAT16 handles hardware related crashes better. Journalling, b-trees, transparent compression and encryption is already in NTFS but is at proof-of-concept stage in reiserfs. So, I really don't see your point about reiser making NTFS look like a toy. Quite the opposite, in fact. If and when reiser4 gets the plugins it was designed for and really make use of its small file optimization, it will be superior to NTFS. It is not there yet, and v3 can never be there. That said, I'm perfectly happy with ext3. Missing all the whiz-bang in NTFS doesn't make it any less useful.
And that would solve the problem how? They could have reserved any piece of the memory, as long as OS could access whole memory, there would have been no 640k limit. It is not like OS has to use a single continous block. The problem was that the 20bit scheme was with 16bit segments and 16bit offsets and could not be transparently (and in a sane way) extended to 32 bits. Even if they reserved no memory for IO and stuff, they could not address more than 1MB of memory and be compatible with older DOS programs without using awful kludges.
I'm pretty sure it was after introduction of the 56k modems. I don't remember hearing it before also 56k modems don't bonk when they connect other home dial-up modems either (since the connection is limited to 33.6 downstream too in that case.)
What if OS X and its bundled applications are every bit as bad as windows' and its market share is equal? What if we had 20 different and exploit-wise incompatible, nevertheless as insecure as each other , versions of Windows and each used as often as other? In both cases the exploits in the wild will be much less frequent. There will be less vectors to carry a particular worm/virus/exploit and they will be much easier to contain. So I believe preaching alternative, however insecure they may be, makes sense from a security POV.
That is only true if the computer can't interrupt and reliably store the state of calculation. If it can, the running the same computation only(!) twice gives the correct answer. Say the average rate of 1 bit errors is 1 per hour, if the computer checks its state every one minute, one has to do triple calculations only once every sixty minutes. There should be an optimum value in which the sum of overhead of storing and retrieving data and overhead of occasionally doing more than two calculations will be lowest.
300 teraflops is the upper bound of human brains' computational capacity based on neuron count and a very generous estimate of how much a neuron can represent and communicate, assuming no redundancy and no calculation outside neurons occur (both assumptions are known to be false, they are there for simplification.) 500 teraflops is the processing power of human brains' based on pretty accurate estimate of computational capacity of retina and extrapolating that to whole brain.
If you have a connection to the source, why not?
You are looking at the wrong level. Here, this device A is doing computation C and when doing so it creates k possible outputs from n possible inputs. If k is less than n it must be destroying data while operating (pigeon holes etc) on some or all of its inputs. It is the computation C that loses data, not the device A. But device A uses energy because it implements a data losing computation C.
The problem is, while the rich countries' labor force is much more efficient relative to poorer countries, they are not as efficient as their wealth suggests (again relative to poor countries.) In a world of completly free and fair trade, you Americans can't possibly ask half the wages you now get. That is doubly true for Europe. It goes without saying rich countries won't give up their relative wealth just because. Restriction of trade is one of the more humane ways of keeping it that way, all alternatives -short of actually making rich people as productive as they should have been, IMHO an impossible feat- involve some sort of destruction of competitiveness of others. Sabotage, terrorism and outright war are time proven ways of doing that.