There are several reasons why software keeps crashing, and they aren't going away any time soon. These reasons are:
You can't prove that most software works.
Except for a restricted set of cases, you can't prove that a given piece of code works or doesn't work. A truly exhaustive set of tests would be impractical to perform, and formal proofs of correctness place strong limits on the type of code you can write and the environment in which you can write it.
The result is that code is assumed correct when no bugs are found. This only means that there probably aren't _many_ bugs left. Thus, it may still crash (or have a security hole, or what-have-you).
Software is very complex.
Software has been complex for a long time. It just tends to be bigger now. A larger system has more opportunities for unexpected high-level interactions between components, but even a smaller system will have enough twists and turns that formulating a really good test suite, or checking the code by inspection, is very difficult. Bugs will be missed. As was discussed above, many of these missed bugs will slip through testing and reach the world.
Nobody wants to pay for perfect software.
As more effort is applied, you can get asymptotically closer to a bug-free system. However, this is far past the point of diminishing returns on the cost/benefit curve. For sufficiently constrained systems, you can even try proving it correct, but this tends to lead to cutting out a lot of functionality, speed, or both.
In situations where reliability must be had at any cost - aerospace control systems, vehicle control systems, medical equipment - the money will exist to produce near-perfect code, but even then there are bugs that occasionally bite. With commercial software, the buyer would rather have an application that crashes now and then than an application that costs ten times as much and comes out several years later.
Free and/or open software avoids some of this by staying in development longer, which allows more of the bugs to be caught, but even free and/or open software evolves. Every change brings new bugs to be squashed. As long as there are new types of software that we want, it isn't going to end.
The ultimate solution to the problem is to let computers write the software themselves. Give them a goal, set up evolutionary and genetic algorithms, and let them go at it on a supercomputer cluster for a few months.
Unfortunately, this only works if you can distinguish between buggy and non-buggy code produced by the algorithm. You can do tests, but no test suite will be exhaustive (otherwise we'd just use it on human-developed code to find the bugs).
Perfect software can only be produced if a formal proof of correctness is possible. Even then, you're limited by the assumptions the proof makes.
""Squishing" only occurs when there is a difference in pressure between two parts of the craft."
No, it doesn't. Water at depth is denser than water at the surface, just as your body tissue is denser at 1 atm than in a vacuum.
I am aware of this, and stated as much in the next paragraph of my message.
Volume does not change drastically - certainly not enough to destroy a probe. It's pressure _gradients_ and shear stresses that you have to worry about, as they are what will tear apart, crush, or otherwise mutilate the craft. A pressure gradient occurs if you try to keep a low-pressure cavity intact in the craft, and shear stress comes if you have two materials with different Young's modulus in the craft's structure. Avoid both, and you're fine.
We're not talking about 1000 atm, though. 10 m of sea water is about 1 atm, so you hit 1000 atm at only 10 km below the surface of the ocean. The earth's radius is ~6000 km, and the vast majority of it is orders of magnitude denser than water.
The "1000 atm" number was chosen for the example because it's about the point at which air would liquify at room temperature. Talking about air at a million atm would be silly, as it would no longer be a gas.
Your density numbers also seem to be way off. Far from "orders of magnitude", density's only about double what it is at earth's surface (density of the lower mantle's about 6 g/cm^3, compared to a 1 atm density of around 3 for olivine, and density of the core's about 10-13 g/cm^3, compared to a 1 atm density of about 8 for iron). Given a corresponding volume change for a given amount of material, this represents a change in dimensions of about 25%.
Solid matter is *difficult* to compress. Packing atoms much more tightly than their natural spacing means smaller electron wavelengths, which means much higher electron energies. That energy has to come from somewhere, which means you're going to have to put in a lot of work to get it to that compressed a state. With relatively small deformation, that means a lot of pressure to perform that amount of work.
Besides, I think the more immediate problem is "squeeze something, and it gets hotter."
You're thinking of gases, which have a greater volume change (heat buildup is related to the amount of work done compressing the gas). You're also thinking of adiabatic systems - where there is little time for energy exchange with the environment. When compression occurs slowly, heating is minimal, because the heat that does build up quickly leaks away into the environment.
The earth has a temperature gradient because there's a strong heat source in the core, not because of pressure. What the heat source is is up for debate (most think radioactive decay, some still think latent heat of crystallization), but as the heat energy deposited flows out towards the surface, you get a temperature gradient.
We have yet to be able to make a probe that can even survive the Venutian atmosphere for any length of time longer than a few hours. And this from the people who built supersonic fighters capable of taking off and landing in grass fields.
Venus's atmosphere is highly corrosive. The mantle is less so (though still pretty reactive at those temperatures). The atmosphere of Jupiter is much less so. This presents a bigger barrier to long-term survival that pressure, and is my first guess at why the Venus landers stopped functioning.
I'd be more worried about the device being able to withstand that pressure. I fail to see how surrounding the probe with molten iron (or any other fluid, for that matter) will prevent the weight of the planet from squishing it like a bug. Or does he plan on violating the laws of physics at the same time?
"Squishing" only occurs when there is a difference in pressure between two parts of the craft. If the craft is all at one pressure, it feels nothing.
The real problems you have to worry about are volume and crystal structure changes with pressure. Air at 1000 atmospheres has a very different volume than air at 1 atmosphere. Iron at 1000 atmospheres, on the other hand, has a not-too-different volume. Build your craft out of materials that react similarly to pressure and make it tolerant of volume changes, and no "squishing" occurs.
Crystal structure changes are a tougher problem. Squeeze carbon, and it becomes diamond. Squeeze other materials, and they go through their own allotrope phases. Mechanical properties will change, and more importantly, electrical properties in semiconductors will also change.
There's also the small matter of keeping semiconductors working above the melting point of iron, but that may be an attainable goal (certainly not with silicon, but IIRC diamond's temperature limits were far, far higher, and other high-temperature semiconductors doubtless exist). Refrigeration is not an option - that big a heat gradient will be next to impossible to maintain with solid-state refrigeration systems, and you're not going to use refrigeration based on changing fluid pressures for materials strength reasons pointed out by another poster.
In summary, the real "real" problem is keeping the probe's electrical system working with the heating and material-structure changes that go on as the probe descends. Keeping the probe's physical structure intact is not a problem as long as you don't try to maintain voids of low pressure material.
Probes that can survive high temperatures and high pressures would have a number of interesting applications. Dropping one into Jupiter's deeper layers would be at least as much fun as dropping one into Earth's mantle, for instance.
Those who support Solar, Wind etc as electrical energy sources fail completely to understand the problem. We could save 1/2 the energy if we did not use electricity at all. If we were to directly burn the fuel or move the mills etc without electricity the energy would be more efficiently used.
Electricity is very easily stored, transmitted, and converted into other types of energy. Fuels are easily stored and moved around, but are not as easily converted; the best fuel-burning motor you can build will have an efficiency far lower than an electric motor. The only thing a fuel is good at producing is heat. Other forms of stored power are even worse.
You would likely try to argue that this efficiency in use is made up for by inefficiencies in electricity generation. For fuel-burning power plants, this is probably true. It is not true for things like hydroelectric or wind power; mechanical to electrical conversion is very efficient. It is not true for solar or nuclear, as the "wasted" energy is energy that could not be captured by any other means.
In short, your argument about efficiency appears to be based on incorrect assumptions.
The real issue in energy will be dealt with when we face up to what we are doing and adapt our designs and technology to the reality and not our imagined situation.
How?
You need light when it's dark out, so using sunlight when it's being produced doesn't work.
Most other cases of energy consumption occur far from the energy source, or similarly can't be used as they are produced.
Situations where you want to turn power into heat already use fossil fuels, which appear to be your favourite energy source.
How exactly would you set things up to use energy more efficiently?
It really gets cool in supernovae, because as much as 40% of a supernova's energy is in the form of neutrinos. I believe that this can be detected in theory, but I don't remember if it ever has been.
They have been.
The neutrino burst from Supernova 1987A was detected and found to coincide with the optical burst to within an hour (an hour before the optical burst, IIRC). This provided a direct demonstration that high-energy neutrinos travel at or extremely close to the speed of light, which in turn placed an upper limit on their mass (a very small value, but neutrino masses measured to date have been very small).
Another neat thing is that there may be a 4th neutrino that does not interact via the weak force. Imagine that!
No such neutrino exists, as far as anyone can tell. Neutrinos, leptons, and quarks are grouped into families. The first familiy - the up and down quarks, the electron, and the electron neutrino - are what normal matter is made of (or produces in nuclear reactions, in the case of the neutrino). The other two families contain much more massive particles, and so are only seen in exotic situations (high-energy collisions, and possibly as "strange matter" in neutron stars). The existence of higher-energy quark/lepton families has a measurable effect on lower-energy reactions (as the high energy flavours show up as virtual particles). All measurements to date indicate that there are only three families - the expected effects of higher families have not been seen.
Perhaps your source was confusing neutrino families with supersymmetric particles, which are strongly hinted to exist and which may qualify as weakly-interacting heavy particle candidates. None that I've heard of would have the properties you describe, however.
Intel's chips are simply horribly inefficient, which is why hyperthreading works.
This does not reflect the efficiency of Intel's chip - it reflects the fact that the most aggressively superscalar chip anyone could hope to build would still be lucky to execute even three instructions per clock on average.
Modern microprocessors are optimized to handle the peak load case - where control and data dependencies go on vacation and you have a string of instructions that you can burn through that also happen to perfectly match the number and type of functional units you have available.
In the real world, this rarely occurs. If you're very lucky, you'll have a tight loop that's reliably predicted operating on data that's mostly in L1, and keep _some_ of the functional units busy _most_ of the time.
The rest of the chip sits idle, waiting for the load profile to change.
This happens on *any* general-purpose processor. Intel, AMD, and everyone else suffer from it.
Symmetric Multi-Threading was proposed quite some time ago to cope with this. You'd have multiple instruction fetch units on the chip and run multiple processes. As these have totally independent instruction streams and mostly-independent workloads, they can be scheduled at the same time without conflict. This deluge of unrelated instructions lets the scheduler fill all slots and issue instructions to all functional units that *any* process has demand for. This drastically boosts utilization, with far fewer units being idle, at a cost of increased memory bandwidth load (mostly).
Intel got to market with this first, and called it "hyperthreading". IBM went the CMP route and didn't mess with SMT. Sun announced a future processor with both CMP and SMT features.
The problem occurs for everyone, and techniques to address it are well-known and being applied.
Claiming that Intel was forced to implement it due to inefficient processors is very silly.
Wind power is shaping up to be the best alternative. There was a recent article - maybe it was on/. but I don't remember - about the new turbines, and how more megawatts of wind generating capacity has been installed recently than all the installations in the last 50 years.
The problem with wind power is that energy density is low (requiring large plant areas) and the energy source is both very unreliable and very specific to location.
My best bet is on solar. For small-scale generation, thin-film photovoltaic cells will be very cheap and cost very little in terms of pollution to make (the main complaint about thick cells). For large-scale generation, you'd use concentrating mirrors to heat a working fluid in pipes or in a resovoir, and draw power off of that the same way you'd draw it from any other power-plant-scale heat source. Aluminum mirrors are cheap and durable, especially when they don't have to be optics-grade.
We'll see what actually happens in a few decades (when fabrication costs drop a bit more and when power storage via fuel cell gets cheap enough to install intermittent power systems like this more widely).
I assume you'll agree that in a 4-d universe (one that is otherwise similar to ours) that matter will collapse into some sort of bodies? Call them planets, stars, neutron stars, whatever. They collapse into some sort of body. Well there you go, you have a reasonably stable collection of matter.
By the arguments outlined in my previous message, any such collection would have a very disorganized internal structure, that would actively resist attempts at organizing it and rapidly degrade should any organization be magically imposed.
Postulating about clusters of such matter collections doesn't work either, as any such cluster would be unstable, resulting in either a merged blob of matter or the component blobs scattering out of interaction range.
Your arguments about emergent properties and patterns in our universe are a great illustration of how conducive our universe is to such organization. Universes with other dimensionalities seem to be less so.
stable orbits are not possible in 4-space or greater... So, assuming that is the case it is reasonable to assume that life would not arrise if solar systems could not even exist.
No, that is not "reasonable to assume". It just means earth-like life could not exist. If the laws of physics are radically different isn't it reasonable to expect the life would be radically different?
You can still set bounds on it.
The poster you are replying to understated the case - in universes with dimensionality other than three (greater or less than), stable orbits don't exist Stable orbits, be they planets about stars or electrons about atomic nuclei, require an inverse square force relation about charge or mass monopoles, which only happens in a universe with exactly three macroscopic dimensions.
It is reasonable to assume that life of *any* kind is a more or less stable, more or less orderly (structured) configuration of matter with finite size. This does not occur without inverse square force rules - anything else produces a highly chaotic system. Nothing structured enough to be a creature.
Before you mention energy beings, "energy" in the physics sense is a number used for bookkeeping - it's a quality of a system, not a "thing' in and of itself. Light is a type of matter - a collection of photons. Things like gravity or sound waves are disturbances - one in matter, and one in the fabric of space itself. None of the above are "energy" themselves (they _have_ energy, instead, just like everything else).
Before you object to the word "creature" and posit sentient stars or what-have-you, remember that I placed no bounds on size, other than requiring it to be finite. An object of infinite extent would have great difficulty communicating with itself (required for long-range structure, and arguably for things like thought as well).
In summary, you can make a pretty good argument for life being impossible in universes with more or less than three dimensions, regardless of the nature of the life in question.
I didn't get his calculations. He gives the number of protons that could fit into our observable universe ("Hubble space") and then calculates the permutations of present/absent for each to get the total number of possible configurations. But wouldn't there be more configurations than a present/absent calculation would account for? Such as variants arising from the momentum and quantum state of each proton?
And from the fact that the universe consists of more than just protons, yes. I think (or at least hope) it was intended to show the kind of calculation that needed to be done, as opposed to provide a useful number.
I think that the best way to figure out the number of system states is to try to measure the entropy of the volume of the universe that you're interested in. This tells you the amount of information that must be duplicated. You can place an upper bound on this by assuming an event horizon around that volume and figuring out the information content of the resulting black hole.
The May issue of Scientific American contains a much more in-depth article on parallel universes, which has enough points in common that it might have inspired the op-ed piece.
Teaser for the article is here. To get the whole thing, you either have to have a subscription or wait until next month.
The gist of it is as follows:
The first type of "parallel universe" is just another part of this universe. Because the universe appears to be infinitely large, any configuration of matter - be it Earth, our galaxy, or our entire currently-observable universe - must be duplicated somewhere out there. Back-of-the-envelope statistics math is used to figure out how far away (hint: really, really far).
In principle, these other "universes" can interact with our own, but in practice they're far enough away that it doesn't matter. Physical laws are likely similar.
Re. an infinite universe, the article states that a finite universe would leave artifacts in the cosmic microwave background that weren't seen.
The second type of parallel universe is other post-inflationary regions in the still-infationary space that holds our own universe. These universes would have different physical laws, and possibly different numbers of uncurled/macroscopic dimensions, some or all of which are set by symmetry breaking as the expanding universe cools.
These parallel universes are utterly unreachable, as the space between them a) exists in a different coordinate system that puts it in our past from our point of view, and b) is expanding exponentially quickly, dragging other universes away from ours at mind-boggling speed.
The third type of parallel universe scenario is the familiar "multiple histories" interpretation of quantum mechanics - the idea that all possible outcomes to any event occur, in their own universes.
As far as I understand it, interaction between these universes wouldn't be possible without violating some of the ground rules involved (the history tree could be thought of as a state transition diagram for all possible states of a closed system; if it's closed, it can't interact with anything else).
The fourth type of parallel universe discussed is, as far as I can tell, imaginary universes. The idea is to consider an arbitrary mathematical description of an object a spacetime diagram, and to consider the result of interpretation of this diagram to be a universe.
If you call this a "real" universe, then Everquest and the reality hosting the United Federation of Planets are also real universes. It depends on your point of view (and what you mean by "real" in this context).
The existance of "universes" of the first type is certain if the universe is infinite, from information theory arguments. The infinite or non-infinite nature of the universe is something that can be empirically tested (though the final test - waiting for every part of it to come within our observation horizon - is impractical).
The existance of the second type of universe hinges on the nature of the scalar fields proposed in the various inflationary models. In principle, this is testable, either by recreating the energies required or by observing distant parts of the universe that are undergoing inflation.
The existance of the third type of universe is not testable, due to the requirement for closed systems. So it's pretty much a moot point.
The existance of the fourth type of universe is a metaphysical question, whose answer depends on what you mean by "exist".
The full article has a lot of additional discussion, and pretty pictures. By all means pick up a copy, if the topic interests you.
The most I've seen ringworlds move is fixing orbits around their component star
The Ringworld is capable of moving by firing the defense laser perpendicular to the plane of the ring, moving the star by photon pressure and pulling the ring along. This was proposed by both Nessus and Teela Brown, if I recall correctly, as a means of leaving the galaxy if the ring material was deemed unsuitable for shielding against the Core explosion.
This was covered in "The Ringworld Engineers".
An alternative approach, discussed in one of Niven's non-fiction articles, is to use magnetic manipulation of the star to cause a plasma jet perpendicular to the plane, instead of using a photon drive. When you run out of star, you're at a speed that makes a Bussard-style ramscoop feasible, using the ring's (presumed) magnetic field. Solving the deflection problem is left as an exercise for the reader.
Another limitation is the large magnetic fields SMES (Superconducting Magnetic Electricity Storage) installations create, and not just from a technical point of view. There is almost no research into the effects of prolonged exposure to large magnetic fields on humans, but given the fuss about the 'emanations' from power lines (valid or not) this would probably become a big issue.
You could shield against this easily enough by using a coil geometry that doesn't have an external field (a toroidal coil is one example; the field exists only inside the donut).
This is really totally unworkable. VIA is trying to manufacture a cheap, cheap chip. Why would they want to mess with integrating radioactive material and detectors into their processor, when a simple overloaded transistor is just as random?
I had to explain to them that for any species to survive, IN THE WILD, there must be a population of sufficient size and more importantly sufficient genetic diversity. We can clone 1000 dodo's (insert politician joke here) but it will still only be ONE dodo.
I've heard about the 50/500 rule, but I still don't quite understand why having a starting population with identical genes is a death sentence.
As long as the original genes were good, none of the first generation will have crippling deficiencies. Yes, recessive traits will show up in the second generation, but as long as you breed individuals that don't have two copies of the bad gene (or better yet, select for individuals with two copies of the good one), this doesn't kill your population.
In the short term, you should always have enough normal individuals for the population to survive, and in the long term, mutation will build up genetic diversity again anyways. The 50/500 rule, AFAICT, is geared towards making sure there's a population big enough that healthy individuals do exist. Starting with that many distinct individuals seems like overkill.
What am I missing about this scenario? Genetics isn't my field of expertise.
Yeah, I agree that some modification is needed to fire protons from the CRTs instead of electrons. However, if you have two CRTs pointed at each other you can double the energy at the collision site and power duetrium-duetrium fusion.
My point is that power supplies are easy enough to build that I'm not sure why you'd do this, given the much lower reaction rate. A good place to search for high-voltage power supply designs is laser hobbyist sites.
When I finally get around to building it I'll post, I promise.
Make sure you use an account suitable for slashdotting:).
I don't know that the random number generator that they've described could ever be "just as random" as radioactive decay, but it looks like it can probably be made "good enough."
As long as neither system has unwanted noise sources, both are perfectly random. That matches my definition of "just as random":).
As for unwanted noise, both systems are suceptible to noise from many sources.
use one sample, two detectors, one on the top and another on the bottom, AFAIK what triggers one can't trigger the other, esp if the sample emits beta instead of gamma rays.
And that is a close variant the system that I proposed for radioisotope random number generation.
The original poster suggested counting the number of events that occurred within a predefined period and looking at the least significant bit.
Problems with the two-detector system with one sample are in making sure that both detectors are equally close to the sample (no variation in intermediate laters), and that they are treated in the same way electrically (which is difficult to guarantee, though you have the same problem - as mentioned in my post - with purely electrical RNGs).
A better system would be to use radioactive decay to generate random numbers. Very easy to implement using existeng technology, one of the few things that is completely random
Your proposed method would be slightly skewed, as the half-life of the material would give you an "expected" number of events in your sampling period, which would cause the result to lean towards either even or odd. The effect would be small, but present.
An alternative approach is to have two detectors, and see which one triggers first. While that method would have no systemic bias, removing intrinsic bias from differences in the samples would be difficult.
The system in the new C3 chip, though, is also completely random if they designed it well (i.e. amplified thermal noise and rejected other noise sources). You have biasing problems, as with any other system where matching is important, but these can be overcome. Noise injection from other parts of the system is the thing to watch out for here.
In summary, purely electrical random number generators can be just as random as your proposed scheme, and your proposed scheme is not significantly easier to implement.
Good luck getting your hands on tritium. Deuterium can be bought, or produced yourself with patience. Other reactions have very high threshold energies.
Note that this energy still isn't enough to penetrate the Coulomb barrier - it's the best tradeoff point between getting the particles close together and keeping them nearby long enough for there to be a reasonable chance of quantum tunnelling taking you through the barrier. So, most collisions will still just cause scattering.
Also note that any system involving a lot of scattering becomes Maxwellian (has a Maxwell-style temperature distribution). The fusor functions best in non-Maxwellian regimes. When the plasma thermalizes, it gets much colder due to the presence of cold ions (or cold, neutral molecules) from the source gas.
Heck, the cathode ray tubes in a TV pack enough punch to fuse to protons together, if you point one down the barrel of the other.
Regrettably, the electrons they are accelerating won't fuse.
You'd have to a) mod one to ionize and accelerate protons or deuterons (pretty extensive mod), and b) crank up the voltage. Typical CRT beam energy is on the order of 10 keV, which isn't enough to overcome the coulomb barrier, or even get close enough for tunnelling to allow fusion. 100 keV is about the minimum for that.
Most sensible approach to the apparatus is to fire a deuteron beam at a metal target that's absorbed deuterium. Colliding beams will give you a very low reaction rate (density is low, so collision rate will be low).
Be sure to post photos when you're done, and be sure to do this somewhere that won't lock you up for years for trying. Enjoy.
Slashdot Mirror
Except for a restricted set of cases, you can't prove that a given piece of code works or doesn't work. A truly exhaustive set of tests would be impractical to perform, and formal proofs of correctness place strong limits on the type of code you can write and the environment in which you can write it.
The result is that code is assumed correct when no bugs are found. This only means that there probably aren't _many_ bugs left. Thus, it may still crash (or have a security hole, or what-have-you).
Software has been complex for a long time. It just tends to be bigger now. A larger system has more opportunities for unexpected high-level interactions between components, but even a smaller system will have enough twists and turns that formulating a really good test suite, or checking the code by inspection, is very difficult. Bugs will be missed. As was discussed above, many of these missed bugs will slip through testing and reach the world.
As more effort is applied, you can get asymptotically closer to a bug-free system. However, this is far past the point of diminishing returns on the cost/benefit curve. For sufficiently constrained systems, you can even try proving it correct, but this tends to lead to cutting out a lot of functionality, speed, or both.
In situations where reliability must be had at any cost - aerospace control systems, vehicle control systems, medical equipment - the money will exist to produce near-perfect code, but even then there are bugs that occasionally bite. With commercial software, the buyer would rather have an application that crashes now and then than an application that costs ten times as much and comes out several years later.
Free and/or open software avoids some of this by staying in development longer, which allows more of the bugs to be caught, but even free and/or open software evolves. Every change brings new bugs to be squashed. As long as there are new types of software that we want, it isn't going to end.
The ultimate solution to the problem is to let computers write the software themselves. Give them a goal, set up evolutionary and genetic algorithms, and let them go at it on a supercomputer cluster for a few months.
Unfortunately, this only works if you can distinguish between buggy and non-buggy code produced by the algorithm. You can do tests, but no test suite will be exhaustive (otherwise we'd just use it on human-developed code to find the bugs).
Perfect software can only be produced if a formal proof of correctness is possible. Even then, you're limited by the assumptions the proof makes.
""Squishing" only occurs when there is a difference in pressure between two parts of the craft."
No, it doesn't. Water at depth is denser than water at the surface, just as your body tissue is denser at 1 atm than in a vacuum.
I am aware of this, and stated as much in the next paragraph of my message.
Volume does not change drastically - certainly not enough to destroy a probe. It's pressure _gradients_ and shear stresses that you have to worry about, as they are what will tear apart, crush, or otherwise mutilate the craft. A pressure gradient occurs if you try to keep a low-pressure cavity intact in the craft, and shear stress comes if you have two materials with different Young's modulus in the craft's structure. Avoid both, and you're fine.
We're not talking about 1000 atm, though. 10 m of sea water is about 1 atm, so you hit 1000 atm at only 10 km below the surface of the ocean. The earth's radius is ~6000 km, and the vast majority of it is orders of magnitude denser than water.
The "1000 atm" number was chosen for the example because it's about the point at which air would liquify at room temperature. Talking about air at a million atm would be silly, as it would no longer be a gas.
Your density numbers also seem to be way off. Far from "orders of magnitude", density's only about double what it is at earth's surface (density of the lower mantle's about 6 g/cm^3, compared to a 1 atm density of around 3 for olivine, and density of the core's about 10-13 g/cm^3, compared to a 1 atm density of about 8 for iron). Given a corresponding volume change for a given amount of material, this represents a change in dimensions of about 25%.
Solid matter is *difficult* to compress. Packing atoms much more tightly than their natural spacing means smaller electron wavelengths, which means much higher electron energies. That energy has to come from somewhere, which means you're going to have to put in a lot of work to get it to that compressed a state. With relatively small deformation, that means a lot of pressure to perform that amount of work.
Besides, I think the more immediate problem is "squeeze something, and it gets hotter."
You're thinking of gases, which have a greater volume change (heat buildup is related to the amount of work done compressing the gas). You're also thinking of adiabatic systems - where there is little time for energy exchange with the environment. When compression occurs slowly, heating is minimal, because the heat that does build up quickly leaks away into the environment.
The earth has a temperature gradient because there's a strong heat source in the core, not because of pressure. What the heat source is is up for debate (most think radioactive decay, some still think latent heat of crystallization), but as the heat energy deposited flows out towards the surface, you get a temperature gradient.
We have yet to be able to make a probe that can even survive the Venutian atmosphere for any length of time longer than a few hours. And this from the people who built supersonic fighters capable of taking off and landing in grass fields.
Venus's atmosphere is highly corrosive. The mantle is less so (though still pretty reactive at those temperatures). The atmosphere of Jupiter is much less so. This presents a bigger barrier to long-term survival that pressure, and is my first guess at why the Venus landers stopped functioning.
I'd be more worried about the device being able to withstand that pressure. I fail to see how surrounding the probe with molten iron (or any other fluid, for that matter) will prevent the weight of the planet from squishing it like a bug. Or does he plan on violating the laws of physics at the same time?
"Squishing" only occurs when there is a difference in pressure between two parts of the craft. If the craft is all at one pressure, it feels nothing.
The real problems you have to worry about are volume and crystal structure changes with pressure. Air at 1000 atmospheres has a very different volume than air at 1 atmosphere. Iron at 1000 atmospheres, on the other hand, has a not-too-different volume. Build your craft out of materials that react similarly to pressure and make it tolerant of volume changes, and no "squishing" occurs.
Crystal structure changes are a tougher problem. Squeeze carbon, and it becomes diamond. Squeeze other materials, and they go through their own allotrope phases. Mechanical properties will change, and more importantly, electrical properties in semiconductors will also change.
There's also the small matter of keeping semiconductors working above the melting point of iron, but that may be an attainable goal (certainly not with silicon, but IIRC diamond's temperature limits were far, far higher, and other high-temperature semiconductors doubtless exist). Refrigeration is not an option - that big a heat gradient will be next to impossible to maintain with solid-state refrigeration systems, and you're not going to use refrigeration based on changing fluid pressures for materials strength reasons pointed out by another poster.
In summary, the real "real" problem is keeping the probe's electrical system working with the heating and material-structure changes that go on as the probe descends. Keeping the probe's physical structure intact is not a problem as long as you don't try to maintain voids of low pressure material.
Probes that can survive high temperatures and high pressures would have a number of interesting applications. Dropping one into Jupiter's deeper layers would be at least as much fun as dropping one into Earth's mantle, for instance.
Those who support Solar, Wind etc as electrical energy sources fail completely to understand the problem. We could save 1/2 the energy if we did not use electricity at all. If we were to directly burn the fuel or move the mills etc without electricity the energy would be more efficiently used.
Electricity is very easily stored, transmitted, and converted into other types of energy. Fuels are easily stored and moved around, but are not as easily converted; the best fuel-burning motor you can build will have an efficiency far lower than an electric motor. The only thing a fuel is good at producing is heat. Other forms of stored power are even worse.
You would likely try to argue that this efficiency in use is made up for by inefficiencies in electricity generation. For fuel-burning power plants, this is probably true. It is not true for things like hydroelectric or wind power; mechanical to electrical conversion is very efficient. It is not true for solar or nuclear, as the "wasted" energy is energy that could not be captured by any other means.
In short, your argument about efficiency appears to be based on incorrect assumptions.
The real issue in energy will be dealt with when we face up to what we are doing and adapt our designs and technology to the reality and not our imagined situation.
How?
You need light when it's dark out, so using sunlight when it's being produced doesn't work.
Most other cases of energy consumption occur far from the energy source, or similarly can't be used as they are produced.
Situations where you want to turn power into heat already use fossil fuels, which appear to be your favourite energy source.
How exactly would you set things up to use energy more efficiently?
It really gets cool in supernovae, because as much as 40% of a supernova's energy is in the form of neutrinos. I believe that this can be detected in theory, but I don't remember if it ever has been.
They have been.
The neutrino burst from Supernova 1987A was detected and found to coincide with the optical burst to within an hour (an hour before the optical burst, IIRC). This provided a direct demonstration that high-energy neutrinos travel at or extremely close to the speed of light, which in turn placed an upper limit on their mass (a very small value, but neutrino masses measured to date have been very small).
Another neat thing is that there may be a 4th neutrino that does not interact via the weak force. Imagine that!
No such neutrino exists, as far as anyone can tell. Neutrinos, leptons, and quarks are grouped into families. The first familiy - the up and down quarks, the electron, and the electron neutrino - are what normal matter is made of (or produces in nuclear reactions, in the case of the neutrino). The other two families contain much more massive particles, and so are only seen in exotic situations (high-energy collisions, and possibly as "strange matter" in neutron stars). The existence of higher-energy quark/lepton families has a measurable effect on lower-energy reactions (as the high energy flavours show up as virtual particles). All measurements to date indicate that there are only three families - the expected effects of higher families have not been seen.
Perhaps your source was confusing neutrino families with supersymmetric particles, which are strongly hinted to exist and which may qualify as weakly-interacting heavy particle candidates. None that I've heard of would have the properties you describe, however.
Intel's chips are simply horribly inefficient, which is why hyperthreading works.
This does not reflect the efficiency of Intel's chip - it reflects the fact that the most aggressively superscalar chip anyone could hope to build would still be lucky to execute even three instructions per clock on average.
Modern microprocessors are optimized to handle the peak load case - where control and data dependencies go on vacation and you have a string of instructions that you can burn through that also happen to perfectly match the number and type of functional units you have available.
In the real world, this rarely occurs. If you're very lucky, you'll have a tight loop that's reliably predicted operating on data that's mostly in L1, and keep _some_ of the functional units busy _most_ of the time.
The rest of the chip sits idle, waiting for the load profile to change.
This happens on *any* general-purpose processor. Intel, AMD, and everyone else suffer from it.
Symmetric Multi-Threading was proposed quite some time ago to cope with this. You'd have multiple instruction fetch units on the chip and run multiple processes. As these have totally independent instruction streams and mostly-independent workloads, they can be scheduled at the same time without conflict. This deluge of unrelated instructions lets the scheduler fill all slots and issue instructions to all functional units that *any* process has demand for. This drastically boosts utilization, with far fewer units being idle, at a cost of increased memory bandwidth load (mostly).
Intel got to market with this first, and called it "hyperthreading". IBM went the CMP route and didn't mess with SMT. Sun announced a future processor with both CMP and SMT features.
The problem occurs for everyone, and techniques to address it are well-known and being applied.
Claiming that Intel was forced to implement it due to inefficient processors is very silly.
Wind power is shaping up to be the best alternative. There was a recent article - maybe it was on /. but I don't remember - about the new turbines, and how more megawatts of wind generating capacity has been installed recently than all the installations in the last 50 years.
The problem with wind power is that energy density is low (requiring large plant areas) and the energy source is both very unreliable and very specific to location.
My best bet is on solar. For small-scale generation, thin-film photovoltaic cells will be very cheap and cost very little in terms of pollution to make (the main complaint about thick cells). For large-scale generation, you'd use concentrating mirrors to heat a working fluid in pipes or in a resovoir, and draw power off of that the same way you'd draw it from any other power-plant-scale heat source. Aluminum mirrors are cheap and durable, especially when they don't have to be optics-grade.
We'll see what actually happens in a few decades (when fabrication costs drop a bit more and when power storage via fuel cell gets cheap enough to install intermittent power systems like this more widely).
Thanks for posting an intelligent response, btw. You probably already know how rare this kind of thing is :).
I assume you'll agree that in a 4-d universe (one that is otherwise similar to ours) that matter will collapse into some sort of bodies? Call them planets, stars, neutron stars, whatever. They collapse into some sort of body. Well there you go, you have a reasonably stable collection of matter.
By the arguments outlined in my previous message, any such collection would have a very disorganized internal structure, that would actively resist attempts at organizing it and rapidly degrade should any organization be magically imposed.
Postulating about clusters of such matter collections doesn't work either, as any such cluster would be unstable, resulting in either a merged blob of matter or the component blobs scattering out of interaction range.
Your arguments about emergent properties and patterns in our universe are a great illustration of how conducive our universe is to such organization. Universes with other dimensionalities seem to be less so.
stable orbits are not possible in 4-space or greater... So, assuming that is the case it is reasonable to assume that life would not arrise if solar systems could not even exist.
No, that is not "reasonable to assume". It just means earth-like life could not exist. If the laws of physics are radically different isn't it reasonable to expect the life would be radically different?
You can still set bounds on it.
The poster you are replying to understated the case - in universes with dimensionality other than three (greater or less than), stable orbits don't exist Stable orbits, be they planets about stars or electrons about atomic nuclei, require an inverse square force relation about charge or mass monopoles, which only happens in a universe with exactly three macroscopic dimensions.
It is reasonable to assume that life of *any* kind is a more or less stable, more or less orderly (structured) configuration of matter with finite size. This does not occur without inverse square force rules - anything else produces a highly chaotic system. Nothing structured enough to be a creature.
Before you mention energy beings, "energy" in the physics sense is a number used for bookkeeping - it's a quality of a system, not a "thing' in and of itself. Light is a type of matter - a collection of photons. Things like gravity or sound waves are disturbances - one in matter, and one in the fabric of space itself. None of the above are "energy" themselves (they _have_ energy, instead, just like everything else).
Before you object to the word "creature" and posit sentient stars or what-have-you, remember that I placed no bounds on size, other than requiring it to be finite. An object of infinite extent would have great difficulty communicating with itself (required for long-range structure, and arguably for things like thought as well).
In summary, you can make a pretty good argument for life being impossible in universes with more or less than three dimensions, regardless of the nature of the life in question.
I didn't get his calculations. He gives the number of protons that could fit into our observable universe ("Hubble space") and then calculates the permutations of present/absent for each to get the total number of possible configurations. But wouldn't there be more configurations than a present/absent calculation would account for? Such as variants arising from the momentum and quantum state of each proton?
And from the fact that the universe consists of more than just protons, yes. I think (or at least hope) it was intended to show the kind of calculation that needed to be done, as opposed to provide a useful number.
I think that the best way to figure out the number of system states is to try to measure the entropy of the volume of the universe that you're interested in. This tells you the amount of information that must be duplicated. You can place an upper bound on this by assuming an event horizon around that volume and figuring out the information content of the resulting black hole.
The May issue of Scientific American contains a much more in-depth article on parallel universes, which has enough points in common that it might have inspired the op-ed piece.
Teaser for the article is here. To get the whole thing, you either have to have a subscription or wait until next month.
The gist of it is as follows:
In principle, these other "universes" can interact with our own, but in practice they're far enough away that it doesn't matter. Physical laws are likely similar.
Re. an infinite universe, the article states that a finite universe would leave artifacts in the cosmic microwave background that weren't seen.
These parallel universes are utterly unreachable, as the space between them a) exists in a different coordinate system that puts it in our past from our point of view, and b) is expanding exponentially quickly, dragging other universes away from ours at mind-boggling speed.
As far as I understand it, interaction between these universes wouldn't be possible without violating some of the ground rules involved (the history tree could be thought of as a state transition diagram for all possible states of a closed system; if it's closed, it can't interact with anything else).
If you call this a "real" universe, then Everquest and the reality hosting the United Federation of Planets are also real universes. It depends on your point of view (and what you mean by "real" in this context).
The existance of "universes" of the first type is certain if the universe is infinite, from information theory arguments. The infinite or non-infinite nature of the universe is something that can be empirically tested (though the final test - waiting for every part of it to come within our observation horizon - is impractical).
The existance of the second type of universe hinges on the nature of the scalar fields proposed in the various inflationary models. In principle, this is testable, either by recreating the energies required or by observing distant parts of the universe that are undergoing inflation.
The existance of the third type of universe is not testable, due to the requirement for closed systems. So it's pretty much a moot point.
The existance of the fourth type of universe is a metaphysical question, whose answer depends on what you mean by "exist".
The full article has a lot of additional discussion, and pretty pictures. By all means pick up a copy, if the topic interests you.
This was proposed by both Nessus and Teela Brown, if I recall correctly
Hindmost, not Nessus; my mistake.
The most I've seen ringworlds move is fixing orbits around their component star
The Ringworld is capable of moving by firing the defense laser perpendicular to the plane of the ring, moving the star by photon pressure and pulling the ring along. This was proposed by both Nessus and Teela Brown, if I recall correctly, as a means of leaving the galaxy if the ring material was deemed unsuitable for shielding against the Core explosion.
This was covered in "The Ringworld Engineers".
An alternative approach, discussed in one of Niven's non-fiction articles, is to use magnetic manipulation of the star to cause a plasma jet perpendicular to the plane, instead of using a photon drive. When you run out of star, you're at a speed that makes a Bussard-style ramscoop feasible, using the ring's (presumed) magnetic field. Solving the deflection problem is left as an exercise for the reader.
Another limitation is the large magnetic fields SMES (Superconducting Magnetic Electricity Storage) installations create, and not just from a technical point of view. There is almost no research into the effects of prolonged exposure to large magnetic fields on humans, but given the fuss about the 'emanations' from power lines (valid or not) this would probably become a big issue.
You could shield against this easily enough by using a coil geometry that doesn't have an external field (a toroidal coil is one example; the field exists only inside the donut).
This is really totally unworkable. VIA is trying to manufacture a cheap, cheap chip. Why would they want to mess with integrating radioactive material and detectors into their processor, when a simple overloaded transistor is just as random?
I believe this was my _point_.
Grouse at the parent poster, not me.
I had to explain to them that for any species to survive, IN THE WILD, there must be a population of sufficient size and more importantly sufficient genetic diversity. We can clone 1000 dodo's (insert politician joke here) but it will still only be ONE dodo.
I've heard about the 50/500 rule, but I still don't quite understand why having a starting population with identical genes is a death sentence.
As long as the original genes were good, none of the first generation will have crippling deficiencies. Yes, recessive traits will show up in the second generation, but as long as you breed individuals that don't have two copies of the bad gene (or better yet, select for individuals with two copies of the good one), this doesn't kill your population.
In the short term, you should always have enough normal individuals for the population to survive, and in the long term, mutation will build up genetic diversity again anyways. The 50/500 rule, AFAICT, is geared towards making sure there's a population big enough that healthy individuals do exist. Starting with that many distinct individuals seems like overkill.
What am I missing about this scenario? Genetics isn't my field of expertise.
Yeah, I agree that some modification is needed to fire protons from the CRTs instead of electrons. However, if you have two CRTs pointed at each other you can double the energy at the collision site and power duetrium-duetrium fusion.
:).
My point is that power supplies are easy enough to build that I'm not sure why you'd do this, given the much lower reaction rate. A good place to search for high-voltage power supply designs is laser hobbyist sites.
When I finally get around to building it I'll post, I promise.
Make sure you use an account suitable for slashdotting
I don't know that the random number generator that they've described could ever be "just as random" as radioactive decay, but it looks like it can probably be made "good enough."
:).
As long as neither system has unwanted noise sources, both are perfectly random. That matches my definition of "just as random"
As for unwanted noise, both systems are suceptible to noise from many sources.
intrinsic bias?
use one sample, two detectors, one on the top and another on the bottom, AFAIK what triggers one can't trigger the other, esp if the sample emits beta instead of gamma rays.
And that is a close variant the system that I proposed for radioisotope random number generation.
The original poster suggested counting the number of events that occurred within a predefined period and looking at the least significant bit.
Problems with the two-detector system with one sample are in making sure that both detectors are equally close to the sample (no variation in intermediate laters), and that they are treated in the same way electrically (which is difficult to guarantee, though you have the same problem - as mentioned in my post - with purely electrical RNGs).
A better system would be to use radioactive decay to generate random numbers. Very easy to implement using existeng technology, one of the few things that is completely random
Your proposed method would be slightly skewed, as the half-life of the material would give you an "expected" number of events in your sampling period, which would cause the result to lean towards either even or odd. The effect would be small, but present.
An alternative approach is to have two detectors, and see which one triggers first. While that method would have no systemic bias, removing intrinsic bias from differences in the samples would be difficult.
The system in the new C3 chip, though, is also completely random if they designed it well (i.e. amplified thermal noise and rejected other noise sources). You have biasing problems, as with any other system where matching is important, but these can be overcome. Noise injection from other parts of the system is the thing to watch out for here.
In summary, purely electrical random number generators can be just as random as your proposed scheme, and your proposed scheme is not significantly easier to implement.
Note that several varieties of digital watch (e.g. Timex IndiGlo) contain tritium.
n ous_dials.htm
Not as far as I can tell - they use electroluminescent chemicals.
Radioactively-driven phosphor watches went out of vogue about the time Radium did.
For a good source of information about this, check http://www.watchprince.com/Rolexreport/rolex_lumi
From http://home.earthlink.net/~jimlux/nuc/reactions.h
Good luck getting your hands on tritium. Deuterium can be bought, or produced yourself with patience. Other reactions have very high threshold energies.
Note that this energy still isn't enough to penetrate the Coulomb barrier - it's the best tradeoff point between getting the particles close together and keeping them nearby long enough for there to be a reasonable chance of quantum tunnelling taking you through the barrier. So, most collisions will still just cause scattering.
Also note that any system involving a lot of scattering becomes Maxwellian (has a Maxwell-style temperature distribution). The fusor functions best in non-Maxwellian regimes. When the plasma thermalizes, it gets much colder due to the presence of cold ions (or cold, neutral molecules) from the source gas.
Heck, the cathode ray tubes in a TV pack enough punch to fuse to protons together, if you point one down the barrel of the other.
Regrettably, the electrons they are accelerating won't fuse.
You'd have to a) mod one to ionize and accelerate protons or deuterons (pretty extensive mod), and b) crank up the voltage. Typical CRT beam energy is on the order of 10 keV, which isn't enough to overcome the coulomb barrier, or even get close enough for tunnelling to allow fusion. 100 keV is about the minimum for that.
Most sensible approach to the apparatus is to fire a deuteron beam at a metal target that's absorbed deuterium. Colliding beams will give you a very low reaction rate (density is low, so collision rate will be low).
Be sure to post photos when you're done, and be sure to do this somewhere that won't lock you up for years for trying. Enjoy.