What happens when you *cut* a cable that is under tension?
The cable is only under tension because there is load upon it. Jettison the load - OK, so a satellite may be lost. Big deal. At that point the cable is no longer under tension, and it begins to settle. Spooling out a few km of cable from the top will recover the cable, though moved significantly.
Again, this type of "annoyance" is not beyond them now. They could easily crash a plane into a launch pad. They don't, however, because the impact is minimal. Ditto here.
Further, what if they *hijack* the elevator.
If terrorists have the capability to hijack any installation they wish, there are far better targets than a space elevator anchor.
Say, a nuclear munitions facility. Or even just a nuclear power plant. Hell, even just the Golden Gate Bridge would be a better target.
First, the notion that the tether from the moon to the top of the elevator would seems a little shaky. If I understand correctly the cable can provide tension but is not rigid. As such, if it were strong enough (probably not unreasonable) it could prevent the elevator from drifting to the Earth side of L1. But L1 is unstable in the Lunar direction too, and the cable's tension would do nothing to prevent it from drifting to the Lunar side of L1; so, the cable alone couldn't keep the elevator centered on L1.
If the elevator starts falling (moves towards the Moon), it starts to lean to the east. To keep it stable, you pull to the west at the anchor point, and it doesn't move.
GEO isn't stable either - it's not a minimum in the potential at all. But the competing accelerations keep the cable taut. It's exactly the same in this manner.
You need thrusters to remain at GEO, too - obviously any slight acceleration pushes you away from it. But a space elevator doesn't need thrusters to sit at GEO, because slight perturbations cause fluctuations in the tension (which are compensated at the anchor point), not in the orbit.
The dynamics of an L1 elevator are pretty much identical to that of a GEO elevator. It's just simple orbital mechanics.
I shouldn't've mentioned that the tether "cancels the radial instability" - that's unimportant, as any orbital perturbations get translated into tension perturbations.
The interesting part is that since unpowered orbital paths naturally lead away from L1, debris will have a harder time hitting that portion of the elevator. The other portions will still be vulnerable, though.
and I'm not sure if the perpendicular direction is unstable as well.
Sigh, I could've tried reading the All Knowing Wikipedia : L1 is stable in the perpendicular direction, so an L1 elevator would be completely stable. Interesting!
It also orbits the Earth once every thirty days - in other words, it's locked in a 1:1 resonance with the Earth. This is of great benefit in this case!
A little background:
Space elevators need to remain stationary with respect to the body that they'll be attached to. With the Earth, that means you need to be rotating at the same speed as the Earth - in other words, GEO.
With the Moon, however, you can either be rotating as fast as the Moon, or orbiting as fast as the moon, because the moon rotates at the same angular speed as it orbits! And there are 5 places where that occurs - the Lagrange points.
From any one of those points, the Earth and the Moon are stationary in the sky - that is, you don't see either of them moving with respect to each other. Since the Moon's rotation is defined by its orientation with respect to the Earth, therefore, you don't see the Moon rotating. That is, you're in something that's exactly the same as GEO.
Of the 5 Lagrange points, obviously L3 is silly - it's on the opposite side of the Earth as the Moon. So that won't work.
L2 is similarly silly - it's on the opposite side of the Moon as the Earth. Could be useful for sending things to interstellar space, but not for Earth-Moon transits.
So you're left with L1, L4, and L5. Obviously if you're talking about getting things from the Earth to the Moon, you'd want the one that's deepest into Earth's gravity well - and that's L1, "gravitationally halfway" between the Earth and the Moon. And that's what's being proposed.
One helpful thing is that L1 is unstable - orbits tend to drift away from there. However, an elevator tethered to the moon at least anchors one of the unstable directions (radially towards the moon/away from the moon), and I'm not sure if the perpendicular direction is unstable as well. So it may be that an elevator in that position is stable, and that unpowered objects will tend to move away from the elevator. You'd have a natural deflection mechanism. Pretty interesting!
Actually, a combination of an L1 and an L2 elevator could be quite interesting, though you'd have to build something like a railway around the Moon. Once you do that, though, you could go out past the L2 point, and you could sling yourself into interplanetary space. I'd have to work out how much of a boost you could get, but Mars orbit seems quite reasonable.
It's not as good as a terrestrial elevator (because the Earth rotates so quickly, so you can steal more of the Earth's angular momentum), but it's certainly currently feasible.
All modern designs must address not just the engineering (is the cable strong enough) and the economics (is moon dust worth the cost of the elevator) but also acts of terrorism.
A terrestrial cable would be 100,000 km long.
Planes fly at about 10 km.
SpaceShipOne just climbed to 100 km.
Anything a terrorist does to a cable will be done to less than 0.01% of the cable.
Get a clue - terrorists can't do anything more than annoy a space elevator. Anything they can do is recoverable.
Clockmakers have used such a material for a long time; it's a complex alloy called invar.
While invar is useful, it just has a small (and controllable, through different alloying) coefficient. It's used many times to put metal through glass becaue they tune it to have the same thermal expansion as the glass. Therefore you can have a gas-tight glass container with metal coming out of it (used mostly for phototubes).
The "controllable" nature is the most important - in most situations you just want a matched thermal expansion.
There are probably instances where you want to create an object with zero thermal expansion. This stuff will allow you to do it, presumedly over some range (which is all you care about, really).
the one where the 4000 Bush votes happened being more an exception than a rule.
Franklin County isn't using "e-voting" machines. They're electronic, but it's stuff from 1992 that's likely implemented in bare logic. No touchscreens, databases, or any of that. The votes are recorded simultaneously on 8 cartridges, one of which is removable to be taken to the counting area.
The cartridge apparently read out wrong on Election Day. Later, when it was checked, it was fine. Hmm. I wouldn't suspect the machine - I would suspect the person who put the cartridge in. It's hard to imagine that this couldn't also have been done with paper ballots with the people who actually read off the values or record the ballots.
# Evoting is a factor that "swings" voters toward (or away from) Bush # Evoting is a factor that influences turn-out # Evoting is a factor that influences change in turn-out w.r.t. the 2000 election # Franklin county's change in relative turnout is unusual
Franklin County wasn't an "e-voting" district. They're machines from 1992. They're electronic, but not computerized. And distinctly not from Diebold.
They're not walking out on the entire process. They're walking out on the Kyoto Protocol.
China will be putting out just as much CO2 as the U.S. by 2013 and then by 2050 will have put out a greater cumulative total.
I'd believe that this was the true concern if the US was currently pushing for a new treaty. Given the fact that Kyoto will have taken 8 years from start to enactment, from your numbers, it seems like the US should be pushing for a new treaty now. Except... they aren't.
If everyone agrees to the Protocol, and we meet the 5% mark, CO2 emissions will be up 70% by 2050.
As with population growth predictions, believing that China will continue to increase its CO2 emissions at the same rate it is now is quite silly. They won't. And if you expect that they will, then Bush should be pushing even harder for a new treaty with China limiting their CO2 emissions immediately.
Incidentally, if Kyoto didn't even exist, then wouldn't CO2 emissions will be up even more by 2050? Like - almost 100%?
So... if China is such a threat, where's the international concern?
Oh, yeah - because the rest of the world is trying to convince the current largest emitter of CO2 to cut back. Not the possible future largest emitter.
Malaysia could become the largest emitter of CO2 by 2050. Who knows. That was 50 years away at the time.
Assuming it takes 1000 joules to raise the temperature of one kilogram of air by one kelvin, the the energy is 5 × 10^21 joules.
That's a tiny amount of energy, on a global scale. Let's just play with numbers a bit.
If the Earth was a simple radiator with an albedo of 0.3, its temperature would be -17.6 degrees C.
Obviously, it isn't a simple radiator - that's due to CO2, water vapor, methane, and other greenhouse gasses.
It currently has an average temperature of about 14 deg C. That means that CO2, methane, and water vapor have stored 800,000 nuclear bombs worth of energy in our atmosphere.
To make it even more fun: CO2 is 350 ppm of the Earth's atmosphere, and water vapor is something like 10,000 ppm. Methane is like 2 ppm, so we'll ignore it. Let's just take the stupid and wrong assumption that water and CO2 are equal greenhouse gasses, and the effect's linear (also wrong).
CO2 is currently rising at 2 ppm/year, and already has risen ~50-60 ppm. That means (with this stupid and wrong model) we're adding the equivalent of 160 nuclear bombs each year. And we've already detonated about 5000. So we'll raise the temperature of the Earth by a degree in about 30 years.
Thankfully the physics behind all of this is slightly more complicated (the Earth isn't a blackbody, for one). Especially when you consider that in 1995, the world expended 3.33 x 10^20 joules.
Or 1500 Tsar Bombas each year.
If the Earth didn't reradiate, we'd raise the temperature of the Earth by 1 degree in about 15 years.
And if everyone on Earth was American, we'd be expending enough energy to raise the temperature of the Earth by 1 degree every other year.
If the earth did not radiate any energy to space, the surface would equalize with the sun's surface within a few years at 5000 degrees.
Um... no.
If the Earth's albedo is 0, then it would radiate all of the energy that impinges on it: which is 1380 W/m^2, the solar constant.
The Earth radiates this over 4 times the area that it receives it over (surface area of a sphere: 4 pi r^2: cross-sectional area of a sphere - pi r^2) so the Earth would radiate 345 W/m^2.
Google says sqrt(sqrt((1380 W/m^2) divided by 4 divided by the Stefan Boltzmann constant)) is 6 degrees C.
It should also be noted that the Earth is not a blackbody radiator, and so the albedo is not the deciding factor for the Earth's temperature. Earth's average albedo is 0.30, so its blackbody temperature would be sqrt(sqrt((1380 W/m^2) times 0.7 divided by 4 divided by the Stefan Boltzmann constant)), which Google says is -17.6 C.
Clearly other processes are at work maintaining Earth's global average temperature of 14 C - almost 32 degrees above its blackbody temperature.
Interestingly, this implies that CO2 and water vapor (and other greenhouse gasses) are collectively responsible for trapping almost 1 million of the atomic bombs that the grandparent poster mentioned.
The Kyoto accord only applies to developed countries, so enforcing it around the world drives industry into developing countries, such as China and India.
So basically, what you're saying is that only the US should have been allowed to dump arbitrarily large amounts of CO2 into the atmosphere while it was developing?
Give me a break. China is expected to be an Annex I country before the end of the decade, and if the US really wanted to complain about it, they could put pressure on that rather than walking out on the entire process.
My hunch is that there is significant overlap between them; thus wireless attitudes are implicitly contained within the poll results and there will not be a signficant difference come tomorrow...
That's assuming you poll 18-24 year olds with phone polling. In general, you don't. You significantly underrepresent the 18-24 bracket, which means you need to post-weight the result correctly so you represent the electorate right. What would be ideal is if they could find a way to poll the situation more equally, so that there'd be less of a systematic error introduced in your poll.
This shows (with high statistics!) that there most likely is a clear preference (about 10 points!) in that age bracket, which means if you don't get the correction right, you'll introduce a systematic bias in all your other polls. Considering you basically can't get the correction right (you'd need to know precisely what the voter turnout will be), you'd much prefer to make sure your sample is less skewed, so the effect of the (known) bias is lessened.
So, it basically comes down to: if young voters turn out like they used to, it'll look exactly like the polls say. However, if young voters turn out higher than they have previously, then the polls are all biased low, and Kerry will win many states by larger margins than expected.
This is what a statistical vs. a systematic error gets you. Most people are convinced that the election will be dead on close because the polls have constantly said "they're close", "they're close", "they're close". Sure they only have a 4% margin of error, but many polls should resolve it, right? Nope - systematics do not clear up with more polls (they just become more obvious).
So the main thing that this tells you is that people should expect that the polls really may not be accurate, and you may see non-"battleground states" decide this election as well.
Look at the respondant distribution - it isn't representative of the voting distribution, nor the population distribution. There's a heavy bias towards older voters (they're home more often) and female voters (ditto).
To correct it, they need to post-weight the results. They use the previous election's results (or a likelihood distribution and the registered voter distribution) to determine the weighting. Both of those methods are imperfect: therefore they generate systematic errors, which they don't quote.
And I'm not sure if you understood that right - randomly calling cellphones is illegal in most states that I know of, including both my current state and the state that my cellphone is from.
You may be right, you may be wrong. One thing people tend to overlook is that we DO NOT have a national election in America, we have 50 state elections on the same day. Most polls are national polls, which as last year shows us, mean pretty much nothing. Just because you have a majority in the national poll doesn't mean squat - I'm much more interested in polls done in "swing states."
His argument applies for swing states, as well (it's a dead heat in all of those states, and thankfully, those states have a large young population as well). The polls are likely to be wrong in that case, as well. Naively, given the estimated direction of the bias, it bodes very well for Kerry in the electoral result.
The polls are bunk - the worst thing about them is that people don't realize that the systematic component is likely to be comparable to the statistical, especially with such a tiny sample size. And systematic components do not improve with the number of polls you do, so even though, for instance, there have been 25 polls in Florida, and if you average them, he has a tiny lead. However, systematic errors don't improve with averaging (statistical errors do). So, you know nothing more.
Anyway, I thought the basic assumption underlying all random sampling is that the sample is, well, random.
Two things:
1) You don't get a random selection of people via any sampling method. Calling people at home, for instance, heavily biases you towards older populations, as they tend to be home more, and hang up less. So you have to correct for those biases, because you can't do a perfect sampling. That correction is called "post-weighting", because it happens after the poll.
2) You aren't trying to get a random sample of people, you're trying to get a random sample of people who will vote on Election Day. The only time you will know what the demographic fraction of people who vote are is after the vote happens. To compensate this, they either try to determine voter likelihood, and then postweight by this likelihood fraction, or they base it on previous year's results. They do not quote the systematic error that results from this, though, which is terrifically bad.
So yah, the polls are bunk. Why do you think all the polls from last election were not spread around the final result - 8/10 predicted Gore? Not enough data points, but it still is a pretty significant bias.
It is true that the polls can get it very wrong, but you forget that it is the pollsters JOB to get it right and so they have an incentive to design a method that will capture the right answer.
Not if you think about it. Most politicians don't know statistics - if they did, they'd be asking for the systematic error as well as the statistical. It's in the pollster's best interests to have polls that jump all over the place, and to have as many biases as possible. That way the politicians will be encouraged to buy more polls.
Here's a simple example:
Suppose you know the distribution of male/female people in your country: let's say it's 50/50, perfectly. Also assume that there are 10,000 people in your country.
Now let's assume you go out and get a random sample of the people in the country - you want to find out who likes A, and who likes B. You get 600 women, and 400 men. 100% of the women like A, and 100% of the men like B.
Naively, you'd conclude that A is favored 60% to 40%, with a sampling error of 2.9%. So a 50/50 split is excluded at 6 sigma, which is huge.
The problem is that something apparently was wrong with your polling method, because the distribution of women and men doesn't match the distribution of people. So you post-weight your results to have the sample accurately represent the country. So that sample of 600 is weighted by 4/6 (2/3), so you then get a 50%/50% split again. You have to do more work to calculate the error properly, so I won't do it here.
The problem comes when you don't know how to post-weight the results properly: suppose that you polled to get that 50% male/50% female number, and the poll was accurate +/-10%? Then your post-weighting becomes 0.66+/-0.44, which means that your poll now has an error term which is huge - your result is now 50%/50% with a 2.94% statistical error and a 10% systematic error (actually, it's more like +10%/-6.6%, but I'm not taking the time to work it out fully). And the problem is that this error term is systematic and not statistical, which means more polls don't help you at all - all those polls will be wrong in exactly the same way.
This gets worse for election polls. The relative percentage of men and women is deterministic: there is an absolute answer, and it is determinable. But election polls don't want a sample of people, they want a sample of voters who will actually vote. So they need some way of figuring out who is going to vote - there is, of course, no way to do that until after the election has happened.
Yes, polling companies post-weight their results. Yes, some post-weight it based on previous years' election data. Some do even worse - they determine the likelihood that a person is going to vote via an independent poll, and use that result (coupled with the registered voter demographic) to determine a "likelihood estimate" for voters at a certain age bracket, and use that to post-weight their data.
There is no way to do this right. You can only guess at the distribution of who will vote, and that guess contains a systematic error. That none of the polling companies tell you about.
So no, the pollsters do not "get it right". They can't. They don't even do a good job - if they did, they'd show systematic errors as well. Then you could possibly compare multiple polls and get a feel for what it truly might be, as well.
Some people consider local polls more accurate than state polls conducted by national entities.
The main problem with polls is the fact that they only quote statistical errors, when the people who actually perform the analysis know there are systematic errors as well.
Systematic errors don't go away when you do multiple polls, so averaging tricks don't work, and even the "well, my eye kinda tells me that Bush is ahead" don't work either. A 2% lead in a state with a 5% systematic error is not a lead at all, not even if you see it steady for an entire year.
Most polls are post-weighted for skewed age distribution (most respondents are old) and skewed geographical distribution. This post-weighting, though, depends on information either from the last election (which, honestly, has absolutely no reason to be correct again this election) or polls which necessarily contain error terms as well. Thus, the corrections introduce a systematic, which simply has to be ~ the statistical.
Statistics geek talk aside, don't ever think your state is a throwaway due to polls. Polls lie. You don't know who's going to vote, therefore you can't know the proper voter distribution to post-weight correctly. Maybe vote trading might be useful if your state has something like a 20-30% spread (like Alaska for Bush, or Kerry for DC) but 5%?
I think half the problem is that they don't come out after the election and say "hey, guess what, state such-and-such! You proved us wrong! Congratulations to all the voters!" Truth is that voting matters, in every state.
I think babies learn everything better than adults. I will stick to my 'brain is still empty' theory:) As we grow, we have more spyware/adware installed, and things tend to go more slowly.
Keep in mind your brain is still growing when you are a child. Once you hit the late teens, your brain's done growing, and it has to live with just rewiring its existing neurons to adapt to things quickly.
Children, honestly, are far smarter than adults are - it's too bad that our most brilliant years are wasted due to having extremely limited information. It's also important for parents to realize that their kids are far more capable than they think they are - lack of knowledge should never be construed as lack of intelligence. Parents often tell children "you wouldn't understand" when, in truth, the children probably would understand, possibly even better than the parents.
With these new findings, maybe a super computer can be built with these analytical and statistical skills, then this computer can learn to speak like HAL.
I'm really interested in the idea that children classify things via phoneme classification and statistical analysis. This sounds remarkably like a "universal translator" from Star Trek. I think a lot of work should be done in this area - it could be exceptionally useful in understanding the way communication works, and also the ability of computers to understand human speech.
Re:Whatever happened to ...
on
Broadband Bits
·
· Score: 3, Informative
Media Fusion. Was a fraud, as most on Slashdot expected. Here is what happened to the founder, thank goodness.
In fluid mechanics you don't call the carriers phonons. You call them molecules.:D And the wave itself is called a sound wave.:)
I'm not talking about the carrier particles, which are electrons in a conductor, atoms in a lattice, and of course, molecules in a liquid. I'm talking about the quanta of the wave itself.
The wave propagates through intermolecular forces. It's quantized, just like any other propagation of energy. What those quanta are called for a liquid or a gas, I don't know.
And the carriers are either particles (atoms/molecules/electrons/ions)
Again, I'm talking about the quanta of the wave itself. They're not real particles, just like the photons aren't real either (they're virtual - they don't need to be on mass-shell), but they are quantized.
Phonons are also typically confined to solid materials. In liquids and gases, the correct term (and way of thinking of the wave) is a sound wave.
A phonon is a sound wave - it's just the quantum of it. It's exactly the same as a "photon" is a quantum of an electromagnetic wave.
Um - any electromagnetic signal can be viewed as being transmitted by photons, whether it's an electromagnetic pulse down a waveguide, light propagating through free space, or someone changing a voltage on an electrical trace, causing it to switch.
In the voltage-switch case, the photons are virtual. In the free-space case, they're real. But they're still photons.
Electron-electron interactions are caused by (virtual) photons. Electrons can't interact with each other directly - there's no such interaction vertex (a three or four point electron vertex).
This is why electrical signals in a conductor propagate at (basically) the speed of light. The electric field is transmitting the signal, and the electric field is just virtual photon exchanges.
But no, you do not need to feed extra power to keep the wave propagating.
Electrical signals take a certain number of particles per second to generate a detectable signal. Photonic signals also would take a certain number of particles per second to generate a detectable signal.
For an equivalent signal/noise ratio, photons have the capability of requiring far, far less particles/second (because the noise floor is so much lower, and virtually everything is transparent enough to have virtually no "resistance"). Since particles/second is proportional to power in both cases (save in the case of a superconductor, but no one's going to suggest superconducting computers).
That's why I said it takes more power to transmit a signal electrically than photonically.
OK, ok, so it was a somewhat poor analogy. But it is true that resistive (or absorption) losses for an electronic chip are going to be significantly higher than for a photonic chip. Granted, resistive losses don't contribute much, but it doesn't make the argument wrong.
The *wave* in the lake, however, is much faster, carried by particles that bounce around each other much faster.
Actually, the wave in the lake is carried by something akin to phonons (heck, they might be phonons - I hate fluid mech). That is, the wave is "transmitted" by quanta of the intermolecular forces, not by any particles in the medium itself.
Strangely enough... as you suggested, the exact same thing happens in electrical signals, except there, the wave is "transmitted" by the inter-electron forces, which we call "electromagnetic" forces. Quanta of the electromagnetic field are, of course, photons, and the reason that electrical signals travel at 75% the speed of light is because that is the speed of light in that material, roughly.
So, in a very real way, signals on chips have always been carried by photons. It takes power to shove electrons around, though, whereas photons will just propagate. So transmitting a signal purely by photons (rather than by photons through electrons) is lower power.
Via already puts the CPU package on the mobo; it saves PCB space and power leakage. We're not going for a powerful system; just a cheap one.
The Via systems with integrated CPU chips are more expensive than socketed counterparts, partially for the reasons I stated. They have to make them more expensive because they make less of them - they make less because their shelf life is shorter.
Note that there are two sets of Via integrated CPU boards - there's the Eden set, which exist for homebrew PVR and somewhat of the embedded system market. But they have a higher feature set - this justifies the fact that they're underpowered. There was also an older one which sold disastrously (Syntax is the only one I knew that actually sold it - the S8601MP and similars).
The point is that Via can't expect to sell them very cheap, because if they've got the CPU and memory integrated, then the product depreciates much faster than the board without an integrated CPU. It should also be noted that if you buy the chips in solderable packages, you're buying more specialty parts rather than commodity, though if you're the chip manufacturer, it's not such a big deal.
And yes, you do save on the connectors.
For a company which buys connectors in volume, you save nothing on connectors. Even when I've bought in very small volumes, I've been able to pull DB9s down to under $1 per. In the larger volumes, you can get it much cheaper. In fact, if you make things other than the motherboard, then they're virtually free, as it's just a fluid item. Unless you do it for space concerns, the cost of a DB9 is just too cheap to even bother not putting on. Which is, of course, why it still exists on most things.
One other thing is the fact that the LPC port is very useful for debugging and board testing/prep, as well as hardware monitoring. There's enough of a demand for low-cost motherboards that if you could save money by dropping the legacy ports, you would.
Even the several legacy-free motherboards have the legacy ports on the board.
The board could also be small with no legacy stuff -- smaller than ITX form factor.
You'd be better off, cost wise, sticking with the largest market - micro-ATX. Volume, volume, volume rules all.
If the volume is high enough, you can design a southbridge that doesn't have the legacy support.
That's the problem - it's not the volume that's important, it's the volume*margins - that is, the profit. If you're starting off by saying "this is going to be a super-cheap platform", your margins will start off being thin, and as the platform ages, either the demand falls off, or your margins get thinner as you drop the price. Now, half the problem is you still need to pay for the development cost.
I don't think you're going to see someone come out and design something like this to be cheap. It wouldn't become profitable. We're almost at the point where the commodity items are near that price point as it is.
Basically, I don't think you're going to shave much cost off from any of the methods you're suggesting. You might be able to get it started on a size argument, but I think the small form factor PCs have shown that that design can command a premium, and so they'll charge it.
One other concern that you do fail to note is that the smaller the motherboard, usually the more expensive its design, regardless of how simple it is - you simply crowd the design, and the routing becomes very, very complicated. I have no doubt that the nano-ITX boards took a few iterations to get the signal integrity reliable enough.
I think VIA might do it - but I think you'll see far more of a commodity PC than you would expect. If it's specialty, it's expensive - if it's commodity, you don't have to pay the development, and risk having stock without any demand.
Slashdot Mirror
What happens when you *cut* a cable that is under tension?
The cable is only under tension because there is load upon it. Jettison the load - OK, so a satellite may be lost. Big deal. At that point the cable is no longer under tension, and it begins to settle. Spooling out a few km of cable from the top will recover the cable, though moved significantly.
Again, this type of "annoyance" is not beyond them now. They could easily crash a plane into a launch pad. They don't, however, because the impact is minimal. Ditto here.
Further, what if they *hijack* the elevator.
If terrorists have the capability to hijack any installation they wish, there are far better targets than a space elevator anchor.
Say, a nuclear munitions facility. Or even just a nuclear power plant. Hell, even just the Golden Gate Bridge would be a better target.
Obviously they don't have that capability.
First, the notion that the tether from the moon to the top of the elevator would seems a little shaky. If I understand correctly the cable can provide tension but is not rigid. As such, if it were strong enough (probably not unreasonable) it could prevent the elevator from drifting to the Earth side of L1. But L1 is unstable in the Lunar direction too, and the cable's tension would do nothing to prevent it from drifting to the Lunar side of L1; so, the cable alone couldn't keep the elevator centered on L1.
If the elevator starts falling (moves towards the Moon), it starts to lean to the east. To keep it stable, you pull to the west at the anchor point, and it doesn't move.
GEO isn't stable either - it's not a minimum in the potential at all. But the competing accelerations keep the cable taut. It's exactly the same in this manner.
You need thrusters to remain at GEO, too - obviously any slight acceleration pushes you away from it. But a space elevator doesn't need thrusters to sit at GEO, because slight perturbations cause fluctuations in the tension (which are compensated at the anchor point), not in the orbit.
The dynamics of an L1 elevator are pretty much identical to that of a GEO elevator. It's just simple orbital mechanics.
I shouldn't've mentioned that the tether "cancels the radial instability" - that's unimportant, as any orbital perturbations get translated into tension perturbations.
The interesting part is that since unpowered orbital paths naturally lead away from L1, debris will have a harder time hitting that portion of the elevator. The other portions will still be vulnerable, though.
and I'm not sure if the perpendicular direction is unstable as well.
Sigh, I could've tried reading the All Knowing Wikipedia : L1 is stable in the perpendicular direction, so an L1 elevator would be completely stable. Interesting!
The moon rotates once every thirty days.
It also orbits the Earth once every thirty days - in other words, it's locked in a 1:1 resonance with the Earth. This is of great benefit in this case!
A little background:
Space elevators need to remain stationary with respect to the body that they'll be attached to. With the Earth, that means you need to be rotating at the same speed as the Earth - in other words, GEO.
With the Moon, however, you can either be rotating as fast as the Moon, or orbiting as fast as the moon, because the moon rotates at the same angular speed as it orbits! And there are 5 places where that occurs - the Lagrange points.
From any one of those points, the Earth and the Moon are stationary in the sky - that is, you don't see either of them moving with respect to each other. Since the Moon's rotation is defined by its orientation with respect to the Earth, therefore, you don't see the Moon rotating. That is, you're in something that's exactly the same as GEO.
Of the 5 Lagrange points, obviously L3 is silly - it's on the opposite side of the Earth as the Moon. So that won't work.
L2 is similarly silly - it's on the opposite side of the Moon as the Earth. Could be useful for sending things to interstellar space, but not for Earth-Moon transits.
So you're left with L1, L4, and L5. Obviously if you're talking about getting things from the Earth to the Moon, you'd want the one that's deepest into Earth's gravity well - and that's L1, "gravitationally halfway" between the Earth and the Moon. And that's what's being proposed.
One helpful thing is that L1 is unstable - orbits tend to drift away from there. However, an elevator tethered to the moon at least anchors one of the unstable directions (radially towards the moon/away from the moon), and I'm not sure if the perpendicular direction is unstable as well. So it may be that an elevator in that position is stable, and that unpowered objects will tend to move away from the elevator. You'd have a natural deflection mechanism. Pretty interesting!
Actually, a combination of an L1 and an L2 elevator could be quite interesting, though you'd have to build something like a railway around the Moon. Once you do that, though, you could go out past the L2 point, and you could sling yourself into interplanetary space. I'd have to work out how much of a boost you could get, but Mars orbit seems quite reasonable.
It's not as good as a terrestrial elevator (because the Earth rotates so quickly, so you can steal more of the Earth's angular momentum), but it's certainly currently feasible.
All modern designs must address not just the engineering (is the cable strong enough) and the economics (is moon dust worth the cost of the elevator) but also acts of terrorism.
A terrestrial cable would be 100,000 km long.
Planes fly at about 10 km.
SpaceShipOne just climbed to 100 km.
Anything a terrorist does to a cable will be done to less than 0.01% of the cable.
Get a clue - terrorists can't do anything more than annoy a space elevator. Anything they can do is recoverable.
Clockmakers have used such a material for a long time; it's a complex alloy called invar.
While invar is useful, it just has a small (and controllable, through different alloying) coefficient. It's used many times to put metal through glass becaue they tune it to have the same thermal expansion as the glass. Therefore you can have a gas-tight glass container with metal coming out of it (used mostly for phototubes).
The "controllable" nature is the most important - in most situations you just want a matched thermal expansion.
There are probably instances where you want to create an object with zero thermal expansion. This stuff will allow you to do it, presumedly over some range (which is all you care about, really).
Ohio: Bush by 136,483 votes, 2.5%
More like ~132,000 or so. 4000 of those votes for Bush in Gahanna in Columbus, OH were some election worker (or machine)'s fantasy, remember?
the one where the 4000 Bush votes happened being more an exception than a rule.
Franklin County isn't using "e-voting" machines. They're electronic, but it's stuff from 1992 that's likely implemented in bare logic. No touchscreens, databases, or any of that. The votes are recorded simultaneously on 8 cartridges, one of which is removable to be taken to the counting area.
The cartridge apparently read out wrong on Election Day. Later, when it was checked, it was fine. Hmm. I wouldn't suspect the machine - I would suspect the person who put the cartridge in. It's hard to imagine that this couldn't also have been done with paper ballots with the people who actually read off the values or record the ballots.
# Evoting is a factor that "swings" voters toward (or away from) Bush
# Evoting is a factor that influences turn-out
# Evoting is a factor that influences change in turn-out w.r.t. the 2000 election
# Franklin county's change in relative turnout is unusual
Franklin County wasn't an "e-voting" district. They're machines from 1992. They're electronic, but not computerized. And distinctly not from Diebold.
They're not walking out on the entire process. They're walking out on the Kyoto Protocol.
China will be putting out just as much CO2 as the U.S. by 2013 and then by 2050 will have put out a greater cumulative total.
I'd believe that this was the true concern if the US was currently pushing for a new treaty. Given the fact that Kyoto will have taken 8 years from start to enactment, from your numbers, it seems like the US should be pushing for a new treaty now. Except... they aren't.
If everyone agrees to the Protocol, and we meet the 5% mark, CO2 emissions will be up 70% by 2050.
As with population growth predictions, believing that China will continue to increase its CO2 emissions at the same rate it is now is quite silly. They won't. And if you expect that they will, then Bush should be pushing even harder for a new treaty with China limiting their CO2 emissions immediately.
Incidentally, if Kyoto didn't even exist, then wouldn't CO2 emissions will be up even more by 2050? Like - almost 100%?
So... if China is such a threat, where's the international concern?
Oh, yeah - because the rest of the world is trying to convince the current largest emitter of CO2 to cut back. Not the possible future largest emitter.
Malaysia could become the largest emitter of CO2 by 2050. Who knows. That was 50 years away at the time.
Assuming it takes 1000 joules to raise the temperature of one kilogram of air by one kelvin, the the energy is 5 × 10^21 joules.
That's a tiny amount of energy, on a global scale. Let's just play with numbers a bit.
If the Earth was a simple radiator with an albedo of 0.3, its temperature would be -17.6 degrees C.
Obviously, it isn't a simple radiator - that's due to CO2, water vapor, methane, and other greenhouse gasses.
It currently has an average temperature of about 14 deg C. That means that CO2, methane, and water vapor have stored 800,000 nuclear bombs worth of energy in our atmosphere.
To make it even more fun: CO2 is 350 ppm of the Earth's atmosphere, and water vapor is something like 10,000 ppm. Methane is like 2 ppm, so we'll ignore it. Let's just take the stupid and wrong assumption that water and CO2 are equal greenhouse gasses, and the effect's linear (also wrong).
CO2 is currently rising at 2 ppm/year, and already has risen ~50-60 ppm. That means (with this stupid and wrong model) we're adding the equivalent of 160 nuclear bombs each year. And we've already detonated about 5000. So we'll raise the temperature of the Earth by a degree in about 30 years.
Thankfully the physics behind all of this is slightly more complicated (the Earth isn't a blackbody, for one). Especially when you consider that in 1995, the world expended 3.33 x 10^20 joules.
Or 1500 Tsar Bombas each year.
If the Earth didn't reradiate, we'd raise the temperature of the Earth by 1 degree in about 15 years.
And if everyone on Earth was American, we'd be expending enough energy to raise the temperature of the Earth by 1 degree every other year.
Doesn't seem so much now, does it?
If the earth did not radiate any energy to space, the surface would equalize with the sun's surface within a few years at 5000 degrees.
Um... no.
If the Earth's albedo is 0, then it would radiate all of the energy that impinges on it: which is 1380 W/m^2, the solar constant.
The Earth radiates this over 4 times the area that it receives it over (surface area of a sphere: 4 pi r^2: cross-sectional area of a sphere - pi r^2) so the Earth would radiate 345 W/m^2.
Google says sqrt(sqrt((1380 W/m^2) divided by 4 divided by the Stefan Boltzmann constant)) is 6 degrees C.
It should also be noted that the Earth is not a blackbody radiator, and so the albedo is not the deciding factor for the Earth's temperature. Earth's average albedo is 0.30, so its blackbody temperature would be sqrt(sqrt((1380 W/m^2) times 0.7 divided by 4 divided by the Stefan Boltzmann constant)), which Google says is -17.6 C.
Clearly other processes are at work maintaining Earth's global average temperature of 14 C - almost 32 degrees above its blackbody temperature.
Interestingly, this implies that CO2 and water vapor (and other greenhouse gasses) are collectively responsible for trapping almost 1 million of the atomic bombs that the grandparent poster mentioned.
The Kyoto accord only applies to developed countries, so enforcing it around the world drives industry into developing countries, such as China and India.
So basically, what you're saying is that only the US should have been allowed to dump arbitrarily large amounts of CO2 into the atmosphere while it was developing?
Give me a break. China is expected to be an Annex I country before the end of the decade, and if the US really wanted to complain about it, they could put pressure on that rather than walking out on the entire process.
My hunch is that there is significant overlap between them; thus wireless attitudes are implicitly contained within the poll results and there will not be a signficant difference come tomorrow...
That's assuming you poll 18-24 year olds with phone polling. In general, you don't. You significantly underrepresent the 18-24 bracket, which means you need to post-weight the result correctly so you represent the electorate right. What would be ideal is if they could find a way to poll the situation more equally, so that there'd be less of a systematic error introduced in your poll.
This shows (with high statistics!) that there most likely is a clear preference (about 10 points!) in that age bracket, which means if you don't get the correction right, you'll introduce a systematic bias in all your other polls. Considering you basically can't get the correction right (you'd need to know precisely what the voter turnout will be), you'd much prefer to make sure your sample is less skewed, so the effect of the (known) bias is lessened.
So, it basically comes down to: if young voters turn out like they used to, it'll look exactly like the polls say. However, if young voters turn out higher than they have previously, then the polls are all biased low, and Kerry will win many states by larger margins than expected.
This is what a statistical vs. a systematic error gets you. Most people are convinced that the election will be dead on close because the polls have constantly said "they're close", "they're close", "they're close". Sure they only have a 4% margin of error, but many polls should resolve it, right? Nope - systematics do not clear up with more polls (they just become more obvious).
So the main thing that this tells you is that people should expect that the polls really may not be accurate, and you may see non-"battleground states" decide this election as well.
Look at the respondant distribution - it isn't representative of the voting distribution, nor the population distribution. There's a heavy bias towards older voters (they're home more often) and female voters (ditto).
To correct it, they need to post-weight the results. They use the previous election's results (or a likelihood distribution and the registered voter distribution) to determine the weighting. Both of those methods are imperfect: therefore they generate systematic errors, which they don't quote.
And I'm not sure if you understood that right - randomly calling cellphones is illegal in most states that I know of, including both my current state and the state that my cellphone is from.
You may be right, you may be wrong. One thing people tend to overlook is that we DO NOT have a national election in America, we have 50 state elections on the same day. Most polls are national polls, which as last year shows us, mean pretty much nothing. Just because you have a majority in the national poll doesn't mean squat - I'm much more interested in polls done in "swing states."
His argument applies for swing states, as well (it's a dead heat in all of those states, and thankfully, those states have a large young population as well). The polls are likely to be wrong in that case, as well. Naively, given the estimated direction of the bias, it bodes very well for Kerry in the electoral result.
The polls are bunk - the worst thing about them is that people don't realize that the systematic component is likely to be comparable to the statistical, especially with such a tiny sample size. And systematic components do not improve with the number of polls you do, so even though, for instance, there have been 25 polls in Florida, and if you average them, he has a tiny lead. However, systematic errors don't improve with averaging (statistical errors do). So, you know nothing more.
Anyway, I thought the basic assumption underlying all random sampling is that the sample is, well, random.
Two things:
1) You don't get a random selection of people via any sampling method. Calling people at home, for instance, heavily biases you towards older populations, as they tend to be home more, and hang up less. So you have to correct for those biases, because you can't do a perfect sampling. That correction is called "post-weighting", because it happens after the poll.
2) You aren't trying to get a random sample of people, you're trying to get a random sample of people who will vote on Election Day. The only time you will know what the demographic fraction of people who vote are is after the vote happens. To compensate this, they either try to determine voter likelihood, and then postweight by this likelihood fraction, or they base it on previous year's results. They do not quote the systematic error that results from this, though, which is terrifically bad.
So yah, the polls are bunk. Why do you think all the polls from last election were not spread around the final result - 8/10 predicted Gore? Not enough data points, but it still is a pretty significant bias.
It is true that the polls can get it very wrong, but you forget that it is the pollsters JOB to get it right and so they have an incentive to design a method that will capture the right answer.
Not if you think about it. Most politicians don't know statistics - if they did, they'd be asking for the systematic error as well as the statistical. It's in the pollster's best interests to have polls that jump all over the place, and to have as many biases as possible. That way the politicians will be encouraged to buy more polls.
Here's a simple example:
Suppose you know the distribution of male/female people in your country: let's say it's 50/50, perfectly. Also assume that there are 10,000 people in your country.
Now let's assume you go out and get a random sample of the people in the country - you want to find out who likes A, and who likes B. You get 600 women, and 400 men. 100% of the women like A, and 100% of the men like B.
Naively, you'd conclude that A is favored 60% to 40%, with a sampling error of 2.9%. So a 50/50 split is excluded at 6 sigma, which is huge.
The problem is that something apparently was wrong with your polling method, because the distribution of women and men doesn't match the distribution of people. So you post-weight your results to have the sample accurately represent the country. So that sample of 600 is weighted by 4/6 (2/3), so you then get a 50%/50% split again. You have to do more work to calculate the error properly, so I won't do it here.
The problem comes when you don't know how to post-weight the results properly: suppose that you polled to get that 50% male/50% female number, and the poll was accurate +/-10%? Then your post-weighting becomes 0.66+/-0.44, which means that your poll now has an error term which is huge - your result is now 50%/50% with a 2.94% statistical error and a 10% systematic error (actually, it's more like +10%/-6.6%, but I'm not taking the time to work it out fully). And the problem is that this error term is systematic and not statistical, which means more polls don't help you at all - all those polls will be wrong in exactly the same way.
This gets worse for election polls. The relative percentage of men and women is deterministic: there is an absolute answer, and it is determinable. But election polls don't want a sample of people, they want a sample of voters who will actually vote. So they need some way of figuring out who is going to vote - there is, of course, no way to do that until after the election has happened.
Yes, polling companies post-weight their results. Yes, some post-weight it based on previous years' election data. Some do even worse - they determine the likelihood that a person is going to vote via an independent poll, and use that result (coupled with the registered voter demographic) to determine a "likelihood estimate" for voters at a certain age bracket, and use that to post-weight their data.
There is no way to do this right. You can only guess at the distribution of who will vote, and that guess contains a systematic error. That none of the polling companies tell you about.
So no, the pollsters do not "get it right". They can't. They don't even do a good job - if they did, they'd show systematic errors as well. Then you could possibly compare multiple polls and get a feel for what it truly might be, as well.
Some people consider local polls more accurate than state polls conducted by national entities.
The main problem with polls is the fact that they only quote statistical errors, when the people who actually perform the analysis know there are systematic errors as well.
Systematic errors don't go away when you do multiple polls, so averaging tricks don't work, and even the "well, my eye kinda tells me that Bush is ahead" don't work either. A 2% lead in a state with a 5% systematic error is not a lead at all, not even if you see it steady for an entire year.
Most polls are post-weighted for skewed age distribution (most respondents are old) and skewed geographical distribution. This post-weighting, though, depends on information either from the last election (which, honestly, has absolutely no reason to be correct again this election) or polls which necessarily contain error terms as well. Thus, the corrections introduce a systematic, which simply has to be ~ the statistical.
Statistics geek talk aside, don't ever think your state is a throwaway due to polls. Polls lie. You don't know who's going to vote, therefore you can't know the proper voter distribution to post-weight correctly. Maybe vote trading might be useful if your state has something like a 20-30% spread (like Alaska for Bush, or Kerry for DC) but 5%?
I think half the problem is that they don't come out after the election and say "hey, guess what, state such-and-such! You proved us wrong! Congratulations to all the voters!" Truth is that voting matters, in every state.
I think babies learn everything better than adults. I will stick to my 'brain is still empty' theory :) As we grow, we have more spyware/adware installed, and things tend to go more slowly.
Keep in mind your brain is still growing when you are a child. Once you hit the late teens, your brain's done growing, and it has to live with just rewiring its existing neurons to adapt to things quickly.
Children, honestly, are far smarter than adults are - it's too bad that our most brilliant years are wasted due to having extremely limited information. It's also important for parents to realize that their kids are far more capable than they think they are - lack of knowledge should never be construed as lack of intelligence. Parents often tell children "you wouldn't understand" when, in truth, the children probably would understand, possibly even better than the parents.
With these new findings, maybe a super computer can be built with these analytical and statistical skills, then this computer can learn to speak like HAL.
I'm really interested in the idea that children classify things via phoneme classification and statistical analysis. This sounds remarkably like a "universal translator" from Star Trek. I think a lot of work should be done in this area - it could be exceptionally useful in understanding the way communication works, and also the ability of computers to understand human speech.
Media Fusion. Was a fraud, as most on Slashdot expected. Here is what happened to the founder, thank goodness.
In fluid mechanics you don't call the carriers phonons. You call them molecules. :D :)
And the wave itself is called a sound wave.
I'm not talking about the carrier particles, which are electrons in a conductor, atoms in a lattice, and of course, molecules in a liquid. I'm talking about the quanta of the wave itself.
The wave propagates through intermolecular forces. It's quantized, just like any other propagation of energy. What those quanta are called for a liquid or a gas, I don't know.
And the carriers are either particles (atoms/molecules/electrons/ions)
Again, I'm talking about the quanta of the wave itself. They're not real particles, just like the photons aren't real either (they're virtual - they don't need to be on mass-shell), but they are quantized.
Phonons are also typically confined to solid materials. In liquids and gases, the correct term (and way of thinking of the wave) is a sound wave.
A phonon is a sound wave - it's just the quantum of it. It's exactly the same as a "photon" is a quantum of an electromagnetic wave.
Hence the name phonon.
And the wave can be viewed as a PHOTON
.
Um - any electromagnetic signal can be viewed as being transmitted by photons, whether it's an electromagnetic pulse down a waveguide, light propagating through free space, or someone changing a voltage on an electrical trace, causing it to switch.
In the voltage-switch case, the photons are virtual. In the free-space case, they're real. But they're still photons.
Electron-electron interactions are caused by (virtual) photons. Electrons can't interact with each other directly - there's no such interaction vertex (a three or four point electron vertex).
This is why electrical signals in a conductor propagate at (basically) the speed of light. The electric field is transmitting the signal, and the electric field is just virtual photon exchanges.
But no, you do not need to feed extra power to keep the wave propagating.
Electrical signals take a certain number of particles per second to generate a detectable signal. Photonic signals also would take a certain number of particles per second to generate a detectable signal.
For an equivalent signal/noise ratio, photons have the capability of requiring far, far less particles/second (because the noise floor is so much lower, and virtually everything is transparent enough to have virtually no "resistance"). Since particles/second is proportional to power in both cases (save in the case of a superconductor, but no one's going to suggest superconducting computers)
That's why I said it takes more power to transmit a signal electrically than photonically.
OK, ok, so it was a somewhat poor analogy. But it is true that resistive (or absorption) losses for an electronic chip are going to be significantly higher than for a photonic chip. Granted, resistive losses don't contribute much, but it doesn't make the argument wrong.
The *wave* in the lake, however, is much faster, carried by particles that bounce around each other much faster.
Actually, the wave in the lake is carried by something akin to phonons (heck, they might be phonons - I hate fluid mech). That is, the wave is "transmitted" by quanta of the intermolecular forces, not by any particles in the medium itself.
Strangely enough... as you suggested, the exact same thing happens in electrical signals, except there, the wave is "transmitted" by the inter-electron forces, which we call "electromagnetic" forces. Quanta of the electromagnetic field are, of course, photons, and the reason that electrical signals travel at 75% the speed of light is because that is the speed of light in that material, roughly.
So, in a very real way, signals on chips have always been carried by photons. It takes power to shove electrons around, though, whereas photons will just propagate. So transmitting a signal purely by photons (rather than by photons through electrons) is lower power.
Via already puts the CPU package on the mobo; it saves PCB space and power leakage. We're not going for a powerful system; just a cheap one.
The Via systems with integrated CPU chips are more expensive than socketed counterparts, partially for the reasons I stated. They have to make them more expensive because they make less of them - they make less because their shelf life is shorter.
Note that there are two sets of Via integrated CPU boards - there's the Eden set, which exist for homebrew PVR and somewhat of the embedded system market. But they have a higher feature set - this justifies the fact that they're underpowered. There was also an older one which sold disastrously (Syntax is the only one I knew that actually sold it - the S8601MP and similars).
The point is that Via can't expect to sell them very cheap, because if they've got the CPU and memory integrated, then the product depreciates much faster than the board without an integrated CPU. It should also be noted that if you buy the chips in solderable packages, you're buying more specialty parts rather than commodity, though if you're the chip manufacturer, it's not such a big deal.
And yes, you do save on the connectors.
For a company which buys connectors in volume, you save nothing on connectors. Even when I've bought in very small volumes, I've been able to pull DB9s down to under $1 per. In the larger volumes, you can get it much cheaper. In fact, if you make things other than the motherboard, then they're virtually free, as it's just a fluid item. Unless you do it for space concerns, the cost of a DB9 is just too cheap to even bother not putting on. Which is, of course, why it still exists on most things.
One other thing is the fact that the LPC port is very useful for debugging and board testing/prep, as well as hardware monitoring. There's enough of a demand for low-cost motherboards that if you could save money by dropping the legacy ports, you would.
Even the several legacy-free motherboards have the legacy ports on the board.
The board could also be small with no legacy stuff -- smaller than ITX form factor.
You'd be better off, cost wise, sticking with the largest market - micro-ATX. Volume, volume, volume rules all.
If the volume is high enough, you can design a southbridge that doesn't have the legacy support.
That's the problem - it's not the volume that's important, it's the volume*margins - that is, the profit. If you're starting off by saying "this is going to be a super-cheap platform", your margins will start off being thin, and as the platform ages, either the demand falls off, or your margins get thinner as you drop the price. Now, half the problem is you still need to pay for the development cost.
I don't think you're going to see someone come out and design something like this to be cheap. It wouldn't become profitable. We're almost at the point where the commodity items are near that price point as it is.
Basically, I don't think you're going to shave much cost off from any of the methods you're suggesting. You might be able to get it started on a size argument, but I think the small form factor PCs have shown that that design can command a premium, and so they'll charge it.
One other concern that you do fail to note is that the smaller the motherboard, usually the more expensive its design, regardless of how simple it is - you simply crowd the design, and the routing becomes very, very complicated. I have no doubt that the nano-ITX boards took a few iterations to get the signal integrity reliable enough.
I think VIA might do it - but I think you'll see far more of a commodity PC than you would expect. If it's specialty, it's expensive - if it's commodity, you don't have to pay the development, and risk having stock without any demand.