YOu would think that, but it's actuallty still a bit short term and short range. 20-30 years would be better.
The famous case recently of a climate scientist who said that temperatures had not risen "significantly" over some ten year period meant exactly that. He did a few calculation and concluded that to get an acceptable significance level he would need a fifteen year period so he answered the question correctly, but got quoted out of context.
Given the amount the US owes to China, I am reminded of the Ankh-Morpork anthem, which goes, in part:
Let others boast of martial dash For we have boldly fought with cash We own all your helmets, we own all your shoes. We own all your generals - touch us and you'll lose
The correct answer, as mentioned previously, is to consider on-call time as utilized and to not schedule the surgery after on-call hours. Truck drivers have more sensible rules.
Shouldn't be hard to measure how likely a given "on-call" role is to actually get called. If it's 10% you (rarely) upset the elective patient the next morning by posponing their surgery. If it's 80% it's probably better making it "on-shift" time an using the 1 in 5 quiet nights for admin. Somewhere in between lie the trickty cases.
Just needs a different type of game. Maybe something like threading a needle (could use actual needle and thread, or a Wii or something). Or slowly guiding a point of light down a narrow twisty track with the mouse.
The only element of this which really needs any non-standard thought is the media server, and that depends. If you're just archiving stuff, even that isn't a problem, but if you have multiple people doing video editing, for instance, you will need some serious power in the server and it's network connection. You also need to assess what level of reliability you need in that media server -- for instance can you afford to lose a few hours updates if something bad happens. If so, a standard server plus (say) mightly backups to another machine with a big RAID will do fine, if not, you need mirrored servers, and other complications.
A question bank isn't the same thing as a sample test. Publishers provide banks of questions that are only provided to authenticated college faculty who sign an agreement to keep them confidential. Makes a lot more sense than hundreds of people making up tests on the same subject.
Either someone stole a copy or bought one from a dodgy faculty member somewhere or is related to a faculty member or an employee of the publisher or....
IPv6 doesn't provide 2^128 independently routable addresses. The design of the system is such that routing decisions may only be made on the first 64 bits of address, the remaining 64 bits being reserved for local network addressing.
Something similar was true of the original IPv4 spec, but the limitation was circumvented by changes to routing protocols. Could something similar be done with IPv6 if necessary?
3/4 x 10^38 addresses is only enough to address the atoms in about 10^15kg of material -- hardly anything at all, really. Slightly less unrealistically, if we made IP-addressible robots the size and mass of a smallish bacterium (1 pg), that is enough addresses for 10^23 kg or about 1% of the mass of the Earth. Doesn't seem unreasonable that we could need more addresses than that, although probably not within 100 years.
Another scenario I can imagine is addressing points in space-time, in the context of route this message to whatever is there at that time. If we wanted to handle a cylinder reaching out to the orbit of Pluto and 1AU high, at say 1 second time resolution for 100 years and 1m spacial resolution, we'd need at least 160 address bits.
I conclude that IPv6 designers are being shortsighted and we will go through all this again, quite possibly in my lifetime. Why couldn't they use (say) 1024 address bits and simply declare a short packet format in which the last 896 bits are implied as zeroes?
It's pretty inconceivable that 128-bits will be exhausted in any foreseeable timeframe. 128 bits is a really astronomically (literally) huge number.
It seemed inconceivable that IPv4 would run out, or that 640K wouldn't be enough (well maybe not that one). I'm curious -- can anyone see a way that IPv6 address space could run out in (say) 50-100 years? It really does seem inconceivable to me, but I have learnt to be very skeptical about this feeling.
Most rocks that come this close to Earth are in orbits tied to Earth's and will come close again every few years. If we wanted to put a probe (or a manned lander) onto one of these, we could target one we've spotted before and arrange to intercept it on a future visit. There's no obvious incentive to visit one the moment we spot it.
If I recall correctly, Callisto is outsaide the main radiation belts and has a much less harsh environment. Another possible target would be a manned base INSIDE Europa, protected by a few km of water and ice.
Actually I know quite a bit about orbital mechanics. You start IN GSO. The cable is 100 000 km long, more than twice the distance from sea level to GSO (which is why you don't need a counterweight). As for unrolling, after about 10m tidal forces keep everything nicely aligned for you. Probably lots of issues with oscillation and stability to worry about, but they are second order.
We'll obviously have to agree to differ, but when I saw this plan, which basically requires a couple of heavy rocket launches and some careful unrolling, rather than the Clarke plan with asteroid capture and automated factories making the elevator cable in space, I began to believe that (a) making the nanotubes really was the only possibly unsolvable problem left and (b) that I might live to see one.
Read my post. No anchor mass at all, and construction is done on the ground. This is not Fountains of Paradise and you do not get a huge great beanstalk and trains getting you to orbit in a few hours, but you do get ton lots of cargo to orbit in a few weeks for next-to-no operating costs.
Making the nanotubes is the hardest part. While we don't have a specific matrix, we do know a lot about composite materials and the matrix doesn't have to do much. Testing the ribbon will be tricky, but it can be built in (say) 100km sections, so long as they are joined together carefully and it doesn't have to be especially automated.
I'm not expecting this in ten years, but I wouldn't be surprised to see it in twenty.
unrolling: this one really is easy, you fold the cable in half in the middle (not too tighly, so you don't kink it) and then roll it up. Now both ends are at the outside of the roll. I think my garden strimmer does this. You have to guide them in the appropriate directions for the first few tens of meters. After that tidal forces are your friend.
terminal guidance is an interesting problem, but you can have a few kg of whatever you like attached to the end of the cable to help. Outside the atmosphere maybe an ion rocket or something of that nature would be appropriate. In the atmosphere, jets or propellers powered remotely with a laser. You can do the swap from a sounding rocket or something of that kind.
You start with spools of tapered ribbon perhaps 1cm wide and 10 microns or so thick on average (more in the middle, less at the ends) and 100 000 km long. The ribbon is made of 90+% carbon nanotubes a few meters long plus a little bit of matrix to keep them in place. They're held together mainly by van der Waals forces between the fibres. This is about 10 m^3 of ribbon, which a Falcon 9 Heavy or Ariane 5 or similar can launch into GTO with enough room and mass to spare for an apogee motor.
Now you unroll the ribbon in both directions, slowly and carefully until one end hits the ground and the other end sticks 60 000 km beyond GSO. Tie down the ground end.
Might be worth doing three or four of these and having the ribbons in parallel.
Now you start sending crawlers up the ribbon, carrying spools of carbon nanotube fibre which they unroll at the edge of the ribbon, or between two of the ribbons and bond in place. Initially the crawlers might need to be quite small because of the limited strength of the cable, but a 100 000 km fibre at this thickness is only 30kg or so.
Probably you actually give the cable a somewhat more complex structure with multiple sub-cables and cross-braces, so it can survive meteorite strikes.
The used crawlers end up at the top of the cable as an anchor mass.
After a few years of work, you have a cable roughly as wide and thick as a piece of a letter paper (but very very much longer and stronger) up which you can operate crawlers in the 1 ton mass range and you're in business.
Lots of engineering challenges, most notably the nanotubes, but no obvious impossibilities.
Obvious steps along the road are to practice on the Moon (there are plenty of interesting parts of the surface we've never sampled and we can try as often as we want) and on Phobos and Deimos.
Yes. And so far that's been the choice, but we could certainly do a lot more analysis of the pebbles back on Earth than anything we could do on Mars. With modern instruments we could come scarily close to listing every atom in the sample by isotope and position.
The ascent stage (ie the Mars to Low Mars Orbit transport) needs about 4.1 km s^-1 of delta-V. About twice what you need from the Moon, but less than half what you need from Earth. The Mars orbit return vehicle, which doesn't need to land on Mars needs about 2.3 km s^-1 to get into a transfer orbit back to Earth. (figures from wikipedia).
This is definitely challenging, ascent stage most challenging of all. We need a rocket that can survive launch from Earth, 9 months coasting, aerocapture and aerobraking at Mars, impact with the Mars surface, a few months sitting on Mars and then take off with no support systems, deliver 4+kms^-1 of delta-V and automatically dock with the orbiting component of the system. The durability requirement pretty much rules out cryogenic fuels, and even relatively stable liquid fuels like kerosene/nitric acid might give trouble, both due to the cold conditions on Mars and the extra mass of tankage robust enough to survive the journey, so you're probably looking at solids.
A few quick checks reveal that good solid rockets have an ISP of maybe 265s, giving a mass ratio of perhaps 10 for Mars to Low Mars orbit, so we need an ascent stage roughly 90% of which is solid rocket propellant (or multiple ascent stages, adding complexity). Suppose we can get the payload capsule + docking system down to 10kg and our solid rocket motors are 95% propellant (5% nozzles and casing -- this might be optimistic) we get a mass of 200kg launching from Mars. This is actually less bad than I'd feared. Seems that a Viking sized lander could probably do it. A 1 ton or so lander includes a digging tool and maybe a mini-rover to collect 5kg of rock and load them into a 5kg capsule with some tiny thrusters on it. That sits on a 200kg solid fuel rocket that gets into Low Mars Orbit and drops the capsule, which docks with a similar sized vehicle with 100kg of solid fuel some batteries and electronics and a heat shield for Earth reentry.
So we need two launches to Mars transfer, each about 1 ton payload, plus heat shields for aero-braking/aero-capture on Mars. Should be doable as two medium large launches from Earth
We know something about the cratering history -- for instance if one crater is partly on top of another, we know it formed later, and the extent of "weathering" may give us further information. We can also compare it with the much studied cratering history of the Moon.
Slashdot Mirror
YOu would think that, but it's actuallty still a bit short term and short range. 20-30 years would be better.
The famous case recently of a climate scientist who said that temperatures had not risen "significantly" over some ten year period meant exactly that. He did a few calculation and concluded that to get an acceptable significance level he would need a fifteen year period so he answered the question correctly, but got quoted out of context.
Given the amount the US owes to China, I am reminded of the Ankh-Morpork anthem, which goes, in part:
Let others boast of martial dash
For we have boldly fought with cash
We own all your helmets, we own all your shoes.
We own all your generals - touch us and you'll lose
See also this version
The correct answer, as mentioned previously, is to consider on-call time as utilized and to not schedule the surgery after on-call hours. Truck drivers have more sensible rules.
Shouldn't be hard to measure how likely a given "on-call" role is to actually get called. If it's 10% you (rarely) upset the elective patient the next morning by posponing their surgery. If it's 80% it's probably better making it "on-shift" time an using the 1 in 5 quiet nights for admin. Somewhere in between lie the trickty cases.
Just needs a different type of game. Maybe something like threading a needle (could use actual needle and thread, or a Wii or something). Or slowly guiding a point of light down a narrow twisty track with the mouse.
The only element of this which really needs any non-standard thought is the media server, and that depends. If you're just archiving stuff, even that isn't a problem, but if you have multiple people doing video editing, for instance, you will need some serious power
in the server and it's network connection. You also need to assess what level of reliability you need in that media server -- for instance can you afford to lose a few hours updates if something bad happens. If so, a standard server plus (say) mightly backups to another machine with a big RAID will do fine, if not, you need mirrored servers, and other complications.
A question bank isn't the same thing as a sample test. Publishers provide banks of questions that are only provided to authenticated college faculty who sign an agreement to keep them confidential. Makes a lot more sense than hundreds of people making up tests on the same subject.
Either someone stole a copy or bought one from a dodgy faculty member somewhere or is related to a faculty member or an employee of the publisher or ....
So that's obstruct then!
IPv6 doesn't provide 2^128 independently routable addresses. The design of the system is such that routing decisions may only be made on the first 64 bits of address, the remaining 64 bits being reserved for local network addressing.
Something similar was true of the original IPv4 spec, but the limitation was circumvented by changes to routing protocols. Could something similar be done with IPv6 if necessary?
3/4 x 10^38 addresses is only enough to address the atoms in about 10^15kg of material -- hardly anything at all, really. Slightly less unrealistically, if we made IP-addressible robots the size and mass of a smallish bacterium (1 pg), that is enough addresses for 10^23 kg or about 1% of the mass of the Earth. Doesn't seem unreasonable that we could need more addresses than that, although probably not within 100 years.
Another scenario I can imagine is addressing points in space-time, in the context of route this message to whatever is there at that time. If we wanted to handle a cylinder reaching out to the orbit of Pluto and 1AU high, at say 1 second time resolution for 100 years and 1m spacial resolution, we'd need at least 160 address bits.
I conclude that IPv6 designers are being shortsighted and we will go through all this again, quite possibly in my lifetime. Why couldn't they use (say) 1024 address bits and simply declare a short packet format in which the last 896 bits are implied as zeroes?
It's pretty inconceivable that 128-bits will be exhausted in any foreseeable timeframe. 128 bits is a really astronomically (literally) huge number.
It seemed inconceivable that IPv4 would run out, or that 640K wouldn't be enough (well maybe not that one). I'm curious -- can anyone see a way that IPv6 address space could run out in (say) 50-100 years?
It really does seem inconceivable to me, but I have learnt to be very skeptical about this feeling.
Most rocks that come this close to Earth are in orbits tied to Earth's and will come close again every few years. If we wanted to put a probe (or a manned lander) onto one of these, we could target one we've spotted before and arrange to intercept it on a future visit. There's no obvious incentive to visit one the moment we spot it.
If I recall correctly, Callisto is outsaide the main radiation belts and has a much less harsh environment. Another possible target would be a manned base INSIDE Europa, protected by a few km of water and ice.
Lead is also too soft. It might not survive launch.
Actually it's 1000 times...
a) solar cells are almost pure silicon these days, which is not toxic.
b) the proposal is for solar-thermal plant which doesn't involve solar cells at all.
Actually I know quite a bit about orbital mechanics. You start IN GSO. The cable is 100 000 km long, more than twice the distance from sea level to GSO (which is why you don't need a counterweight). As for unrolling, after about 10m tidal forces keep everything nicely aligned for you. Probably lots of issues with oscillation and stability to worry about, but they are second order.
We'll obviously have to agree to differ, but when I saw this plan, which basically requires a couple of heavy rocket launches and some careful unrolling, rather than the Clarke plan with asteroid capture and automated factories making the elevator cable in space, I began to believe that (a) making the nanotubes really was the only possibly unsolvable problem left and (b) that I might live to see one.
Read my post. No anchor mass at all, and construction is done on the ground. This is not Fountains of Paradise and you do not get a huge great beanstalk and trains getting you to orbit in a few hours, but you do get ton lots of cargo to orbit in a few weeks for next-to-no operating costs.
Making the nanotubes is the hardest part. While we don't have a specific matrix, we do know a lot about composite materials and the matrix doesn't have to do much. Testing the ribbon will be tricky, but it can be built in (say) 100km sections, so long as they are joined together carefully and it doesn't have to be especially automated.
I'm not expecting this in ten years, but I wouldn't be surprised to see it in twenty.
Two specific issues:
unrolling: this one really is easy, you fold the cable in half in the middle (not too tighly, so you don't kink it) and then roll it up. Now both ends are at the outside of the roll. I think my garden strimmer does this. You have to guide them in the appropriate directions for the first few tens of meters. After that tidal forces are your friend.
terminal guidance is an interesting problem, but you can have a few kg of whatever you like attached to the end of the cable to help. Outside the atmosphere maybe an ion rocket or something of that nature would be appropriate. In the atmosphere, jets or propellers powered remotely with a laser. You can do the swap from a sounding rocket or something of that kind.
This has been worked out. Here's one scenario:
You start with spools of tapered ribbon perhaps 1cm wide and 10 microns or so thick on average (more in the middle, less at the ends) and 100 000 km long. The ribbon is made of 90+% carbon nanotubes a few meters long plus a little bit of matrix to keep them in place. They're held together mainly by van der Waals forces between the fibres. This is about 10 m^3 of ribbon, which a Falcon 9 Heavy or Ariane 5 or similar can launch into GTO with enough room and mass to spare for an apogee motor.
Now you unroll the ribbon in both directions, slowly and carefully until one end hits the ground and the other end sticks 60 000 km beyond GSO. Tie down the ground end.
Might be worth doing three or four of these and having the ribbons in parallel.
Now you start sending crawlers up the ribbon, carrying spools of carbon nanotube fibre which they unroll at the edge of the ribbon, or between two of the ribbons and bond in place. Initially the crawlers might need to be quite small because of the limited strength of the cable, but a 100 000 km fibre at this thickness is only 30kg or so.
Probably you actually give the cable a somewhat more complex structure with multiple sub-cables and cross-braces, so it can survive meteorite strikes.
The used crawlers end up at the top of the cable as an anchor mass.
After a few years of work, you have a cable roughly as wide and thick as a piece of a letter paper (but very very much longer and stronger) up which you can operate crawlers in the 1 ton mass range and you're in business.
Lots of engineering challenges, most notably the nanotubes, but no obvious impossibilities.
Muon spin resonance is real technology.
Obvious steps along the road are to practice on the Moon (there are plenty of interesting parts of the surface we've never sampled and we can try as often as we want) and on Phobos and Deimos.
Yes. And so far that's been the choice, but we could certainly do a lot more analysis of the pebbles back on Earth than anything we could do on Mars. With modern instruments we could come scarily close to listing every atom in the sample by isotope and position.
Let's do the numbers.
The ascent stage (ie the Mars to Low Mars Orbit transport) needs about 4.1 km s^-1 of delta-V. About twice what you need from the Moon, but less than half what you need from Earth. The Mars orbit return vehicle, which doesn't need to land on Mars needs about 2.3 km s^-1 to get into a transfer orbit back to Earth. (figures from wikipedia).
This is definitely challenging, ascent stage most challenging of all. We need a rocket that can survive launch from Earth, 9 months coasting, aerocapture and aerobraking at Mars, impact with the Mars surface, a few months sitting on Mars and then take off with no support systems, deliver 4+kms^-1 of delta-V and automatically dock with the orbiting component of the system. The durability requirement pretty much rules out cryogenic fuels, and even relatively stable liquid fuels like kerosene/nitric acid might give trouble, both due to the cold conditions on Mars and the extra mass of tankage robust enough to survive the journey, so you're probably looking at solids.
A few quick checks reveal that good solid rockets have an ISP of maybe 265s, giving a mass ratio of perhaps 10 for Mars to Low Mars orbit, so we need an ascent stage roughly 90% of which is solid rocket propellant (or multiple ascent stages, adding complexity). Suppose we can get the payload capsule + docking system down to 10kg and our solid rocket motors are 95% propellant (5% nozzles and casing -- this might be optimistic) we get a mass of 200kg launching from Mars. This is actually less bad than I'd feared. Seems that a Viking sized lander could probably do it. A 1 ton or so lander includes a digging tool and maybe a mini-rover to collect 5kg of rock and load them into a 5kg capsule with some tiny thrusters on it. That sits on a 200kg solid fuel rocket that gets into Low Mars Orbit and drops the capsule, which docks with a similar sized vehicle with 100kg of solid fuel some batteries and electronics and a heat shield for Earth reentry.
So we need two launches to Mars transfer, each about 1 ton payload, plus heat shields for aero-braking/aero-capture on Mars. Should be doable as two medium large launches from Earth
We know something about the cratering history -- for instance if one crater is partly on top of another, we know it formed later, and the extent of "weathering" may give us further information. We can also compare it with the much studied cratering history of the Moon.