I agree with most of your analysis, but I think you're lowballing the 1985 state of the art in control systems a bit. My first job out of college was building a distributed HVAC control system for a large General Dynamics manufacturing plant. We used a VAX 11/730 hub talking to Analog Devices MicroMAC-5000 process control computers mounted right in the air handlers, cooling towers, and so forth. The whole setup was digital right to the switches for on/off controls, and up to the DACs for continuous outputs. We started building all this with standard, relatively inexpensive, off-the-shelf hardware in the fall of 1985, and had first (small) building cut over to use the new system exclusively a few months later.
Note that I wrote "to and from the surface of the moon", not to and from lunar orbit. Getting to and from the moon, overcoming lunar gravity up and down from the surface, and stopping at the surface requires more delta-v than travelling to L2, stopping, and coming home.
Since the JWST will be at the L2 point, servicing will be impossible.
Certainly impossible using the Shuttle or similar low-earth-orbit transportation. However, if the Bush plan moves forward, we'll have lunar-capable transportation coming online right around the time the JWST is deployed. If you can get people to and from the surface of the Moon, getting to and from Sun-Earth L2 is arguably easier and safer. So we can service JWST with our lunar fleet.
What of any conceivable use can you send up a cable with a 100 ton counterweight? Assuming Forward and Moravec got their mass ratios right (and if anybody can be trusted on that score, it's them), that implies a climber payload of around 20 grams.
From Forward and Moravec's 1980 paper on space elevators:
An earth skyhook would be an engineering marvel. The job of
building the 36,000 kilometer section down to the earth would be
equivalent to building a suspension bridge around the equator. In
order to lift appreciable loads, say 100 tons at a time, the skyhook
would have to weigh 600 thousand tons. Fortunately, the carbon needed
for the graphite fibers can be found in special kinds of asteroids
called carbonaceous chondrites. After the carbon was extracted from
the asteroid, the remaining slag could be used as the counterweight.
Note that the mass requirement is not dependent on choice of materials; rather, it depends on payload size -- what can safely crawl up the cable without pulling the whole thing down instead.
For comparison, the International Space Station, which is the largest structure ever built in space thus far, masses around 112 tons. The Shuttle orbiter masses around 100 tons.
So building the space elevator can be estimated as being roughly 6,000 times as hard as building ISS, or as being like launching 6,000 shuttles on one-way trips (and this neglects the fact that you need to go to geosync orbit rather than LEO, and also that you need to grab an asteroid for the counterweight, and that you need to do on-orbit fabrication of a material barely out of the laboratory right now).
Funny, all the old space probes had nuclear powerplants and that all worked out just fine.
No spacecraft to date has ever used nuclear propulsion. Many have used nuclear power to generate electricity to power onboard systems, but that's a very different kettle of fish; RTGs are simple, self-contained, and safe compared to nuclear propulsion systems.
I strongly support moving forward with nuclear propulsion, but it really is something new (and hence fraught with potential engineering problems).
Also, building the space elevator involves moving insane amounts of mass around, both up from earth and in from elsewhere in the solar system (e.g., Luna or the asteroids). The space elevator would be several orders of magnitude more massive than the combined total of everything ever sent into space to date, and that's even if you count each Shuttle launch separately. There's no reasonable way to build a space elevator without nuclear propulsion.
Due to the low triple point of water on Mars, and the theory that it's just coming out of an ice-age, there's every chance there is no liquid left around there to melt, but there's certainly a chance there is.
The triple point (at which solid, gas, and liquid phases are in equilibrium) doesn't change from planet to planet; it's a fixed temperature and pressure pair for any given material.
For water, the triple point is 273.16 K at 611.2 Pa. That pressure is about twice the highest found in the lowest parts of the Martian surface. As a result, any liquid water on the surface will very quickly change phase to ice, vapor, or (most likely) some of both phases.
The nice thing for would-be Martian terraformers is that you only have to double Mars's surface pressure to begin to make liquid water stable in low-lying parts of the surface. Even there, it would freeze solid every night and most days, but you'd get *some* periods where the water might stay liquid for hours at a time during the local afternoon.
A bigger leak would definitely be easier to find. For a medium-sized leak (the kind that would drain the station over the course of a day or two), the sound would be loud and easy to find; most likely a whistling, rushing sound.
For a very large leak, just look for the impact damage where equipment and bodies slammed hardest against the surrounding hull sections.
I wonder how they justify making this a separate and more serious offense from copyright infringement in general. Or do they even bother with justifications beyond the implicit "they paid me a boatload of money to pass this law" these days?
Two points regarding your "anything similar with a crew onboard would be fatal" assertion.
First, all these robot probes are built far more cheaply than a hypothetical manned Mars vessel. To be blunt, because they're machines, they're more expendable, so the budgets are lower. The risk/reward equation is different.
Second, many probe-loss accidents could probably have been resolved on the scene by a quick-witted pilot or a technician with a screwdriver. Humans are far better at responding effectively to unforeseen circumstances than robots are. (So far, anyway.)
A solar physicist walks into a bar, gets the bartender's attention, and says "I'd like a Mexican beer, please."
The bartender immediately begins shouting "OK, everybody out! Right now! Everyone out of the bar!" And he heards all the patrons out into the street, slamming the door behind them.
The solar physicist shakes his head ruefully. "Darn," he says, "I should have seen that Corona mass ejection coming!"
(By the way, it goes without saying that the bar is in SoHo.)
I think the two of them are arguing past each other. No, I don't think Drexler's simplified thought-experiment version of nanotech -- conceptually, little nanoscale forklifts and welders -- is physically realizable, for all the chemical/QM reasons given. However, useful fine-tuned molecular assembly and even self-replication can be accomplished; in fact, it's happening right now inside the organism typing this message. An existence proof is the strongest kind of proof.
So the question is not "is nanotech possible?", but rather "will useful human-created nanotech look more like industrial engineering or applied biochemistry?" Personally, I believe it will emerge from a fusion of the two approaches, but still looking more like advanced biochemistry than scaled-down industrial engineering.
Actually, the problem was that Julus was a 14th level Emperor-General combined class with about 200 hit points, wearing a +2 toga of blade turning, and with daggers only doing 1d4 per hit, it took a lot of work to even make him notice he was being attacked, much less kill him.
First, there was plenty of life in 5000 BC. In fact, there was plenty of life in 1000000000 BC.
Second, Vega is only 25 lightyears away, meaning that the horrid bug-eyed Vegans are peering through their observoscopes and lusting after Farrah Fawcett.
I did some work with MS's media player team a couple of years ago, as part of a partnership dev effort. I learned a lot about good software practices working with them; their coding discipline, QC and QA techniques, attention to metrics, adherence to a dev plan and structure, and every other "good dev team" attribute I can think of were among the best I've ever encountered.
What ends up kneecapping Microsoft (IMHO) are three things:
1) They used to rush to market without adequate testing. This has been less true the last few years, but it ruined their quality rep, and also saddled them with half-cooked code they now have to support forever.
2) They committed to building everything for total, open interoperability before the net made that into a giant security hazard. For an isolated, single-user machine, or even one on a trusted, isolated LAN, having your word processor able to quietly modify stylesheets or even OS behavior is a very useful metafeature. Put hostile users/programs into the mix, and it's a gaping security hole. There's no easy way for them to back out of the total-interoperability model, so they're stuck hacking kludged security patches over each new exploit as it's found.
3) They have a giant base of legacy code APIs which they can only phase out very slowly. A lot of those APIs are inherently insecure or poorly thought out (in retrospect), but all new code has to wrap around them somehow. It's as if you built a perfectly fine shack, then decided to build a skyscraper instead, but had to retain the shack as the ground floor.
So be fair to the hard-working and clueful code monkeys at MS. They are (in my experience) a good bunch, laboring under impossible constraints just like the rest of us do.
All good points, but all amenable to technical and economic solutions -- if you don't like spam, learn to fight it, or pay someone to fight it by choosing an appropriate ISP. For that matter, if people would stop responding to spam with their wallets, the problem would go away this week.
Your point 5 is commonly used as a bludgeon by would-be regulators, which makes it especially infuriating to me. I refuse to support making the Net safe for children. Parents should learn to make their own computers safe for children, and far more important, to make their children safe for the net. And I am a parent myself, so I'm not blowing smoke here.
I am extremely reluctant to let government in to solve a problem which individuals can readily solve for themselves. Government regulation is like a Viking or a vampire -- once you let it in, you're already doomed, so the trick is not to let it in.
Slashdot Mirror
I agree with most of your analysis, but I think you're lowballing the 1985 state of the art in control systems a bit. My first job out of college was building a distributed HVAC control system for a large General Dynamics manufacturing plant. We used a VAX 11/730 hub talking to Analog Devices MicroMAC-5000 process control computers mounted right in the air handlers, cooling towers, and so forth. The whole setup was digital right to the switches for on/off controls, and up to the DACs for continuous outputs. We started building all this with standard, relatively inexpensive, off-the-shelf hardware in the fall of 1985, and had first (small) building cut over to use the new system exclusively a few months later.
Note that I wrote "to and from the surface of the moon", not to and from lunar orbit. Getting to and from the moon, overcoming lunar gravity up and down from the surface, and stopping at the surface requires more delta-v than travelling to L2, stopping, and coming home.
And I'm sure that, in the face of pressing national interest, this treaty will prove just as sturdy as the Kellogg-Briand Pact.
Unless these guys specialize in validating identity on signed code escrow documents or something, I imagine the desired term was notables.
Jeez, what part of "attempt no landing there" didn't they get?
Oh, wait, that's Europa. Never mind.
What of any conceivable use can you send up a cable with a 100 ton counterweight? Assuming Forward and Moravec got their mass ratios right (and if anybody can be trusted on that score, it's them), that implies a climber payload of around 20 grams.
For comparison, the International Space Station, which is the largest structure ever built in space thus far, masses around 112 tons. The Shuttle orbiter masses around 100 tons.
So building the space elevator can be estimated as being roughly 6,000 times as hard as building ISS, or as being like launching 6,000 shuttles on one-way trips (and this neglects the fact that you need to go to geosync orbit rather than LEO, and also that you need to grab an asteroid for the counterweight, and that you need to do on-orbit fabrication of a material barely out of the laboratory right now).
I strongly support moving forward with nuclear propulsion, but it really is something new (and hence fraught with potential engineering problems).
Also, building the space elevator involves moving insane amounts of mass around, both up from earth and in from elsewhere in the solar system (e.g., Luna or the asteroids). The space elevator would be several orders of magnitude more massive than the combined total of everything ever sent into space to date, and that's even if you count each Shuttle launch separately. There's no reasonable way to build a space elevator without nuclear propulsion.
The triple point (at which solid, gas, and liquid phases are in equilibrium) doesn't change from planet to planet; it's a fixed temperature and pressure pair for any given material.
For water, the triple point is 273.16 K at 611.2 Pa. That pressure is about twice the highest found in the lowest parts of the Martian surface. As a result, any liquid water on the surface will very quickly change phase to ice, vapor, or (most likely) some of both phases.
The nice thing for would-be Martian terraformers is that you only have to double Mars's surface pressure to begin to make liquid water stable in low-lying parts of the surface. Even there, it would freeze solid every night and most days, but you'd get *some* periods where the water might stay liquid for hours at a time during the local afternoon.
"It's piss, Jim, but not as we know it."
Damn that Hubble data embargo!
A bigger leak would definitely be easier to find. For a medium-sized leak (the kind that would drain the station over the course of a day or two), the sound would be loud and easy to find; most likely a whistling, rushing sound.
For a very large leak, just look for the impact damage where equipment and bodies slammed hardest against the surrounding hull sections.
I wonder how they justify making this a separate and more serious offense from copyright infringement in general. Or do they even bother with justifications beyond the implicit "they paid me a boatload of money to pass this law" these days?
Two points regarding your "anything similar with a crew onboard would be fatal" assertion.
First, all these robot probes are built far more cheaply than a hypothetical manned Mars vessel. To be blunt, because they're machines, they're more expendable, so the budgets are lower. The risk/reward equation is different.
Second, many probe-loss accidents could probably have been resolved on the scene by a quick-witted pilot or a technician with a screwdriver. Humans are far better at responding effectively to unforeseen circumstances than robots are. (So far, anyway.)
(Yes, as a matter of fact I did steal that from The Onion, why do you ask?)
A solar physicist walks into a bar, gets the bartender's attention, and says "I'd like a Mexican beer, please."
The bartender immediately begins shouting "OK, everybody out! Right now! Everyone out of the bar!" And he heards all the patrons out into the street, slamming the door behind them.
The solar physicist shakes his head ruefully. "Darn," he says, "I should have seen that Corona mass ejection coming!"
(By the way, it goes without saying that the bar is in SoHo.)
I think the two of them are arguing past each other. No, I don't think Drexler's simplified thought-experiment version of nanotech -- conceptually, little nanoscale forklifts and welders -- is physically realizable, for all the chemical/QM reasons given. However, useful fine-tuned molecular assembly and even self-replication can be accomplished; in fact, it's happening right now inside the organism typing this message. An existence proof is the strongest kind of proof.
So the question is not "is nanotech possible?", but rather "will useful human-created nanotech look more like industrial engineering or applied biochemistry?" Personally, I believe it will emerge from a fusion of the two approaches, but still looking more like advanced biochemistry than scaled-down industrial engineering.
Tacere, Senatus Romanus?
Actually, the problem was that Julus was a 14th level Emperor-General combined class with about 200 hit points, wearing a +2 toga of blade turning, and with daggers only doing 1d4 per hit, it took a lot of work to even make him notice he was being attacked, much less kill him.
Yes, but in the Roman era D&D, which is set in a mythic version of the medieval period, was considered a cutting-edge hyperfuturistic sci-fi game.
First, there was plenty of life in 5000 BC. In fact, there was plenty of life in 1000000000 BC.
Second, Vega is only 25 lightyears away, meaning that the horrid bug-eyed Vegans are peering through their observoscopes and lusting after Farrah Fawcett.
Fair standards department:
I did some work with MS's media player team a couple of years ago, as part of a partnership dev effort. I learned a lot about good software practices working with them; their coding discipline, QC and QA techniques, attention to metrics, adherence to a dev plan and structure, and every other "good dev team" attribute I can think of were among the best I've ever encountered.
What ends up kneecapping Microsoft (IMHO) are three things:
1) They used to rush to market without adequate testing. This has been less true the last few years, but it ruined their quality rep, and also saddled them with half-cooked code they now have to support forever.
2) They committed to building everything for total, open interoperability before the net made that into a giant security hazard. For an isolated, single-user machine, or even one on a trusted, isolated LAN, having your word processor able to quietly modify stylesheets or even OS behavior is a very useful metafeature. Put hostile users/programs into the mix, and it's a gaping security hole. There's no easy way for them to back out of the total-interoperability model, so they're stuck hacking kludged security patches over each new exploit as it's found.
3) They have a giant base of legacy code APIs which they can only phase out very slowly. A lot of those APIs are inherently insecure or poorly thought out (in retrospect), but all new code has to wrap around them somehow. It's as if you built a perfectly fine shack, then decided to build a skyscraper instead, but had to retain the shack as the ground floor.
So be fair to the hard-working and clueful code monkeys at MS. They are (in my experience) a good bunch, laboring under impossible constraints just like the rest of us do.
All good points, but all amenable to technical and economic solutions -- if you don't like spam, learn to fight it, or pay someone to fight it by choosing an appropriate ISP. For that matter, if people would stop responding to spam with their wallets, the problem would go away this week.
Your point 5 is commonly used as a bludgeon by would-be regulators, which makes it especially infuriating to me. I refuse to support making the Net safe for children. Parents should learn to make their own computers safe for children, and far more important, to make their children safe for the net. And I am a parent myself, so I'm not blowing smoke here.
I am extremely reluctant to let government in to solve a problem which individuals can readily solve for themselves. Government regulation is like a Viking or a vampire -- once you let it in, you're already doomed, so the trick is not to let it in.