IIRC the Titan boosters used the rather poisonous Nitrogen Tetroxide as the oxidizer. That stuff is mighty bad for human lungs if it gets into your air. Also the computer was cooled by liquid mercury.
Also I had the pleasure of taking apart an one of these Titan guidance computers. It was about the size of a big suitcase. Built to take many G's-- it had a aluminum case about 3/4 inch thick. All thge modules inside were potted in a tough pink styrofoam.
An amazing device with about 300 credit-card sized PC boards all plugged in and soldered into a backplane. Each PC board had what looked like four to six Westinghouse flat-pack IC's, probably DTL logic, maybe four gates max per chip. Amazing what they could do with that little hardware. The memory was some PC-board version of magnetic wire memory, as cores probably couldnt take the g's and vibration. Sobering to be poking through a device designed to land 9 Megatons on the Ruskies.
A quantum channel is only good for as far as one photon is likely to survive above the noise level-- maybe a kilometer at most?
Anything farther than that will involve detecting the photon and relaying it-- a chancy proposition which adds waay too much noise and of course one can intercept the signal at the repeater.
Exactly how long can you make a diamond cable? Splices are very unreliable and lossy.
What is that 120 km number doing in there? Mighty unlikely.
>>And the weight of copper is rarely a problem.
>Unless you're sending stuff into space.
Sending stuff into space *is* mighty rare. And one might estimate copper makes up somewhat less than 1% of the weight of the ISS. Much bigger weight concerns are the continuing need for several pounds of water per person per day. Figure out a way to dehydrate wather and then you're onto something.
" 10 times better than traditional copper cables at one-sixth the weight."
Writing "10 times better" is mighty misleading. If a superconductor is "1", and copper is "0.95", then this whizzy new wire would be 0.995, only 4% better than copper.
It would be nice if they mentioned the acceleration of this thingy.
That's one heavy mothership to be pushed by shooting out tiny amounts of ions!
At a first approximation, the acceleration's gotta be somewhere under 0.00001G's, which makes the whole scheme rather slothful and boring.
And if there's going to be pumps and turbines involved, how're they going to keep the thing from spinning wildly in several axes? Surely not with reaction thrusters. The more I think about this thing, the more it looks like some gigantic delayed April Fools joke.
It just works*
* For relatively small values of "works". Such as:
Worked once when the programmer tested it, 5 minutes before code check-in time.
Works fine, just takes 20 times longer than grepping for it yourself.
Works, if you have ports 147, 373, 8288.9, and 51599 open.
Works, you just have to hand-patch the registry.
Works, but breaks three unrelated things.
Works, but to get to it you have to do Start//Control Panel//Admin Tools//System Maintenance//Resize window so you can see it//Squint at 4-point tree//Users and Groups//Users//Click on a User//Advanced//Startup Shares//Add//Then you have to correctly type in the path to a network resource, with no browsing ability, no hints, no error checking//Toss machine out the window.
Works, but requires 1.3Gb of disk space to install 630k of system patches, and doesnt check first off for enough room, or that there's 88 free gigabytes on another disk.
Works, but the progress bars *always* lie.
Works, but you can never be sure.
Works, but the interface to the web server is through a web page, which is basically nuts.
Works, if you ignore the fact that System restore pops up even though you have firmly turned it off.
Works, but the installer for Office pops up at random times when you open a folder.
Works, but the error messages never tell you what program or module is complaining, or about what file.
Works, if you're happy with help files that tell you you have a problem, which you already know.
Works for you, and for most any hacker that wants something to "work" on your computer.
As a guy that's programmed 4K machines, I'm in favor of less code. But:
Let's say the kernel bloats by 5 megabytes.
At today's memory prices, that's about 75 cents US.
How long would it take me to trim out and reboot and test a smaller kernel? Oh, at least $10 of my time. Plus the hard to quantify costs and risks of changing the OS and rebooting during the workday.
By this argument kernel bloat, just the memory consumption aspect, shouldnt be a big cause for concern. As to the potential for less stability when there are unused modules in memory, I don't see how code that's never called can cause stability problems. Any surmises?
I'd like to buy one of these SkyCars, but they have to work out a few teensy details:
How are these things going to be licensed? The poor FAA can barely manage things as they are. If even one person in 100 got a SkyCar, that's 2 million planes to register, 2 million junior pilots.
Very few people have the tenacity and skill level to fly on instruments. That means whenever the fog or low clouds move in, hundreds of these SkyCars are going to be falling out of the sky.
Any consumer-maintained device where your life depends on FOUR internal-combustion engines working flawlessly probably has a MTTCrash of not too many hours.
What's the insurance going to be like on one of these things? Sky high, at the very least.
A quick peek at the task manager in Windows XP shows that when AVG anti-virus is running, even on this old 600 MHz box, the CPU is only about 27% busy. The real slowdown is due to the disk cache getting flushed by all the anti-virus disk activity. So anotehr CPU isnt going to help. It would be nice if AVG could ask the OS to not fill up the fisk cache with files that are going to be read just once. AVG, are you listening?:)
Well, if you're going to correct things, pls try correcting with correct facts. Edison came up with the carbon-filament, high-vacuum lamp. These were not too bad, but far from the tungsten-filament, argon-filled bulbs of just a few years later, which are what we still use now almost 100 years later.
There's nothing inherent in "high voltage" that is any better for bulbs, or more suited to use in parallel. Series lighting circuits are quite practical-- in fact most streetlights until the 1960's were series circuits. Many fluorescent lamps use the bulbs in series too.
Hate to bring facts into this, but while your raw red green or blue LED does have a near-infinite life, the typical white LED does not-- it has phosphors in there that get tired after a few years.
Also while LED's have made great strides in efficiency over the last 20 years, they're still no better than a fluorescent tube.
The average amount of cloud cover for the month of March is around 43%
The chances of a pilot getting lsot and straying is much higher when they can't see landmarks.
If it's cloudy the pilot better be "IFR", on instruments. This usually means he has the seat cranked down and eyes flitting between the altimeter, rate of climb, gyro compass, and the attitude indicators-- i.e. eyes in the cockpit.
So the whole laser concept works only in clear weather, when it's needed the least, and doesnt work at all when needed the most. Only something the govt could dream up.
IIRC water cuts gamma rays by a factor of TEN for every 3.4 feet. So anything below 34 feet is going to be protected by a factor of 10^10. Pretty good safety factor.
It seems rather unlikely that marine animals are going to be harder hit when they have that kind of protection.
Compressing air is a pitifully inefficient way to store energy. When you compress a gas, a considerable part of the energy goes into heat. That heat then goes into the storage tank, where it artificaially increases the back-pressure on the compressor, making it even harder to pump. the only up-side is when yu let out the compressed air,the air will be mighty cool, so it might make a fine auto air-conditioner .
It was a mildly thrilling book, as long as you ignored the fact they could have killed the whole microscopic swarm by doing something as simple as spraying some water on them. Surface tension beats nanites every time!
RTFA and learn there's no nanotech involved yet. And there's the not so small glitch involving the laws of physics. Like scale. Everytime you shrink a device by a factor of 10, its horsepower goes down by a factor of 1,000, but friction and surface tension only by a factor of 100. Do the shrinking a time or three and the thingy can't spin its motors or even lift itself off a surface.
As far as I can tell from the mediocre article, all they've done is gotten a tiny LED to flash at a rate of 10 giga flashes per second. I guess you could call that a "modulator". But it's just a fast LED. And they havent explained how it's going to be a economical and compact way to shuttle data around.
SO I'm expecting something with rubber bands, shock absorbers, integrating accelerometers, gyroscopes, microcontrollers doing video pattern tracking. Instead we get counterweights. And comments that say it works a lot better if you hold it steady.
OF COURSE IT WORKS BETTER IF YOU HOLD IT STEADY!
That's like all the diet pill commercials that say you'll lose weight if you take the pills and diet and exercise and catch giardia and aomebic dysentery.
Also their web page shows some distressing proofreading. "Get a sneak Peak", and they don't mean adopt-a-mountain.
I'm all for a new "Popular Electronics" for this century, but really now....
I suspect the glitches may be due to the mostly brain-dead API for setting up the codecs. Let's say a program wants to play a certain video clip that's in format F, bit-rate R, bit depth D, size X by Y. You call some API to inquire: is there a codec that can play [F,R,D,X,Y]?. Reply: No, but codec C can do [F,R] and codec C2 can convert [F,R] to [F,R2]. And so on, just like the "Who's on first" routine. A very coy API, where you never quite get a straight answer. LIke the guy said, don't attribute to maliciousness what you could attribute to stupidity.
What is this guy a professor of?
As others have noted, this isnt very likely to work in practice. It's not even good enough for an answering machine if it compresses the audio. Any good compression method is likely to be tripped up by even one bad bit. After all the goal of compression is to make every bit count!
In the case of CPU's, it doesnt seem likely that a random stuck bit is going to be innocuous. The quoted example of a LSB stuck on an adder is very contrived-- The arithmetic adder is probably less than 1% of a CPU's real estate. And again, even a LSB error is going to be unacceptable if any compressed or encrypted data goes through the adder, which is extremely likely these days.
And let's not forget programs like compilers and linkers, which use the adder to calculate things like addresses. Off by a bit isnt going to cut it for avery large range of applications. And this guy got $1M to research this hare-brain idea? Sheesh.
I suspect this is an early April Fools article preprint. IMHO it ISNT SCIENCE IF:
It only happened once (WOW signal, Viking methane, placebo trial).
It just doesnt make sense (homeopathy, placebo).
The original paper is riddled with ludicrous errors (enough radiation to kill the experimenters, in the cold fusion paper).
The effects are sub-sub-subminiscule (cosmic ray fraction, Alpha shift, Pioneer acceleration).
The effects are easily explained (cold fusion is likely to be a well-known palladium reaction, Wow signal is ubiquitous heterodyning).
And those are just the ones a disinterested dweeb like me can pick out. How many of the others are easily refuted by someone that knows a bit about astronomy?
Only the teensy tiny $$$$$ prob remains
on
Sunlight in a Tube
·
· Score: 1
Nice in theory. But has anybody costed out the $$$$ to make and install a solor collector of the required size? I'm estimating my office which has 20 80-watt fixtures would need about would need about TEN 1-meter diameter collectors. I can't see these going for much less than $1K installed each. That's an awful lot of $$$, especially when you still need the old lighting system for cloudy times and at night. The current lighting costs about 7 cents an hour, 56 cents a day, about $100 a year. There's no way to pay for the cost of the collectors, not even within a factor or ten.
It would be nice to have more than hype. IIRC the Intel hyperthreading documents were mostly hype, plus a few very unimpressive benchmarks. When benchmarks by the original company are borderline, a little bell should go off.
So now Sun has something similar. We're supposed to buy their new proprietary hardware and rewrite our programs and introduce concurrency bugs? And for what, a few percent improvement? Hmmmm.... Pass..
Yes indeedy,but error correction slows things down, as does narrowing the bandwidth. And these folks are claiming very high data rates. Something's mighty fishy.
Slashdot Mirror
Also I had the pleasure of taking apart an one of these Titan guidance computers. It was about the size of a big suitcase. Built to take many G's-- it had a aluminum case about 3/4 inch thick. All thge modules inside were potted in a tough pink styrofoam.
An amazing device with about 300 credit-card sized PC boards all plugged in and soldered into a backplane. Each PC board had what looked like four to six Westinghouse flat-pack IC's, probably DTL logic, maybe four gates max per chip. Amazing what they could do with that little hardware. The memory was some PC-board version of magnetic wire memory, as cores probably couldnt take the g's and vibration. Sobering to be poking through a device designed to land 9 Megatons on the Ruskies.
>>And the weight of copper is rarely a problem. >Unless you're sending stuff into space. Sending stuff into space *is* mighty rare. And one might estimate copper makes up somewhat less than 1% of the weight of the ISS. Much bigger weight concerns are the continuing need for several pounds of water per person per day. Figure out a way to dehydrate wather and then you're onto something.
And the weight of copper is rarely a problem.
Sorry to bring facts into this...
That's one heavy mothership to be pushed by shooting out tiny amounts of ions!
At a first approximation, the acceleration's gotta be somewhere under 0.00001G's, which makes the whole scheme rather slothful and boring.
And if there's going to be pumps and turbines involved, how're they going to keep the thing from spinning wildly in several axes? Surely not with reaction thrusters. The more I think about this thing, the more it looks like some gigantic delayed April Fools joke.
Go to a hardwar store and ask for a "humidifier belt". For $4.99 you'll get about a square yard of exactly the right material-- a lifetime supply.
Let's say the kernel bloats by 5 megabytes.
At today's memory prices, that's about 75 cents US.
How long would it take me to trim out and reboot and test a smaller kernel? Oh, at least $10 of my time. Plus the hard to quantify costs and risks of changing the OS and rebooting during the workday.
By this argument kernel bloat, just the memory consumption aspect, shouldnt be a big cause for concern. As to the potential for less stability when there are unused modules in memory, I don't see how code that's never called can cause stability problems. Any surmises?
A quick peek at the task manager in Windows XP shows that when AVG anti-virus is running, even on this old 600 MHz box, the CPU is only about 27% busy. The real slowdown is due to the disk cache getting flushed by all the anti-virus disk activity. So anotehr CPU isnt going to help. It would be nice if AVG could ask the OS to not fill up the fisk cache with files that are going to be read just once. AVG, are you listening? :)
There's nothing inherent in "high voltage" that is any better for bulbs, or more suited to use in parallel. Series lighting circuits are quite practical-- in fact most streetlights until the 1960's were series circuits. Many fluorescent lamps use the bulbs in series too.
Also while LED's have made great strides in efficiency over the last 20 years, they're still no better than a fluorescent tube.
Just a few gotchas:
It seems rather unlikely that marine animals are going to be harder hit when they have that kind of protection.
Compressing air is a pitifully inefficient way to store energy. When you compress a gas, a considerable part of the energy goes into heat. That heat then goes into the storage tank, where it artificaially increases the back-pressure on the compressor, making it even harder to pump. the only up-side is when yu let out the compressed air,the air will be mighty cool, so it might make a fine auto air-conditioner .
It was a mildly thrilling book, as long as you ignored the fact they could have killed the whole microscopic swarm by doing something as simple as spraying some water on them. Surface tension beats nanites every time!
RTFA and learn there's no nanotech involved yet. And there's the not so small glitch involving the laws of physics. Like scale. Everytime you shrink a device by a factor of 10, its horsepower goes down by a factor of 1,000, but friction and surface tension only by a factor of 100. Do the shrinking a time or three and the thingy can't spin its motors or even lift itself off a surface.
As far as I can tell from the mediocre article, all they've done is gotten a tiny LED to flash at a rate of 10 giga flashes per second. I guess you could call that a "modulator". But it's just a fast LED. And they havent explained how it's going to be a economical and compact way to shuttle data around.
OF COURSE IT WORKS BETTER IF YOU HOLD IT STEADY!
That's like all the diet pill commercials that say you'll lose weight if you take the pills and diet and exercise and catch giardia and aomebic dysentery.
Also their web page shows some distressing proofreading. "Get a sneak Peak", and they don't mean adopt-a-mountain.
I'm all for a new "Popular Electronics" for this century, but really now....
I suspect the glitches may be due to the mostly brain-dead API for setting up the codecs. Let's say a program wants to play a certain video clip that's in format F, bit-rate R, bit depth D, size X by Y. You call some API to inquire: is there a codec that can play [F,R,D,X,Y]?. Reply: No, but codec C can do [F,R] and codec C2 can convert [F,R] to [F,R2]. And so on, just like the "Who's on first" routine. A very coy API, where you never quite get a straight answer. LIke the guy said, don't attribute to maliciousness what you could attribute to stupidity.
What is this guy a professor of? As others have noted, this isnt very likely to work in practice. It's not even good enough for an answering machine if it compresses the audio. Any good compression method is likely to be tripped up by even one bad bit. After all the goal of compression is to make every bit count! In the case of CPU's, it doesnt seem likely that a random stuck bit is going to be innocuous. The quoted example of a LSB stuck on an adder is very contrived-- The arithmetic adder is probably less than 1% of a CPU's real estate. And again, even a LSB error is going to be unacceptable if any compressed or encrypted data goes through the adder, which is extremely likely these days. And let's not forget programs like compilers and linkers, which use the adder to calculate things like addresses. Off by a bit isnt going to cut it for avery large range of applications. And this guy got $1M to research this hare-brain idea? Sheesh.
- It only happened once (WOW signal, Viking methane, placebo trial).
- It just doesnt make sense (homeopathy, placebo).
- The original paper is riddled with ludicrous errors (enough radiation to kill the experimenters, in the cold fusion paper).
- The effects are sub-sub-subminiscule (cosmic ray fraction, Alpha shift, Pioneer acceleration).
- The effects are easily explained (cold fusion is likely to be a well-known palladium reaction, Wow signal is ubiquitous heterodyning).
And those are just the ones a disinterested dweeb like me can pick out. How many of the others are easily refuted by someone that knows a bit about astronomy?Nice in theory. But has anybody costed out the $$$$ to make and install a solor collector of the required size? I'm estimating my office which has 20 80-watt fixtures would need about would need about TEN 1-meter diameter collectors. I can't see these going for much less than $1K installed each. That's an awful lot of $$$, especially when you still need the old lighting system for cloudy times and at night. The current lighting costs about 7 cents an hour, 56 cents a day, about $100 a year. There's no way to pay for the cost of the collectors, not even within a factor or ten.
It would be nice to have more than hype. IIRC the Intel hyperthreading documents were mostly hype, plus a few very unimpressive benchmarks. When benchmarks by the original company are borderline, a little bell should go off. So now Sun has something similar. We're supposed to buy their new proprietary hardware and rewrite our programs and introduce concurrency bugs? And for what, a few percent improvement? Hmmmm.... Pass..
Yes indeedy,but error correction slows things down, as does narrowing the bandwidth. And these folks are claiming very high data rates. Something's mighty fishy.