If you had RTFA you would have known that they kept the computer on for 30 seconds every 2 hours, and powered it through winter on 5kg of Lithium batteries.
Which means we can grow these nanotube fibers almost ten thousand times faster than cotton fibers!
That's a very interesting way to think about it. Of course a research lab does take up much more space than a single cotton plant, and a single cotton plant makes more than one fiber, but still...
True, but the combined length of the fibers must be as long as the rope. (Well, longer, since there will be overlap). Before you say "just make lots at the same time", let me remind you that an elevator made of a single strand will be unable to lift anything significant. Lots must to be made at the same time just to get a sufficiently wide cable. Not that it matters really, it would be highly surprising if the first long carbon monofilament was made by an efficient process. A better process will be needed, and I am confident that it will happen.
1/10th of a centimeter is a millimeter. 1/10000th of a centimeter is a micrometer. Anyway, at 11um/s, 4cm takes an hour. Geosynch takes a hundred thousand years. Better get started.
In reality, they would in fact most likely use farmland. The trick is that the land can still be farmed. There isn't a sufficient amount of tall buildings to put turbines on top of, and I am sure most of the existing buildings would be unsuitable for it anyway. (Wind forces are already a major problem for tall buildings. Putting a big fan on top would tend to make the problem worse.)
People generally overestimate transmission losses anyway. It is not like most conventional or nuclear power plants are built in city centres.
If they are building turbines within half a mile of housing, they are crazy. Assuming that not to be the case, a lot of your objections become moot.
Anyway, Denmark is an important route for migratory birds. It also has a lot of wind turbines. Yet the wind turbines fail to kill any significant amount of birds. Buildings are much more of a concern.
The ice thing is interesting. I have never heard of it being a problem, even back when people were silly enough to build turbines close to houses. My guess would be that ice is most likely to get thrown off when the blade vibrates because of passing the tower. That would limit the throwing distance. Also, people don't generally spend much time in the vicinity of turbines in conditions of freezing rain.
Oh and if the landowners end up with 400 foot towers that aren't paid for, I'd suggest taking them down and selling them as scrap metal. That ought to cover the lost rent for many many years.
Wind energy is only really interesting with at least a 250kW turbine, and preferably something like 1 or 2MW. I am afraid you would find the cost rather prohibitive if you are an average joe.
Chernobyl was planned to be a negative-feedback reactor. The simulations at the time had shown it to be. Unfortunately the simulations were too simplistic; new simulations done (with vastly increased computing power, of course) after the accident showed that there were operating conditions with positive feedback.
Hopefully the simulations are good enough these days to avoid such mistakes.
Convincing the major labels to sell their songs for 70c in a format without DRM is probably impossible. Yes, we all know that DRM is a waste of energy, but music industry executives still have dreams. Or nightmares, perhaps.
And you agree with the other guy that linux loads application only once when 2 users are running it.....
Of course it loads applications only once. It is a fairly trivial optimization on systems with virtual memory, and it can be done even on systems without it. There are no security problems with it as long as memory isn't writable.
Incidentally, Windows does the same optimization for all applications, no matter which company they company from. You really need to get a clue.
I could ask you the same. Linux loads applications incrementally, and will drop unneeded pages of code without writing them to swap. Pages of code are certainly shared between all processes using them.
You're not even right about the stuff about Windows. OpenOffice is loaded incrementally too.
Perhaps your user name shouldn't be "badriram" but badRAM, and you should get it fixed?
I have no idea where people got the idea that the serial number is a text field. It is a simple 32 bit integer. However, it is supposed to be compared using "sequence space arithmetic". This has been defined in RFC 1982. Basically it means that overflows are fine, as long as no secondary nameserver keeps really old revisions around. So if you make a secondary for the.com zone now, unplug it for 40 years, and plug it in again, it may fail to get the latest zone.
Factorization is not an NP complete problem, a solution to the travelling salesman problem will not solve factorization.
That is the wrong way around. If you solve travelling salesman, you have solved factorization. On the other hand, if you solve factorization, you have not necessarily solved travelling salesman.
[BTW the gender is intentional, the reason the salesman is visiting the cities in the first place is he has a mistress in each town]
In the test, the filesystem eventually failed to mount, and reiserfsck failed to fix it. Not very nice at all. Some lost data is one thing, a lost filesystem is quite another.
My point about the mechanical system is that it is suboptimal, simply because it is mechanical. The design is constrained because of that. Paper is much more flexible.
And yes, I think that a system based on judgement calls is more secure and more accurate than an electronic system. It is easy to know how large a problem mistaken votes are, you simply count how many ballots were considered "questionable" and had to go in for a second check. With an electronic system, a mistake looks no different from an intentional vote. Experience shows that with paper ballots there are very few mistaken or dubious votes. In case an election is really close, and those votes might actually matter, you simply count very very carefully.
The machine is a one-size-fits-all for all elections. It is completely mechanical, and as such it is not reconfigurable.
And yes, there will be incorrectly filled ballots. That is where humans make judgement calls. Humans are pretty good at that, except when you force them to obey hard rules.
Yes, you can make the tallying arbitrarily secure. But each additional control seems to bring a cost in anonymity and in transparency. For voting, anonymity and transparency are vital.
Mandelbrot is simply pointing out that there is not much fundamental research into market dynamics. In fact, it seems there is much more interest in the "weather prediction" kind of market research, rather than the "climate prediction" kind. It would be very interesting to know just how risky the markets are, in general. It seems right now that everyone predicts trends which are generally pretty correct, but once in a while most everyone is completely wrong. If we could say something about how common such mistakes are, risk management would be easier -- even if each occurence by itself is entirely unpredictable.
Why, pray tell, is it good for the world to pay lots of hedge funds to reinvent the wheel over and over? What would be so horrible about obsoleting the hedge fund industry?
The alignment problem exists because the ballots were formatted for use in a voting machine. Noone sane would format the ballot that way, except to fit the constraints of the machine.
No need for plugging anything in when the voting machines have built-in wireless. It is said that you should never attribute to malice what can be adequately explained by stupidity. However I refuse to believe that stupidity alone can make anyone think that wireless is a good idea in a voting machine.
How do you prove that the system is more secure? How can I, as a concerned citizen, verify the security? Right now voting is transparent (at least around here). You can go watch the count if you don't trust the people counting. The security is obvious, and the breakdown of security is equally obvious and limited. (Someone might tamper with ballots when everyone else happens to look away. Limited damage done, easy to detect, easy to find the guilty.)
Slashdot Mirror
If you had RTFA you would have known that they kept the computer on for 30 seconds every 2 hours, and powered it through winter on 5kg of Lithium batteries.
That's a very interesting way to think about it. Of course a research lab does take up much more space than a single cotton plant, and a single cotton plant makes more than one fiber, but still...
True, but the combined length of the fibers must be as long as the rope. (Well, longer, since there will be overlap). Before you say "just make lots at the same time", let me remind you that an elevator made of a single strand will be unable to lift anything significant. Lots must to be made at the same time just to get a sufficiently wide cable. Not that it matters really, it would be highly surprising if the first long carbon monofilament was made by an efficient process. A better process will be needed, and I am confident that it will happen.
1/10th of a centimeter is a millimeter. 1/10000th of a centimeter is a micrometer. Anyway, at 11um/s, 4cm takes an hour. Geosynch takes a hundred thousand years. Better get started.
People generally overestimate transmission losses anyway. It is not like most conventional or nuclear power plants are built in city centres.
Anyway, Denmark is an important route for migratory birds. It also has a lot of wind turbines. Yet the wind turbines fail to kill any significant amount of birds. Buildings are much more of a concern.
The ice thing is interesting. I have never heard of it being a problem, even back when people were silly enough to build turbines close to houses. My guess would be that ice is most likely to get thrown off when the blade vibrates because of passing the tower. That would limit the throwing distance. Also, people don't generally spend much time in the vicinity of turbines in conditions of freezing rain.
Oh and if the landowners end up with 400 foot towers that aren't paid for, I'd suggest taking them down and selling them as scrap metal. That ought to cover the lost rent for many many years.
Wind energy is only really interesting with at least a 250kW turbine, and preferably something like 1 or 2MW. I am afraid you would find the cost rather prohibitive if you are an average joe.
Hopefully the simulations are good enough these days to avoid such mistakes.
I think a lot of people would have woken up very confused on 2001-09-12 if the day before had been 2001-09-10.
Convincing the major labels to sell their songs for 70c in a format without DRM is probably impossible. Yes, we all know that DRM is a waste of energy, but music industry executives still have dreams. Or nightmares, perhaps.
Of course it loads applications only once. It is a fairly trivial optimization on systems with virtual memory, and it can be done even on systems without it. There are no security problems with it as long as memory isn't writable.
Incidentally, Windows does the same optimization for all applications, no matter which company they company from. You really need to get a clue.
I could ask you the same. Linux loads applications incrementally, and will drop unneeded pages of code without writing them to swap. Pages of code are certainly shared between all processes using them.
You're not even right about the stuff about Windows. OpenOffice is loaded incrementally too.
Perhaps your user name shouldn't be "badriram" but badRAM, and you should get it fixed?
I have no idea where people got the idea that the serial number is a text field. It is a simple 32 bit integer. However, it is supposed to be compared using "sequence space arithmetic". This has been defined in RFC 1982. Basically it means that overflows are fine, as long as no secondary nameserver keeps really old revisions around. So if you make a secondary for the .com zone now, unplug it for 40 years, and plug it in again, it may fail to get the latest zone.
That is the wrong way around. If you solve travelling salesman, you have solved factorization. On the other hand, if you solve factorization, you have not necessarily solved travelling salesman.
[BTW the gender is intentional, the reason the salesman is visiting the cities in the first place is he has a mistress in each town]
No wonder it is difficult for large N.
No, Hz/s is correct here. They are talking about a frequency drift rate.
In the test, the filesystem eventually failed to mount, and reiserfsck failed to fix it. Not very nice at all. Some lost data is one thing, a lost filesystem is quite another.
And yes, I think that a system based on judgement calls is more secure and more accurate than an electronic system. It is easy to know how large a problem mistaken votes are, you simply count how many ballots were considered "questionable" and had to go in for a second check. With an electronic system, a mistake looks no different from an intentional vote. Experience shows that with paper ballots there are very few mistaken or dubious votes. In case an election is really close, and those votes might actually matter, you simply count very very carefully.
And yes, there will be incorrectly filled ballots. That is where humans make judgement calls. Humans are pretty good at that, except when you force them to obey hard rules.
Yes, you can make the tallying arbitrarily secure. But each additional control seems to bring a cost in anonymity and in transparency. For voting, anonymity and transparency are vital.
Mandelbrot is simply pointing out that there is not much fundamental research into market dynamics. In fact, it seems there is much more interest in the "weather prediction" kind of market research, rather than the "climate prediction" kind. It would be very interesting to know just how risky the markets are, in general. It seems right now that everyone predicts trends which are generally pretty correct, but once in a while most everyone is completely wrong. If we could say something about how common such mistakes are, risk management would be easier -- even if each occurence by itself is entirely unpredictable.
Why, pray tell, is it good for the world to pay lots of hedge funds to reinvent the wheel over and over? What would be so horrible about obsoleting the hedge fund industry?
It seems like they cannot. What then will distinguish each server, and what is the point of the whole exercise?
The alignment problem exists because the ballots were formatted for use in a voting machine. Noone sane would format the ballot that way, except to fit the constraints of the machine.
No need for plugging anything in when the voting machines have built-in wireless. It is said that you should never attribute to malice what can be adequately explained by stupidity. However I refuse to believe that stupidity alone can make anyone think that wireless is a good idea in a voting machine.
How do you prove that the system is more secure? How can I, as a concerned citizen, verify the security? Right now voting is transparent (at least around here). You can go watch the count if you don't trust the people counting. The security is obvious, and the breakdown of security is equally obvious and limited. (Someone might tamper with ballots when everyone else happens to look away. Limited damage done, easy to detect, easy to find the guilty.)