It all depends on how much temporal locality you have in your algorithm. If you are trying to get the dot product of two gigantic vectors, you will be limited by memory bandwidth. If you're trying to multiply or invert matrices, there is a great deal of temporal locality, and recent P4s can sustain over 4Gflops on large matrices.
So have the U.S. Attorney charge him with treason if he feels those charges are justified, have a trial, and if he's found guilty, by all means, take him out back and deal with him. America was founded on the concept of public jury trials, not incarcaration based on John Ashcroft's say-so.
Not true. The standard way to reprocess fuel is using the purex process, which produces plutonium. There is another process, however called pyroprocessing which does not produce weapons-usable material.
Chernobyl killed a lot more than 44 people. If you count the effects from a decent fraction of the core going up into the atmosphere and being breathed in by the surrounding population, the expected number of cancers is somewhere around 3000-5000. Furthermore, Three Mile Island was a near disaster. The fuel partially melted, and if a few things had gone differently, there could easily have been a loss of containment. You might say that things are different now, but just last year you had the Davis-Besse situation where there could have been an explosive loss of coolant accident. I am a big supporter of nuclear power in principle. The problem is that safety is not necessarily the #1 driver in day to day operations. Companies want to make as much money as possible, so they will keep plants operating that should be shut down for repairs. They know that in the unlikely event that there is an accident, their insurance and the government will bail them out. It should be a criminal offense to be an officer of a corporation that has a major nuclear accident. Safely operating a nuclear plant is very complicated and requires a great deal of dedication. Unfortunately the incentives are not right in the deregulated electricity industry for this to happen.
It's a primitive cruise missile. This thing was designed to be launched out of the back of a pickup truck going down the highway. There are plenty of lonely highways within 100 miles of Washington, and a number of targets that a terrorist would find appealing. More range is pointless as it just increases the flight time and thus the chance of intercept.
The halting problem says that it is impossible to write a program that will decide halting for every single possible program. The real problem is a certain class of pathological problems. There's no reason to believe that a specific program devised by a GA could not be proven to halt. It is also quite possible to prove that certain optimizaitons will preserve certain properties, such as halting. It is not possible to prove whether any given optimization will preserve halting.
I really don't think Friendster is going to have a problem, as it is virtually the same product as Sixdegrees was in 1997, except Sixdegrees didn't have the dating angle.
The way this works on FX is this: You want to move money from player A to player B. You find a thinly traded claim, say currently trading at a price of 50. Player A offers to sell a bunch at 20. Immediately, most of the other orders on the board will execute, but you picked a thinly traded claim, so they should represent a relatively small fraction of the ammount you wished to move. Now there is a big offer to sell on the board for 20, which player B scoops up. Then the process is reversed, with player A offering to buy at 70. Player B then unloads the claims he bought at 20 for 70 to player A. There are a lot of variations on this theme with multiple accounts and such, that you could think up.
In the stock market, it's a lot harder to do this type of thing, because of several factors. First, there aren't any stocks that are as thinly traded as the thinnest claims on FX. Second, even if you pick the smallest stock off the pink sheets, you will be dealing with a broker. If you offer to buy at a very high price or sell at a very low one, the broker may just execute the transaction for himself, or with another market maker. If you execute on an ECN, someone else may snag the trade before you have a chance to yourself. Furthermore, in its pure form, you need to be able to sell short. You usually can't do this on a penny stock. Furthermore, these transactions are logged, and this type of transaction would be likely to arouse suspicion if you were way out of line with the market. Therefore, it would be tough to do this with good efficiency, because the other traders and commissions would eat into the money you were trying to transfer.
Microsoft just got hit with a $500M damage award for past infringement. You can buy a lot of lawyers for that much money, even discounting the cost of modifying IE. I'm sure they did their best to fight the patent.
You might want to look into getting a cellsocket. It's not exactly what you are asking for, but it's a cradle for your cell phone that charges the battery, provides an external antenna hookup, and plugs into your phone jack. You can then use regular phones in your house, and it will send the call through the cell phone.
The laserjet 4 is prone to having the "accordion" jams where the paper gets jammed, and when you pull it out from the back, it is folded in an accordion shape. You can fix this yourself by installing new upper rollers. The kits are about $25, and it takes about a half hour to install them. Since I did this to my printer, it hasn't jammed.
Yeah I know all of that, but if you fly a lot, you'll be spending at least 800 a year in tow fees. Believe me, I do, and I haven't found a glider simulator that gives me the same dumbass grin you get from finding yourself exactly 500' below cloudbase...
With a couple hours of instruction, any idiot can land a small airplane if you don't care about landing on the numbers. A license is the easiest thing in the world to forge -- it's just a piece of paper. You could run one off on any laser printer. A logbook isn't much harder. You'll want fake instructor and examiner signoffs, but they probably won't check the numbers, and a lot of those guys write illegibly anyhow, so in a lot of ways it's easier than faking a driver's license.
Most people can only soar on the weekends, in good weather, when the tow pilot is out, durring the day, when the towplane and glider have had their annuals and everything is in working order, when the thermals are popping. Then, if you want to go x-c, you have to have a crew, a trailer, and a truck. Then you have to watch out for boulder-strewn fields because it's important that you bring it in in one piece. Granted, you aren't going to get the shit-eating grin of soaring for real, but there are advantages.
Yeah, but you have to pay most loans back. If the album only sells a few copies, the artist still keeps the advance. If the album makes money, then it retroactively becomes a loan. Otherwise it is just payment for making the album.
The DF-5A can hit almost any landmass in the world except South America and little bits of Africa. Not all of China's missiles are upgraded to the new standard, but the definitely have at least a few that are capable of hitting anywhere in the U.S. If they can wipe out 4-5 eastern cities and more western ones, that is enough of a deterrent.
All you need is a universally acceptable micropayment infrastructure. Oh yeah, and to send email now you need an account with money. You have to be able to deposit this somehow. If you don't have a credit card, you are out of luck. This shuts down all free email services. Payment based spam solutions should be a very last resort, as it is the end of email as we know it. I want to be able to communicate with people that have no prior relationship with me. I just don't want my email box full of spam.
Any company can grow exponentially -- it's just a question of the exponent. I could start a company today and buy t-bonds and it would grow a couple of percent every year.
Lockheed is a partner with Boeing in United Space Alliance, which is the primary shuttle contractor since they essentally privatized the program a few years back. I really don't think it's in their interests to have the program cancelled.
I'm not an aeronautical engineer, but I do know of a few reasons why a 747 doesn't fly at 60,000 feet. There are tons of tradeoffs in design, but a few of the simpler ones are these:
1) The speed of sound goes down as you go higher. Since drag goes way up when you start approaching the speed of sound, and aerodynamics change in other ways, airliners are generally limited to in the neighborhood of.8 mach. Going higher means you go slower.
2) To make an airplane go fast, you generally want the smallest wings you can get away with, as big wings will produce more drag and add more weight. At high altitudes, the air is thinner, so if you want the plane to fly at an efficient angle of attack, you need bigger wings. This means that you have to sacrifice performance on the way up and on the way down, and you have to haul more weight in wings.
3) Pressurization is an issue as you mentioned.
4) Engines are designed to be most efficient at a certain altitude. If you spend a lot of time climbing and descending, you are spending a lot of time operating the engines inefficiently
5) There is a certain degree of energy loss climbing and descending (The wing is not perfectly efficient, so you don't get back all the energy you spent on the climb), and so it's only worth climbing if you're going to spend some time at the higher altitudes.
So all of these factors (and probably a bunch more that I don't know about) combine to make subsonic passenger aviation most efficient at around 30,000 feet. Now that's certainly not to say that technology limits your altitude, as demonstrated by the U-2, Concorde, SR-71, etc. However, it's not economical to go that high with passengers on a normal flight, and the airline business is all about econcomics.
The point is that it's significant, but would aviation really be any different if no one had ever crossed the atlantic solo? Every passenger flight has multiple pilots, and the vast majority of military flights across oceans have more than one pilot. Getting across in an airplane is a much more significant milestone than doing it with only one person on board.
Slashdot Mirror
It all depends on how much temporal locality you have in your algorithm. If you are trying to get the dot product of two gigantic vectors, you will be limited by memory bandwidth. If you're trying to multiply or invert matrices, there is a great deal of temporal locality, and recent P4s can sustain over 4Gflops on large matrices.
So have the U.S. Attorney charge him with treason if he feels those charges are justified, have a trial, and if he's found guilty, by all means, take him out back and deal with him. America was founded on the concept of public jury trials, not incarcaration based on John Ashcroft's say-so.
Not true. The standard way to reprocess fuel is using the purex process, which produces plutonium. There is another process, however called pyroprocessing which does not produce weapons-usable material.
Chernobyl killed a lot more than 44 people. If you count the effects from a decent fraction of the core going up into the atmosphere and being breathed in by the surrounding population, the expected number of cancers is somewhere around 3000-5000. Furthermore, Three Mile Island was a near disaster. The fuel partially melted, and if a few things had gone differently, there could easily have been a loss of containment. You might say that things are different now, but just last year you had the Davis-Besse situation where there could have been an explosive loss of coolant accident. I am a big supporter of nuclear power in principle. The problem is that safety is not necessarily the #1 driver in day to day operations. Companies want to make as much money as possible, so they will keep plants operating that should be shut down for repairs. They know that in the unlikely event that there is an accident, their insurance and the government will bail them out. It should be a criminal offense to be an officer of a corporation that has a major nuclear accident. Safely operating a nuclear plant is very complicated and requires a great deal of dedication. Unfortunately the incentives are not right in the deregulated electricity industry for this to happen.
It's a primitive cruise missile. This thing was designed to be launched out of the back of a pickup truck going down the highway. There are plenty of lonely highways within 100 miles of Washington, and a number of targets that a terrorist would find appealing. More range is pointless as it just increases the flight time and thus the chance of intercept.
About a year and a half ago. Lookup Jose Padilla.
The halting problem says that it is impossible to write a program that will decide halting for every single possible program. The real problem is a certain class of pathological problems. There's no reason to believe that a specific program devised by a GA could not be proven to halt. It is also quite possible to prove that certain optimizaitons will preserve certain properties, such as halting. It is not possible to prove whether any given optimization will preserve halting.
Didn't notice that. I think Friendster has a problem then. As I mentioned before, it's basically the same service.
I really don't think Friendster is going to have a problem, as it is virtually the same product as Sixdegrees was in 1997, except Sixdegrees didn't have the dating angle.
The way this works on FX is this: You want to move money from player A to player B. You find a thinly traded claim, say currently trading at a price of 50. Player A offers to sell a bunch at 20. Immediately, most of the other orders on the board will execute, but you picked a thinly traded claim, so they should represent a relatively small fraction of the ammount you wished to move. Now there is a big offer to sell on the board for 20, which player B scoops up. Then the process is reversed, with player A offering to buy at 70. Player B then unloads the claims he bought at 20 for 70 to player A. There are a lot of variations on this theme with multiple accounts and such, that you could think up.
In the stock market, it's a lot harder to do this type of thing, because of several factors. First, there aren't any stocks that are as thinly traded as the thinnest claims on FX. Second, even if you pick the smallest stock off the pink sheets, you will be dealing with a broker. If you offer to buy at a very high price or sell at a very low one, the broker may just execute the transaction for himself, or with another market maker. If you execute on an ECN, someone else may snag the trade before you have a chance to yourself. Furthermore, in its pure form, you need to be able to sell short. You usually can't do this on a penny stock. Furthermore, these transactions are logged, and this type of transaction would be likely to arouse suspicion if you were way out of line with the market. Therefore, it would be tough to do this with good efficiency, because the other traders and commissions would eat into the money you were trying to transfer.
Microsoft just got hit with a $500M damage award for past infringement. You can buy a lot of lawyers for that much money, even discounting the cost of modifying IE. I'm sure they did their best to fight the patent.
You might want to look into getting a cellsocket. It's not exactly what you are asking for, but it's a cradle for your cell phone that charges the battery, provides an external antenna hookup, and plugs into your phone jack. You can then use regular phones in your house, and it will send the call through the cell phone.
The laserjet 4 is prone to having the "accordion" jams where the paper gets jammed, and when you pull it out from the back, it is folded in an accordion shape. You can fix this yourself by installing new upper rollers. The kits are about $25, and it takes about a half hour to install them. Since I did this to my printer, it hasn't jammed.
Yeah I know all of that, but if you fly a lot, you'll be spending at least 800 a year in tow fees. Believe me, I do, and I haven't found a glider simulator that gives me the same dumbass grin you get from finding yourself exactly 500' below cloudbase...
I'm not. Clearly successful artists are subsidizing unsuccessful ones, and that subsidy is part of the label's cut.
Yeah but the expenses of maintaining and flying it are going to be a hell of a lot more than 2500 pounds over a few years.
With a couple hours of instruction, any idiot can land a small airplane if you don't care about landing on the numbers. A license is the easiest thing in the world to forge -- it's just a piece of paper. You could run one off on any laser printer. A logbook isn't much harder. You'll want fake instructor and examiner signoffs, but they probably won't check the numbers, and a lot of those guys write illegibly anyhow, so in a lot of ways it's easier than faking a driver's license.
Most people can only soar on the weekends, in good weather, when the tow pilot is out, durring the day, when the towplane and glider have had their annuals and everything is in working order, when the thermals are popping. Then, if you want to go x-c, you have to have a crew, a trailer, and a truck. Then you have to watch out for boulder-strewn fields because it's important that you bring it in in one piece. Granted, you aren't going to get the shit-eating grin of soaring for real, but there are advantages.
Yeah, but you have to pay most loans back. If the album only sells a few copies, the artist still keeps the advance. If the album makes money, then it retroactively becomes a loan. Otherwise it is just payment for making the album.
The DF-5A can hit almost any landmass in the world except South America and little bits of Africa. Not all of China's missiles are upgraded to the new standard, but the definitely have at least a few that are capable of hitting anywhere in the U.S. If they can wipe out 4-5 eastern cities and more western ones, that is enough of a deterrent.
All you need is a universally acceptable micropayment infrastructure. Oh yeah, and to send email now you need an account with money. You have to be able to deposit this somehow. If you don't have a credit card, you are out of luck. This shuts down all free email services. Payment based spam solutions should be a very last resort, as it is the end of email as we know it. I want to be able to communicate with people that have no prior relationship with me. I just don't want my email box full of spam.
Any company can grow exponentially -- it's just a question of the exponent. I could start a company today and buy t-bonds and it would grow a couple of percent every year.
Lockheed is a partner with Boeing in United Space Alliance, which is the primary shuttle contractor since they essentally privatized the program a few years back. I really don't think it's in their interests to have the program cancelled.
I'm not an aeronautical engineer, but I do know of a few reasons why a 747 doesn't fly at 60,000 feet. There are tons of tradeoffs in design, but a few of the simpler ones are these:
.8 mach. Going higher means you go slower.
1) The speed of sound goes down as you go higher. Since drag goes way up when you start approaching the speed of sound, and aerodynamics change in other ways, airliners are generally limited to in the neighborhood of
2) To make an airplane go fast, you generally want the smallest wings you can get away with, as big wings will produce more drag and add more weight. At high altitudes, the air is thinner, so if you want the plane to fly at an efficient angle of attack, you need bigger wings. This means that you have to sacrifice performance on the way up and on the way down, and you have to haul more weight in wings.
3) Pressurization is an issue as you mentioned.
4) Engines are designed to be most efficient at a certain altitude. If you spend a lot of time climbing and descending, you are spending a lot of time operating the engines inefficiently
5) There is a certain degree of energy loss climbing and descending (The wing is not perfectly efficient, so you don't get back all the energy you spent on the climb), and so it's only worth climbing if you're going to spend some time at the higher altitudes.
So all of these factors (and probably a bunch more that I don't know about) combine to make subsonic passenger aviation most efficient at around 30,000 feet. Now that's certainly not to say that technology limits your altitude, as demonstrated by the U-2, Concorde, SR-71, etc. However, it's not economical to go that high with passengers on a normal flight, and the airline business is all about econcomics.
The point is that it's significant, but would aviation really be any different if no one had ever crossed the atlantic solo? Every passenger flight has multiple pilots, and the vast majority of military flights across oceans have more than one pilot. Getting across in an airplane is a much more significant milestone than doing it with only one person on board.