The reason nuclear power plants are very frequently along the costs is that they need huge amounts of cooling water. That means either the sea, or a very large river. There are not many rivers large enough.
Solar of the direct thermal, molten-salt variety, storing heat in an underground salt tank for overnight, does not need batteries. I agree that pure photovoltaic solar will never cut it, but that is not the only form of solar around.
According to New Scientist, coal kills about 13,000 Americans per annum. In a chart in their most recent edition, coal is by far the most lethal power source per billion GWh generated.
I wouldn't say so. A particle is just an identifiable thing, something which retains its identity over time. You can identify that the thingy you saw at one position at one time is the same thingy as you saw in another position at another time. You have some reason for saying that e.g. you tracked it through several intermediate positions. In 3S1T spacetime it takes 3 values to identify its position, in 1S1T it takes only 1. But the key fact about a particle is its persistence, not its dimensionality,
Backwards and forwards along the line. Each particle, if such exists, is bouncing back and forward against its neighbour. Or, perhaps more likely, there are no particles, and the universe is one ginormous string twanging along its length with compression waves which along though and interfere with each other. A bit like Gods organ pipe (God's bong?)
It is a word - to me. I read the summary and received what I believed to be the author's meaning without a stutter in my reading. Therefore it achieved what words are meant to do - carry meaning. It may be an ugly neologism, but you can only fight that by using what are, in your opinion, the correct alternatives, and have them carry the day by their superior usefulness.
When I left the industry three years ago, broadcasters were using either 720p for fast moving stuff (sports) or 1080i for hi res stuff (drama, documentary), both using data compressed to 100Mbit/sec. Generally, studio infrastructure is standardizing on 100Mbit/sec for post-production, so you aren't going to get more underlying data even if they do upgrade to 1080p. It would be cheaper to fit an upscaler in your TV.
Buying local food puts a cap on the price rise. If local food costs 25% more than imported food, once increased fuel costs have raised the price 30% the price rise will stop as local production (after a delay) replaces imported. This stops a runaway effect where rising costs everywhere put up prices on things consumed everywhere.
So, in the case of Airbus, it would be the CEO of EADS which owns Airbus, which at the time of the accident was two political appointees, one French and one German, who probably know little about engineering. Or possibly the heads of state of France, Germany and Spain which have controlling shareholdings in EADS. And in the case if the Herald of Free Enterprise, a CEO who had had responsibility for the ship for three weeks, when the ship had been operating/apparently/ safely for many years.
Nice for vengeance to see such a suit going to prison, but in terms of preventing future disasters, it will mean only that competent people will refuse such jobs out of fear of taking the blame for a disaster they could not prevent, and be replaced by less-competent chancers who will take high pay in the hope that no disaster will happen on their watch, thus increasing the likelihood of problems.
Another case of the US feeling that it is perfectly acceptable to treat foreigners in a way that would bring outrage if it tried it on its own citizens. Such things started, and were found acceptable: in cases of outright declared war (World Wars, Korea, Vietnam), it is OK to deceive and manipulate enemy civilians. But that has been translated to being OK to do the same to all except strong allies - and it would be a small step to apply it to them. The philosophy has become that war justifies lethal force, and non-war justifies anything except lethal force.
What would the US do if another country started doing the same within the US?
Hey - maybe they have. Maybe all the appalling right-wing bigots are really Chinese sock-puppets intended to discredit the US. I have never met an American who offered those views: all the ones I meet are nice people. Perhaps the bad side of America that I hear of but never see is just a huge cyber sock puppet! Damn!
That is because the pilot is the ultimate catch-all for errors. There is a general presumption that, if the automatic systems fail, the pilot will fix the problem manually. And if, by hindsight, we can show that there is something the pilot might possibly have done and didn't, then pilot error is a contributing factor.
But AI may be taught, rather than programmed. Google Translate "learns" by taking known translations of documents (particularly EU directives for which machine-readable versions in several languages are available) and correlating two versions to find out how to transform one language into another. If it mistranslates, is it the fault of the programmer, who need not necessarily have spoken either of the languages concerned? Is the computer "doing the boring labour" of a task the programmer was incapable of?
Not that the software gave bad readings: that the software reacted badly when hardware gave it bad readings, possibly because of stormy weather. The suggestion is that the air speed readings were faulty, and the software throttled back until the aircraft fell out of the sky.
And an A3330 has two engines. But it is not the engine control software which is suspect, but the main flight control software, which may have ordered the engines to reduce power when it shouldn't.
What we need is strong, forensically reliable audit trails for every policy and decision.
The problem is that in many cases of corporate manslaughter, is is lack of decision which is the problem. Nobody thought to run the tests which would have found the fault. And that may be up to a corporate structure which split responsibility so that everybody thought somebody else was taking care of it. That is why an offence of Corporate Manslaughter is useful. Often, it is impossible to say which individual made the mistake that caused the death, but it is possible to say that the corporation as a whole was flawed in a way that caused the accident,
Consider the Herald of Free Enterprise disaster: the ship, shuttling backwards and forwards several times each day, had six captains. Each has probably taken up the flawed working practice from predecessors who had been doing it for years. The Director of Safety at corporate level had only had responsibility for three weeks since the company had take over the ships. Tracking back to who originated the dodgy practice years ago would probably be impossible. And they never thought they were doing anything unusual, so there was nothing to write down to be audited.
Manslaughter is vary often a crime of omission. Somebody didn't think through what they were doing. And audit trails are bad at catching omissions.
I don't actually have much problem with computers knowing everything about me - it is when people get information that I worry. I use GMail, and it posts "targeted" ads alongside my mails. Very occasionally, they are of interest, mostly I can see the keyword they are responding to, sometimes I just wonder WIHIH. But it doesn't worry me. And the same applies more generally. I intend to obey speed limits: I have no problem with a computer checking that I do so. If I have strange sexual tastes, i have no problem with a computer knowing it - it may even be able to guide me ("people who viewed what you just viewed also..."
It is purely when it gets into the hands of people that I get nervous. And that includes "legitimate" users such as law authorities, Because they will judge me: it is what people do. And their values will not be my values, so they will judge me by values I do not understand. (As a geek, I am not much of a "people person", so I often don't understand other people's values).
So if this 100% surveillance is totally automated, I don't have a problem. By all means, track where I go for traffic purposes. Measure how much I pee - what do I care? Correlate automatically my this against my that - feel free. But DON'T TELL ANYBODY! Because they WILL judge me (as I, if the situations were reversed, would judge them. Nobody, but nobody, can help being at least somewhat judgmental.
Of course, there have to be exceptions for investigating crimes. And I grudgingly accept that - a system of warrants etc. The trouble is, once the data is in a computer, it is too easy to release. In physical searches, the process of getting a warrant and doing the search was laborious and obvious enough that it tended to be self-limiting: you knew you were being searched, and could protest. Not that there were no abuses, but that the abuses were small enough that the benefits of the system outweighed the costs of the abuses.
But with purely electronic searches, it is too easy to set the search too wide, and to easy to do searches without the object of the search knowing. about it. And I am cynic enough to say with absolute confidence that if the system is capable of being abused, it will be abused - by the over-zealous, by the nosy, and by the criminal.
Since the report shows that the flaw was detected in testing, they are not doing so. No doubt they were designed not to be susceptible, but being human the design did not achieved that goal. Rather then assuming that they were designed right, they were testing them - which necessarily requires production units to test. And the test showed a flaw, and they are not rolling out the systems because of the flaw.
This is the system working properly. Proper testing before aircraft delivery has shown a problem, and steps are being taken to rectify the problem. The short term fix is to outlaw WiFi; when they have found and tested a proper fix, they may (assuming tests are passed) then enable the equipment.
There is no point in getting alarmed when the proper test show the faults before the aircraft leave the ground.
But then you have to stock different sizes of envelope. How about pre-printing a small sticky label representing the amount paid which you attach to the envelope of your choice? Can I patent that idea?
It would all be solved if we could microcharge for email. At $0.001 per email, no sensible use of email would build up a significant cost, even for the very poor. But if your message to the whole company cost $5, you might think twice. And the spam industry, of course, would become much less viable. I don't know how you could do it safely, reliably, unfraudably, but it would be nice to try.
One of the reasons for encrypting the accurate GPS and adding an error to the public version was that in order to take out a hardened silo, an offensive missile needs to strike the ground within two meters outside the hardened lid of the the silo. It then penetrates the ground and explodes alongside the body if the silo, which is softer than the lid. That accuracy might be hard to achieve purely ballistically,
We don't need liquidity at picosecond resolutions. We need liquidity at human reaction times. At the bottom of everything is human investment decisions - even if it is sometimes the decision to entrust your money to a bot. We need a market fluid enough for people to be able to get their money in and out at rates that are, to them, fast: perhaps a few minutes. Which means that the underlying market has to ace maybe an order of magnitude faster to handle peaks and avoid frictional problems. So a trading speed of a few tens of seconds is entirely adequate for a market that serves societies needs. Anything else is traders exploiting unintended opportunities which are not part of the normal functioning of the market. Which is not, in itself, wrong: freedom to exploit such opportunities is fundamental to innovation. But the cost of their high speed computer systems and the profits they take are a cost to the market for (as far as I can see) no corresponding gain. So I don't like this sort of trading, and its speeding up. But "I don't like it" is not a sufficient reason to ban something. If we want to damp such (in my opinion) excessive speed-freakery, we need to do so with great care, being well aware of the Law of Unintended Consequences. By bet would be for a micro-levy on transactions. Say 0.01%. The same computers which are speed trading could easily charge the levy. It would be small enough that it would be negligible for "real" transactions, but would damp out ultra-high-speed churn.
The trouble with all these schemes is that they are fine for those who are already criminals, but they push those who want to campaign peacefully for political change into criminal or criminal-like behaviour. You should have an assumption of privacy: until you appear to be something nefarious, your comings and goings should be your private business. This means that the authorities can, at their choice, track your recent movements and get a good idea of your contacts. Of course, this can be used for good, but history shows that good uses are soon overwhelmed by authoritarian control.
That is always true, and NaCl makes little difference to this. Of course, the more different ways you use an OS, the more different places you get to probe for bugs. If you never try anything new, you will not find new bugs, but you will not get the benefit of new capabilities. Take this to an extreme, and you end up Amish: they have decided not to take up new technology because it brings new risks.
No, you are still running in User mode rather than Kernel mode. The OS still gets to trap and inspect all your accesses, so that you can only look at the HD in the same way as any user program can.
Not necessarily "perfectly good". A structurally sound house in the wrong place is not perfect and not really good. In a way, this mirrors the soviet failure, rather than capitalist problems. The soviets assumed that if a factory was working at full speed producing whatever had been specified, it was doing good work. But producing obsolete or excessive goods is a net loss. If you could move houses from the rustbelt to the sunbelt, your observation might be true. But you cannot, and it is better to scale back the shrinking communities to a functional size than continue to mimic a city with four times the population.
(Or you can try to relocate jobs to where the houses are. if you succeed in that, your fortune is made, just on the lecture circuit).
Slashdot Mirror
The reason nuclear power plants are very frequently along the costs is that they need huge amounts of cooling water. That means either the sea, or a very large river. There are not many rivers large enough.
The plant survived the 9.0 earthquake perfectly well. The automatic shutdown systems worked properly. It was the Tsunami that caused the problems.
Solar of the direct thermal, molten-salt variety, storing heat in an underground salt tank for overnight, does not need batteries. I agree that pure photovoltaic solar will never cut it, but that is not the only form of solar around.
According to New Scientist, coal kills about 13,000 Americans per annum. In a chart in their most recent edition, coal is by far the most lethal power source per billion GWh generated.
I wouldn't say so. A particle is just an identifiable thing, something which retains its identity over time. You can identify that the thingy you saw at one position at one time is the same thingy as you saw in another position at another time. You have some reason for saying that e.g. you tracked it through several intermediate positions. In 3S1T spacetime it takes 3 values to identify its position, in 1S1T it takes only 1. But the key fact about a particle is its persistence, not its dimensionality,
Backwards and forwards along the line. Each particle, if such exists, is bouncing back and forward against its neighbour. Or, perhaps more likely, there are no particles, and the universe is one ginormous string twanging along its length with compression waves which along though and interfere with each other. A bit like Gods organ pipe (God's bong?)
It is a word - to me. I read the summary and received what I believed to be the author's meaning without a stutter in my reading. Therefore it achieved what words are meant to do - carry meaning. It may be an ugly neologism, but you can only fight that by using what are, in your opinion, the correct alternatives, and have them carry the day by their superior usefulness.
When I left the industry three years ago, broadcasters were using either 720p for fast moving stuff (sports) or 1080i for hi res stuff (drama, documentary), both using data compressed to 100Mbit/sec. Generally, studio infrastructure is standardizing on 100Mbit/sec for post-production, so you aren't going to get more underlying data even if they do upgrade to 1080p. It would be cheaper to fit an upscaler in your TV.
Buying local food puts a cap on the price rise. If local food costs 25% more than imported food, once increased fuel costs have raised the price 30% the price rise will stop as local production (after a delay) replaces imported. This stops a runaway effect where rising costs everywhere put up prices on things consumed everywhere.
So, in the case of Airbus, it would be the CEO of EADS which owns Airbus, which at the time of the accident was two political appointees, one French and one German, who probably know little about engineering. Or possibly the heads of state of France, Germany and Spain which have controlling shareholdings in EADS. And in the case if the Herald of Free Enterprise, a CEO who had had responsibility for the ship for three weeks, when the ship had been operating /apparently/ safely for many years.
Nice for vengeance to see such a suit going to prison, but in terms of preventing future disasters, it will mean only that competent people will refuse such jobs out of fear of taking the blame for a disaster they could not prevent, and be replaced by less-competent chancers who will take high pay in the hope that no disaster will happen on their watch, thus increasing the likelihood of problems.
Another case of the US feeling that it is perfectly acceptable to treat foreigners in a way that would bring outrage if it tried it on its own citizens. Such things started, and were found acceptable: in cases of outright declared war (World Wars, Korea, Vietnam), it is OK to deceive and manipulate enemy civilians. But that has been translated to being OK to do the same to all except strong allies - and it would be a small step to apply it to them. The philosophy has become that war justifies lethal force, and non-war justifies anything except lethal force.
What would the US do if another country started doing the same within the US?
Hey - maybe they have. Maybe all the appalling right-wing bigots are really Chinese sock-puppets intended to discredit the US. I have never met an American who offered those views: all the ones I meet are nice people. Perhaps the bad side of America that I hear of but never see is just a huge cyber sock puppet! Damn!
That is because the pilot is the ultimate catch-all for errors. There is a general presumption that, if the automatic systems fail, the pilot will fix the problem manually. And if, by hindsight, we can show that there is something the pilot might possibly have done and didn't, then pilot error is a contributing factor.
But AI may be taught, rather than programmed. Google Translate "learns" by taking known translations of documents (particularly EU directives for which machine-readable versions in several languages are available) and correlating two versions to find out how to transform one language into another. If it mistranslates, is it the fault of the programmer, who need not necessarily have spoken either of the languages concerned? Is the computer "doing the boring labour" of a task the programmer was incapable of?
Not that the software gave bad readings: that the software reacted badly when hardware gave it bad readings, possibly because of stormy weather. The suggestion is that the air speed readings were faulty, and the software throttled back until the aircraft fell out of the sky.
And an A3330 has two engines. But it is not the engine control software which is suspect, but the main flight control software, which may have ordered the engines to reduce power when it shouldn't.
What we need is strong, forensically reliable audit trails for every policy and decision.
The problem is that in many cases of corporate manslaughter, is is lack of decision which is the problem. Nobody thought to run the tests which would have found the fault. And that may be up to a corporate structure which split responsibility so that everybody thought somebody else was taking care of it. That is why an offence of Corporate Manslaughter is useful. Often, it is impossible to say which individual made the mistake that caused the death, but it is possible to say that the corporation as a whole was flawed in a way that caused the accident,
Consider the Herald of Free Enterprise disaster: the ship, shuttling backwards and forwards several times each day, had six captains. Each has probably taken up the flawed working practice from predecessors who had been doing it for years. The Director of Safety at corporate level had only had responsibility for three weeks since the company had take over the ships. Tracking back to who originated the dodgy practice years ago would probably be impossible. And they never thought they were doing anything unusual, so there was nothing to write down to be audited.
Manslaughter is vary often a crime of omission. Somebody didn't think through what they were doing. And audit trails are bad at catching omissions.
I don't actually have much problem with computers knowing everything about me - it is when people get information that I worry. I use GMail, and it posts "targeted" ads alongside my mails. Very occasionally, they are of interest, mostly I can see the keyword they are responding to, sometimes I just wonder WIHIH. But it doesn't worry me. And the same applies more generally. I intend to obey speed limits: I have no problem with a computer checking that I do so. If I have strange sexual tastes, i have no problem with a computer knowing it - it may even be able to guide me ("people who viewed what you just viewed also..."
It is purely when it gets into the hands of people that I get nervous. And that includes "legitimate" users such as law authorities, Because they will judge me: it is what people do. And their values will not be my values, so they will judge me by values I do not understand. (As a geek, I am not much of a "people person", so I often don't understand other people's values).
So if this 100% surveillance is totally automated, I don't have a problem. By all means, track where I go for traffic purposes. Measure how much I pee - what do I care? Correlate automatically my this against my that - feel free. But DON'T TELL ANYBODY! Because they WILL judge me (as I, if the situations were reversed, would judge them. Nobody, but nobody, can help being at least somewhat judgmental.
Of course, there have to be exceptions for investigating crimes. And I grudgingly accept that - a system of warrants etc. The trouble is, once the data is in a computer, it is too easy to release. In physical searches, the process of getting a warrant and doing the search was laborious and obvious enough that it tended to be self-limiting: you knew you were being searched, and could protest. Not that there were no abuses, but that the abuses were small enough that the benefits of the system outweighed the costs of the abuses.
But with purely electronic searches, it is too easy to set the search too wide, and to easy to do searches without the object of the search knowing. about it. And I am cynic enough to say with absolute confidence that if the system is capable of being abused, it will be abused - by the over-zealous, by the nosy, and by the criminal.
Since the report shows that the flaw was detected in testing, they are not doing so. No doubt they were designed not to be susceptible, but being human the design did not achieved that goal. Rather then assuming that they were designed right, they were testing them - which necessarily requires production units to test. And the test showed a flaw, and they are not rolling out the systems because of the flaw.
This is the system working properly. Proper testing before aircraft delivery has shown a problem, and steps are being taken to rectify the problem. The short term fix is to outlaw WiFi; when they have found and tested a proper fix, they may (assuming tests are passed) then enable the equipment.
There is no point in getting alarmed when the proper test show the faults before the aircraft leave the ground.
But then you have to stock different sizes of envelope. How about pre-printing a small sticky label representing the amount paid which you attach to the envelope of your choice? Can I patent that idea?
It would all be solved if we could microcharge for email. At $0.001 per email, no sensible use of email would build up a significant cost, even for the very poor. But if your message to the whole company cost $5, you might think twice. And the spam industry, of course, would become much less viable. I don't know how you could do it safely, reliably, unfraudably, but it would be nice to try.
One of the reasons for encrypting the accurate GPS and adding an error to the public version was that in order to take out a hardened silo, an offensive missile needs to strike the ground within two meters outside the hardened lid of the the silo. It then penetrates the ground and explodes alongside the body if the silo, which is softer than the lid. That accuracy might be hard to achieve purely ballistically,
We don't need liquidity at picosecond resolutions. We need liquidity at human reaction times. At the bottom of everything is human investment decisions - even if it is sometimes the decision to entrust your money to a bot. We need a market fluid enough for people to be able to get their money in and out at rates that are, to them, fast: perhaps a few minutes. Which means that the underlying market has to ace maybe an order of magnitude faster to handle peaks and avoid frictional problems. So a trading speed of a few tens of seconds is entirely adequate for a market that serves societies needs. Anything else is traders exploiting unintended opportunities which are not part of the normal functioning of the market. Which is not, in itself, wrong: freedom to exploit such opportunities is fundamental to innovation. But the cost of their high speed computer systems and the profits they take are a cost to the market for (as far as I can see) no corresponding gain. So I don't like this sort of trading, and its speeding up. But "I don't like it" is not a sufficient reason to ban something. If we want to damp such (in my opinion) excessive speed-freakery, we need to do so with great care, being well aware of the Law of Unintended Consequences. By bet would be for a micro-levy on transactions. Say 0.01%. The same computers which are speed trading could easily charge the levy. It would be small enough that it would be negligible for "real" transactions, but would damp out ultra-high-speed churn.
The trouble with all these schemes is that they are fine for those who are already criminals, but they push those who want to campaign peacefully for political change into criminal or criminal-like behaviour. You should have an assumption of privacy: until you appear to be something nefarious, your comings and goings should be your private business. This means that the authorities can, at their choice, track your recent movements and get a good idea of your contacts. Of course, this can be used for good, but history shows that good uses are soon overwhelmed by authoritarian control.
That is always true, and NaCl makes little difference to this. Of course, the more different ways you use an OS, the more different places you get to probe for bugs. If you never try anything new, you will not find new bugs, but you will not get the benefit of new capabilities. Take this to an extreme, and you end up Amish: they have decided not to take up new technology because it brings new risks.
No, you are still running in User mode rather than Kernel mode. The OS still gets to trap and inspect all your accesses, so that you can only look at the HD in the same way as any user program can.
Not necessarily "perfectly good". A structurally sound house in the wrong place is not perfect and not really good. In a way, this mirrors the soviet failure, rather than capitalist problems. The soviets assumed that if a factory was working at full speed producing whatever had been specified, it was doing good work. But producing obsolete or excessive goods is a net loss. If you could move houses from the rustbelt to the sunbelt, your observation might be true. But you cannot, and it is better to scale back the shrinking communities to a functional size than continue to mimic a city with four times the population.
(Or you can try to relocate jobs to where the houses are. if you succeed in that, your fortune is made, just on the lecture circuit).