I'm not a climatologist, so I am speaking as a non-expert. That said, I do not see "everything hanging together beautifully". I see a lot of data that supports climate change, being compared with a lot of models that predict change. None of this though is "clear".
Looking for trends in a system with low frequency noise is very tricky: Imagine it is March in the norther hemisphere, and a bunch of people make sacrifices to the flying spaghetti monster for warmer weather. Being scientists (like most FSM worshipers), they take careful data. The show that the temperatures in late march are clearly statistically higher than they were in early march, and in fact the warmest they have been since records started in January. They conclued at many-sigma that the sacrifices have caused warming. The statistics are right, but because the system has low frequency noise their conclusion are wrong - they are just seeing the seasonal warming.
Now, I think that human generated CO2 is very likely to cause global climate change, but I object when anyone suggests a high level of certainty on this. This is not orbital mechanics, the predictions are not very accurate or reliable. We need to act on imperfect information - which is fine, as a society we do that all the time. When people start to claim unreasonable accuracy though, they just set themselves up to be defeated by their political opponents.
Climate modeling is exceptionally difficult - it is a nonlinear,not well understood system with hysteresis, that does not allow for controlled experiments. The system contains a lot of low frequency noise which makes statistical analysis of past data unreliable. The studies are complex and difficult, trying to condense the assumptions and results into a form where even a well educated non-expert can understand them is difficult.
The results of the models are not simple: they do not predict "global warming", they predict widespread climate change that varies by region, though the overall average may be warming, that isn't the most important part.
Even if you knew the resulting climate change, the effect on society are not easy to model, and are highly variable. Trying to decide what resources should be spent to avoid climate change is difficult even if you knew with absolute certainty what its effects ere.
The results have a very large economic impact - trillions of dollars may be directed from some industries to others based on these predictions. Lots of people have a financial motivation to push one answer or the other.
Science is not done by surveying scientists and taking a vote. The only people in a position to judge the work are those who are directly involved in climate modeling. (I could easily sell a non-expert in my field on a technology that isn't really viable).
The only other scientific field with similar problems is medical research, and it also has very serious problems.
People like this are the reason that scientists need to be very careful to present their data in an unbiased fashion. The temptation to show "simple" or "clear" data that supports something they are sure is true needs to be resisted. Any evidence that the scientists are in any way biasing their data can be used politically to discredit the entire field.
There are huge maintenance advantages to having your SCADA system connected to the internet: as an example it allows your experts to debug problems from home, rather than wasting time to drive in. This leads to the (quite reasonable) desire to connect the systems but implement various types of security to prevent unauthorized access. Of course even if that is done correctly, the "authorized" person may themselves be the unintentional source of the hack - just because someone knows how to tune the parameters of an oil refinery doesn't mean that they are an expert in resisting hacks.
Some classified computer systems really are kept isolated, but it is a huge hit to the efficiency of work done there, and in many cases the added cost isn't worth the extra security. Even that won't protect against hacks at the manufacturer end......
You need to be very careful in using people's searches in court. As an example, I run role-playing games, some of which are set in the modern world. I've searched for all sorts of information on criminal activity.
There are some emails that should go to a large group, but not many. UI changes can help - put the reply all far from the reply button so people don't hit it accidentally.
Even better (if technologically practical) would be to have different thresholds of numbers of people who will receive an email. For more than say 10, you need an extra confirmation, for more than 100, something more complex.
I'd also like to see different spelling / language checks for different recipients, the sort of language you use with friends may no be appropriate to an email to your entire department.
I wonder if the airlines will use the higher cabin pressure. The extra weight of the air will decrease fuel efficiency and increase operating costs (both direct and in aircraft pressurization lifetime).
You can easily get slightly more leg room, (but not more width), for modestly more money, and I do that. On United at least, if you want significantly more space, or better food, or electrical outlets, or in-seat video the price is very much higher (~10X) - much more than the extra aircraft interior space taken up by the large seat.
Airlines are of course free to sell their services at whatever price they want, but I'm not going to get excited (give them flying preference) for minor issues like exactly which model aircraft they are flying. From inside a 767, 777, 747, A330, A340 all are so similar that if you don't count seats its hard to tell which you are on.
As a frequent passenger, the 787 looks like a slightly more efficient 767 to me.
Just saw one at Beijing in ANA colors - looked a lot like a mid-sized twin engine airliner.
I spend a lot of my life on airliners. The things that matter to me are: Big overhead bins: 787 has them, but so do lat model 747s and 777s. AC power sockets: These could be put on any plane, but usually airlines only have them in business. Legroom: Entirely up to the airline to set the seat spacing, nothing to do with the airplane. In seat video with a selection of movies: Again up to the airline for the interior configuration. Sufficient restrooms: Again, an airline configuration issue.
The improved fuel efficiency will reduce costs some - which is nice, but that is an ongoing trend. Presumably the airbus A350 will be the next step, followed by a 797 or something. I occasionally look out the windows, but most of the time there isn't much to see - so big windows are only a minor change. If they really operate the plane at lower cabin altitude that would be nice, but the extra weight burns more fuel - I doubt they actually operate that way for long. I couldn't care less about the multi-colored lighting.
It looks like a nice plane, but not in any way a game changer. Give me a Mach 3 SST, or a sub-orbital that can do Shanghai to SFO in 40 minutes and we'll talk. We've had Mach.85 airliners for >50 years now.
If the X-rays have a spectrum that looks at all like a blackbody then temperature is a reasonable way to describe them. If you have a box of material at some temperature, the inside of that box will be filled with electromagnetic radiation who's spectrum matches that temperature and it is reasonable to describe that radiation as having a temperature (even from a technical thermodynamic point if view). As a very rough guide, you get radiation with photon energies in the range of 1eV for 10,000 degrees Kelvin. So, 100 million kelvin would give photons vaguely in the 10KeV range which are X-rays.
If the X-rays were only over a very narrow range in frequencies, then temperature would not be a good description, but it sounds like these were nearly thermal.
I haven't seen how any of these deal with the causality problem : if you move between two points travel faster than light in any frame and if relativistic frame in variance holds (darned well tested), you can reach your own past. It doesn't matter how you get between those points or if you do it in this or some other universe. One can probably write down the math for non-causal physics but I can't imagine what it would mean.
I'd say serious work includes: engineering design, simulations, software development, graphic design, photo / video processing, accounting, process control / automation, financial trading, and a bunch of other things I'm certainly missing. In all of these occupations having a large screen and powerful processor improves efficiency, and the cost of a desktop computer is small compared with the annual salary of the employee.
I can see mobile devices as useful for jobs where you need to constantly move around - sales, inventory, inspection, on-site management etc, but my feeling is that those do not represent the majority of computer users at work.
If you are doing the sort of project that has been done many times before it may be possible to have an idea of the expected effort to meet specifications. If you are doing something new, then it can be very difficult. To take a real example: Someone in my group is tasked with building a system to synchronize two clocks a kilometer apart to 100 femtoseconds drift with a stabilized fiber system. This has been done twice before using 2 different techniques, both very complex and expensive. We have a concept that *should* be much less expensive. How long should it take? Does he get a bad performance review if it doesn't work?
Or how about a self-driving car application component that can recognize the difference between a solid object that has moved in front of a car, and a lightweight harmless item like a garbage bag based on video of the object's motion. I could imagine defining this application, but I have no idea how long it might take to write.
I think part of the problem is that many children just do not understand how serious the consequences of some actions can be - this seems to develop in different people at different ages. Until a child reaches that maturity, there are a whole range of items (guns, lasers, motorized transportation etc) that should not be available to them.
I don't disagree, but take it up with your city council / elected officials. This is not an excuse to shine a laser at them (I'm not suggesting that you did, but it might be the reason some people do)
Reckless endangerment seems about right. The public does need to be made aware of how serious this sort of thing can be before we start slapping on very heavy penalties.
One could imagine building very high reliability accelerators, probably a cluster of lower power machines. Existing large accelerators can have pretty good uptimes (>95%), and there has not been a big push to do better.
It is an idea that has been kicked around for decades but now with the operation of SNS (a 1MW machine), and the European spallation neutron source (~5MW?) there are accelerators with enough power to make this feasible.
It has some great advantages: If you add a lot of neutrons to a reactor you can operate below criticality so you do not need to worry about the stability of the fuel mix (a reactor relies on delayed neutron production to have the reaction rate be controllable). This will let you safety burn fuel much more thoroughly than can be done in a conventional reactor.
The down side is the significant added cost of the accelerator and associated technical systems may make commercial designs uneconomical. It may be possible to solve this, other than average power, the beam requirements are much simpler than for most scientific accelerators and it may be possible to realize substantial cost savings.
LibreOffice is fine by itself, but it is not sufficiently compatible with MS products to allow you to share files in a MS dominated environment. I made several attempts over the years to get by with StarOffice, OpenOffice, LibreOffice and just couldn't manage it - I wasted too much of my and my coworkers time dealing with file incompatibility issues.
Software is a strong natural monopoly and MS has a very strong position in the desktop office market. I don't blame them for trying to milk this position for as much money as they can (that is the function of a company).
I could see either answer as being better. Do you REALLY want employees who will report any sign of crime to the authorities, or do you want employees who display LOYALTY
There may not be a "simple" correct way to answer the test questions, but how do you keep potential employees from learning the type of responses that are desired. Jobs are valuable so as this sort of testing becomes more popular there will be a market in training people to take personality tests. This will quickly invalidate the predictive power of the tests.
There is also the serious issue of the tests resulting in ethnic or gender bias that could open a company up to lawsuits.
The final problem is whether companies actually know how to measure performance at all. In some simple manufacturing jobs throughput and error rates are easy to measure, but its much more difficult in jobs where there is interaction with customers, or where technical expertise is required.
It still doesn't help to have an op-in. The carriers will just require you to op-in before using any of the features of the phone. Since all carriers will have nearly identical EULAs you will be required to op-in if you want to communicate in the modern world.
There is a lot of good science that sounds boring to the public, and more importantly to the funding agencies, so the investigators try to make it sound more exciting.
An interesting physics question about how a wave function collapses when a measurement is made becomes "quantum teleportation". Using X-ray pulses to saturated a L-shell (M?) transition in Aluminum becomes "transparent aluminum". A new technique for measuring fast chemical reactions on a surface becomes a 'breakthrough for hydrogen power". The list goes on and on.
This is of course very common in industry. A new "breakthrough microprocessor that introduces a new paradigm in computing", is the previous generation processor done in a smaller process.
There are a huge number of people working in research, and their collective effort has made the world a completely different place. Since there are so many projects, the impact of any one project is generally so small that it seems uninteresting.
Does the FMS give you all the information you need (frequencies etc) so the charts are just needed when you program? If so, then I agree - I only fly GA airplanes.
That sort of works, but not as well as you might like. I had an Ipad with my charts on it overhead during an instrument departure from a busy Los Angeles area airport. Frequencies are very crowded, things happening quickly, no time to ask again at what DME should I change course. I had paper charts as a backup and was very glad to have them.
I'm not saying its a terrible idea, but the "contact ATC" backup just doesn't work as well as you might think.
For an airliner, the best bet would be for each pilot to have an IPAD. Not many things would take both offline at the same time.
So, what is a high confidence level? 90% 99%? 1 part per million error? What is considered good enough statistics for a conviction??? What does "beyond a reasonable doubt" mean? If I'm only 99% sure someone did it, should I convict?
Or a Barge. You could have a barge on a canal that you flood or drain. Barges of 10,000 tons are not unreasonable, and canal filling / emptying technology is well established. Its possible the center of mass of the rocket is so high that a barge would need to be too wide to be stable without an unreasonable amount of ballast.
Slashdot Mirror
I'm not a climatologist, so I am speaking as a non-expert. That said, I do not see "everything hanging together beautifully". I see a lot of data that supports climate change, being compared with a lot of models that predict change. None of this though is "clear".
Looking for trends in a system with low frequency noise is very tricky: Imagine it is March in the norther hemisphere, and a bunch of people make sacrifices to the flying spaghetti monster for warmer weather. Being scientists (like most FSM worshipers), they take careful data. The show that the temperatures in late march are clearly statistically higher than they were in early march, and in fact the warmest they have been since records started in January. They conclued at many-sigma that the sacrifices have caused warming. The statistics are right, but because the system has low frequency noise their conclusion are wrong - they are just seeing the seasonal warming.
Now, I think that human generated CO2 is very likely to cause global climate change, but I object when anyone suggests a high level of certainty on this. This is not orbital mechanics, the predictions are not very accurate or reliable. We need to act on imperfect information - which is fine, as a society we do that all the time. When people start to claim unreasonable accuracy though, they just set themselves up to be defeated by their political opponents.
Several reasons.
Climate modeling is exceptionally difficult - it is a nonlinear,not well understood system with hysteresis, that does not allow for controlled experiments. The system contains a lot of low frequency noise which makes statistical analysis of past data unreliable. The studies are complex and difficult, trying to condense the assumptions and results into a form where even a well educated non-expert can understand them is difficult.
The results of the models are not simple: they do not predict "global warming", they predict widespread climate change that varies by region, though the overall average may be warming, that isn't the most important part.
Even if you knew the resulting climate change, the effect on society are not easy to model, and are highly variable. Trying to decide what resources should be spent to avoid climate change is difficult even if you knew with absolute certainty what its effects ere.
The results have a very large economic impact - trillions of dollars may be directed from some industries to others based on these predictions. Lots of people have a financial motivation to push one answer or the other.
Science is not done by surveying scientists and taking a vote. The only people in a position to judge the work are those who are directly involved in climate modeling. (I could easily sell a non-expert in my field on a technology that isn't really viable).
The only other scientific field with similar problems is medical research, and it also has very serious problems.
People like this are the reason that scientists need to be very careful to present their data in an unbiased fashion. The temptation to show "simple" or "clear" data that supports something they are sure is true needs to be resisted. Any evidence that the scientists are in any way biasing their data can be used politically to discredit the entire field.
There are huge maintenance advantages to having your SCADA system connected to the internet: as an example it allows your experts to debug problems from home, rather than wasting time to drive in. This leads to the (quite reasonable) desire to connect the systems but implement various types of security to prevent unauthorized access. Of course even if that is done correctly, the "authorized" person may themselves be the unintentional source of the hack - just because someone knows how to tune the parameters of an oil refinery doesn't mean that they are an expert in resisting hacks.
Some classified computer systems really are kept isolated, but it is a huge hit to the efficiency of work done there, and in many cases the added cost isn't worth the extra security. Even that won't protect against hacks at the manufacturer end......
You need to be very careful in using people's searches in court. As an example, I run role-playing games, some of which are set in the modern world. I've searched for all sorts of information on criminal activity.
There are some emails that should go to a large group, but not many. UI changes can help - put the reply all far from the reply button so people don't hit it accidentally.
Even better (if technologically practical) would be to have different thresholds of numbers of people who will receive an email. For more than say 10, you need an extra confirmation, for more than 100, something more complex.
I'd also like to see different spelling / language checks for different recipients, the sort of language you use with friends may no be appropriate to an email to your entire department.
I wonder if the airlines will use the higher cabin pressure. The extra weight of the air will decrease fuel efficiency and increase operating costs (both direct and in aircraft pressurization lifetime).
You can easily get slightly more leg room, (but not more width), for modestly more money, and I do that. On United at least, if you want significantly more space, or better food, or electrical outlets, or in-seat video the price is very much higher (~10X) - much more than the extra aircraft interior space taken up by the large seat.
Airlines are of course free to sell their services at whatever price they want, but I'm not going to get excited (give them flying preference) for minor issues like exactly which model aircraft they are flying. From inside a 767, 777, 747, A330, A340 all are so similar that if you don't count seats its hard to tell which you are on.
As a frequent passenger, the 787 looks like a slightly more efficient 767 to me.
Just saw one at Beijing in ANA colors - looked a lot like a mid-sized twin engine airliner.
I spend a lot of my life on airliners. The things that matter to me are:
Big overhead bins: 787 has them, but so do lat model 747s and 777s.
AC power sockets: These could be put on any plane, but usually airlines only have them in business.
Legroom: Entirely up to the airline to set the seat spacing, nothing to do with the airplane.
In seat video with a selection of movies: Again up to the airline for the interior configuration.
Sufficient restrooms: Again, an airline configuration issue.
The improved fuel efficiency will reduce costs some - which is nice, but that is an ongoing trend. Presumably the airbus A350 will be the next step, followed by a 797 or something. I occasionally look out the windows, but most of the time there isn't much to see - so big windows are only a minor change. If they really operate the plane at lower cabin altitude that would be nice, but the extra weight burns more fuel - I doubt they actually operate that way for long. I couldn't care less about the multi-colored lighting.
It looks like a nice plane, but not in any way a game changer. Give me a Mach 3 SST, or a sub-orbital that can do Shanghai to SFO in 40 minutes and we'll talk. We've had Mach .85 airliners for >50 years now.
If the X-rays have a spectrum that looks at all like a blackbody then temperature is a reasonable way to describe them. If you have a box of material at some temperature, the inside of that box will be filled with electromagnetic radiation who's spectrum matches that temperature and it is reasonable to describe that radiation as having a temperature (even from a technical thermodynamic point if view). As a very rough guide, you get radiation with photon energies in the range of 1eV for 10,000 degrees Kelvin. So, 100 million kelvin would give photons vaguely in the 10KeV range which are X-rays.
If the X-rays were only over a very narrow range in frequencies, then temperature would not be a good description, but it sounds like these were nearly thermal.
I haven't seen how any of these deal with the causality problem : if you move between two points travel faster than light in any frame and if relativistic frame in variance holds (darned well tested), you can reach your own past. It doesn't matter how you get between those points or if you do it in this or some other universe. One can probably write down the math for non-causal physics but I can't imagine what it would mean.
I'd say serious work includes: engineering design, simulations, software development, graphic design, photo / video processing, accounting, process control / automation, financial trading, and a bunch of other things I'm certainly missing. In all of these occupations having a large screen and powerful processor improves efficiency, and the cost of a desktop computer is small compared with the annual salary of the employee.
I can see mobile devices as useful for jobs where you need to constantly move around - sales, inventory, inspection, on-site management etc, but my feeling is that those do not represent the majority of computer users at work.
If you are doing the sort of project that has been done many times before it may be possible to have an idea of the expected effort to meet specifications. If you are doing something new, then it can be very difficult. To take a real example: Someone in my group is tasked with building a system to synchronize two clocks a kilometer apart to 100 femtoseconds drift with a stabilized fiber system. This has been done twice before using 2 different techniques, both very complex and expensive. We have a concept that *should* be much less expensive. How long should it take? Does he get a bad performance review if it doesn't work?
Or how about a self-driving car application component that can recognize the difference between a solid object that has moved in front of a car, and a lightweight harmless item like a garbage bag based on video of the object's motion. I could imagine defining this application, but I have no idea how long it might take to write.
I think part of the problem is that many children just do not understand how serious the consequences of some actions can be - this seems to develop in different people at different ages. Until a child reaches that maturity, there are a whole range of items (guns, lasers, motorized transportation etc) that should not be available to them.
I don't disagree, but take it up with your city council / elected officials. This is not an excuse to shine a laser at them (I'm not suggesting that you did, but it might be the reason some people do)
Reckless endangerment seems about right. The public does need to be made aware of how serious this sort of thing can be before we start slapping on very heavy penalties.
One could imagine building very high reliability accelerators, probably a cluster of lower power machines. Existing large accelerators can have pretty good uptimes (>95%), and there has not been a big push to do better.
It is an idea that has been kicked around for decades but now with the operation of SNS (a 1MW machine), and the European spallation neutron source (~5MW?) there are accelerators with enough power to make this feasible.
It has some great advantages: If you add a lot of neutrons to a reactor you can operate below criticality so you do not need to worry about the stability of the fuel mix (a reactor relies on delayed neutron production to have the reaction rate be controllable). This will let you safety burn fuel much more thoroughly than can be done in a conventional reactor.
The down side is the significant added cost of the accelerator and associated technical systems may make commercial designs uneconomical. It may be possible to solve this, other than average power, the beam requirements are much simpler than for most scientific accelerators and it may be possible to realize substantial cost savings.
LibreOffice is fine by itself, but it is not sufficiently compatible with MS products to allow you to share files in a MS dominated environment. I made several attempts over the years to get by with StarOffice, OpenOffice, LibreOffice and just couldn't manage it - I wasted too much of my and my coworkers time dealing with file incompatibility issues.
Software is a strong natural monopoly and MS has a very strong position in the desktop office market. I don't blame them for trying to milk this position for as much money as they can (that is the function of a company).
I could see either answer as being better. Do you REALLY want employees who will report any sign of crime to the authorities, or do you want employees who display LOYALTY
There may not be a "simple" correct way to answer the test questions, but how do you keep potential employees from learning the type of responses that are desired. Jobs are valuable so as this sort of testing becomes more popular there will be a market in training people to take personality tests. This will quickly invalidate the predictive power of the tests.
There is also the serious issue of the tests resulting in ethnic or gender bias that could open a company up to lawsuits.
The final problem is whether companies actually know how to measure performance at all. In some simple manufacturing jobs throughput and error rates are easy to measure, but its much more difficult in jobs where there is interaction with customers, or where technical expertise is required.
It still doesn't help to have an op-in. The carriers will just require you to op-in before using any of the features of the phone. Since all carriers will have nearly identical EULAs you will be required to op-in if you want to communicate in the modern world.
There is a lot of good science that sounds boring to the public, and more importantly to the funding agencies, so the investigators try to make it sound more exciting.
An interesting physics question about how a wave function collapses when a measurement is made becomes "quantum teleportation". Using X-ray pulses to saturated a L-shell (M?) transition in Aluminum becomes "transparent aluminum". A new technique for measuring fast chemical reactions on a surface becomes a 'breakthrough for hydrogen power". The list goes on and on.
This is of course very common in industry. A new "breakthrough microprocessor that introduces a new paradigm in computing", is the previous generation processor done in a smaller process.
There are a huge number of people working in research, and their collective effort has made the world a completely different place. Since there are so many projects, the impact of any one project is generally so small that it seems uninteresting.
Does the FMS give you all the information you need (frequencies etc) so the charts are just needed when you program? If so, then I agree - I only fly GA airplanes.
That sort of works, but not as well as you might like. I had an Ipad with my charts on it overhead during an instrument departure from a busy Los Angeles area airport. Frequencies are very crowded, things happening quickly, no time to ask again at what DME should I change course. I had paper charts as a backup and was very glad to have them.
I'm not saying its a terrible idea, but the "contact ATC" backup just doesn't work as well as you might think.
For an airliner, the best bet would be for each pilot to have an IPAD. Not many things would take both offline at the same time.
So, what is a high confidence level? 90% 99%? 1 part per million error? What is considered good enough statistics for a conviction??? What does "beyond a reasonable doubt" mean? If I'm only 99% sure someone did it, should I convict?
(this is why I'm never picked for a jury).
Or a Barge. You could have a barge on a canal that you flood or drain. Barges of 10,000 tons are not unreasonable, and canal filling / emptying technology is well established. Its possible the center of mass of the rocket is so high that a barge would need to be too wide to be stable without an unreasonable amount of ballast.