Yup - for starters just look at the National Electric Code. EVERYBODY is required to follow it, and yet it must be purchased from a private entity.
IMHO any document referenced in legislation should automatically enter the public domain. People shouldn't have to pay to read the laws they are subject to.
I don't get the point of having a new gTLD if it just ends up being a perfect replica of existing TLDs. Why do we need harvard.cool when we already have harvard.edu?
If anything there should be an outright ban on owning the same name in more than one TLD.
It just seems like another way to get everybody with a domain name to fork over a chunk of change to get another one.
I suspect that 10kg of welt is probably weaker than 10kg of intact steel. However, when weld are used most likely a lot more metal is put down in the weld than the materials being bonded.
It is like saying "which is stronger - crazy glue or a steel girder." Well, the steel girder is probably stronger, but if you compare a 1" diameter girder to a blob of crazy glue two feet thick, then the crazy glue might just win out.
As others indicated, welds in serious projects tend to get a lot of scrutiny and quality control, so this sort of thing shouldn't happen if they were properly specified. I wouldn't be surprised if welds on a structure like this would be spot-checked with x-rays.
You missed the other side of the reaction. It is charge neutral because the +/- charges in the two side pools are then balanced with -/+ charges from the water to be desalinated.
The charges flow apart in the first place because the central pool is highly concentrated - so it contains far more + and - charges than anything else in the system.
This kind of approach would never yield completely drinkable water, but that isn't the point. The goal is to get rid of a lot of the ion load before using more expensive processes to get rid of the rest.
Desalination is a marvel of process optimization. Multiple stages of purification are used - each one being more expensive than the last but more effective. The early steps get rid of a huge mass of dissolved matter for dirt cheap, so even if their product isn't drinkable it GREATLY reduced the cost of the later stages.
If you don't care about cost then desalination is trivially easy. Just run any kind of water you like through a H+ exchange resin followed by an OH- exchange resin, and then run it trough activated charcoal. The resulting water will be as clean as clean can be and the system would be remarkably simple. The catch is that those resins cost a small fortune to make, and if you run seawater into them then they're probably going to last all of 5 minutes. It might be a good approach for a camper to use to obtain water (the resin is a lot lighter than the amount of water that it could clean), but it is not a cost-effective method overall. Also - the purity it would achieve would be massive overkill. This is drinking water - we're not manufacturing CPUs.
First - I'm not convinced that the nudity ban on TV is nearly as critical as it is made. That said, I'm fine with the FCC enforcing content restrictions on broadcasters even when they weren't the party directly responsible. Otherwise there is no real incentive for networks to police their content - in fact they're better off just collecting ad revenue and not even watching their own programming lest they be found responsible.
The solution in that case was to go ahead and fine the wazoo out of the network, and then let the network sue the wazoo out of the artist and their record label.
Some of the best movies I have seen were more independent movies with unknown actors. Im sure they still received good salaries relative to your average job, but they werent astronomical.
Some of the worst movies I have seen were more independent movies with unknown actors. They might have been cheaper failures, but the investors still lost out.
Actually, on the whole I tend to agree with you - I'd rather see an innovator reap the rewards, than see a typical retread movie. However, there is a reason people pay big money for big name talent - audience draw. Stick Robin Williams in a movie and lots of people will buy tickets. To some degree, the fact that Robin Williams will even take a part helps to assure the investors that the producers are doing their job right.
It is the same reason that a company will hire a big consulting firm when chances are that some guy in their company could do the same thing for 1/100th the cost, or they could easily maintain a more diverse staff if they weren't spending that money on consultants. On a recent project a consulting firm was making 16X my salary for the duration of the project, and yet I probably contributed the majority of the innovation. However, the fact is that if I had messed up and there were no consultants then my boss would take a big hit, but if the consultants messed up then rather than everybody yelling at the guy who hired them for some reason everybody would obsess over arguing with the consulting firm. When you hire a big-name consulting firm then you're shielded from responsibility for your actions. Managers shudder at the though of actually having to find and hire people who can do a job right, and would rather contract it out so that when it doesn't go right they can blame an external company. Overall companies are more focused on blaming people when things go wrong than making them go right in the first place, and then we wonder why we're in the current state...
Studios are no different - they make the conservative choice, which is to do what worked before even if it costs more. When it goes wrong everybody just says "well, nobody could have seen that coming."
I tend to agree with a lot of your points - a specialized device will generally give a better experience than a general-purpose device. However, a LOT of people will stop buying specialized devices when the general-purpose device is "good enough." I didn't replace my mp3 player the last time it broke because while android isn't nearly as capable as Rockbox, it is "good enough" and is one less thing to carry around.
Another thing to consider is that Japan is not a typical market where GPS units are concerned. In much of rural Japan many streets are not even named. Sure, but ones are, but my rule of thumb is that if a road in the US has a number than the equivalent road in Japan would have a name, and if a road in the US doesn't have a number than the Japanese equivalent will just be pavement leading in some random direction. At best you'd have names for roads that would have numbers in New Jersey (a state that likes to number roads). As a result almost EVERYBODY in Japan owns a GPS. My understanding is that before GPS became popular most rural directions relied heavily on landmarks (turn right at the red house after the big tree, etc).
Having been there I have to say that Japan has the most curious rural/suburban culture. A road through town might have no names, and will alternatively lead past residential houses, rice fields, and tire dumps. And somewhere right in the middle of all of that (and nowhere near a crossroads or anything resembling a busy part of town) there will be a line of several vending machines that appear to get heavy use. I wonder if Japanese prefectures have any kind of semblance of a zoning board...
I agree with your points about the problems with automation - I avoid driving much with cruise control for the same reason - having to micro-manage my speed helps keep me alert. Sure, I could pay more attention to other things without having to do it, but in reality I'd pay less attention.
I'm not sure that firing pilots who fall asleep is going to help either.
I liked something I saw for airport security screeners - who face similar problems (they screen thousands of packages and 99.99999% of the time there is nothing to see). It was an x-ray machine that would add in images of contraband for the operator to spot - if an operator didn't hit a button when one was spotted then it would alert a supervisor. It gave the operator something to actually do, and thus it kept them alert.
Maybe the plane needs to trigger a random simulated failure (caution light or whatever) that requires a button to be hit to clear the condition (FMS would avoid triggering it at critical moments, and the operator would have plenty of time to deal with other stuff first). Or, maybe the cockpit should have officially-sanctioned ways to do things like check email/etc which will do things like pop up occasional messages to do a visual scan and which will blank the screen the instant an alert of some kind occurs. Or, maybe there is some task the crew could perform that is more mentally stimulating than staring out the window at blue sky.
The human brain is a machine - a complex one, but a machine nonetheless. It has certain limitations. In particular, it gets bored if you don't give it something to do. This is biology, and simply telling pilots not to be bored doesn't fix the problem. Likewise, the human brain requires sleep, and if people are overworked they won't get enough of that.
That's a BIG stretch. That's like saying that a mayor who invests in anything shouldn't be allowed to pass a law that might improve the economy (heaven forbid!).
I'm all for requiring those holding significant political power to put their investments in some kind of a blind trust. I'm also all for declarations and avoidances of conflicts of interest.
However, the idea that anything that helps any company could be a conflict of interest because your wife works at a bank that may or may not invest in that company is REALLY far fetched. For all the guy knows Goldman might have invested in a major competitor to the company making the H1N1 vaccine and that they'll suffer as a result.
I wasn't trying to suggest that this wasn't in keeping with the letter of the law. I was trying to suggest that the law is dumb if this is the case, and I doubt this was the original intent of those who wrote the law.
If a software needs to be installed for it to be used, then that installation should not be considered copying/distribution under copyright law. Yes, I realize that it does require bits to be copied. However, copyright law doesn't exist because copying is evil - it exists to ensure that those who make creative works are able to profit from them reasonably. If I buy a copy of microsoft office, and then install it on my computer, I haven't deprived microsoft of revenue despite having "copied" it. If I make a backup of the original media it doesn't cost them a dime either. If I make 10,000 backups of the original media that also doesn't cost them a dime. If I then sell those copies on ebay for $9.95 or pass them out for free on the street, then that does impact them financially. At the same time, some forms of non-profitable distibution should still be permitted.
The a legitimate of law is to facilitate commerce. Sane copyright policies can help do that. Insane policies do not.
That is a pretty ridiculous argument (spare me the citations, I'm sure there are 14 judges out there that have upheld it nonetheless).
When I listen to music from the radio it ends up being embedded in my memory - and yet I don't need a license to listen to the music.
I don't really have a major problem with copyright, but it is something that should be exclusively applied to people distributing things, not people who receive them. P2P technology is a gray area - I'm actually not opposed to liability there but the statutory fines need to be several orders of magnitude lower.
There are a couple of optimizations that are possible when you cross layers. Sure, it could be done in a way that retains the ability to swap layers, but the individual layers would need to expose a LOT more information to each other about their implementations. It would be difficult to abstract as well.
You asked for a good reason - here is an example:
Suppose I want to modify a 10kb text file. Since ZFS uses copy-on-write it picks a new allocation unit on the disk to write to. Now, it could just pick any old empty area on the disk. However, instead it can find a stripe on the RAID that is completely empty, and fit that 10kb file in with another few MB of other writes and put them all in the same stripe. Then it can pick a stripe that happens to be near where the cache was recently flushed. Since the whole stripe is being written only a single disk write to each component drive is needed.
Suppose you did the same thing with ext4 over lvm and md. Now the filesystem is going to overwrite the existing file in place. From the filesystem's perspective it isn't a big deal - either way we're overwriting one block with another. However, the filesystem doesn't realize that there is RAID underneath, so the md layer eventually gets a request to modify a few blocks in the middle of a stripe. The md layer of course does what it can to combine operations, but ext4 isn't doing anything to make this easier. Chances are that the whole stripe is going to get read (using all but one of the disks for raid 5), a small area inside will be modified, and then the stripe will get written again. While the stripe is being written the old stripe is being destroyed, so if anything goes wrong the previous contents of everything on that stripe will be lost - most of it EVEN if there is full data journaling at the ext4 layer (although the md layer could implement is own journal).
By crossing layers you can minimize the amount of work the drives need to do, and at the same time provide a higher level of data security by avoiding overwrites whenever possible.
Now, you could abstract these layers a bit more to make them more interchangeable, but the behavior of the filesystem is going to need to change based upon whether the underlying layer is striped or not, and of course that is already muddying the waters.
The set-top boxes are run directly to the router - on the world-facing side. They're all attached to the same coax line, and they communicate using an odd ethernet-over-coax protocol.
The other reply did point out a valid alternative.
Yup - same issue here with FIOS - I really didn't want to have to mess with getting their router to bridge (it can be done, but it is a real pain and if you need to make a change you need to reset and reconfigure). So, now I have a box sitting on my LAN that I have no control over. In theory all the devices on the LAN are routinely scanned with nessus/etc, but it isn't ideal.
Many of these integrated services have all kinds of tie-ins that make bridging the router painful. For example, on FIOS the network link from set-top boxes to the internet is via the outward-facing port on the router. If you bridge the router then it has no internet connectivity of its own and can't route packets from the set-top boxes. Plus, when it is bridged you can't get into the router's web-based admin console, so to change a setting you need to hard-reset it. I guess if you don't mind having your own router NATed that is an easy option. Sometimes I'm tempted to go IPv6 with a tunnel provider just to get past all that stuff...
FIOS is a bit of an unusual case since they run the network over coax. Where standard ethernet is used you have more hope of just bypassing the router entirely.
Yup - insurance only works in the absence of knowledge. If you could predict with 80% accuracy whether somebody's house would burn down, then almost everybody could get dirt-cheap fire insurance (which they wouldn't buy anyway since they wouldn't need it), and a small number of people wouldn't be able to afford it and would lose everything they have in a fire. The insurance companies would go out of business since nobody would bother buying insurance either way.
The only thing that would work once genetic testing becomes reliable is a system that has these attributes: 1. No denial of claims for pre-existing conditions. 2. No differential charging based upon genetic factors. 3. All people must pay in - coverage is not voluntary.
If you don't have all three the system breaks down. Either people rip off the insurers, or insurers rip off the people. Neither works.
Most of what people consider "health insurance" isn't really insurance anyway - it is more of a buyer's club for health services. IMHO a major area of reform should be to eliminate this aspect of health insurance entirely. There is no reason that a poor person without insurance should have to pay $100 for a doctor's visit that costs $30 for Aetna (even if the poor person can haggle them down to $50 - assuming they are in the condition to haggle BEFORE the services are rendered). This alone wouldn't fix health care for the poor, but it would certainly make that problem a lot easier to solve.
Then simply make a policy of no belt = no healthcare bailout. Generally speaking the sorts of people who advocate no seatbelt laws are also the sorts of people who advocate no tax-funded brain scraping. I can't imagine that the belt is going to make a big difference in how long you end up waiting stuck in traffic.
Nanny state policies tend to be the result of a nanny state. When the government bails you out when you can't afford your rent, then suddenly you have regulations on what kind of home you're allowed to rent. When the government pays for your doctor's bills, then suddenly your body doesn't really belong to you anymore.
Health care is a whole mess that I don't want to get into. I'm not totally opposed to a little bit of socialism for those who are down on their luck, but the whole system is in need of a massive overhaul. A catastrophic care insurance system coupled with clearly published provider fees and a requirement that EVERYBODY pay the same fee for the same service would go a long way (but there are no simple solutions to the whole mess).
Waste discharges are an economic externality, and many libertarian-minded people would argue that it should be regulated (preferably in a market-based manner).
Financial regulations is a pretty big category - most libertarian-minded people would be in favor of regulation that maximizes transparency and openness, and which prevents the creation of monopolies. So, CEOs misstating profits would be fair game - that is fraud.
Seat belts are a different matter. People should be perfectly free to drive without wearing seat belts - if people want to kill themselves why should that matter to me? However, I'm all for requiring standardized safety tests with clear disclosure of their results. The goal is to enable consumers to make their own decisions.
Finally you mention food safety. This is also an area where regulation is perfectly acceptable, because it is not straightforward for a consumer to determine whether a piece of meat they want to buy is contaminated. Now, regulation should be primarily about disclosure. For example, mandating disclosure of use of GMOs seems fine to me - if consumers care they can choose appropriately. Likewise, if there is some big debate about nitrate levels in meat, then require disclosure of nitrate concentration and let consumers decide whether they care or not.
Most libertarians do not advocate a world free of all regulation - just those which essentially protect people from themselves and often create perverse incentives.
If you want an example of a perverse incentive - instead of spending $10 more on a TV that uses less power when it runs, maybe some consumers with children would rather spend $10 on a TV with an IR sensor that detects when nobody is in the room and shuts itself off. Such a TV might actually save more power overall, but consumers wouldn't be able to buy it under the proposed regulations. They could only buy a TV that cost $20 more that did this (being forced to spend money on a feature that actually only has a minor improvement for them).
Others have the right answer - just raise electrical rates to reflect the true cost including all externalities. Then people can figure out what kinds of efficiency improvements make the most sense for them.
Would this even do anything in a quench? Suppose that power line is carrying 1 GA (giga-amps) during normal use. The FCL would prevent damage to the line if for some reason it shorted and tried to pull 2 GA.
However, if there were a quench along the line, the current would actually drop, not rise (resistance increases dramatically). An FCL seems to protect against a short, but during normal operation a transmission line IS a short. The issue isn't that the overall line is carrying too much current. The problem is that the point that overheated and quenched is carrying WAY too much current by several orders of magnitude, but the overall line is below design capacity.
During a quench the current would generate heat at an incredible rate.
Maybe if you had some lightning-fast detector that could spot a voltage rise on either side of the line and cut the circuit completely that might work. I'm guessing that voltage would rise since the line now has resistance.
Disclaimer - I'm not an expert on superconduting transmission lines, nor am I an electrical engineer. I do have a half-decent grasp of Ohm's law, however, and have seen the results of superconducting magnet quenches.
Yup - if I owned a huge company my first rule would be that I would need to personally approve any consultant engagement.
When a manager comes to me and points out that if we don't hire the consultant we could end up losing more through a bad decision, I'd say, "fine - but if it happens I'll make sure you never get another job, unless you admit you're not qualified to do your job and recommend a replacement."
There really are situations where companies probably should engage consultants. The problem is that SO many of them are brought in to do stuff that the company should be able to do in the first place.
I've worked with some fairly competant consultants but even they tend to have knowledge gaps. For example, in the IT world many of these guys are great at deploying systems, but they don't have any idea how to deploy them in a way that makes them cheaper to maintain in the long term. In fact, they have an incentive to make them harder to maintain to get repeat business. When you do it yourself you have incentive to make sure that you won't have to spend six months out of the year recoding stuff that should be configurable...
Look, despite being fairly conservative I like a lot of things about Obama.
However, he doesn't deserve a Nobel, and I don't like what has happened to the Peace Prize in recent years. It is more a political statement than an award for actually accomplishing things. Perhaps Obama will save the world from itself, but he hasn't actually done it yet.
We don't give the Physics Prize to promising grad students in the hope that they will use the prestige to do amazing things. The Prize is given to people who already have done amazing things, in recognition of what they have done. In fact, it usually isn't given in other fields until the accomplishment has stood the test of time. Almost every science nobel has been awarded for some major advance that scientists are likely to consider important even a century later. Just look a the prizes from 50 years ago and how they've stood the test of time.
Has so little actually been accomplished to further world peace in the last 30 years that we must give the prize to people who are merely likely to further world peace?
You don't need nearly as long an antenna to receive an AM signal (granted, I still suspect they're a bit big for pocket-sized phones). If you allowed that the phone could transmit in one band in receive in another then the AM bands could still be useful.
Well, the energy the rocket would end up with would be:.5*1E6ton*(0.5c)^2 = 1E25J.
The equivalent mass is 1E25J/c^2 = 125kT (hmm - all those c's cancel out).
This assumes that your energy source is completely transformed into kinetic energy, and that the exhaust contains no kinetic energy. I suspect to figure that all out I'd need to get into impulse and conservation of momentum and all that, and I'm too lazy. However, the mass isn't actually unreasonable considering a typical rocket going to vastly slower speeds is 99% fuel anyway. Of course, in addition to your super-efficient engine you also need a way to extract 100% of the energy out of matter to get away with so little of it. If you carried matter/antimatter you might get in the ballpark. Of course, if you crashed your ship with 60kT of antimatter aboard I think I'll like to be on the other side of the solar system. For that matter, the same applies if you crashed your ship into anything solid once you were at speed (granted you'd probably just punch a hole in the earth and come right out the other side with a big exit crater).
Slashdot Mirror
Yup - for starters just look at the National Electric Code. EVERYBODY is required to follow it, and yet it must be purchased from a private entity.
IMHO any document referenced in legislation should automatically enter the public domain. People shouldn't have to pay to read the laws they are subject to.
I don't get the point of having a new gTLD if it just ends up being a perfect replica of existing TLDs. Why do we need harvard.cool when we already have harvard.edu?
If anything there should be an outright ban on owning the same name in more than one TLD.
It just seems like another way to get everybody with a domain name to fork over a chunk of change to get another one.
Well, you're probably both right.
I suspect that 10kg of welt is probably weaker than 10kg of intact steel. However, when weld are used most likely a lot more metal is put down in the weld than the materials being bonded.
It is like saying "which is stronger - crazy glue or a steel girder." Well, the steel girder is probably stronger, but if you compare a 1" diameter girder to a blob of crazy glue two feet thick, then the crazy glue might just win out.
As others indicated, welds in serious projects tend to get a lot of scrutiny and quality control, so this sort of thing shouldn't happen if they were properly specified. I wouldn't be surprised if welds on a structure like this would be spot-checked with x-rays.
You missed the other side of the reaction. It is charge neutral because the +/- charges in the two side pools are then balanced with -/+ charges from the water to be desalinated.
The charges flow apart in the first place because the central pool is highly concentrated - so it contains far more + and - charges than anything else in the system.
This kind of approach would never yield completely drinkable water, but that isn't the point. The goal is to get rid of a lot of the ion load before using more expensive processes to get rid of the rest.
Desalination is a marvel of process optimization. Multiple stages of purification are used - each one being more expensive than the last but more effective. The early steps get rid of a huge mass of dissolved matter for dirt cheap, so even if their product isn't drinkable it GREATLY reduced the cost of the later stages.
If you don't care about cost then desalination is trivially easy. Just run any kind of water you like through a H+ exchange resin followed by an OH- exchange resin, and then run it trough activated charcoal. The resulting water will be as clean as clean can be and the system would be remarkably simple. The catch is that those resins cost a small fortune to make, and if you run seawater into them then they're probably going to last all of 5 minutes. It might be a good approach for a camper to use to obtain water (the resin is a lot lighter than the amount of water that it could clean), but it is not a cost-effective method overall. Also - the purity it would achieve would be massive overkill. This is drinking water - we're not manufacturing CPUs.
First - I'm not convinced that the nudity ban on TV is nearly as critical as it is made. That said, I'm fine with the FCC enforcing content restrictions on broadcasters even when they weren't the party directly responsible. Otherwise there is no real incentive for networks to police their content - in fact they're better off just collecting ad revenue and not even watching their own programming lest they be found responsible.
The solution in that case was to go ahead and fine the wazoo out of the network, and then let the network sue the wazoo out of the artist and their record label.
Some of the best movies I have seen were more independent movies with unknown actors. Im sure they still received good salaries relative to your average job, but they werent astronomical.
Some of the worst movies I have seen were more independent movies with unknown actors. They might have been cheaper failures, but the investors still lost out.
Actually, on the whole I tend to agree with you - I'd rather see an innovator reap the rewards, than see a typical retread movie. However, there is a reason people pay big money for big name talent - audience draw. Stick Robin Williams in a movie and lots of people will buy tickets. To some degree, the fact that Robin Williams will even take a part helps to assure the investors that the producers are doing their job right.
It is the same reason that a company will hire a big consulting firm when chances are that some guy in their company could do the same thing for 1/100th the cost, or they could easily maintain a more diverse staff if they weren't spending that money on consultants. On a recent project a consulting firm was making 16X my salary for the duration of the project, and yet I probably contributed the majority of the innovation. However, the fact is that if I had messed up and there were no consultants then my boss would take a big hit, but if the consultants messed up then rather than everybody yelling at the guy who hired them for some reason everybody would obsess over arguing with the consulting firm. When you hire a big-name consulting firm then you're shielded from responsibility for your actions. Managers shudder at the though of actually having to find and hire people who can do a job right, and would rather contract it out so that when it doesn't go right they can blame an external company. Overall companies are more focused on blaming people when things go wrong than making them go right in the first place, and then we wonder why we're in the current state...
Studios are no different - they make the conservative choice, which is to do what worked before even if it costs more. When it goes wrong everybody just says "well, nobody could have seen that coming."
I tend to agree with a lot of your points - a specialized device will generally give a better experience than a general-purpose device. However, a LOT of people will stop buying specialized devices when the general-purpose device is "good enough." I didn't replace my mp3 player the last time it broke because while android isn't nearly as capable as Rockbox, it is "good enough" and is one less thing to carry around.
Another thing to consider is that Japan is not a typical market where GPS units are concerned. In much of rural Japan many streets are not even named. Sure, but ones are, but my rule of thumb is that if a road in the US has a number than the equivalent road in Japan would have a name, and if a road in the US doesn't have a number than the Japanese equivalent will just be pavement leading in some random direction. At best you'd have names for roads that would have numbers in New Jersey (a state that likes to number roads). As a result almost EVERYBODY in Japan owns a GPS. My understanding is that before GPS became popular most rural directions relied heavily on landmarks (turn right at the red house after the big tree, etc).
Having been there I have to say that Japan has the most curious rural/suburban culture. A road through town might have no names, and will alternatively lead past residential houses, rice fields, and tire dumps. And somewhere right in the middle of all of that (and nowhere near a crossroads or anything resembling a busy part of town) there will be a line of several vending machines that appear to get heavy use. I wonder if Japanese prefectures have any kind of semblance of a zoning board...
I agree with your points about the problems with automation - I avoid driving much with cruise control for the same reason - having to micro-manage my speed helps keep me alert. Sure, I could pay more attention to other things without having to do it, but in reality I'd pay less attention.
I'm not sure that firing pilots who fall asleep is going to help either.
I liked something I saw for airport security screeners - who face similar problems (they screen thousands of packages and 99.99999% of the time there is nothing to see). It was an x-ray machine that would add in images of contraband for the operator to spot - if an operator didn't hit a button when one was spotted then it would alert a supervisor. It gave the operator something to actually do, and thus it kept them alert.
Maybe the plane needs to trigger a random simulated failure (caution light or whatever) that requires a button to be hit to clear the condition (FMS would avoid triggering it at critical moments, and the operator would have plenty of time to deal with other stuff first). Or, maybe the cockpit should have officially-sanctioned ways to do things like check email/etc which will do things like pop up occasional messages to do a visual scan and which will blank the screen the instant an alert of some kind occurs. Or, maybe there is some task the crew could perform that is more mentally stimulating than staring out the window at blue sky.
The human brain is a machine - a complex one, but a machine nonetheless. It has certain limitations. In particular, it gets bored if you don't give it something to do. This is biology, and simply telling pilots not to be bored doesn't fix the problem. Likewise, the human brain requires sleep, and if people are overworked they won't get enough of that.
That's a BIG stretch. That's like saying that a mayor who invests in anything shouldn't be allowed to pass a law that might improve the economy (heaven forbid!).
I'm all for requiring those holding significant political power to put their investments in some kind of a blind trust. I'm also all for declarations and avoidances of conflicts of interest.
However, the idea that anything that helps any company could be a conflict of interest because your wife works at a bank that may or may not invest in that company is REALLY far fetched. For all the guy knows Goldman might have invested in a major competitor to the company making the H1N1 vaccine and that they'll suffer as a result.
I wasn't trying to suggest that this wasn't in keeping with the letter of the law. I was trying to suggest that the law is dumb if this is the case, and I doubt this was the original intent of those who wrote the law.
If a software needs to be installed for it to be used, then that installation should not be considered copying/distribution under copyright law. Yes, I realize that it does require bits to be copied. However, copyright law doesn't exist because copying is evil - it exists to ensure that those who make creative works are able to profit from them reasonably. If I buy a copy of microsoft office, and then install it on my computer, I haven't deprived microsoft of revenue despite having "copied" it. If I make a backup of the original media it doesn't cost them a dime either. If I make 10,000 backups of the original media that also doesn't cost them a dime. If I then sell those copies on ebay for $9.95 or pass them out for free on the street, then that does impact them financially. At the same time, some forms of non-profitable distibution should still be permitted.
The a legitimate of law is to facilitate commerce. Sane copyright policies can help do that. Insane policies do not.
That is a pretty ridiculous argument (spare me the citations, I'm sure there are 14 judges out there that have upheld it nonetheless).
When I listen to music from the radio it ends up being embedded in my memory - and yet I don't need a license to listen to the music.
I don't really have a major problem with copyright, but it is something that should be exclusively applied to people distributing things, not people who receive them. P2P technology is a gray area - I'm actually not opposed to liability there but the statutory fines need to be several orders of magnitude lower.
There are a couple of optimizations that are possible when you cross layers. Sure, it could be done in a way that retains the ability to swap layers, but the individual layers would need to expose a LOT more information to each other about their implementations. It would be difficult to abstract as well.
You asked for a good reason - here is an example:
Suppose I want to modify a 10kb text file. Since ZFS uses copy-on-write it picks a new allocation unit on the disk to write to. Now, it could just pick any old empty area on the disk. However, instead it can find a stripe on the RAID that is completely empty, and fit that 10kb file in with another few MB of other writes and put them all in the same stripe. Then it can pick a stripe that happens to be near where the cache was recently flushed. Since the whole stripe is being written only a single disk write to each component drive is needed.
Suppose you did the same thing with ext4 over lvm and md. Now the filesystem is going to overwrite the existing file in place. From the filesystem's perspective it isn't a big deal - either way we're overwriting one block with another. However, the filesystem doesn't realize that there is RAID underneath, so the md layer eventually gets a request to modify a few blocks in the middle of a stripe. The md layer of course does what it can to combine operations, but ext4 isn't doing anything to make this easier. Chances are that the whole stripe is going to get read (using all but one of the disks for raid 5), a small area inside will be modified, and then the stripe will get written again. While the stripe is being written the old stripe is being destroyed, so if anything goes wrong the previous contents of everything on that stripe will be lost - most of it EVEN if there is full data journaling at the ext4 layer (although the md layer could implement is own journal).
By crossing layers you can minimize the amount of work the drives need to do, and at the same time provide a higher level of data security by avoiding overwrites whenever possible.
Now, you could abstract these layers a bit more to make them more interchangeable, but the behavior of the filesystem is going to need to change based upon whether the underlying layer is striped or not, and of course that is already muddying the waters.
The set-top boxes are run directly to the router - on the world-facing side. They're all attached to the same coax line, and they communicate using an odd ethernet-over-coax protocol.
The other reply did point out a valid alternative.
Yup - same issue here with FIOS - I really didn't want to have to mess with getting their router to bridge (it can be done, but it is a real pain and if you need to make a change you need to reset and reconfigure). So, now I have a box sitting on my LAN that I have no control over. In theory all the devices on the LAN are routinely scanned with nessus/etc, but it isn't ideal.
Many of these integrated services have all kinds of tie-ins that make bridging the router painful. For example, on FIOS the network link from set-top boxes to the internet is via the outward-facing port on the router. If you bridge the router then it has no internet connectivity of its own and can't route packets from the set-top boxes. Plus, when it is bridged you can't get into the router's web-based admin console, so to change a setting you need to hard-reset it. I guess if you don't mind having your own router NATed that is an easy option. Sometimes I'm tempted to go IPv6 with a tunnel provider just to get past all that stuff...
FIOS is a bit of an unusual case since they run the network over coax. Where standard ethernet is used you have more hope of just bypassing the router entirely.
Yup - insurance only works in the absence of knowledge. If you could predict with 80% accuracy whether somebody's house would burn down, then almost everybody could get dirt-cheap fire insurance (which they wouldn't buy anyway since they wouldn't need it), and a small number of people wouldn't be able to afford it and would lose everything they have in a fire. The insurance companies would go out of business since nobody would bother buying insurance either way.
The only thing that would work once genetic testing becomes reliable is a system that has these attributes:
1. No denial of claims for pre-existing conditions.
2. No differential charging based upon genetic factors.
3. All people must pay in - coverage is not voluntary.
If you don't have all three the system breaks down. Either people rip off the insurers, or insurers rip off the people. Neither works.
Most of what people consider "health insurance" isn't really insurance anyway - it is more of a buyer's club for health services. IMHO a major area of reform should be to eliminate this aspect of health insurance entirely. There is no reason that a poor person without insurance should have to pay $100 for a doctor's visit that costs $30 for Aetna (even if the poor person can haggle them down to $50 - assuming they are in the condition to haggle BEFORE the services are rendered). This alone wouldn't fix health care for the poor, but it would certainly make that problem a lot easier to solve.
Then simply make a policy of no belt = no healthcare bailout. Generally speaking the sorts of people who advocate no seatbelt laws are also the sorts of people who advocate no tax-funded brain scraping. I can't imagine that the belt is going to make a big difference in how long you end up waiting stuck in traffic.
Nanny state policies tend to be the result of a nanny state. When the government bails you out when you can't afford your rent, then suddenly you have regulations on what kind of home you're allowed to rent. When the government pays for your doctor's bills, then suddenly your body doesn't really belong to you anymore.
Health care is a whole mess that I don't want to get into. I'm not totally opposed to a little bit of socialism for those who are down on their luck, but the whole system is in need of a massive overhaul. A catastrophic care insurance system coupled with clearly published provider fees and a requirement that EVERYBODY pay the same fee for the same service would go a long way (but there are no simple solutions to the whole mess).
Waste discharges are an economic externality, and many libertarian-minded people would argue that it should be regulated (preferably in a market-based manner).
Financial regulations is a pretty big category - most libertarian-minded people would be in favor of regulation that maximizes transparency and openness, and which prevents the creation of monopolies. So, CEOs misstating profits would be fair game - that is fraud.
Seat belts are a different matter. People should be perfectly free to drive without wearing seat belts - if people want to kill themselves why should that matter to me? However, I'm all for requiring standardized safety tests with clear disclosure of their results. The goal is to enable consumers to make their own decisions.
Finally you mention food safety. This is also an area where regulation is perfectly acceptable, because it is not straightforward for a consumer to determine whether a piece of meat they want to buy is contaminated. Now, regulation should be primarily about disclosure. For example, mandating disclosure of use of GMOs seems fine to me - if consumers care they can choose appropriately. Likewise, if there is some big debate about nitrate levels in meat, then require disclosure of nitrate concentration and let consumers decide whether they care or not.
Most libertarians do not advocate a world free of all regulation - just those which essentially protect people from themselves and often create perverse incentives.
If you want an example of a perverse incentive - instead of spending $10 more on a TV that uses less power when it runs, maybe some consumers with children would rather spend $10 on a TV with an IR sensor that detects when nobody is in the room and shuts itself off. Such a TV might actually save more power overall, but consumers wouldn't be able to buy it under the proposed regulations. They could only buy a TV that cost $20 more that did this (being forced to spend money on a feature that actually only has a minor improvement for them).
Others have the right answer - just raise electrical rates to reflect the true cost including all externalities. Then people can figure out what kinds of efficiency improvements make the most sense for them.
Would this even do anything in a quench? Suppose that power line is carrying 1 GA (giga-amps) during normal use. The FCL would prevent damage to the line if for some reason it shorted and tried to pull 2 GA.
However, if there were a quench along the line, the current would actually drop, not rise (resistance increases dramatically). An FCL seems to protect against a short, but during normal operation a transmission line IS a short. The issue isn't that the overall line is carrying too much current. The problem is that the point that overheated and quenched is carrying WAY too much current by several orders of magnitude, but the overall line is below design capacity.
During a quench the current would generate heat at an incredible rate.
Maybe if you had some lightning-fast detector that could spot a voltage rise on either side of the line and cut the circuit completely that might work. I'm guessing that voltage would rise since the line now has resistance.
Disclaimer - I'm not an expert on superconduting transmission lines, nor am I an electrical engineer. I do have a half-decent grasp of Ohm's law, however, and have seen the results of superconducting magnet quenches.
Yup - just think about what would happen if any part of this conduit warmed up - talk about a MASSIVE heat dump!
Or, for that matter think about what would happen if somebody took the cable and twisted it into a coil - now you suddenly have a HUGE electromagnet.
Yup - if I owned a huge company my first rule would be that I would need to personally approve any consultant engagement.
When a manager comes to me and points out that if we don't hire the consultant we could end up losing more through a bad decision, I'd say, "fine - but if it happens I'll make sure you never get another job, unless you admit you're not qualified to do your job and recommend a replacement."
There really are situations where companies probably should engage consultants. The problem is that SO many of them are brought in to do stuff that the company should be able to do in the first place.
I've worked with some fairly competant consultants but even they tend to have knowledge gaps. For example, in the IT world many of these guys are great at deploying systems, but they don't have any idea how to deploy them in a way that makes them cheaper to maintain in the long term. In fact, they have an incentive to make them harder to maintain to get repeat business. When you do it yourself you have incentive to make sure that you won't have to spend six months out of the year recoding stuff that should be configurable...
Look, despite being fairly conservative I like a lot of things about Obama.
However, he doesn't deserve a Nobel, and I don't like what has happened to the Peace Prize in recent years. It is more a political statement than an award for actually accomplishing things. Perhaps Obama will save the world from itself, but he hasn't actually done it yet.
We don't give the Physics Prize to promising grad students in the hope that they will use the prestige to do amazing things. The Prize is given to people who already have done amazing things, in recognition of what they have done. In fact, it usually isn't given in other fields until the accomplishment has stood the test of time. Almost every science nobel has been awarded for some major advance that scientists are likely to consider important even a century later. Just look a the prizes from 50 years ago and how they've stood the test of time.
Has so little actually been accomplished to further world peace in the last 30 years that we must give the prize to people who are merely likely to further world peace?
You don't need nearly as long an antenna to receive an AM signal (granted, I still suspect they're a bit big for pocket-sized phones). If you allowed that the phone could transmit in one band in receive in another then the AM bands could still be useful.
Dalvik (the android "Java" interpreter) does not do JIT. So, it is quite slow. Biggest limitation on the platform I'd have to say.
I think a license like this would work well for Flash.
Dunno, but a license like this certainly would work well for Gentoo. :)
It's the one distro that can modify Firefox for its users AND leave the branding intact.
Well, the energy the rocket would end up with would be: .5*1E6ton*(0.5c)^2 = 1E25J.
The equivalent mass is 1E25J/c^2 = 125kT (hmm - all those c's cancel out).
This assumes that your energy source is completely transformed into kinetic energy, and that the exhaust contains no kinetic energy. I suspect to figure that all out I'd need to get into impulse and conservation of momentum and all that, and I'm too lazy. However, the mass isn't actually unreasonable considering a typical rocket going to vastly slower speeds is 99% fuel anyway. Of course, in addition to your super-efficient engine you also need a way to extract 100% of the energy out of matter to get away with so little of it. If you carried matter/antimatter you might get in the ballpark. Of course, if you crashed your ship with 60kT of antimatter aboard I think I'll like to be on the other side of the solar system. For that matter, the same applies if you crashed your ship into anything solid once you were at speed (granted you'd probably just punch a hole in the earth and come right out the other side with a big exit crater).