Apparently some fool filled up the body of the thing with methane and liquid oxygen. It's almost like they WANTED fire to come shooting out of that rocket!
The biggest that the US ever actually popped was Castle Bravo. Design yield: 4-6Mt. Actual yield: 15Mt, resulting in the dry bit of island it was sitting on turning into a deep spot in the reef and destruction of the monitoring equipment two islands over, not to mention dropping fallout all over the local civilians. Oops. The Castle-* designs were weaponized into the Mk-17/Mk-21/Mk-24 with a 5-15 Mt range.
The biggest the US ever deployed was the B41, at a perfectly practical 25Mt.
Yields peaked in the 60s because the complete assemblies were huge and if you could only cart around one bomb on your plane or missile, it might as well be a big one. Since then the trend in big bombs has been toward the 0.5-1 Mt range, like the B83. The reason doesn't really have much to do with "arsenal reduction"; the real story is they figured out how to shrink midsize ones down to a much smaller package, and it's simply more efficient (more stuff blown up per kg of plutonium) to drop a half dozen 1Mt bombs in a pattern than to drop a single 25Mt one and having most of the energy end up in a stratosphere-bumping mushroom cloud.
Of course that Soviet triple-stage monster takes the cake. There's simply no possible use for a larger one, even as a national dick waving status symbol. 50Mt is basically the most you can ever drop from a plane and live to tell about it, and you HAVE to drop it from a plane because a ground burst would create stupid amounts of fallout while not even being that impressive (air bursts work better); and no one's going to bother building a missile big enough to carry a 27,000 Kg firework just to show off. I hope.
So now you know, and knowing is how we get the next generation interested in one upmanship.
Now, arbitrage, where HFT makes money, is when Alice wants to sell for 100, and Bob wants to buy at 110. They don't know about each other. (WTF is the point of an exchange then!?)
If Alice and Bob are on the same exchange, the trade goes through directly.
Arbitrage comes into play when they are on two different exhcanges - Alice is asking 1@100 on MarketA, Bob is bidding 1@110 on MarketB. Since they're on different markets they never see each other, but HFT Harry is watching both markets and sees the opportunity, and is able to take advantage of it because he has available currency on one market and the security on the other. Harry takes both Alice and Bob's orders simultaneously, and after the trades settle, pockets the difference.
End result: Both Alice and Bob got the deal that they wanted, and the prices on both markets get pulled toward each other. That's good all around. Harry takes a little risk: his assets are tied up for a while during the settlement; someone else might take Bob's deal just after Harry's purchase of Alice's security went through, leaving Harry holding goods without a buyer (in reality he'll just sell to the next available offer on MarketB, say 1@109 - or perhaps 1@95, taking a net loss on the deal).
Because Harry's not the only HFT in town, MarketA and MarketB constantly get pulled closer together until the arbitrageurs decide there isn't enough margin to be worth the risks. Whichever HFT can operate most efficiently (IE, still managing to make a profit despite the diminishing returns, eg by reducing their risks by being able to reduce their latency to the markets) gets the most business.
Hope that helps. Ask away if any of it is unclear.:)
If someone is reading at a company with a filter so invasive that it's counting the number of fucks on a random web site, then perhaps they shouldn't be reading slashdot at work. I'm generally not very profane, but I will not bowdlerize myself for the convenience of people fucking around on paid time.
You're assuming that the advantage has no disadvantages. What if it is possible, but it requires a large expenditure of energy, or increases the rate of psychological problems? It may therefore be triggered by a recessive gene which is only common in populations where it has a benefit, much like sickle cell (big disadvantage generally) is common where malaria is widespread (it provides protection).
Or another way: flying is such an advantage that animals with wings would rapidly displace those without. In reality it's only enough of an advantage to occupy a niche due to the high costs.
None of this is to say I believe in remote viewing or anything, but you're rejecting it on faulty logic.
Trucks vs luxury cars doesn't really capture it though. Top notch stealth is a huge force multiplier. Two F-22s can take on several dozen non-stealth fighters: the conventionals swarm around not finding anything to shoot at while the F-22s pick them off one by one. It's really that dramatic of a difference.
In limited-scale wars, the F-22 is just overwhelmingly better even in small numbers. In a protracted world war things would be different: eventually the other guys would take them out through blind luck, or a significant portion of the small fleet would be down for maintenance; then the "lots of cheap fighters" strategy wins (and the US still wins, since it still has tousands of cheap F-15s, F-16s, and to a lesser degree F/A-18s). The thing is that every war fought in the last 50 years has been the former sort where having a small number of F-22s IS the better plan, and I don't see anything on the horizon that will change it.
How so? It's just OFDM, much like 802.11A/G/N and DSL. The technique isn't really new - it's just that we're starting to get more efficient chips which can run a continuous FFT while still getting reasonable battery life.
It depends how you measure efficiency. It would solve the problem faster in many cases, but that doesn't mean it would use less of your time. Both you and the customer can multitask much more effectively in chat. You're off helping someone else while they reboot, instead of just racking up minutes of dead air. I consider that to be more efficient, even if it takes longer.
It's also much more efficient when you have a rambley customer. Instead of cutting him off continuously or waiting it out, you do something else while he types up his whole story, then you skim it to find the bits you wanted to know.
I personally find it much more enjoyable to use chat as a customer. I'll call in if I need something fixed RIGHT NOW, but most of the time chat is much less frustrating than waiting on hold.
Yes. They solved the meltdown problem but created a bunch of different problems, which is why this is a potential future technology instead of a current one. It was just an example of inherent safety. These problems are just some of the reasons that I think LFTR is a better technology to pursue, though LFTR also needs a ton of engineering before it will be ready.
... thought that all modern reactors were protected by passive safety mechanisms...
Kori-1 is a PWR from 1978 and like most reactors currently in service, it is a generation II design. These have many passive safety features, but they are generally not fully passively safe.
Some Gen-III reactors (1990s tech) and most Gen-III+ reactors (2000s tech) have full passive safety, in addition to many other safety improvements like simplified designs, better containment, better redundancy, etc. Gen-IV designs (future) step it up to inherent safety - for example pebble beds (meltdown is impossible because thermal expansion stops the reaction even if all cooling strategies fail) or my favorite, the LFTR (liquid fuel - you can't melt down when you're already melted, and thermal runaway just drains it into a basin in a noncritical configuration; and it's an unpressurized design, which eliminates tons of problems).
So does this actually pose a safety risk?
Yes - like everything worth doing in life, there is always risk. In this case the risk is that the reactor requires water to be actively pumped for a while after shutdown. The risk is generally acceptable: a failure like the one that happened a few months back doesn't cause a sudden catastrophic failure. There is considerable thermal mass in the water inside the reactor so the temperature rise is gradual. In event of a failure you have days to get some power back online. In the case of the failure a few months back, the power was back within minutes and there was little chance that they wouldn't be able to manage it in time. The incident wasn't an acute safety risk; but it was a failure that shouldn't have happened in the first place, so it's somewhat worrying.
The problem is if you have a disaster like the Japan earthquake: if the power is knocked out and all your infrastructure is too crippled to fix the grid or truck in some generators; then things go sideways. These are low frequency events, but they happen, which is why in my opinion we need to start building modern reactors so we can decommission these old Gen-II relics.
Or the much easier way that I use: Buy a block of dry ice. Put it in a cooler. Drop the hard drive on the dry ice and close the lid, dangling the cables out. The CO2 drives out any water vapor. When you're done you just leave the drive inside until the rest of the dry ice sublimes away and the drive gets back to room temperature while still sitting in the dry chest.
It's also much more effective at cooling - the hard drive in the freezer will quickly warm up above room temperature when it starts operating. The cold air in the freezer doesn't conduct heat quickly enough. Sitting on a block of dry ice often lets you run nonstop. For the same reason, if you DO go with the freezer method, pack it in between a couple bags of peas or something instead of just sitting on the rack.
Also, there's no point in unplugging the freezer if you have the drive in a plastic bag. Any moisture that was in there originally will still be there tomorrow. All the new moisture condenses on the outside of the bag.
No one expects these "safes" to be secure against a burgular or fire. The threat model is young children pilfering the top dresser drawer. That SHOULD be possible to defend with a locking steel box, and I would expect that level of security even for the very low price, but the lock is so incompetently designed that a drop from a few inches unlocks it.
No, I mean QoS, as in prioritizing which packets get to go first to fulfill certain delivery guarantees. Traffic shaping is a QoS technique - without help from your ISP you have to manage the receive buffer with RED or similar.
There is a reason why firearm manufacturers create parts out of single bar stock aluminum or steel metal pieces and machine them out, that is to maintain durability during firing process of the gun so it doesn't explode in your hands and injure you or worse and that it will last over repeated uses.
In the case of the AR-15, all the fast moving, high stress parts (the chamber, bolt, bolt carrier, etc) are in the upper receiver. The lower receiver is all the low performance stuff: the trigger group, magazine latch, etc. Injection molded plastic lowers are already common. I wouldn't want to take one into combat where you're likely to bash your rifle while diving for cover or using the stock as an improvised battering ram, but for normal use they don't have any reliability or safety problems.
it's unlikely to make much difference unless you're planning to host a reasonably heavy server
For bandwidth, yes, but there's a big advantage in having such a surplus: you don't have to do aggressive QOS to prevent latency spikes and loss. Wanna game while someone else watches Netflix? No problem.
The technology is already available - OpenID and several other standards are ready to go.
The trouble is that everyone wants to be the ID provider, but no one wants to accept other providers. Passport is a great example - Microsoft wants to be the central gatekeeper. Well thanks, but no, I'd rather run my own, but of course MS won't accept it.
I understand engineering is hard, but engineering an electric car is quite a bit easier than a gas one. Any car company could do it in a heartbeat if they felt like it, and plenty of people do conversions in their own garage.
Everyone isn't doing it because gas cars are at a more profitable price point for the mass market, not because it's a particularly difficult engineering problem.
To me it appears to be a straightforward application of the idea: "Why isn't anyone trying to make a desirable electric car? Why don't we make an electric sports car? Fuck the people who say it can't be done, let's do it and see what happens!"
I know it's still a hard sell, but I think conservatives will find it much more palatable to have a "city-run utility providing access to a free market" than "more regulation".
It's called "regulation".. aka law and order for corporations.
I prefer a non-regulatory way to solve the problem: Municipal fiber. Have your city drag fiber from a central data center to everyone's homes. They rent you a pair for $10/month or so. At the data center you get cross-connected to any of a dozen providers who will sell you TV, IP, and telephone service in any combination you desire.
There's no need to regulate once you take away the natural monopoly held by the telcos/cablecos. Once you have a bunch of companies competing to have the other end of your fiber you win.
Unfortunately, history is full of cranks with over-unity devices reeling in literally millions from people willing to invest in a hypothetical free energy source without even being allowed to see what's in the box or get a vaguely plausible explanation for how it would work. There was one particular guy that managed to get himself in the news every couple of years with rigged demos, each time luring in a few new suckers, going on for decades.
But how can that possibly work out? The customers who get screwed are learning: if you want updates, buy a Nexus. The other brands are going to lose out with this policy.
I jumped ship from my iPhone after I got sick of Apple's BS, but honestly this was one of the BEST things they had going: my phone regularly got updates the very day they were released; my only delays were waiting for someone to jailbreak the new version before I could upgrade (which was the BS I got sick of; every damn update was a hassle for me, but for non-jailbreakers life was good). I got a Nexus specifically because I want regular updates and community support, but the entire rest of the Android landscape is pretty grim: updates are a "Maybe we'll get around to it next year" thing for all the non-Nexus phones.
Slashdot Mirror
Apparently some fool filled up the body of the thing with methane and liquid oxygen. It's almost like they WANTED fire to come shooting out of that rocket!
The biggest that the US ever actually popped was Castle Bravo. Design yield: 4-6Mt. Actual yield: 15Mt, resulting in the dry bit of island it was sitting on turning into a deep spot in the reef and destruction of the monitoring equipment two islands over, not to mention dropping fallout all over the local civilians. Oops. The Castle-* designs were weaponized into the Mk-17/Mk-21/Mk-24 with a 5-15 Mt range.
The biggest the US ever deployed was the B41, at a perfectly practical 25Mt.
Yields peaked in the 60s because the complete assemblies were huge and if you could only cart around one bomb on your plane or missile, it might as well be a big one. Since then the trend in big bombs has been toward the 0.5-1 Mt range, like the B83. The reason doesn't really have much to do with "arsenal reduction"; the real story is they figured out how to shrink midsize ones down to a much smaller package, and it's simply more efficient (more stuff blown up per kg of plutonium) to drop a half dozen 1Mt bombs in a pattern than to drop a single 25Mt one and having most of the energy end up in a stratosphere-bumping mushroom cloud.
Of course that Soviet triple-stage monster takes the cake. There's simply no possible use for a larger one, even as a national dick waving status symbol. 50Mt is basically the most you can ever drop from a plane and live to tell about it, and you HAVE to drop it from a plane because a ground burst would create stupid amounts of fallout while not even being that impressive (air bursts work better); and no one's going to bother building a missile big enough to carry a 27,000 Kg firework just to show off. I hope.
So now you know, and knowing is how we get the next generation interested in one upmanship.
Now, arbitrage, where HFT makes money, is when Alice wants to sell for 100, and Bob wants to buy at 110. They don't know about each other. (WTF is the point of an exchange then!?)
If Alice and Bob are on the same exchange, the trade goes through directly.
Arbitrage comes into play when they are on two different exhcanges - Alice is asking 1@100 on MarketA, Bob is bidding 1@110 on MarketB. Since they're on different markets they never see each other, but HFT Harry is watching both markets and sees the opportunity, and is able to take advantage of it because he has available currency on one market and the security on the other. Harry takes both Alice and Bob's orders simultaneously, and after the trades settle, pockets the difference.
End result: Both Alice and Bob got the deal that they wanted, and the prices on both markets get pulled toward each other. That's good all around. Harry takes a little risk: his assets are tied up for a while during the settlement; someone else might take Bob's deal just after Harry's purchase of Alice's security went through, leaving Harry holding goods without a buyer (in reality he'll just sell to the next available offer on MarketB, say 1@109 - or perhaps 1@95, taking a net loss on the deal).
Because Harry's not the only HFT in town, MarketA and MarketB constantly get pulled closer together until the arbitrageurs decide there isn't enough margin to be worth the risks. Whichever HFT can operate most efficiently (IE, still managing to make a profit despite the diminishing returns, eg by reducing their risks by being able to reduce their latency to the markets) gets the most business.
Hope that helps. Ask away if any of it is unclear. :)
If someone is reading at a company with a filter so invasive that it's counting the number of fucks on a random web site, then perhaps they shouldn't be reading slashdot at work. I'm generally not very profane, but I will not bowdlerize myself for the convenience of people fucking around on paid time.
You're assuming that the advantage has no disadvantages. What if it is possible, but it requires a large expenditure of energy, or increases the rate of psychological problems? It may therefore be triggered by a recessive gene which is only common in populations where it has a benefit, much like sickle cell (big disadvantage generally) is common where malaria is widespread (it provides protection).
Or another way: flying is such an advantage that animals with wings would rapidly displace those without. In reality it's only enough of an advantage to occupy a niche due to the high costs.
None of this is to say I believe in remote viewing or anything, but you're rejecting it on faulty logic.
Sorry guys, I created a monster. :)
Trucks vs luxury cars doesn't really capture it though. Top notch stealth is a huge force multiplier. Two F-22s can take on several dozen non-stealth fighters: the conventionals swarm around not finding anything to shoot at while the F-22s pick them off one by one. It's really that dramatic of a difference.
In limited-scale wars, the F-22 is just overwhelmingly better even in small numbers. In a protracted world war things would be different: eventually the other guys would take them out through blind luck, or a significant portion of the small fleet would be down for maintenance; then the "lots of cheap fighters" strategy wins (and the US still wins, since it still has tousands of cheap F-15s, F-16s, and to a lesser degree F/A-18s). The thing is that every war fought in the last 50 years has been the former sort where having a small number of F-22s IS the better plan, and I don't see anything on the horizon that will change it.
How so? It's just OFDM, much like 802.11A/G/N and DSL. The technique isn't really new - it's just that we're starting to get more efficient chips which can run a continuous FFT while still getting reasonable battery life.
... but unfortunately that doesn't mean that they ARE.
It would be much more efficient...
It depends how you measure efficiency. It would solve the problem faster in many cases, but that doesn't mean it would use less of your time. Both you and the customer can multitask much more effectively in chat. You're off helping someone else while they reboot, instead of just racking up minutes of dead air. I consider that to be more efficient, even if it takes longer.
It's also much more efficient when you have a rambley customer. Instead of cutting him off continuously or waiting it out, you do something else while he types up his whole story, then you skim it to find the bits you wanted to know.
I personally find it much more enjoyable to use chat as a customer. I'll call in if I need something fixed RIGHT NOW, but most of the time chat is much less frustrating than waiting on hold.
Yes. They solved the meltdown problem but created a bunch of different problems, which is why this is a potential future technology instead of a current one. It was just an example of inherent safety. These problems are just some of the reasons that I think LFTR is a better technology to pursue, though LFTR also needs a ton of engineering before it will be ready.
... thought that all modern reactors were protected by passive safety mechanisms ...
Kori-1 is a PWR from 1978 and like most reactors currently in service, it is a generation II design. These have many passive safety features, but they are generally not fully passively safe.
Some Gen-III reactors (1990s tech) and most Gen-III+ reactors (2000s tech) have full passive safety, in addition to many other safety improvements like simplified designs, better containment, better redundancy, etc. Gen-IV designs (future) step it up to inherent safety - for example pebble beds (meltdown is impossible because thermal expansion stops the reaction even if all cooling strategies fail) or my favorite, the LFTR (liquid fuel - you can't melt down when you're already melted, and thermal runaway just drains it into a basin in a noncritical configuration; and it's an unpressurized design, which eliminates tons of problems).
So does this actually pose a safety risk?
Yes - like everything worth doing in life, there is always risk. In this case the risk is that the reactor requires water to be actively pumped for a while after shutdown. The risk is generally acceptable: a failure like the one that happened a few months back doesn't cause a sudden catastrophic failure. There is considerable thermal mass in the water inside the reactor so the temperature rise is gradual. In event of a failure you have days to get some power back online. In the case of the failure a few months back, the power was back within minutes and there was little chance that they wouldn't be able to manage it in time. The incident wasn't an acute safety risk; but it was a failure that shouldn't have happened in the first place, so it's somewhat worrying.
The problem is if you have a disaster like the Japan earthquake: if the power is knocked out and all your infrastructure is too crippled to fix the grid or truck in some generators; then things go sideways. These are low frequency events, but they happen, which is why in my opinion we need to start building modern reactors so we can decommission these old Gen-II relics.
Or the much easier way that I use: Buy a block of dry ice. Put it in a cooler. Drop the hard drive on the dry ice and close the lid, dangling the cables out. The CO2 drives out any water vapor. When you're done you just leave the drive inside until the rest of the dry ice sublimes away and the drive gets back to room temperature while still sitting in the dry chest.
It's also much more effective at cooling - the hard drive in the freezer will quickly warm up above room temperature when it starts operating. The cold air in the freezer doesn't conduct heat quickly enough. Sitting on a block of dry ice often lets you run nonstop. For the same reason, if you DO go with the freezer method, pack it in between a couple bags of peas or something instead of just sitting on the rack.
Also, there's no point in unplugging the freezer if you have the drive in a plastic bag. Any moisture that was in there originally will still be there tomorrow. All the new moisture condenses on the outside of the bag.
No one expects these "safes" to be secure against a burgular or fire. The threat model is young children pilfering the top dresser drawer. That SHOULD be possible to defend with a locking steel box, and I would expect that level of security even for the very low price, but the lock is so incompetently designed that a drop from a few inches unlocks it.
No, I mean QoS, as in prioritizing which packets get to go first to fulfill certain delivery guarantees. Traffic shaping is a QoS technique - without help from your ISP you have to manage the receive buffer with RED or similar.
There is a reason why firearm manufacturers create parts out of single bar stock aluminum or steel metal pieces and machine them out, that is to maintain durability during firing process of the gun so it doesn't explode in your hands and injure you or worse and that it will last over repeated uses.
In the case of the AR-15, all the fast moving, high stress parts (the chamber, bolt, bolt carrier, etc) are in the upper receiver. The lower receiver is all the low performance stuff: the trigger group, magazine latch, etc. Injection molded plastic lowers are already common. I wouldn't want to take one into combat where you're likely to bash your rifle while diving for cover or using the stock as an improvised battering ram, but for normal use they don't have any reliability or safety problems.
it's unlikely to make much difference unless you're planning to host a reasonably heavy server
For bandwidth, yes, but there's a big advantage in having such a surplus: you don't have to do aggressive QOS to prevent latency spikes and loss. Wanna game while someone else watches Netflix? No problem.
The technology is already available - OpenID and several other standards are ready to go.
The trouble is that everyone wants to be the ID provider, but no one wants to accept other providers. Passport is a great example - Microsoft wants to be the central gatekeeper. Well thanks, but no, I'd rather run my own, but of course MS won't accept it.
So we're now in a standoff.
I understand engineering is hard, but engineering an electric car is quite a bit easier than a gas one. Any car company could do it in a heartbeat if they felt like it, and plenty of people do conversions in their own garage.
Everyone isn't doing it because gas cars are at a more profitable price point for the mass market, not because it's a particularly difficult engineering problem.
To me it appears to be a straightforward application of the idea: "Why isn't anyone trying to make a desirable electric car? Why don't we make an electric sports car? Fuck the people who say it can't be done, let's do it and see what happens!"
There is an entire profession for this: Lawyers.
I know it's still a hard sell, but I think conservatives will find it much more palatable to have a "city-run utility providing access to a free market" than "more regulation".
It's called "regulation" .. aka law and order for corporations.
I prefer a non-regulatory way to solve the problem: Municipal fiber. Have your city drag fiber from a central data center to everyone's homes. They rent you a pair for $10/month or so. At the data center you get cross-connected to any of a dozen providers who will sell you TV, IP, and telephone service in any combination you desire.
There's no need to regulate once you take away the natural monopoly held by the telcos/cablecos. Once you have a bunch of companies competing to have the other end of your fiber you win.
Unfortunately, history is full of cranks with over-unity devices reeling in literally millions from people willing to invest in a hypothetical free energy source without even being allowed to see what's in the box or get a vaguely plausible explanation for how it would work. There was one particular guy that managed to get himself in the news every couple of years with rigged demos, each time luring in a few new suckers, going on for decades.
But how can that possibly work out? The customers who get screwed are learning: if you want updates, buy a Nexus. The other brands are going to lose out with this policy.
I jumped ship from my iPhone after I got sick of Apple's BS, but honestly this was one of the BEST things they had going: my phone regularly got updates the very day they were released; my only delays were waiting for someone to jailbreak the new version before I could upgrade (which was the BS I got sick of; every damn update was a hassle for me, but for non-jailbreakers life was good). I got a Nexus specifically because I want regular updates and community support, but the entire rest of the Android landscape is pretty grim: updates are a "Maybe we'll get around to it next year" thing for all the non-Nexus phones.