If every miner stopped today, they network would adjust the difficulty down, down, down, to the point where a single miner on a desktop PC could quickly process the low volume of transactions for the whole network and start farming up BTC.
If every miner stopped today, Bitcoin would die. Difficulty isn't adjusted based on time, but on blocks mined. Specifically, after every 2016 blocks, the difficulty is adjusted up or down by no more than a factor of 4. With nobody mining, the difficulty is never adjusted, and the required computing power to re-start the network is the same as it was when the shutdown happened.
Bitcoin can survive gradual changes in available computing power, but not rapid shifts.
It's not even remotely practical. An airplane's wings and fuselage are made out of a combination of high strength-to-weight materials and clever geometry. Batteries have pretty much zero structural strength, and seriously restrict what sort of geometry you can use.
A human can't maintain the level of vigilance needed to take over in an emergency for more than a few minutes. It's better to lift the "human driver" requirement explicitly than it is to pretend that the passenger sitting in the front-left seat is a safety measure.
None of the N-1 explosions killed anybody. Maybe you're thinking of the Nedelin disaster, where the pad explosion of an R-16 killed about a hundred people. The Soviet Union didn't have just one space program, it had three, and the N-1 and R-16 were from different programs.
And although it would turn up a lot of false positives, it would also catch a lot of cancers very, very early, making survival rates higher and costs lower.
Forty years ago, prostate cancer was considered a fast-moving, highly lethal cancer. PSA screening was introduced, with a heavy push for annual screening of men over the age of 50. Large number of men tested positive, and a metric shitload of tumors were found and treated. And the death rate didn't go down.
For the past several decades, there's been a heavy push for frequent breast examinations and mammograms for women, the idea being that if breast cancer is detected in the early stages, it's easy to treat and many lives will be saved. The detection rate has certainly gone up, as has the number of tumors treated. The death rate? Hardly budged.
It's something of an article of faith among anti-cancer activists that screening and early treatment save lives. In practice, the vast majority of improvement in cancer survival has come from improved treatments, not improved detection. Most people with early-stage cancer either have something so slow-growing that it can be safely treated at a later date (or not at all), or something so fast-growing that they'll die despite treatment. The percent of cancers where early treatment will improve the outcome is believed to be in the low single digits.
There are three threat situations involved here: one process attacking another process, one process attacking the kernel, and sandboxed code (think: Javascript) attacking parts of the process outside the sandbox.
Intel's fix is a partial fix for the first two situations (cache contamination is not the only way for information to leak, it's just the easiest and most reliable one to exploit), and is important for cloud and shared-hosting providers. It does absolutely nothing to mitigate the third situation, where sandboxed code is trying to read memory outside the sandbox, yet that situation is the one that desktop users are most vulnerable to.
All Intel CPUs that support speculative execution are vulnerable. That means the Pentium Pro and newer, all Celeron and Xeon CPUs, all Core CPUs, and all Atom CPUs except the early "Bonnell" architecture. If you've got an original-flavor Pentium or earlier, you're fine. If you've got a first-generation Atom CPU, you're fine. If you're one of those suckers who bet on Itanium, you're fine. Anything else, you need an update.
I don't think you quite understand how steel works. Steel gets dramatically weaker as it gets hotter. At a temperature of merely "too hot to hold", it's already lost a measurable amount of strength. By the time you hit the ignition point of wood, your typical structural steel will have lost about half its strength (and will stretch like taffy, making materials testing an exciting proposition).
This would be a good thing to sell policies for, though... the odds of the company not also being obliterated in either the impact itself or subsequent extinction event/societal collapse and having to make payouts would be very small!
Actually, it would be a pretty poor thing to sell policies for. Most asteroids aren't dinosaur-killers. They're 20-meter city-shattering rocks, and if one of them actually hits a city, the resulting payouts would bankrupt most insurance companies.
Forget Ryzen. I'd like to see one of the latest CPUs benchmarked against a Core i7-3960X. 6C/12T, 3.3GHz base clock, 15MB of cache, fully buzzword-compliant. Oh, and it's almost six years old.
Honestly, it's hard to get excited about "bringing the heat" when we're talking about single-digit percentage gains. There hasn't been a breakthrough in either clock speed or IPC in years, and even core counts have remained pretty much the same.
First, the nervous system uses both positive ("there's something happening") and negative ("there's nothing happening") signals. If you amputate a limb, the brain stops receiving both types of signals, and the absence of negative signals is interpreted as sensations from the limb.
Second, the boundaries between the parts of the brain controlling different parts of the body isn't sharp. If you cut off somebody's hand, signals from other areas such as the "arm" part of the brain will spill over, and there won't be stronger "hand" signals to override them. Since the signals don't come with tags indicating their source, the "hand" part of the brain sees them as coming from the hand.
Congratulations -- you found the edge cases, the few works that continue to bring in substantial profits for a long time.
For literature as a whole, 99% of profits are made within the first decade of initial release. For music, within a year. Magazines make their profit within a month, and newspaper articles, within a day. Movies probably fall into the "one year" bucket, but Hollywood accounting makes it impossible to tell.
The single greatest threat to most creators is copyright terms. Most people aren't the next Shakespeare, or even the next Douglas Adams. Their best bet for keeping their works in circulation isn't a company raking in the millions, but communities of dedicated fans, and copyright terms -- even if they were a simple "author's lifespan" duration -- prevent that.
The newest StreetView data around here is ten months old, the oldest ten years. Most small towns have coverage of exactly one road; some don't even have that. Even in cities, there are gaps in coverage, and not just minor dead-ends: I'm aware of a three-block gap in a major arterial.
So with all that in mind, is the barrier to selling a lot of supersonic jet tickets really the sonic booms? The Concorde didn't go under because it was annoying too many transatlantic tanker crews with sonic booms. I may not be fully informed, but this seems like a dumb business plan if you are just trying to solve the audio problem.
The sonic-boom problem locked the Concorde out of most of the really long-haul runs: things like Europe-West Coast US, Asia-East Coast US, and Anywhere->Chicago/Atlanta. New York-London in two hours versus six hours isn't much of an advantage; New York-Tokyo in four hours rather than fourteen is huge.
A solar-powered electric car is an attractive idea, but the numbers don't actually work out.
Let's assume a best-case scenario: 100% efficient panels on a car that manages to somehow combine the roof area of a full-sized van with the aerodynamics of a Model S. A solar panel facing directly into the sun produces 1 kW per square meter, so the 10 square meters of panels on your roof are producing a constant 10 kW. Feed that into a Model S's 32 kWh per 100 miles, and you get a sustainable speed of around 30 mph with the sun directly overhead. Now, aerodynamic effects are non-linear, so your actual sustainable speed is probably around 40-45 mph, but still not impressive.
And that's with unrealistically idealistic assumptions. The Model S looks like it's got maybe 3-4 square meters of non-window surface. Coat that with the best panels a lab has ever produced (46% efficient) and de-rate by a factor of 2 to deal with real-world solar incidence angles, and you're looking at a power supply of 1 kW or so, giving a sustainable speed of perhaps 5 mph. Hey, it beats walking.
These are things like the bespoke industrial-control software running a 50-ton press built in 1943, or an EDM cutter that still thinks it's speaking to an ASR-33 terminal at the other end of its RS-232 cable. These are hardware peripherals that would cost upwards of half a million dollars to replace, where reverse-engineering is nearly as expensive and considerably riskier.
The cost of developing a 16-bit-compatible version of Windows is peanuts next to the cost of upgrading.
It's possible to create a shaped charge occupying part but not all of a laptop's battery compartment. This leaves enough battery to power the laptop for a "turn it on to prove it's a laptop" check, so if you dress it up right, the laptop bomb could theoretically get through security.
Thing is, this shaped charge can't be made very large, because you need to leave room for the truncated battery. If you hold the laptop up against the side of the airplane while detonating it, you might be able to make a hole big enough to bring down the plane. If you detonate the laptop in the middle of the cargo hold, the blast will diffuse itself against the surrounding luggage.
Now, Aviation Herald reports an in-flight battery fire every few weeks. It reports a terrorist attack bringing down an airplane once every decade or so. Given the relative frequencies, I know which threat I'd prefer the TSA protect us against.
Okay, so a Comcast tech is calling to tell you he'll be arriving in half an hour to fix your cable.
How many phone lines do you think a big company like Comcast has? Are you aware that each of those phone lines has its own number? Are you willing to give up the ability to identify business callers, in exchange for being able to reliably identify private callers?
Are you still sure that caller ID masking is a bad thing?
Because "Caller ID" is just a bit of user-controlled data tacked on to the call. Call routing and billing are handled through a different, internally-controlled number.
There are good reasons to do this: consider an office with a hundred phone lines routing into a PBX, and then fanning out to a thousand internal extensions. Each of those phone lines has its own number for billing and routing purposes, and it's unreasonable to expect that office's contacts to memorize all hundred of those phone numbers. Instead, the office PBX can be set up to send out a single, "official" phone number in the caller ID data, so that people just need to remember that calls from 1-800-547-7277 are from the company in question.
Averaged over significant periods of time, heart rate is proportional to breath rate, breath rate is proportional to oxygen consumption, and oxygen consumption is proportional to calories burned. So in theory, starting from heart rate alone, you can measure calorie burn.
The problem is that the constants of proportionality vary from person to person. Further, since the number you're measuring (heart rate) is three steps away from the number you want (calorie burn), errors tend to magnify each other. In practice, any device that calculates calorie burn from heart rate is just making a wild guess.
Slashdot Mirror
If every miner stopped today, Bitcoin would die. Difficulty isn't adjusted based on time, but on blocks mined. Specifically, after every 2016 blocks, the difficulty is adjusted up or down by no more than a factor of 4. With nobody mining, the difficulty is never adjusted, and the required computing power to re-start the network is the same as it was when the shutdown happened.
Bitcoin can survive gradual changes in available computing power, but not rapid shifts.
It's not even remotely practical. An airplane's wings and fuselage are made out of a combination of high strength-to-weight materials and clever geometry. Batteries have pretty much zero structural strength, and seriously restrict what sort of geometry you can use.
A human can't maintain the level of vigilance needed to take over in an emergency for more than a few minutes. It's better to lift the "human driver" requirement explicitly than it is to pretend that the passenger sitting in the front-left seat is a safety measure.
None of the N-1 explosions killed anybody. Maybe you're thinking of the Nedelin disaster, where the pad explosion of an R-16 killed about a hundred people. The Soviet Union didn't have just one space program, it had three, and the N-1 and R-16 were from different programs.
It's not something people like to talk about, but early detection doesn't correspond to improved survival (see, for example, Screening for prostate cancer: systematic review and meta-analysis of randomised controlled trials ).
Forty years ago, prostate cancer was considered a fast-moving, highly lethal cancer. PSA screening was introduced, with a heavy push for annual screening of men over the age of 50. Large number of men tested positive, and a metric shitload of tumors were found and treated. And the death rate didn't go down.
For the past several decades, there's been a heavy push for frequent breast examinations and mammograms for women, the idea being that if breast cancer is detected in the early stages, it's easy to treat and many lives will be saved. The detection rate has certainly gone up, as has the number of tumors treated. The death rate? Hardly budged.
It's something of an article of faith among anti-cancer activists that screening and early treatment save lives. In practice, the vast majority of improvement in cancer survival has come from improved treatments, not improved detection. Most people with early-stage cancer either have something so slow-growing that it can be safely treated at a later date (or not at all), or something so fast-growing that they'll die despite treatment. The percent of cancers where early treatment will improve the outcome is believed to be in the low single digits.
Power 7 and later, yes, Power 6 and earlier, no, and I don't know about System/Z.
Skip the fix. It doesn't help you.
There are three threat situations involved here: one process attacking another process, one process attacking the kernel, and sandboxed code (think: Javascript) attacking parts of the process outside the sandbox.
Intel's fix is a partial fix for the first two situations (cache contamination is not the only way for information to leak, it's just the easiest and most reliable one to exploit), and is important for cloud and shared-hosting providers. It does absolutely nothing to mitigate the third situation, where sandboxed code is trying to read memory outside the sandbox, yet that situation is the one that desktop users are most vulnerable to.
All Intel CPUs that support speculative execution are vulnerable. That means the Pentium Pro and newer, all Celeron and Xeon CPUs, all Core CPUs, and all Atom CPUs except the early "Bonnell" architecture. If you've got an original-flavor Pentium or earlier, you're fine. If you've got a first-generation Atom CPU, you're fine. If you're one of those suckers who bet on Itanium, you're fine. Anything else, you need an update.
ARM is largely immune to Meltdown and Spectre. Both attacks require out-of-order execution; most ARM CPUs are strictly in-order devices.
I don't think you quite understand how steel works. Steel gets dramatically weaker as it gets hotter. At a temperature of merely "too hot to hold", it's already lost a measurable amount of strength. By the time you hit the ignition point of wood, your typical structural steel will have lost about half its strength (and will stretch like taffy, making materials testing an exciting proposition).
Actually, it would be a pretty poor thing to sell policies for. Most asteroids aren't dinosaur-killers. They're 20-meter city-shattering rocks, and if one of them actually hits a city, the resulting payouts would bankrupt most insurance companies.
Forget Ryzen. I'd like to see one of the latest CPUs benchmarked against a Core i7-3960X. 6C/12T, 3.3GHz base clock, 15MB of cache, fully buzzword-compliant. Oh, and it's almost six years old.
Honestly, it's hard to get excited about "bringing the heat" when we're talking about single-digit percentage gains. There hasn't been a breakthrough in either clock speed or IPC in years, and even core counts have remained pretty much the same.
What happens with phantom limbs is twofold:
First, the nervous system uses both positive ("there's something happening") and negative ("there's nothing happening") signals. If you amputate a limb, the brain stops receiving both types of signals, and the absence of negative signals is interpreted as sensations from the limb.
Second, the boundaries between the parts of the brain controlling different parts of the body isn't sharp. If you cut off somebody's hand, signals from other areas such as the "arm" part of the brain will spill over, and there won't be stronger "hand" signals to override them. Since the signals don't come with tags indicating their source, the "hand" part of the brain sees them as coming from the hand.
Congratulations -- you found the edge cases, the few works that continue to bring in substantial profits for a long time.
For literature as a whole, 99% of profits are made within the first decade of initial release. For music, within a year. Magazines make their profit within a month, and newspaper articles, within a day. Movies probably fall into the "one year" bucket, but Hollywood accounting makes it impossible to tell.
The single greatest threat to most creators is copyright terms. Most people aren't the next Shakespeare, or even the next Douglas Adams. Their best bet for keeping their works in circulation isn't a company raking in the millions, but communities of dedicated fans, and copyright terms -- even if they were a simple "author's lifespan" duration -- prevent that.
The newest StreetView data around here is ten months old, the oldest ten years. Most small towns have coverage of exactly one road; some don't even have that. Even in cities, there are gaps in coverage, and not just minor dead-ends: I'm aware of a three-block gap in a major arterial.
We know what microwaves do: not much. As long as the emissions remain below a level that causes localized heating, they don't cause problems.
The sonic-boom problem locked the Concorde out of most of the really long-haul runs: things like Europe-West Coast US, Asia-East Coast US, and Anywhere->Chicago/Atlanta. New York-London in two hours versus six hours isn't much of an advantage; New York-Tokyo in four hours rather than fourteen is huge.
A solar-powered electric car is an attractive idea, but the numbers don't actually work out.
Let's assume a best-case scenario: 100% efficient panels on a car that manages to somehow combine the roof area of a full-sized van with the aerodynamics of a Model S. A solar panel facing directly into the sun produces 1 kW per square meter, so the 10 square meters of panels on your roof are producing a constant 10 kW. Feed that into a Model S's 32 kWh per 100 miles, and you get a sustainable speed of around 30 mph with the sun directly overhead. Now, aerodynamic effects are non-linear, so your actual sustainable speed is probably around 40-45 mph, but still not impressive.
And that's with unrealistically idealistic assumptions. The Model S looks like it's got maybe 3-4 square meters of non-window surface. Coat that with the best panels a lab has ever produced (46% efficient) and de-rate by a factor of 2 to deal with real-world solar incidence angles, and you're looking at a power supply of 1 kW or so, giving a sustainable speed of perhaps 5 mph. Hey, it beats walking.
These aren't "apps".
These are things like the bespoke industrial-control software running a 50-ton press built in 1943, or an EDM cutter that still thinks it's speaking to an ASR-33 terminal at the other end of its RS-232 cable. These are hardware peripherals that would cost upwards of half a million dollars to replace, where reverse-engineering is nearly as expensive and considerably riskier.
The cost of developing a 16-bit-compatible version of Windows is peanuts next to the cost of upgrading.
I believe the correct response to this is "Challenge Accepted".
It's possible to create a shaped charge occupying part but not all of a laptop's battery compartment. This leaves enough battery to power the laptop for a "turn it on to prove it's a laptop" check, so if you dress it up right, the laptop bomb could theoretically get through security.
Thing is, this shaped charge can't be made very large, because you need to leave room for the truncated battery. If you hold the laptop up against the side of the airplane while detonating it, you might be able to make a hole big enough to bring down the plane. If you detonate the laptop in the middle of the cargo hold, the blast will diffuse itself against the surrounding luggage.
Now, Aviation Herald reports an in-flight battery fire every few weeks. It reports a terrorist attack bringing down an airplane once every decade or so. Given the relative frequencies, I know which threat I'd prefer the TSA protect us against.
Okay, so a Comcast tech is calling to tell you he'll be arriving in half an hour to fix your cable.
How many phone lines do you think a big company like Comcast has? Are you aware that each of those phone lines has its own number? Are you willing to give up the ability to identify business callers, in exchange for being able to reliably identify private callers?
Are you still sure that caller ID masking is a bad thing?
Because "Caller ID" is just a bit of user-controlled data tacked on to the call. Call routing and billing are handled through a different, internally-controlled number.
There are good reasons to do this: consider an office with a hundred phone lines routing into a PBX, and then fanning out to a thousand internal extensions. Each of those phone lines has its own number for billing and routing purposes, and it's unreasonable to expect that office's contacts to memorize all hundred of those phone numbers. Instead, the office PBX can be set up to send out a single, "official" phone number in the caller ID data, so that people just need to remember that calls from 1-800-547-7277 are from the company in question.
Averaged over significant periods of time, heart rate is proportional to breath rate, breath rate is proportional to oxygen consumption, and oxygen consumption is proportional to calories burned. So in theory, starting from heart rate alone, you can measure calorie burn.
The problem is that the constants of proportionality vary from person to person. Further, since the number you're measuring (heart rate) is three steps away from the number you want (calorie burn), errors tend to magnify each other. In practice, any device that calculates calorie burn from heart rate is just making a wild guess.
It also helps that heart-surgery robots are not under continuous attack from organized crime.