The only reason you have to move your head around with a side mirror is because they are placed very close to where your peripheral vision ends. Consequently, if you're sitting facing forward, for a mirror to show the region between where the rearview mirror's view ends (almost straight back) to where your peripheral vision picks up (almost straight to the side), it has to have a very large field of view.
With a camera, you have the option of mounting it at the front corners of the car instead of by the driver, The display can still be by the driver, but the camera can be way in front. It can then show the same area using a much smaller field of view. The blind spot will still be there, but it'll be pushed out to 2-3 lanes away, making it irrelevant.
No hypocrisy or double standards. The U.S. didn't benefit from the Iraq vote (most of the oil contracts went to non-US companies). Hence there was no conflict of interest. I think the U.S. was wrong to invade without UN approval, but the U.S. lost lives, lost equipment, lost money, lost international respect, and suffered degraded ability to react to matters more pressing to its self interests (Taliban in Afghanistan and Pakistan). The only thing they gained was eliminating a dictator from the world stage (one they helped put there in the first place, but that's another story).
Russia did benefit from the Crimea vote - they annexed a huge amount of territory. Hence there is a huge conflict of interest, and why people are refusing to recognize the vote. If Russian had simply stood by the sidelines, and the people of Crimea had revolted on their own and held their own elections demanding secession from Ukraine, then there might be some international support for what happened. Heck, people might even have supported Russia's invasion. But the way Russian played it out, the results are indistinguishable from if they invaded and held a rigged election.
As the saying goes, to have a crime there has to be a motive, means, and opportunity. While the U.S. had the means and opportunity in Iraq, there was no motive - else we'd still be there. While the means Russia used may have been more benign*, the fact that they also had a motive and opportunity raises a lot of suspicion.
* The fact that no lives were lost in Crimea is I think more attributable to Ukraine deciding not to elevate the situation into a war they knew they had no chance of winning militarily. Not due to Russian benevolence. Had Ukraine fought back as Saddam Hussein did, do you really think no lives would have been lost?
Pants have pockets. Phones fit in pockets. Problem solved.
Computer storage used to be the size of a washing machine, then a desktop PC, then a brick, then half a brick, then a quarter of a brick, then a deck of playing cards, then a stack of credit cards, then half a stack of credit cards, then a postage stamp, and now a microSD card is about the size of my little fingernail. About the time it went to postage stamp size, people started complaining it was too small and too easy to lose.
The same thing is happening to personal computers. That used to be a desktop, then a laptop, now it's becoming a tablet (or a laptop nearly as small as a tablet), soon it'll be your phone. Looking beyond that, it'll be the size of a credit card, then a postage stamp, and eventually the size of your fingernail. Somewhere between those last three stages, your PC will be so small that carrying it around in your pocket will be too risky because you might lose it.
Wearable computing is the obvious solution. First a device like a wristwatch. Then probably a ring or integrated into your glasses. If you're content living in the present, then good for you. But stop trying to drag down those of us trying to bring the future here more quickly.
The Vermont Yankee plant has been in operation since 1972, and generates about 4700 GWh each year. So over its lifetime it has probably generated about 190,000 GWh of electricity. $1 billion in decommissioning costs works out to about 0.5 cents per kWh (vs an industrial/commercial/residential cost of about 10/14/16 cents per kWh in Vermont). It does not substantially change the lifetime cost of the electricity generated.
All the states I know of require nuclear plant operators to collect these decommissioning costs in a fund. That is, part of their revenue from electricity sales must be placed into a trust fund specifically for decommissioning the plant in the future. That the Yankee plant has insufficient funds despite 41 years of operation suggests accounting fraud, regulatory incompetence, or inflated decommissioning costs more than it does non-viability of nuclear power. The San Onofre nuclear plant operated a similar amount of time, had about 3.6x the generating capacity when it was shut down, and was also shut down prematurely. It has a $2.7 billion decommissioning fund which is expected to exceed the costs, so they are planning to refund the excess to customers.
Yeah, this reminds me of the 1980 Moscow Olympics boycott. While it may have had a symbolic meaning (it was in protest of the Soviet invasion of Afghanistan), it ended up hurting the athletes more than the Soviet Union it was intended to hurt. And it seemingly forgot that the 1984 Olympics were in Los Angeles, which of course the Soviets promptly boycotted.
A lot of the police/government are on the cartels' payroll.
That's the problem with endemic corruption. If you try to fix just one part of it, the other parts resist. You have to somehow simultaneously fix all of it.
But not for technical reasons. While the engineering would be difficult to do, it's possible.
The problem is legal. It stems back to a question a friend of mine asked when noise-cancelling headphones first appeared. "If a sound wave has energy, and an equivalent wave 180 degrees out of phase also has energy, and when you combine the two you get no sound, where does the energy go?" Obviously the energy for both waves goes to regions where the two waves don't cancel.
So if you protected an area with this device, after an earthquake you'd be sued into oblivion by neighbors claiming your device increased damage to their property.
I don't understand why we're trying to solve this problem with backup cameras. There's a cheaper and better solution already available. Just mount a fresnel lens on the rear window.
It only costs a few bucks.
It functions whether you're using the rearview mirror or looking over your shoulder.
It functions without depriving you of peripheral vision out the other windows.
It doesn't get blocked by a tiny amount of mud or snow.
It's almost like someone who manufactures the backup camera systems has friends in the NTSB.
Getting a patent is supposed to require coming up with something new and non-obvious - something many of Apple's patent claims lack. (e.g. pinch/spread to compress/zoom).
Pinch to zoom in 1988. Apple's patented version is one of those ridiculous "on a" patents, where you take (steal) a pre-existing idea, and tack the words "on a [something]" to the end of it. In their case, "pinch to zoom on a capacitive screen." The USPTO overturned the patent last year.
- If a smartphone is stolen, the thief gets a free phone, and you have to buy a new phone. Net result = companies sold 0 phones to the thief, 2 phones to you = 2 phones sold.
- If no phone is stolen, you keep your phone, and thief has to buy a new phone. Net result = companies sell 1 phone to the thief, 1 phone to you = 2 phones sold.
From the company's perspective they're the same. The only thing that's changed is who pays for the phone(s). But now look at the kill switch case:
- Thief steals phone, owner bricks it, thief throws it away. Owner has to buy a new phone, thief has to buy a new phone. Net result = companies sell 1 phone to the theif, 2 phones to you = 3 phones sold.
I dunno. I agree it reduces entropy by eliminating adjacent pairs, but the frequency of sequential 16+ bit random numbers being identical is 1/65536. 0.0015% or less. You're losing just a tiny bit of entropy.
OTOH if the RNG breaks for whatever reason and keeps returning the same value, then throwing away identical sequential results would prevent the broken values from passing into the algorithm. It sounds to me more like this is a safeguard against the RNG crapping out, or attack vectors where the RNG is replaced by a call which returns the same value all the time. I mean if you compromised the system's RNG, the encryption software would still function without complaint and pass any binary md5 checksum. But by feeding it a known value instead of a random number, the encryption would be compromised. This requirement makes the software complain (by hanging in a loop) if it's being passed bad random numbers.
Yes, I know it seems like mSATA SSDs just showed up yesterday. They were a shortcut using the existing PCIe mini format (used primarily for wifi cards), but connecting it to a SATA controller instead of / in addition to the PCIe bus.
It's being replaced by the M.2 form factor, which supports multi-lane PCIe, SATA, SATA Express, as well as USB 3.0. It also gives manufacturers the choice of multiple standard lengths. A few of the M.2 SSDs already out operate in PCIe mode instead of SATA, and can top 1 GB/s. The quick and dirty way to tell the SATA vs PCIe M.2 drives apart is that the SATA ones have two notches (three rows of connectors). The PCIe ones have one notch (two rows of connectors).
M.2 is still in its early stages, and some systems with M.2 connectors can take the SATA-type SSDs but not the PCIe-type, or vice versa. But the drives and systems are out there. The PCIe SSD in the Macbook Pro is actually a modified (proprietary) version of Samsung's M.2 SSD (XP941).
The issue right now isn't what we'll be able to print. It's what we won't be able to print. Under current IP laws, the designs you'd want to print on a 3D printer will probably fall under copyright, which means they'll be locked up under license for 100+ years. We'll have the ability to vastly improve lives everywhere by printing stuff cheaply; but we won't have the right to do it until it's 100+ years old (and likely obsolete).
This is why the current debate over copyrights is so important. Making copyright laws with only songs and movies (optional entertainment) in mind is horribly short-sighted. What we decide upon for these laws will affect basic products needed for everyday life in the near future. We'll have the capability to print a keyboard for almost no cost, but it'll still cost $20 to print it because the people who hold the copyright on the designs will insist on being paid as much for rights to use the designs as for a manufactured keyboard.
I suspect either IP law will finally change in the face of this new reality, or pretty much all of the world's population will become criminals because they'll ignore the IP laws and just print designs which are useful regardless of who holds the copyright.
Viking 1 (first lander on Mars) was killed by a software update.
The lander operated for 2245 sols (about 2306 Earth days or 6 years) until November 11, 1982 (sol 2600), when a faulty command sent by ground control resulted in loss of contact. The command was intended to uplink new battery charging software to improve the lander's deteriorating battery capacity, but it inadvertently overwrote data used by the antenna pointing software.
Sears did that in the 1980s. They called it Discover Card. When it got big enough, they found it was more lucrative to partake in the fat profit margins Visa and Mastercard had colluded to create, than it was to try to disrupt the market and come to dominate the credit card industry. Kinda like DeBeers no longer has a diamond monopoly, but the high prices they've conditioned people to accept are so tempting that other diamond produces just follow them and vastly overcharge.
I'm very pro-free market, but this is one of those places where some massive government intervention is needed to undo the damage from decades of market manipulation. Pricing gradient works as a market force when you decrease profit per item by cutting prices, but the lower price results in increased market share which causes a net increase in profit - you make up in volume what you lose in profit margin per item. But if you've managed to fix prices with massive profit margins, then a competitor who tries to gain market share by charging lower prices ends up losing money compared to if they just play along with your high prices. It breaks this crucial free-market mechanism.
It's actually not a problem if you force users to change their password upon their first login. It's stupid, but it's not a problem. The worst that can happen is that someone can hijack an account/username that's never been used before.
Mainly, how prevalent is cell phone use when driving?
Just to make up some numbers to illustrate the point, say 50% of the time drivers were on their cell phones. If cell phones were linked with just 25% of accidents, then that would actually mean cell phones made driving safer. The 50% of cars where the driver used a cell phone accounted for 25% of accidents; the 50% of cars where the driver didn't use a cell phone accounted for 75% of accidents.
I'm pretty sure cell phone use does increase the accident rate. But to neutrally evaluate their effect, you have to compare on the basis of frequency of involvement in accidents / frequency of use in all cases. If you're only given the former, it could mean cell phones are bad for driving or cell phones are good for driving.
The equations for energy of rotational motion are analogous to linear motion. Whereas kinetic energy for linear motion is 0.5mv^2, for rotational energy it's 0.5mw^2 where w is angular velocity. So most of the energy storage actually comes from the high RPM, not the mass. Doubling the mass only doubles the stored energy, whereas doubling the RPM quadruples the stored energy.
The sensible way to design this is to estimate the greatest amount of energy it's going to store, take the max RPM the flywheel can withstand without flying apart, and select a mass a little larger than is needed to store that amount of energy.
#3 can be handled by containing the flywheel and designing it to disintegrate if it breaks loose from its bearing or hits the containment. Its stored energy then goes into destroying itself, rather than into whatever it hits.
#4 can be handled with two counter-rotating flywheels.
How has performance been? I ran zfsonlinux for a few months 2-3 years ago, and on a single redundancy zvol with compression on I could never get sustained data transfer rates over the network (using samba) above about 35 MB/s. I switched to ZFS on FreeBSD and it easily hits 60-100 MB/s. (This is with deduplication off. While dedup is a great idea in theory, I have never seen such a performance hog. It dropped my transfer rates to about 15, 8, and 2 MB/s for best, average, and worst case. And yes before I gave up, I tried giving it 8 GB of RAM for my ~1 TB storage pool devoted to stuff that could benefit from dedup. Using a better compression algorithm has worked better that dedup for me.)
I'm curious how well the project has progressed. I thought of trying btrfs but the "experimental" warnings scared me off - I just wanted something that works. The weird thing was zfsonlinux was faster at reads that writes. ZFS on FreeBSD has been faster at writes than reads.
"Elegant" really depends on what is most important for the task at hand. If you have gobs of CPU time and the code will need to be modified frequently in the future, then yes readability is probably most important. If you're processor-limited (e.g. 3D games), elegant = fast. The best example is probably the fast inverse square root.
Yes, I've heard the 6 kWh figure too. Assuming it's true, I suspect it's the cost to refine a volume of crude oil which yields a gallon of gasoline. So the 6 kWh would actually need to be amortized over the other petroleum products too, not just the gasoline. The EIA says a barrel (42 gallons) of crude oil yields about 19 gallons of gasoline. So if I'm right, only 2.7 kWh is attributable to the gasoline. (This isn't strictly correct because I believe 42 gallons of crude oil yields more than 42 gallons of product - such are the pitfalls of working in volume instead of mass.)
The 300 Wh is also the electrical energy stored in the battery (the Tesla S has an 85 kWh battery rated at 300 miles, so that works out to 283 Wh/mile). If you're going to factor in production costs of gasoline, you also need to factor in production costs of electricity. Charging the battery is about 75% efficient. Transmission to the home is about 98% efficient. And coal plants are about 45% efficient. So to produce the 300 Wh/mile the EV uses, the power company actually has to burn 300/(.75*.98*.45) = 907 Wh/mile. Factor in coal mining and transport costs and you're probably up around 1 kW/mile.
So the energy cost to refine gasoline is probably more likely enough to drive the EV only 2-3 miles.
The two biggest forms of biomass are wood, and methane reclaimed from landfills. They are considered renewable because it's "new" carbon and so is considered carbon-neutral. As opposed to the "old" carbon locked up in fossil fuels which is released when burned. This ignores the pollutants it releases into the atmosphere other than CO2.
Of the electricity generated in the U.S., 7% comes from hydro, 3.5% from wind, 1.4% from biomass, 0.4% from geothermal, and 0.1% from solar. So it's actually the third-biggest form of renewable electricity in the U.S. For overall energy used (i.e. including wood burned for heating), it is the biggest form of renewable energy, accounting for almost half of renewable energy produced in the U.S.
So take the stats saying "we're getting x% of our energy from renewable sources" with a grain of salt. It may not be as clean as you think it is.
Unless you build more dams, there is only a fixed amount of hydroelectric power available. Its total energy generation capacity is limited by the amount of water behind the dam (which in turn is limited by the amount of rain that falls). In a nutshell, it does not scale with use. Over a year of rainfall, it can only generate x GWh. No more.
Since hydro is the cheapest energy source, every attempt is made to use as much of it as we can every year. Water spilling over the top of the dam is wasted energy.
So when you buy a new EV and charge it with electricity from a hydro area, that reduces the amount of hydro electricity available for others to use. If everyone else in the country uses exactly as much energy as before, you now have a shortfall exactly equal to the amount of electricity the EV used. To fulfill that shortfall, a coal plant somewhere has to burn a little more coal, or a gas plant has to burn a little more gas.
Ideally a nuclear plant would burn a little more uranium fuel, but those are already pretty much run at full capacity so act like hydro does in the energy accounting books. Shortfalls are made up primarily by gas and oil-burning plants. Those are expensive so the power companies don't like to run them. But they're the only ones flexible enough to cover for shortfalls.
So yes, plugging in a new EV to a hydroelectric area increases CO2 emissions. At least until new plants can be built.
Then you'd also have to include the energy costs for mining and transporting coal used to make electricity. That's why people don't include that part - it's kinda assumed to be similar for both processes and cancels out. (Though to be honest I don't think I've ever seen an energy cost-analysis of coal mining vs petroleum drilling and refining. Coal is substantially cheaper per Joule than gasoline so its mining costs may be a lot lower too.)
Most electric carts I've used have only one pedal. When you push it, the cart moves forward. When you let go, it brakes. It makes sense to design it this way when you have regenerative braking so both acceleration and deceleration are handled by the same mechanism. If there is a second pedal, it's just the parking brake. Which I suppose you could mash down on if you feel the regenerative braking isn't decelerating you quickly enough.
Mathematically, it's the integral of a regular car pedal. The distance the pedal is pressed correlates to velocity (with allowances for lag to speed up/slow down the cart's mass), rather than acceleration (or more precisely, acceleration minus air resistance and energy transfers due to altitude changes when going up/down hills). Which is why you can control both acceleration and deceleration with a single pedal. The difference is similar to a car steering wheel (angle of turn correlates to angular velocity, or rate of change in orientation) vs an airplane's control yoke (angle it's pushed to the side correlates to rate of change in angular velocity, or angular acceleration).
Slashdot Mirror
The only reason you have to move your head around with a side mirror is because they are placed very close to where your peripheral vision ends. Consequently, if you're sitting facing forward, for a mirror to show the region between where the rearview mirror's view ends (almost straight back) to where your peripheral vision picks up (almost straight to the side), it has to have a very large field of view.
With a camera, you have the option of mounting it at the front corners of the car instead of by the driver, The display can still be by the driver, but the camera can be way in front. It can then show the same area using a much smaller field of view. The blind spot will still be there, but it'll be pushed out to 2-3 lanes away, making it irrelevant.
No hypocrisy or double standards. The U.S. didn't benefit from the Iraq vote (most of the oil contracts went to non-US companies). Hence there was no conflict of interest. I think the U.S. was wrong to invade without UN approval, but the U.S. lost lives, lost equipment, lost money, lost international respect, and suffered degraded ability to react to matters more pressing to its self interests (Taliban in Afghanistan and Pakistan). The only thing they gained was eliminating a dictator from the world stage (one they helped put there in the first place, but that's another story).
Russia did benefit from the Crimea vote - they annexed a huge amount of territory. Hence there is a huge conflict of interest, and why people are refusing to recognize the vote. If Russian had simply stood by the sidelines, and the people of Crimea had revolted on their own and held their own elections demanding secession from Ukraine, then there might be some international support for what happened. Heck, people might even have supported Russia's invasion. But the way Russian played it out, the results are indistinguishable from if they invaded and held a rigged election.
As the saying goes, to have a crime there has to be a motive, means, and opportunity. While the U.S. had the means and opportunity in Iraq, there was no motive - else we'd still be there. While the means Russia used may have been more benign*, the fact that they also had a motive and opportunity raises a lot of suspicion.
* The fact that no lives were lost in Crimea is I think more attributable to Ukraine deciding not to elevate the situation into a war they knew they had no chance of winning militarily. Not due to Russian benevolence. Had Ukraine fought back as Saddam Hussein did, do you really think no lives would have been lost?
Computer storage used to be the size of a washing machine, then a desktop PC, then a brick, then half a brick, then a quarter of a brick, then a deck of playing cards, then a stack of credit cards, then half a stack of credit cards, then a postage stamp, and now a microSD card is about the size of my little fingernail. About the time it went to postage stamp size, people started complaining it was too small and too easy to lose.
The same thing is happening to personal computers. That used to be a desktop, then a laptop, now it's becoming a tablet (or a laptop nearly as small as a tablet), soon it'll be your phone. Looking beyond that, it'll be the size of a credit card, then a postage stamp, and eventually the size of your fingernail. Somewhere between those last three stages, your PC will be so small that carrying it around in your pocket will be too risky because you might lose it.
Wearable computing is the obvious solution. First a device like a wristwatch. Then probably a ring or integrated into your glasses. If you're content living in the present, then good for you. But stop trying to drag down those of us trying to bring the future here more quickly.
The Vermont Yankee plant has been in operation since 1972, and generates about 4700 GWh each year. So over its lifetime it has probably generated about 190,000 GWh of electricity. $1 billion in decommissioning costs works out to about 0.5 cents per kWh (vs an industrial/commercial/residential cost of about 10/14/16 cents per kWh in Vermont). It does not substantially change the lifetime cost of the electricity generated.
All the states I know of require nuclear plant operators to collect these decommissioning costs in a fund. That is, part of their revenue from electricity sales must be placed into a trust fund specifically for decommissioning the plant in the future. That the Yankee plant has insufficient funds despite 41 years of operation suggests accounting fraud, regulatory incompetence, or inflated decommissioning costs more than it does non-viability of nuclear power. The San Onofre nuclear plant operated a similar amount of time, had about 3.6x the generating capacity when it was shut down, and was also shut down prematurely. It has a $2.7 billion decommissioning fund which is expected to exceed the costs, so they are planning to refund the excess to customers.
Yeah, this reminds me of the 1980 Moscow Olympics boycott. While it may have had a symbolic meaning (it was in protest of the Soviet invasion of Afghanistan), it ended up hurting the athletes more than the Soviet Union it was intended to hurt. And it seemingly forgot that the 1984 Olympics were in Los Angeles, which of course the Soviets promptly boycotted.
A lot of the police/government are on the cartels' payroll.
That's the problem with endemic corruption. If you try to fix just one part of it, the other parts resist. You have to somehow simultaneously fix all of it.
But not for technical reasons. While the engineering would be difficult to do, it's possible.
The problem is legal. It stems back to a question a friend of mine asked when noise-cancelling headphones first appeared. "If a sound wave has energy, and an equivalent wave 180 degrees out of phase also has energy, and when you combine the two you get no sound, where does the energy go?" Obviously the energy for both waves goes to regions where the two waves don't cancel.
So if you protected an area with this device, after an earthquake you'd be sued into oblivion by neighbors claiming your device increased damage to their property.
It's almost like someone who manufactures the backup camera systems has friends in the NTSB.
Pinch to zoom in 1988. Apple's patented version is one of those ridiculous "on a" patents, where you take (steal) a pre-existing idea, and tack the words "on a [something]" to the end of it. In their case, "pinch to zoom on a capacitive screen." The USPTO overturned the patent last year.
From the company's perspective they're the same. The only thing that's changed is who pays for the phone(s). But now look at the kill switch case:
I dunno. I agree it reduces entropy by eliminating adjacent pairs, but the frequency of sequential 16+ bit random numbers being identical is 1/65536. 0.0015% or less. You're losing just a tiny bit of entropy.
OTOH if the RNG breaks for whatever reason and keeps returning the same value, then throwing away identical sequential results would prevent the broken values from passing into the algorithm. It sounds to me more like this is a safeguard against the RNG crapping out, or attack vectors where the RNG is replaced by a call which returns the same value all the time. I mean if you compromised the system's RNG, the encryption software would still function without complaint and pass any binary md5 checksum. But by feeding it a known value instead of a random number, the encryption would be compromised. This requirement makes the software complain (by hanging in a loop) if it's being passed bad random numbers.
Yes, I know it seems like mSATA SSDs just showed up yesterday. They were a shortcut using the existing PCIe mini format (used primarily for wifi cards), but connecting it to a SATA controller instead of / in addition to the PCIe bus.
It's being replaced by the M.2 form factor, which supports multi-lane PCIe, SATA, SATA Express, as well as USB 3.0. It also gives manufacturers the choice of multiple standard lengths. A few of the M.2 SSDs already out operate in PCIe mode instead of SATA, and can top 1 GB/s. The quick and dirty way to tell the SATA vs PCIe M.2 drives apart is that the SATA ones have two notches (three rows of connectors). The PCIe ones have one notch (two rows of connectors).
M.2 is still in its early stages, and some systems with M.2 connectors can take the SATA-type SSDs but not the PCIe-type, or vice versa. But the drives and systems are out there. The PCIe SSD in the Macbook Pro is actually a modified (proprietary) version of Samsung's M.2 SSD (XP941).
The issue right now isn't what we'll be able to print. It's what we won't be able to print. Under current IP laws, the designs you'd want to print on a 3D printer will probably fall under copyright, which means they'll be locked up under license for 100+ years. We'll have the ability to vastly improve lives everywhere by printing stuff cheaply; but we won't have the right to do it until it's 100+ years old (and likely obsolete).
This is why the current debate over copyrights is so important. Making copyright laws with only songs and movies (optional entertainment) in mind is horribly short-sighted. What we decide upon for these laws will affect basic products needed for everyday life in the near future. We'll have the capability to print a keyboard for almost no cost, but it'll still cost $20 to print it because the people who hold the copyright on the designs will insist on being paid as much for rights to use the designs as for a manufactured keyboard.
I suspect either IP law will finally change in the face of this new reality, or pretty much all of the world's population will become criminals because they'll ignore the IP laws and just print designs which are useful regardless of who holds the copyright.
Sears did that in the 1980s. They called it Discover Card. When it got big enough, they found it was more lucrative to partake in the fat profit margins Visa and Mastercard had colluded to create, than it was to try to disrupt the market and come to dominate the credit card industry. Kinda like DeBeers no longer has a diamond monopoly, but the high prices they've conditioned people to accept are so tempting that other diamond produces just follow them and vastly overcharge.
I'm very pro-free market, but this is one of those places where some massive government intervention is needed to undo the damage from decades of market manipulation. Pricing gradient works as a market force when you decrease profit per item by cutting prices, but the lower price results in increased market share which causes a net increase in profit - you make up in volume what you lose in profit margin per item. But if you've managed to fix prices with massive profit margins, then a competitor who tries to gain market share by charging lower prices ends up losing money compared to if they just play along with your high prices. It breaks this crucial free-market mechanism.
It's actually not a problem if you force users to change their password upon their first login. It's stupid, but it's not a problem. The worst that can happen is that someone can hijack an account/username that's never been used before.
Mainly, how prevalent is cell phone use when driving?
Just to make up some numbers to illustrate the point, say 50% of the time drivers were on their cell phones. If cell phones were linked with just 25% of accidents, then that would actually mean cell phones made driving safer. The 50% of cars where the driver used a cell phone accounted for 25% of accidents; the 50% of cars where the driver didn't use a cell phone accounted for 75% of accidents.
I'm pretty sure cell phone use does increase the accident rate. But to neutrally evaluate their effect, you have to compare on the basis of frequency of involvement in accidents / frequency of use in all cases. If you're only given the former, it could mean cell phones are bad for driving or cell phones are good for driving.
The equations for energy of rotational motion are analogous to linear motion. Whereas kinetic energy for linear motion is 0.5mv^2, for rotational energy it's 0.5mw^2 where w is angular velocity. So most of the energy storage actually comes from the high RPM, not the mass. Doubling the mass only doubles the stored energy, whereas doubling the RPM quadruples the stored energy.
The sensible way to design this is to estimate the greatest amount of energy it's going to store, take the max RPM the flywheel can withstand without flying apart, and select a mass a little larger than is needed to store that amount of energy.
#3 can be handled by containing the flywheel and designing it to disintegrate if it breaks loose from its bearing or hits the containment. Its stored energy then goes into destroying itself, rather than into whatever it hits.
#4 can be handled with two counter-rotating flywheels.
How has performance been? I ran zfsonlinux for a few months 2-3 years ago, and on a single redundancy zvol with compression on I could never get sustained data transfer rates over the network (using samba) above about 35 MB/s. I switched to ZFS on FreeBSD and it easily hits 60-100 MB/s. (This is with deduplication off. While dedup is a great idea in theory, I have never seen such a performance hog. It dropped my transfer rates to about 15, 8, and 2 MB/s for best, average, and worst case. And yes before I gave up, I tried giving it 8 GB of RAM for my ~1 TB storage pool devoted to stuff that could benefit from dedup. Using a better compression algorithm has worked better that dedup for me.)
I'm curious how well the project has progressed. I thought of trying btrfs but the "experimental" warnings scared me off - I just wanted something that works. The weird thing was zfsonlinux was faster at reads that writes. ZFS on FreeBSD has been faster at writes than reads.
"Elegant" really depends on what is most important for the task at hand. If you have gobs of CPU time and the code will need to be modified frequently in the future, then yes readability is probably most important. If you're processor-limited (e.g. 3D games), elegant = fast. The best example is probably the fast inverse square root.
Yes, I've heard the 6 kWh figure too. Assuming it's true, I suspect it's the cost to refine a volume of crude oil which yields a gallon of gasoline. So the 6 kWh would actually need to be amortized over the other petroleum products too, not just the gasoline. The EIA says a barrel (42 gallons) of crude oil yields about 19 gallons of gasoline. So if I'm right, only 2.7 kWh is attributable to the gasoline. (This isn't strictly correct because I believe 42 gallons of crude oil yields more than 42 gallons of product - such are the pitfalls of working in volume instead of mass.)
The 300 Wh is also the electrical energy stored in the battery (the Tesla S has an 85 kWh battery rated at 300 miles, so that works out to 283 Wh/mile). If you're going to factor in production costs of gasoline, you also need to factor in production costs of electricity. Charging the battery is about 75% efficient. Transmission to the home is about 98% efficient. And coal plants are about 45% efficient. So to produce the 300 Wh/mile the EV uses, the power company actually has to burn 300/(.75*.98*.45) = 907 Wh/mile. Factor in coal mining and transport costs and you're probably up around 1 kW/mile.
So the energy cost to refine gasoline is probably more likely enough to drive the EV only 2-3 miles.
The two biggest forms of biomass are wood, and methane reclaimed from landfills. They are considered renewable because it's "new" carbon and so is considered carbon-neutral. As opposed to the "old" carbon locked up in fossil fuels which is released when burned. This ignores the pollutants it releases into the atmosphere other than CO2.
Of the electricity generated in the U.S., 7% comes from hydro, 3.5% from wind, 1.4% from biomass, 0.4% from geothermal, and 0.1% from solar. So it's actually the third-biggest form of renewable electricity in the U.S. For overall energy used (i.e. including wood burned for heating), it is the biggest form of renewable energy, accounting for almost half of renewable energy produced in the U.S.
So take the stats saying "we're getting x% of our energy from renewable sources" with a grain of salt. It may not be as clean as you think it is.
Unless you build more dams, there is only a fixed amount of hydroelectric power available. Its total energy generation capacity is limited by the amount of water behind the dam (which in turn is limited by the amount of rain that falls). In a nutshell, it does not scale with use. Over a year of rainfall, it can only generate x GWh. No more.
Since hydro is the cheapest energy source, every attempt is made to use as much of it as we can every year. Water spilling over the top of the dam is wasted energy.
So when you buy a new EV and charge it with electricity from a hydro area, that reduces the amount of hydro electricity available for others to use. If everyone else in the country uses exactly as much energy as before, you now have a shortfall exactly equal to the amount of electricity the EV used. To fulfill that shortfall, a coal plant somewhere has to burn a little more coal, or a gas plant has to burn a little more gas.
Ideally a nuclear plant would burn a little more uranium fuel, but those are already pretty much run at full capacity so act like hydro does in the energy accounting books. Shortfalls are made up primarily by gas and oil-burning plants. Those are expensive so the power companies don't like to run them. But they're the only ones flexible enough to cover for shortfalls.
So yes, plugging in a new EV to a hydroelectric area increases CO2 emissions. At least until new plants can be built.
Then you'd also have to include the energy costs for mining and transporting coal used to make electricity. That's why people don't include that part - it's kinda assumed to be similar for both processes and cancels out. (Though to be honest I don't think I've ever seen an energy cost-analysis of coal mining vs petroleum drilling and refining. Coal is substantially cheaper per Joule than gasoline so its mining costs may be a lot lower too.)
Most electric carts I've used have only one pedal. When you push it, the cart moves forward. When you let go, it brakes. It makes sense to design it this way when you have regenerative braking so both acceleration and deceleration are handled by the same mechanism. If there is a second pedal, it's just the parking brake. Which I suppose you could mash down on if you feel the regenerative braking isn't decelerating you quickly enough.
Mathematically, it's the integral of a regular car pedal. The distance the pedal is pressed correlates to velocity (with allowances for lag to speed up/slow down the cart's mass), rather than acceleration (or more precisely, acceleration minus air resistance and energy transfers due to altitude changes when going up/down hills). Which is why you can control both acceleration and deceleration with a single pedal. The difference is similar to a car steering wheel (angle of turn correlates to angular velocity, or rate of change in orientation) vs an airplane's control yoke (angle it's pushed to the side correlates to rate of change in angular velocity, or angular acceleration).