Yes it attenuates the signal. No it's not necessarily bad. The open frequencies were made open specifically because there was high signal attenuation through the air at those frequencies. 2.4 GHz and 5 GHz are absorbed by water molecules (so much at 2.4 GHz that microwave ovens operate at 2.45 GHz). The new 60 GHz band is absorbed by atmospheric oxygen. This makes the frequencies relatively useless for long-distance radio communications, but perfect if you going to have multiple short-range communication hotspots because each hotspot will interfere with neighboring hotspots less. The greater signal attenuation actually becomes an advantage in reducing the noise the devices will see from neighboring hotspots using the same frequency.
The fly in the ointment is that the trend is towards directional radio communications. Things like MIMO direct more radio energy in the direction of the intended recipient of the signal, rather than blasting it at equal strength in 360 degrees. So as this trend progresses, it's going to become harder and harder for a device like this to absorb RF power from the air - it's less likely to intersect with the "meaty part" of a directional radio broadcast.
H-1Bs were created in response to a trend that research turned up. More and more U.S. college and university graduates were accepting jobs overseas, resulting in a net drain of skilled graduates out of the country. The idea behind the H-1B was to make it so that a well-educated foreigner could get a job in the U.S. more easily, countering that trend. Many other countries run a similar visa program for skilled workers. It also gave another option for foreign students who recently graduated from college in the U.S. to get a job here after their student visa expired. So more of those foreign students could stay in the U.S. after graduation instead of returning home, again countering the trend. The long-term idea being that the H-1B would be a first step towards U.S. citizenship (meaning the skilled worker stays here permanently).
Unfortunately the program got exploited by companies trying to (ab)use it to hire cheap foreign workers to replace Americans. Those job listings you've seen with a ridiculously specific list of required qualifications are mostly H-1B visa jobs. The listing was carefully crafted to exclude anyone from qualifying for the job except the person they wanted to get the H-1B visa for. Companies are required to advertise those jobs for a certain length of time to prove that no American is capable of doing the job. Adding skills or certifications which aren't really necessary for the job but possessed by the foreigner they have in mind for the visa is one of the tricks to pass the advertisement requirement without "finding" any qualified Americans.
Favoring graduate degree holders to receive H-1Bs is a step in the right direction. There are a lot fewer of them than graduates with a bachelors degree. And their field of research tends to be a lot more specialized and thus legitimately harder to find a qualifying American.
One is that employees should be free to use whatever devices they want, even if it's contrary to what their employer makes or sells.
The other is that all employees must conform to the wishes and desires of their employer in any and every way, including their "choice" of what devices they can use at work and in their personal lives.
I support the former. I believe the strength of democracy derives from its diversity of opinion, allowing ideas to compete and the better ideas to percolate up to the top. As opposed to dictatorships where someone else decides what can do and think. What about you?
For those of you too young to remember, the Soviet Union had elections too. And the predesignated candidate always won with 99% of the vote (about 1% of people weren't able to make it to the polls). They liked to brag that it was a sign of how unified the population was in supporting the government, while democratic candidates rarely got more than 60% of the popular vote. A friend of mine who had managed to flee the U.S.S.R. told us that if you failed to vote for the designated candidate and didn't have a good excuse, you could be thrown into a gulag.
Ultraviolet reinforced the concept that if you bought a license for a show or movie, that license was universal and entitled you to stream it from anywhere So if you bought a movie on Fandango, you could use that license to view it from Vudu, and vice versa. No more having to dig through a half dozen streaming services trying to figure out which one you used to buy a particular movie. If you paid for the license, you could stream it from any of them (who supported Ultraviolet).
IMHO, a service like this should be a legal requirement for anything that's sold as a license instead of a physical product. It reinforces the concept that the digital licenses you buy belong to you, not to the service which happened to sell it to you.
AFAIK, that exemption is only for non-profit educational use in a face-to-face classroom setting. Being non-profit is not carte blanche to distribute whatever copyrighted works you want to anyone you want. It still has to fall under fair use. A judge may be more likely to determine distribution to be fair use if it's done by a non-profit rather than a for-profit entity. But other guidelines still apply - the amount of the work reproduced, the type of work being distributed, and the effect of the distribution on the copyright holder's ability to profit from it. All of these (with the exception of documentary PBS programs) would still favor a ruling of not fair use.
Copyright law doesn't say commercial distribution is prohibited while free distribution is allowed. It says the copyright holder has complete authority over distribution of their work. In fact that's why Aereo lost. Aereo wasn't actually charging for the broadcast TV content. They were charging you to rent an antenna from them (they went so far as to give each user their own individual antenna with their own encoder to generate their own individual stream, instead of using the signal from a single antenna to encode a single video stream broadcast to all their users). They were technically providing the TV broadcasts for free; you were only paying for equipment rental.
The Supreme Court ruled against them because the copyright holder has ultimate say over how their content is distributed, paid or free.
Civic emergencies (tornado, fire, etc.) aren't a problem. The school will be notified anyway, and will take steps to keep the students safe.
For family emergencies, back in the days before cell phones the family would just call the school's office. The office would then send someone to the classroom to collect the student (cue movie scene of student being told his parents were just killed in a car crash) and bring them to the front office to wait for another family member to come pick them up.
There's no need for students in school to have a cell phone for emergencies. They're in a known location in a known classroom at a known time. Totally different from, say, wandering around in some random location at Disneyland.
We already do this to collect the closing costs of a nuclear plant. For every dollar a customer pays for electricity generated with nuclear power, a few cents go into building up a fund to pay for the cleanup of any accidents. Japan's nuclear plants have produced roughly 200 TWh per year for the last 30 years, or 6000 TWh. The Fukushima cleanup cost is currently estimated at $180 billion. So its cost relative to the amount of power generated is ($180 billion) / (6000 TWh) = $30 million / TWh = $30 / MWh = 3 cents / kWh.
So a surcharge of just 3 cents/kWh on all electricity generated by nuclear power would have paid for the Fukushima cleanup costs. As there have only been two major nuclear accidents, 3 cents/kWh is probably towards the high end. But it's small enough you could just go with it and collect that into a disaster fund. (The third-biggest accident - 3 Mile Island - had a $1 billion cleanup cost. If you amortize that over all nuclear power production in the U.S., it works out to just 0.006 cents/kWh. A Fukushima-sized cleanup here would work out to a 1.1 cent/kWh surcharge.)
And to address AC's comment, Insurance doesn't work because only a small number of nuclear plants are necessary to power the world. The U.S. has about 100 nuclear plants, which generate 20% of all our electricity. About 450 nuclear plants throughout the world provides 10% of the world's electricity. For insurance to work, insurers have to be able to reliably predict what the rate of payout will be year-to-year. This requires a huge number of individual insurance policies.
The greater your sample size (the more individual insurance policies there are), the tighter the probability distribution gets. That's what turns unexpected costs of accidents and disasters into predictable costs. To get a distribution tight enough for insurance to be reliably predictable requires at least ~10,000 individual insured. Fewer than that and it becomes dfficult to make business decisions with a high degree of certainty. (i.e. their profit margin fluctuates by several percent each year based on random chance, swamping out any effects of their actual business decisions, making it difficult for them to determine if a good year was due to good decisions or good luck, or a bad year was due to bad decisions or bad luck.)
This is why insurance on nuclear plants is astronomical. The insurers can't sell enough policies to make the risk predictable. So they end up having to charge a premium several hundred or several thousand times the expected payout to minimize their risk exposure.
Wikipedia has been around for 18 years. 18 years / 3 million edits = 3 min 9 sec between edits. If you figure he has an 40 hour/week job (since Wikipedia doesn't pay him) and sleeps / showers / eats 8 hours a day, that works out to 83 seconds between edits if he did nothing but edit Wikipedia during his free time for 18 years.
Because IE came with Windows, a lot of companies only bothered to test things like device interfaces with IE. Some of my older security cameras have this problem. I can receive their video and change some general settings using a NVR app just fine. But if I want to change some of the more obscure settings, it has to be done via a web interface on the camera. And that interface only works with an older version of IE. I have to fire up an un-updated copy of Windows XP in a virtual machine to change those settings.
The thing you have to watch out for in these PR pieces is that if someone at a company they hire spends half their time filling Apple's order, half their time filling orders for other customers, they'll still count that as one job. When in fact it's only half of a job. To appraise it correctly, you have to multiply the number of individual jobs by the percentage of those people's time which was spent producing stuff for the $60 billion (as opposed to time spent making stuff for other customers).
That will get you a revenue per employee statistic which you can then compare to other companies. If you take $133,000 RPE at face value, that's actually really low. About on par with the level most small businesses operate at, and indicative of very inefficient operation. Which is a pretty good sign that it's just wrong, and the true number of fractional jobs they add to the economy is much less.
And if you want to gripe about what percentage of that is going to the employee, you need to look at profit (net income) per employee. That is, how much more could these companies pay each employee if all profits went to them instead of shareholders. That's why revenue per employee isn't used that often - it mixes together COGS, insurance, equipment costs, taxes, etc with payroll and profit. Net income per employee is a much better metric for judging how much extra each employee could have been paid.
But the tech industry (along with banking, petroleum, and pharmaceuticals is skewed way towards the high end in terms of net income per employee. Most big companies operate at less than $10,000 net income per employee. That is, even if all profits were distributed to employees instead of to shareholders and owners, each employee would only get a few thousand dollars extra.
People willing and wanting to cover more than one side of an issue. It's ok for reporters to have an angle, if you have lots of different reporters so that pretty much all angles are covered. That is, fake news is a symptom, not a problem. The remedy isn't to quash fake news. The proper remedy is to educate the people so that they're better able to determine for themselves what news is fake and what isn't. Then all you have to do is allow all news reports to get through, and the people can decide for themselves what the truth is.
That's what worries me about the current fervor to stamp out "fake news." While I can appreciate the intent, you're basically taking the the power to judge what constitutes real or fake news away from the people, and handing it to a few people at a company or in the government in charge of deciding what's fake and what's real. Those of you who've never lived through armed repression (i.e. the vast majority of the U.S. population) don't truly understand how precious a free media is. The people in Gwangju burned down one of the TV stations because they weren't doing their job and reporting what was happening. The reporters had been intimidated by the government and feared for their lives if they spoke the truth.
Yes, having a commission or department which decides what constitutes fake news can help. But it can also be horribly abused, with no way for the population to tell that it's being abused to repress legitimate information. The system works best when such repression is made more difficult, not easier. You just have to take upon yourself the harder task of properly educating the population, instead of the quick and dirty solution of a committee to decide which "fake news" should be repressed.
It is almost as if market capitalism doesn't work as advertised unless there are more than a hundred vendors to choose from.
It's ironic that you would level that charge against an industry which was nearly crippled by overzealous regulation. The EU mandated all phones must adhere to the GSM standard. A bunch of regulators sat down, and stamped out what they thought would be the ideal way to use the technology, and mandated that everyone use the system they came up with. Most countries outside the EU followed suit. The U.S. did not, and allowed competing phone standards. One of these was CDMA.
GSM worked fine for voice communication, but it turned out to have a fatal flaw when cellular data service became more common. You see, GSM mandated the use of TDMA - each phone is assigned a timeslice and takes turns talking to the tower. That's fine for low-bandwidth applications like voice. But for high-bandwidth data, it results in bandwidth being wasted on phones which needed little or no data during their timeslice.
OTOH, CDMA allows all devices to talk to the tower simultaneously. The tower tells them apart via orthogonal codes (kinda like writing horizontally and vertically on the same sheet of paper - the letters are orthogonal enough for you to distinguish which are supposed to be read horizontally, which vertically). The transmissions of other phones then just raise the noise floor, which automatically divides all the available bandwidth equally between all phones which happen to be transmitting at that moment. If only one phone transmits, it gets all the bandwidth. If a dozen phones transmit simultaneously, they each get 1/12th the bandwidth.
CDMA worked so well for data that within a year, GSM threw in the towel, licensed CDMA, and included it in their UMTS spec for cellular data. That's right, CDMA won the GSM vs CDMA war. Your 3G GSM phone used wideband CDMA for data. That's why CDMA phones got 3G about a year before GSM phones - they just had to flip the existing CDMA radio from voice mode to data mode, while GSM users needed to get a new phone. That's why GSM phones could talk and use data at the same time - they had a TDMA radio for voice, and a W-CDMA radio for data, while CDMA phones only had a single CDMA radio.
If the U.S. regulators had followed the EU and required GSM, we probably would be stuck around 1 Mbps data speeds today. Newer technologies like LTE (which mostly uses OFDMA - like CDMA except using orthogonal frequencies instead of orthogonal codes) likely wouldn't exist or would still be in the research stages. CDMA was the test case which proved that this crazy "every phone transmits at the same time and we use orthogonality to tell them apart" scheme actually worked when scaled up to a nationwide cellular network. So we narrowly averted a regulatory disaster with cell phones. Capitalism saved the industry, and with only two competitors too.
Capitalism can fail. Regulation can fail. The key is not to become so ideologically beholden to one that you consider it to be infallible, and focus only on failure modes of the other while ignoring cases where your chosen ideology fails.
In fact, it's so cheap that recycling plastic is uneconomical,
It's not just a matter of cost. There are basically two types of plastics - thermoplasts and thermosets.
A thermoplast is kinda like metal. You can heat it to soften or melt it, and reform it. These are sort of recyclable.
A thermoset is like a hard-boiled egg. Once it's set in the final shape, it cannot be reformed. They cannot be recycled, except by being shredded and used as filler.
It's so bad that many recycling centers will simply landfill plastic bags and their contents when encountered in the recycling stream.
Throwing away used plastics in landfills is not as bad as it would seem. Since they were originally made from oil, all you're doing is returning the hydrocarbons back underground, sequestering the carbon so it can't enter the atmosphere as CO2. It's the bags that litter the streets and oceans and never make it into landfills which are the problem. This is really a case of the worst of us (people who don't bother throwing trash into trash cans) preventing the rest of us from having nice things.
The atmosphere absorbs enough of that so the amount hitting the ground is about 750-850 Watts/m^2. Figure about 800 W/m^2.
A 22% efficient panel can then pull in 176 W/m^2.
But that's peak - you'll only get that much at noon on summer solstice with clear skies. You have to factor in night, movement of the sun, weather, dust buildup, weather, etc (capacity factor). The average capacity factor for fixed mount PV solar in the U.S. is about 0.145, with the desert Southwest peaking at around 0.195. So that brings your average PV solar power generation down to just 25.5 W/m^2 (34 W/m^2 in the desert Southwest).
That means installation large enough to power the average home (10399 kWh/yr / 8766 hours/yr = 1186 Watts) is about 1866 W / 25.5 W/m^2 = 73 m^2 or a 12.8 kW system (55 m^2 in the desert Southwest, 9.7 kW). Although that's based on an average - you'll have a shortage of power in winter, so will probably need an even bigger PV installation to get you off the grid year-round.
That quantity of PV panels costs several tens of thousands of dollars. On the other hand, a reflector/concentrator that size can be made for a few tens of dollars. Though the added bulk will increase the mounting and maintenance costs. So you can see where the guy in TFA is going with this. (I left out cost of mounting and voltage regulation circuitry since I figured it'd be about the same for both cases.)
The other option is to just allow plants/algae to collect solar energy. They use it to convert sunlight, CO2, and H2O into sugar, starches (sugar molecules glued together), and wood (starch molecules glued together). You can then use those as fuel. The efficiency is much lower (around 1% or less). But plants grow and reproduce on their own, so manufacturing cost is zero (negative in cases like weeds where we're actively trying to prevent the plants from growing). You only have to pay for harvesting and processing. Human global energy consumption is about 170,000 TWh per year, which at 8766 hours/yr is 19.4 TW. The rate at which energy is stored chemically by photosynthesis worldwide is estimated to be several hundred TW.
I'll add that most of HP's desktops and servers are assembled in the U.S. They don't seem to be having any problems getting screws despite being the #2 computer maker by volume, so they probably sell 2-3 orders of magnitude more PCs than Mac Pros.
It included Appletalk, which let you connect computers into a network using a cheap cable. The speed wasn't that great (basically a fast serial port), but it worked for transferring data between Macs.
Sound ability was limited on all computers of the era. Digital signal processors were ungodly expensive (as much as a computer; didn't come down in price until the 1990s), so sound synthesis was unaffordable. Memory was ungodly expensive too, so you couldn't play back high quality sound samples either (the 128 kB of RAM would hold just 5 seconds of a modern 192 kbps MP3 file) . That left some rudimentary synthesis with a square wave generator, some pre-selected samples stored in ROM, and instructing the speaker to generate a single frequency tone for x milliseconds (what we now recognize as 8-bit video game music).
You gotta remember that 13" was considered a big screen for a computer back then. Most were 10" or 11" (the IBM PC had a rather generous 11.5" screen). So a 9" display wasn't that big a step down. Especially since the Mac was designed to be portable (one of the commercials had a guy putting it into a backpack). I never really understood what Jobs insisted on it being portable, but he did. At a 3:2 aspect ratio, that meant the screen was 7.5" x 5" - big enough to display a letter-sized sheet of paper with half inch margins (3/4" or 1" margins were the norm) without side-scrolling.
The other reason for using a fixed screen size is because Apple wanted to make it truly WYSIWYG - your friend's thesis when printed would not only look identical in form, but it would also be identical in size. If the screen was showing an 11 point font, it was exactly the same size as an 11 point font when printed. Regular monitors at the time didn't have a way for the monitor to communicate its physical size and supported resolution back to the video card. So sticking with a fixed monitor was pretty much the only way they could do it for the first iteration. This is why Macs became ubiquitous in the publishing industry.
The irony is that the conformist monoculture controlled by a single authority that the commercial was rallying against, describes iOS, with its walled garden and uniform device design, to a T.
In the study, published Thursday, Rekognition made no errors in recognizing the gender of lighter-skinned men. But it misclassified women as men 19 percent of the time, the researchers said
Obviously this technology needs to be banned from use until it misidentifies men as women as often as it misidentifies women as men. We can't allow anything that yields unequal results from ever being used.
This is what I've been trying to stress to people (without much success - the Dunning-Kreuger effect works both ways). The problem with Monsanto's RoundUp-Ready crops isn't that they're GMO. The problem is that they've legally shed themselves of liability for any problems it causes. Basically they sell the seed and collect the profit from all farmers who want to use it. But if the seed spreads to farmers who don't want it, they just throw their hands up and say, "Not our problem! The courts say you have to pay to clean it up yourself."
For new technologies to be fairly tried, the company introducing it has to reap the profit from selling the new technology, but also need to be liable for damages due to any problems the new technology causes. Separating the risk from the reward causes technologies to "succeed" regardless of any negative problems they cause. The problem shows up in other areas as well.
It's the problem we have with fossil fuel pollution, where the cost of the pollution is shifted from the person who burns the fuel (and thus benefits from its use), onto society overall. Thus artificially lowering its price to just acquisition and refinement. Keep the costs coupled, and fossil fuels become much less attractive
It's what's leading to all the data breaches at companies holding private user identity data. The company benefits from the said data (either by operating more efficiently, or selling it to marketers). But if they should happen to lose the data, the cost of cleaning up the mess falls upon individuals whose identity ends up stolen. If you make the cost of cleaning up identity theft fall on the companies which lost the data, suddenly they will become much more careful about keeping your private data secure.
It also applies to the MPAA and RIAA trying to shift the cost of enforcing their copyrights to ISPs and websites. The whole point of copyright is to create a net benefit to society. The short-term monopoly is a smaller price than the benefit of the works created. But if the cost of enforcing copyright exceeds the revenue you can generate from selling copyrighted materials, then copyright loses its purpose. The expense of enforcing copyright exceeds the benefit to society of copyright, making it a bad proposition overall. But the only way you can truly determine if that has happened is if the beneficiary of the copyright (the IP holder) is also fully responsible for the cost of enforcing that copyright. If they successfully outsource enforcement cost, then they can continue making a profit from copyright long after it's become a net drain on society.
The infrastructure and prices aren't set up for every customer to use 100% of their bandwidth all the time. If everyone did that,
The entire network would have to be redesigned, since cable bandwidth is shared between you and your neighbors. It relies on customers not using 100% of their bandwidth all the time in order to function properly.
Prices would be a lot higher. OC3 fiber optic lines have a data rate of 148 Mbps, and cost upwards of about $10,000/mo. That's a dedicated line so you can saturate it all month if you like. (148 Mbps) * (1 year) / (12 months/year) = about 49 TB/mo. $10,000/mo / 49 TB/mo = $204 per TB.
So if your cable company's average plan costs 1/4 that ($51/mo), they're gambling that the average household usage will be below 250 GB/mo. More than that and they lose money providing service. The 1 TB cap is just to insure that a small number of high-volume users do not blow the average up.
So paradoxically, the fact that more households are hitting 1 TB cap means the cap is even more necessary to maintain pricing. It's like an all-you-can-eat buffet with a 2 hour time limit. The buffet restaurant bases their price on how much the average customer will eat. If customers are eating more than before, then they need to either impose a shorter time limit, or raise prices to stay in business.
(The wild card being that most places in the U.S. have a cable monopoly. So we have no way of knowing if the prices are actually fair, or if these additional surcharges just become extra profit.)
Netflix recommends 5 Mbps for 1080p video streams. That's a max (low-action scenes will require less), but if you assume worst-case and 8 hours/day of continuous streaming, that works out to (5 Mbps) * (8 hours/day) * (30 days/mo) = 540 GB.
The households having problems with the 1 TB cap are typically larger - 5 or more people. Mom, dad, 3+ kids all streaming different shows. It's actually fairer for them to be paying more. If each household pays the same per month, then the homes which use only 100-200 GB/mo are actually subsidizing the homes which use 1+ TB/mo. They're paying the same amount per month, while only putting 1/5 to 1/10 the load on the cable company's resources. The only catch is that because so many places in the U.S. have a cable monopoly, you have no way of knowing if the cable company is charging a fair rate, or if these overage fees just become extra profit.
I'll assume you really understand that what you wrote is nonsense. An output of 1 GW is 1 GW, no matter what the source was.
Let's take the example of night since that's the easiest to grasp. If you have a PV solar installation on your house which churns out 10 kW during the day, during the night it will yield 0 kW. Averaged over 24 hours, the average power production is then only 5 kW.
If you factor in similar reductions in actual production due to clouds, angle of the sun being sub-optimal, dust build-up on the panels, degradation due to age, the panels being taken offline for maintenance, etc., over a year a typical 10 kW PV solar installation will produce as much power as a 1.45 kW PV solar installation with the sun always shining and directly overhead.
In other words, the Wattage rating of these power installations is not their actual average production. It's the maximum they can generate under optimal conditions. Nuclear plants operate at those optimal conditions most of the time, so over a year they produce about 90% of their rated max Wattage. PV solar panels rarely operates at those optimal conditions (basically only during noon in summer on cloudless days), and on average they only product about 14.5% of their rated max Wattage.
It's the same reason you can take a laptop with only 5 hours of battery life, and use it for 8 hours. The power consumption when in use remains the same, but if you suspend it when you're not using it, its power consumption drops to near zero during that time. And thus its average power consumption drops enough to allow the battery to get it through the 8 hour day.
I'll become a big fan of nuclear energy when the radioactive waste from nuclear plants is completely removed from the earth.
Of course, "the best minds in the US" (and elsewhere) have been working on this for many decades, and no solution has been found.
The nuclear waste problem is political, not technological. What we call nuclear "waste" actually still has about 90%-93% of the energy from the original uranium still in it. That's why it stays dangerously radioactive for tens of thousands of years. It's possible to run the waste through a breeder reactor, which uses it for fuel to generate power (called reprocessing). The "waste" from a breeder reactor is usable as fuel in a regular reactor. If you run the waste through this cycle, you can extract about 90% of the energy in the uranium. And the remaining 10% means the final waste will only be dangerously radioactive for a few hundred years.
So why don't we reprocess? It turns out one of the byproducts from a breeder reactor is weapons-grade plutonium. So there's enormous political pressure not to reprocess spent nuclear fuel. President Carter banned reprocessing the spent fuel from commercial reactors in the U.S. in the 1970s. Which is why we're stuck with "waste" which will be dangerous for tens of thousands of years.
However, consider that more and more countries are developing nuclear weapons. At some point in the future, so many countries will be nuclear-armed that it will be pointless trying to stop the proliferation of nuclear weapons. At that point, all the "nuclear waste" we buried or are holding in spent fuel tanks at nuclear plants suddenly becomes precious fuel containing 10x more energy that we extracted from the original uranium fuel.
Nuclear is the safest power source man has ever invented. Even with the disasters at Chernobyl and Fukushima, it has killed fewer people per TWh generated than any other power source.
What's going on is that people are really bad at appraising big but rare risks. Their mind focuses on the magnitude of the risk, exaggerating the larger risks. Simultaneously, their mind glosses over the lower frequency of the risk. Consequently, big, rare events like nuclear disasters get overemphasized in people's minds, while small, common events like maintenance workers falling from wind turbines get overlooked.
It's the same reason plane crashes are splashed over all the TV news, while car crashes rare make the news, even though going to a destination by car is 1-2 orders of magnitude more dangerous than going by plane. The magnitude of the carnage from a plane crash is greater and overwhelms our minds, while the much lower frequency of plane crashes is overlooked. Or on the flip side, it's why people spend money on lottery tickets even though on average they'll lose money. The magnitude of the payoff if you win overwhelms our mind, to where we completely ignore the infinitesimal odds of winning.
Slashdot Mirror
Yes it attenuates the signal. No it's not necessarily bad. The open frequencies were made open specifically because there was high signal attenuation through the air at those frequencies. 2.4 GHz and 5 GHz are absorbed by water molecules (so much at 2.4 GHz that microwave ovens operate at 2.45 GHz). The new 60 GHz band is absorbed by atmospheric oxygen. This makes the frequencies relatively useless for long-distance radio communications, but perfect if you going to have multiple short-range communication hotspots because each hotspot will interfere with neighboring hotspots less. The greater signal attenuation actually becomes an advantage in reducing the noise the devices will see from neighboring hotspots using the same frequency.
The fly in the ointment is that the trend is towards directional radio communications. Things like MIMO direct more radio energy in the direction of the intended recipient of the signal, rather than blasting it at equal strength in 360 degrees. So as this trend progresses, it's going to become harder and harder for a device like this to absorb RF power from the air - it's less likely to intersect with the "meaty part" of a directional radio broadcast.
H-1Bs were created in response to a trend that research turned up. More and more U.S. college and university graduates were accepting jobs overseas, resulting in a net drain of skilled graduates out of the country. The idea behind the H-1B was to make it so that a well-educated foreigner could get a job in the U.S. more easily, countering that trend. Many other countries run a similar visa program for skilled workers. It also gave another option for foreign students who recently graduated from college in the U.S. to get a job here after their student visa expired. So more of those foreign students could stay in the U.S. after graduation instead of returning home, again countering the trend. The long-term idea being that the H-1B would be a first step towards U.S. citizenship (meaning the skilled worker stays here permanently).
Unfortunately the program got exploited by companies trying to (ab)use it to hire cheap foreign workers to replace Americans. Those job listings you've seen with a ridiculously specific list of required qualifications are mostly H-1B visa jobs. The listing was carefully crafted to exclude anyone from qualifying for the job except the person they wanted to get the H-1B visa for. Companies are required to advertise those jobs for a certain length of time to prove that no American is capable of doing the job. Adding skills or certifications which aren't really necessary for the job but possessed by the foreigner they have in mind for the visa is one of the tricks to pass the advertisement requirement without "finding" any qualified Americans.
Favoring graduate degree holders to receive H-1Bs is a step in the right direction. There are a lot fewer of them than graduates with a bachelors degree. And their field of research tends to be a lot more specialized and thus legitimately harder to find a qualifying American.
I support the former. I believe the strength of democracy derives from its diversity of opinion, allowing ideas to compete and the better ideas to percolate up to the top. As opposed to dictatorships where someone else decides what can do and think. What about you?
For those of you too young to remember, the Soviet Union had elections too. And the predesignated candidate always won with 99% of the vote (about 1% of people weren't able to make it to the polls). They liked to brag that it was a sign of how unified the population was in supporting the government, while democratic candidates rarely got more than 60% of the popular vote. A friend of mine who had managed to flee the U.S.S.R. told us that if you failed to vote for the designated candidate and didn't have a good excuse, you could be thrown into a gulag.
Ultraviolet reinforced the concept that if you bought a license for a show or movie, that license was universal and entitled you to stream it from anywhere So if you bought a movie on Fandango, you could use that license to view it from Vudu, and vice versa. No more having to dig through a half dozen streaming services trying to figure out which one you used to buy a particular movie. If you paid for the license, you could stream it from any of them (who supported Ultraviolet).
IMHO, a service like this should be a legal requirement for anything that's sold as a license instead of a physical product. It reinforces the concept that the digital licenses you buy belong to you, not to the service which happened to sell it to you.
AFAIK, that exemption is only for non-profit educational use in a face-to-face classroom setting. Being non-profit is not carte blanche to distribute whatever copyrighted works you want to anyone you want. It still has to fall under fair use. A judge may be more likely to determine distribution to be fair use if it's done by a non-profit rather than a for-profit entity. But other guidelines still apply - the amount of the work reproduced, the type of work being distributed, and the effect of the distribution on the copyright holder's ability to profit from it. All of these (with the exception of documentary PBS programs) would still favor a ruling of not fair use.
Copyright law doesn't say commercial distribution is prohibited while free distribution is allowed. It says the copyright holder has complete authority over distribution of their work. In fact that's why Aereo lost. Aereo wasn't actually charging for the broadcast TV content. They were charging you to rent an antenna from them (they went so far as to give each user their own individual antenna with their own encoder to generate their own individual stream, instead of using the signal from a single antenna to encode a single video stream broadcast to all their users). They were technically providing the TV broadcasts for free; you were only paying for equipment rental.
The Supreme Court ruled against them because the copyright holder has ultimate say over how their content is distributed, paid or free.
Civic emergencies (tornado, fire, etc.) aren't a problem. The school will be notified anyway, and will take steps to keep the students safe.
For family emergencies, back in the days before cell phones the family would just call the school's office. The office would then send someone to the classroom to collect the student (cue movie scene of student being told his parents were just killed in a car crash) and bring them to the front office to wait for another family member to come pick them up.
There's no need for students in school to have a cell phone for emergencies. They're in a known location in a known classroom at a known time. Totally different from, say, wandering around in some random location at Disneyland.
We already do this to collect the closing costs of a nuclear plant. For every dollar a customer pays for electricity generated with nuclear power, a few cents go into building up a fund to pay for the cleanup of any accidents. Japan's nuclear plants have produced roughly 200 TWh per year for the last 30 years, or 6000 TWh. The Fukushima cleanup cost is currently estimated at $180 billion. So its cost relative to the amount of power generated is ($180 billion) / (6000 TWh) = $30 million / TWh = $30 / MWh = 3 cents / kWh.
So a surcharge of just 3 cents/kWh on all electricity generated by nuclear power would have paid for the Fukushima cleanup costs. As there have only been two major nuclear accidents, 3 cents/kWh is probably towards the high end. But it's small enough you could just go with it and collect that into a disaster fund. (The third-biggest accident - 3 Mile Island - had a $1 billion cleanup cost. If you amortize that over all nuclear power production in the U.S., it works out to just 0.006 cents/kWh. A Fukushima-sized cleanup here would work out to a 1.1 cent/kWh surcharge.)
And to address AC's comment, Insurance doesn't work because only a small number of nuclear plants are necessary to power the world. The U.S. has about 100 nuclear plants, which generate 20% of all our electricity. About 450 nuclear plants throughout the world provides 10% of the world's electricity. For insurance to work, insurers have to be able to reliably predict what the rate of payout will be year-to-year. This requires a huge number of individual insurance policies.
The greater your sample size (the more individual insurance policies there are), the tighter the probability distribution gets. That's what turns unexpected costs of accidents and disasters into predictable costs. To get a distribution tight enough for insurance to be reliably predictable requires at least ~10,000 individual insured. Fewer than that and it becomes dfficult to make business decisions with a high degree of certainty. (i.e. their profit margin fluctuates by several percent each year based on random chance, swamping out any effects of their actual business decisions, making it difficult for them to determine if a good year was due to good decisions or good luck, or a bad year was due to bad decisions or bad luck.)
This is why insurance on nuclear plants is astronomical. The insurers can't sell enough policies to make the risk predictable. So they end up having to charge a premium several hundred or several thousand times the expected payout to minimize their risk exposure.
Wikipedia has been around for 18 years. 18 years / 3 million edits = 3 min 9 sec between edits. If you figure he has an 40 hour/week job (since Wikipedia doesn't pay him) and sleeps / showers / eats 8 hours a day, that works out to 83 seconds between edits if he did nothing but edit Wikipedia during his free time for 18 years.
Because IE came with Windows, a lot of companies only bothered to test things like device interfaces with IE. Some of my older security cameras have this problem. I can receive their video and change some general settings using a NVR app just fine. But if I want to change some of the more obscure settings, it has to be done via a web interface on the camera. And that interface only works with an older version of IE. I have to fire up an un-updated copy of Windows XP in a virtual machine to change those settings.
The thing you have to watch out for in these PR pieces is that if someone at a company they hire spends half their time filling Apple's order, half their time filling orders for other customers, they'll still count that as one job. When in fact it's only half of a job. To appraise it correctly, you have to multiply the number of individual jobs by the percentage of those people's time which was spent producing stuff for the $60 billion (as opposed to time spent making stuff for other customers).
That will get you a revenue per employee statistic which you can then compare to other companies. If you take $133,000 RPE at face value, that's actually really low. About on par with the level most small businesses operate at, and indicative of very inefficient operation. Which is a pretty good sign that it's just wrong, and the true number of fractional jobs they add to the economy is much less.
And if you want to gripe about what percentage of that is going to the employee, you need to look at profit (net income) per employee. That is, how much more could these companies pay each employee if all profits went to them instead of shareholders. That's why revenue per employee isn't used that often - it mixes together COGS, insurance, equipment costs, taxes, etc with payroll and profit. Net income per employee is a much better metric for judging how much extra each employee could have been paid.
But the tech industry (along with banking, petroleum, and pharmaceuticals is skewed way towards the high end in terms of net income per employee. Most big companies operate at less than $10,000 net income per employee. That is, even if all profits were distributed to employees instead of to shareholders and owners, each employee would only get a few thousand dollars extra.
People willing and wanting to cover more than one side of an issue. It's ok for reporters to have an angle, if you have lots of different reporters so that pretty much all angles are covered. That is, fake news is a symptom, not a problem. The remedy isn't to quash fake news. The proper remedy is to educate the people so that they're better able to determine for themselves what news is fake and what isn't. Then all you have to do is allow all news reports to get through, and the people can decide for themselves what the truth is.
That's what worries me about the current fervor to stamp out "fake news." While I can appreciate the intent, you're basically taking the the power to judge what constitutes real or fake news away from the people, and handing it to a few people at a company or in the government in charge of deciding what's fake and what's real. Those of you who've never lived through armed repression (i.e. the vast majority of the U.S. population) don't truly understand how precious a free media is. The people in Gwangju burned down one of the TV stations because they weren't doing their job and reporting what was happening. The reporters had been intimidated by the government and feared for their lives if they spoke the truth.
Yes, having a commission or department which decides what constitutes fake news can help. But it can also be horribly abused, with no way for the population to tell that it's being abused to repress legitimate information. The system works best when such repression is made more difficult, not easier. You just have to take upon yourself the harder task of properly educating the population, instead of the quick and dirty solution of a committee to decide which "fake news" should be repressed.
It's ironic that you would level that charge against an industry which was nearly crippled by overzealous regulation. The EU mandated all phones must adhere to the GSM standard. A bunch of regulators sat down, and stamped out what they thought would be the ideal way to use the technology, and mandated that everyone use the system they came up with. Most countries outside the EU followed suit. The U.S. did not, and allowed competing phone standards. One of these was CDMA.
GSM worked fine for voice communication, but it turned out to have a fatal flaw when cellular data service became more common. You see, GSM mandated the use of TDMA - each phone is assigned a timeslice and takes turns talking to the tower. That's fine for low-bandwidth applications like voice. But for high-bandwidth data, it results in bandwidth being wasted on phones which needed little or no data during their timeslice.
OTOH, CDMA allows all devices to talk to the tower simultaneously. The tower tells them apart via orthogonal codes (kinda like writing horizontally and vertically on the same sheet of paper - the letters are orthogonal enough for you to distinguish which are supposed to be read horizontally, which vertically). The transmissions of other phones then just raise the noise floor, which automatically divides all the available bandwidth equally between all phones which happen to be transmitting at that moment. If only one phone transmits, it gets all the bandwidth. If a dozen phones transmit simultaneously, they each get 1/12th the bandwidth.
CDMA worked so well for data that within a year, GSM threw in the towel, licensed CDMA, and included it in their UMTS spec for cellular data. That's right, CDMA won the GSM vs CDMA war. Your 3G GSM phone used wideband CDMA for data. That's why CDMA phones got 3G about a year before GSM phones - they just had to flip the existing CDMA radio from voice mode to data mode, while GSM users needed to get a new phone. That's why GSM phones could talk and use data at the same time - they had a TDMA radio for voice, and a W-CDMA radio for data, while CDMA phones only had a single CDMA radio.
If the U.S. regulators had followed the EU and required GSM, we probably would be stuck around 1 Mbps data speeds today. Newer technologies like LTE (which mostly uses OFDMA - like CDMA except using orthogonal frequencies instead of orthogonal codes) likely wouldn't exist or would still be in the research stages. CDMA was the test case which proved that this crazy "every phone transmits at the same time and we use orthogonality to tell them apart" scheme actually worked when scaled up to a nationwide cellular network. So we narrowly averted a regulatory disaster with cell phones. Capitalism saved the industry, and with only two competitors too.
Capitalism can fail. Regulation can fail. The key is not to become so ideologically beholden to one that you consider it to be infallible, and focus only on failure modes of the other while ignoring cases where your chosen ideology fails.
It's not just a matter of cost. There are basically two types of plastics - thermoplasts and thermosets.
Throwing away used plastics in landfills is not as bad as it would seem. Since they were originally made from oil, all you're doing is returning the hydrocarbons back underground, sequestering the carbon so it can't enter the atmosphere as CO2. It's the bags that litter the streets and oceans and never make it into landfills which are the problem. This is really a case of the worst of us (people who don't bother throwing trash into trash cans) preventing the rest of us from having nice things.
That means installation large enough to power the average home (10399 kWh/yr / 8766 hours/yr = 1186 Watts) is about 1866 W / 25.5 W/m^2 = 73 m^2 or a 12.8 kW system (55 m^2 in the desert Southwest, 9.7 kW). Although that's based on an average - you'll have a shortage of power in winter, so will probably need an even bigger PV installation to get you off the grid year-round.
That quantity of PV panels costs several tens of thousands of dollars. On the other hand, a reflector/concentrator that size can be made for a few tens of dollars. Though the added bulk will increase the mounting and maintenance costs. So you can see where the guy in TFA is going with this. (I left out cost of mounting and voltage regulation circuitry since I figured it'd be about the same for both cases.)
The other option is to just allow plants/algae to collect solar energy. They use it to convert sunlight, CO2, and H2O into sugar, starches (sugar molecules glued together), and wood (starch molecules glued together). You can then use those as fuel. The efficiency is much lower (around 1% or less). But plants grow and reproduce on their own, so manufacturing cost is zero (negative in cases like weeds where we're actively trying to prevent the plants from growing). You only have to pay for harvesting and processing. Human global energy consumption is about 170,000 TWh per year, which at 8766 hours/yr is 19.4 TW. The rate at which energy is stored chemically by photosynthesis worldwide is estimated to be several hundred TW.
I'll add that most of HP's desktops and servers are assembled in the U.S. They don't seem to be having any problems getting screws despite being the #2 computer maker by volume, so they probably sell 2-3 orders of magnitude more PCs than Mac Pros.
It included Appletalk, which let you connect computers into a network using a cheap cable. The speed wasn't that great (basically a fast serial port), but it worked for transferring data between Macs.
Sound ability was limited on all computers of the era. Digital signal processors were ungodly expensive (as much as a computer; didn't come down in price until the 1990s), so sound synthesis was unaffordable. Memory was ungodly expensive too, so you couldn't play back high quality sound samples either (the 128 kB of RAM would hold just 5 seconds of a modern 192 kbps MP3 file) . That left some rudimentary synthesis with a square wave generator, some pre-selected samples stored in ROM, and instructing the speaker to generate a single frequency tone for x milliseconds (what we now recognize as 8-bit video game music).
You gotta remember that 13" was considered a big screen for a computer back then. Most were 10" or 11" (the IBM PC had a rather generous 11.5" screen). So a 9" display wasn't that big a step down. Especially since the Mac was designed to be portable (one of the commercials had a guy putting it into a backpack). I never really understood what Jobs insisted on it being portable, but he did. At a 3:2 aspect ratio, that meant the screen was 7.5" x 5" - big enough to display a letter-sized sheet of paper with half inch margins (3/4" or 1" margins were the norm) without side-scrolling.
The other reason for using a fixed screen size is because Apple wanted to make it truly WYSIWYG - your friend's thesis when printed would not only look identical in form, but it would also be identical in size. If the screen was showing an 11 point font, it was exactly the same size as an 11 point font when printed. Regular monitors at the time didn't have a way for the monitor to communicate its physical size and supported resolution back to the video card. So sticking with a fixed monitor was pretty much the only way they could do it for the first iteration. This is why Macs became ubiquitous in the publishing industry.
The irony is that the conformist monoculture controlled by a single authority that the commercial was rallying against, describes iOS, with its walled garden and uniform device design, to a T.
Obviously this technology needs to be banned from use until it misidentifies men as women as often as it misidentifies women as men. We can't allow anything that yields unequal results from ever being used.
For new technologies to be fairly tried, the company introducing it has to reap the profit from selling the new technology, but also need to be liable for damages due to any problems the new technology causes. Separating the risk from the reward causes technologies to "succeed" regardless of any negative problems they cause. The problem shows up in other areas as well.
So if your cable company's average plan costs 1/4 that ($51/mo), they're gambling that the average household usage will be below 250 GB/mo. More than that and they lose money providing service. The 1 TB cap is just to insure that a small number of high-volume users do not blow the average up.
So paradoxically, the fact that more households are hitting 1 TB cap means the cap is even more necessary to maintain pricing. It's like an all-you-can-eat buffet with a 2 hour time limit. The buffet restaurant bases their price on how much the average customer will eat. If customers are eating more than before, then they need to either impose a shorter time limit, or raise prices to stay in business.
(The wild card being that most places in the U.S. have a cable monopoly. So we have no way of knowing if the prices are actually fair, or if these additional surcharges just become extra profit.)
Netflix recommends 5 Mbps for 1080p video streams. That's a max (low-action scenes will require less), but if you assume worst-case and 8 hours/day of continuous streaming, that works out to (5 Mbps) * (8 hours/day) * (30 days/mo) = 540 GB.
The households having problems with the 1 TB cap are typically larger - 5 or more people. Mom, dad, 3+ kids all streaming different shows. It's actually fairer for them to be paying more. If each household pays the same per month, then the homes which use only 100-200 GB/mo are actually subsidizing the homes which use 1+ TB/mo. They're paying the same amount per month, while only putting 1/5 to 1/10 the load on the cable company's resources. The only catch is that because so many places in the U.S. have a cable monopoly, you have no way of knowing if the cable company is charging a fair rate, or if these overage fees just become extra profit.
Let's take the example of night since that's the easiest to grasp. If you have a PV solar installation on your house which churns out 10 kW during the day, during the night it will yield 0 kW. Averaged over 24 hours, the average power production is then only 5 kW.
If you factor in similar reductions in actual production due to clouds, angle of the sun being sub-optimal, dust build-up on the panels, degradation due to age, the panels being taken offline for maintenance, etc., over a year a typical 10 kW PV solar installation will produce as much power as a 1.45 kW PV solar installation with the sun always shining and directly overhead.
In other words, the Wattage rating of these power installations is not their actual average production. It's the maximum they can generate under optimal conditions. Nuclear plants operate at those optimal conditions most of the time, so over a year they produce about 90% of their rated max Wattage. PV solar panels rarely operates at those optimal conditions (basically only during noon in summer on cloudless days), and on average they only product about 14.5% of their rated max Wattage.
It's the same reason you can take a laptop with only 5 hours of battery life, and use it for 8 hours. The power consumption when in use remains the same, but if you suspend it when you're not using it, its power consumption drops to near zero during that time. And thus its average power consumption drops enough to allow the battery to get it through the 8 hour day.
The nuclear waste problem is political, not technological. What we call nuclear "waste" actually still has about 90%-93% of the energy from the original uranium still in it. That's why it stays dangerously radioactive for tens of thousands of years. It's possible to run the waste through a breeder reactor, which uses it for fuel to generate power (called reprocessing). The "waste" from a breeder reactor is usable as fuel in a regular reactor. If you run the waste through this cycle, you can extract about 90% of the energy in the uranium. And the remaining 10% means the final waste will only be dangerously radioactive for a few hundred years.
So why don't we reprocess? It turns out one of the byproducts from a breeder reactor is weapons-grade plutonium. So there's enormous political pressure not to reprocess spent nuclear fuel. President Carter banned reprocessing the spent fuel from commercial reactors in the U.S. in the 1970s. Which is why we're stuck with "waste" which will be dangerous for tens of thousands of years.
However, consider that more and more countries are developing nuclear weapons. At some point in the future, so many countries will be nuclear-armed that it will be pointless trying to stop the proliferation of nuclear weapons. At that point, all the "nuclear waste" we buried or are holding in spent fuel tanks at nuclear plants suddenly becomes precious fuel containing 10x more energy that we extracted from the original uranium fuel.
Nuclear is the safest power source man has ever invented. Even with the disasters at Chernobyl and Fukushima, it has killed fewer people per TWh generated than any other power source.
What's going on is that people are really bad at appraising big but rare risks. Their mind focuses on the magnitude of the risk, exaggerating the larger risks. Simultaneously, their mind glosses over the lower frequency of the risk. Consequently, big, rare events like nuclear disasters get overemphasized in people's minds, while small, common events like maintenance workers falling from wind turbines get overlooked.
It's the same reason plane crashes are splashed over all the TV news, while car crashes rare make the news, even though going to a destination by car is 1-2 orders of magnitude more dangerous than going by plane. The magnitude of the carnage from a plane crash is greater and overwhelms our minds, while the much lower frequency of plane crashes is overlooked. Or on the flip side, it's why people spend money on lottery tickets even though on average they'll lose money. The magnitude of the payoff if you win overwhelms our mind, to where we completely ignore the infinitesimal odds of winning.