"Retirement" has been a part of most cultures for over a thousand years. I wouldn't be surprised if it's been around for as long as humans have stuck together as a family. In most cultures, when your parents/grandparents get too old and frail to do physical labor, they move into your house and you take care of them until they die. (Though a few cultures took the opposite approach and simply killed the elderly.)
Social Security is the exact same thing, except it expands it from direct descendants taking care of their direct ancestors, to all descendants as a whole (everyone of working age in the country) taking care of all their ancestors as a whole. Young working people pay into Social Security, and the money is given to retired people as a monthly stipend. (No, the government does not put it into a savings account where it waits for you until you retire.)
I've been saying this for over a decade: Put a physical read-only switch on storage drives (and motherboard BIOSes). Then design OSes to boot off a read-only device, with things that need to be written (like logfiles) going to a different drive. Same for programs - the OS should only allow programs on the boot device to run. Double-clicking an executable on another drive should pop up an error (unless the read-only switch of the boot device is off).
Then, once you have the computer set up as you want it with the OS and and desired programs running, you can flip the switch and lock down the system. Anyone who uses the computer, whether remotely or locally cannot change the OS or programs without first physically opening it up to flip the switch. A hack might open up a crack to let a hacker's foot in the door, but they cannot then leverage it to root the entire system. If they got in via a memory overrun exploit, then all the modifications they try to make to the system have to be done through that memory overrun exploit. Malware might be able to take hold, but it cannot write itself to automatically start next time the computer reboots. Malware wouldn't be able to cause computers to fail to boot. In fact a reboot would clear out any such malware, though it might still be attached to a data file if a program is vulnerable to it when the data file is read. (Ransomware wouldn't change since it already leaves the OS and program files alone - it just wouldn't be able to set itself to load and run every time the computer boots - it would need to finish encrypting your data before you rebooted your computer.)
Yes it would make updates a pain. But the need for regular updates would be substantially diminished since it'd be much harder for malware to exploit a known vulnerability. You could make updates a once a month or once every few months thing, instead of needing daily updates like we do today. And the need to shutdown the computer before you opened it up to flip the read-only switch would clear out any malware laying in wait for update day. You'd just have to make sure the update was the first (and only) thing you ran when you turned the computer back on.
There is no justification for making these extensions retroactive. The purpose of copyright is to promote the creation of new works. You cannot encourage the creation of something which has already been created. So there is absolutely zero benefit to extending the copyright term of existing works. Since the only rationale given in the Constitution for copyright is the promotion of the creation of new works, extending the copyright term for existing works is unconstitutional.
If Congress wants to make copyright last 140 years, fine. But the lengthened term should only apply to works created after the term is changed. Those are the only works whose creation could have been encouraged by a lengthened copyright term, and thus fall under the justification outlined in the Constitution.
In gambling, the game is rigged by the house to statistically always be in their favor. The stock market is different because there is no rigging (outside of insider trading).
In cases where a company is obviously doing something right or wrong, this helps it rise or fall in the market, thus encouraging economic optimizations which increase our standard of living for the same amount of effort expended.
It only starts to look like gambling when all the obvious low-lying optimization fruit has already been picked. Then you start to get more into too-small-to-measure and luck-based optimizations, resulting in more random outcomes that look like gambling.
And pretty much all political proposals to penalize "bad" behavior is that they're not designed to respond to feedback. If you're going to do this, the correct way to do it is to designate some desired level of atmospheric carbon. The higher we go above that level, the greater the tax per ton of CO2 becomes. Likewise, the closer the actual CO2 level gets to the desired level, the smaller the tax becomes. (Conceivably it could go negative if it's deemed that too little atmospheric CO2 is also a danger.)
When you don't use a simple feedback mechanism like this to make incentives auto-adjusting, you get the screwed-up situation we currently have with gender "equality" in education. Girls are outperforming boys at every education level. We've reached the goal of gender equality in education, then blown right past it to where girls are now advantaged relative to boys. But the incentives (created when girls lagged behind boys) are still set to favor girls.
If someone gives me a tablet or a computer. How do I know that when I click my vote was marked correctly? Without me personally inspecting the code, then inspecting the hardware, I can never be sure.
The electronic voting machine i used in this past Tuesday's election used a monitor, scroll wheel, and buttons to let you select which candidates you wanted to vote for. After you finished voting, it showed you your votes on the screen and asked you to make sure those votes were correct. Then it printed your votes onto a piece of paper, and asked you to confirm that the printed ballot accurately reflected your choices. (Presumably if you clicked no, it would shred the printed ballot.)
That seemed to me a sensible method of electronic voting. Use a computer to reduce the expense and confusion of custom-printed ballots for every election. But retain the paper ballot for counting and auditing purposes.
The problem is Amtrak competes against a service which is heavily subsidized by the government - cars and trucks run on freeways constructed with tax dollars (fuel taxes only pay for maintenance). If the roads weren't so heavily subsidized, the U.S. would be more like Europe and would make greater use of rail transport for both cargo and people.*
So expecting Amtrak to be profitable on its own is unrealistic. You need to subsidize it to the same degree you're subsidizing freeways just to level the playing field. And when you subsidize a small service to that degree, politicians start to play around with how the money should be spent.
* The idea back when freeways were first made was that trucks could transport goods from endpoint to endpoint, eliminating the need for expensive labor-intensive loading and unloading stages, where people at the railyard have to move cargo from the train onto a truck to make it to its final destination, or vice versa. The labor of the loading/unloading stages was the predominant cost to cargo transport at the time, so eliminating it was an economically sound idea. But since then, fuel costs have increased substantially, and the advent of container transport has reduced loading/unloading costs. But we're still stuck with a cargo transport system built based on the old cost structure, which is artificially keeping trucks competitive with trains for long-distance transport.
The long-term suicide rate for the U.S. has fluctuated between about 10-14 per 100,000 over the last 40 years. 1999 was just the minimum, which the reporter cherry picked to try to turn this into a story. It's been higher in the past.
Also, suicide rates have been climbing up throughout the world, not just in the U.S. Curiously, homicide rates also stopped the long-term decline around 2010 and started creeping up again. So average global temperature may actually be linked (violence and higher temperatures have been correlated in the past). But nobody really knows for sure why, and anyone who claims they do is just trying to spin the story to fit their preconceptions.
Gasoline generates about 8.89 kg of CO2 per gallon. So producing 1 ton of CO2 would require burning (1000 kg)/(8.89 kg/gallon) = 112.5 gallons of gasoline. The current average price of gasoline is $2.934/gallon, so 112.5 gallons of gas would cost (112.5 gal)*($2.934/gal) = $330.
A $100-$200 surcharge per ton of CO2 would thus raise the price of gasoline by just 30%-61%.
Using the same EIA chart, coal generates roughly 2 tons of CO2 per ton of coal. One ton of coal contains roughly 24 Gigajoules of thermal energy, which is 6.67 MWh. If the coal plant is 40% efficient, that means that one ton of coal generates 2.67 MWh of electricity. Since that one ton of coal also emits 2 tons of CO2, we end up with (2 tons CO2) / (2.67 MWh) = 0.75 tons per MWh.
Natural gas generates roughly 53.12 kg of CO2 per thousand cubic feet. A thousand cubic feet of methane contains 1.037 million BTUs of thermal energy = 303.9 kWh. If the gas plant is 60% efficient, this means 53.12 kg of CO2 are emitted per 182.3 kWh, or (0.053 tons CO2) / (0.1823 MWh) = 0.29 tons per MWh.
Coal accounts for 30.1% of U.S. electricity. Natural gas accounts for 31.7%. So the fractional CO2 contribution of these fossil fuels to electricity is (0.75 tons/MWh)*(0.301)+(0.29 tons/MWh)*(0.317) = 0.318 tons of CO2 per MWh. A $100-$200 surcharge per ton of CO2 then ends up costing $31.80-$63.60 per MWh, or 3.2 cents - 6.4 cents per kWh.
Average electricity price in the U.S. is 12 cents/kWh. So a $100-$200 surcharge per ton of CO2 would raise the price of electricity by 27%-53%. Almost exactly the same percentage as gasoline.
Like I keep trying to explain to people: Electric vehicles aren't cheap to operate because they're more energy efficient. They use nearly as much energy as ICE vehicles. They're just cheaper to operate because the coal and natural gas used to generate electricity are roughly an order of magnitude cheaper per MJ than gasoline. If you want to reduce CO2 emissions, buying an EV presently doesn't help. When you replace an ICE vehicleswith an EV without changing the makeup of your electricity sources, all you've done is shift your CO2 emissions from the car's tailpipe to a fossil fuel power plant's smokestack. That's why the claim that EVs are "zero emissions" is BS at present. You need to replace fossil fuel power plants with nuclear and renewable plants to cause a reduction in CO2 emissions.
That is essentially the problem. The way things like this are supposed to happen is that a few local governments get an idea so they pass a law mandating net neutrality. After some time to gauge the results of the law, a few states take notice and say "this looks like a good idea" and pass their own laws. And after more time to gauge how this affects the states, the Federal government takes notice and says "maybe we should make this a national law." These incremental steps, gradually expanding the reach of a law, allow us to properly gauge the law's effect and make modifications to it if problems should arise, before it affects the entire nation.
Instead, Tom Wheeler short-circuited the entire process and unilaterally declared that net neutrality must be the law of the land. Net neutrality isn't the only possible solution to this problem. The problem is actually government-created - local governments granted service monopolies to cable companies. These companies, assured that their customers cannot flee to another ISP, then intentionally degrade online services like Netflix to extort payments from Netflix to restore service.
The way it should work is some areas try net neutrality, some areas try rescinding these government-granted monopolies and allow multiple ISPs to compete, some areas try some different solution that we haven't yet thought of. Let these different solutions play out for a few years. Then we can study actual empirical data, and decide what's best for the entire country. And only then do we pass a national law with a solution to the problem.
The knee-jerk reaction method used to get net neutrality implemented via the FCC is totally the wrong way for government to operate. Heavy-handed decisions like this without first exploring possible solutions is what nearly saddled us with GSM. The original GSM spec was built on TDMA - each phone takes turns talking to the tower. Europe mandated GSM, and most of the rest of the world followed. The U.S. refused to require it, which allowed a competing service based on CDMA to develop. When phones started being used more for data than talking, suddenly the achilles heel of TDMA reared up. TDMA requires each phone to get a full timeslice even if it has little or no data to transmit. This wastes a huge amount of bandwidth. CDMA on the other hand allows all phones to transmit at the same time (they see each others' transmissions as noise, thus reducing the signal-to-noise ratio), and bandwidth is automatically allocated in proportion to how much each phone is transmitting. No wasted bandwidth. Within a year GSM threw in the towel, licensed CDMA, and added wideband CDMA to the GSM spec for data services. If the U.S. had gone along with the "sensible" decision by bureaucrats to impose GSM, then CDMA wouldn't have happened, and our cellular data speeds today would probably down around 1 Mbps. And many of the services we enjoy on our phones today wouldn't yet exist.
Tablets more popular than desktop computers for browsing. The vast majority of desktop computers are used in businesses (where web browsing is usually prohibited) and by gamers. Most home users switched to laptops long ago, and most gamers also own a laptop for their non-gaming computer use.
And tablets are very popular. They just have a really long usable lifespan, which caused sales per year to plummet once the market was saturated. Mine is going on 4 years old and I have no intention of replacing it, whereas I get the itch to replace my laptop after 2 years and usually replace it by 3 years. I'd still be using my 8 year old tablet if the battery hadn't died.
The problem with using datasets over long times (like 19 years) is a myriad of other factors changing over time can influence the economic outcome. I suspect if you'd set the endpoint at the middle of the 2008/2009 recession, you would've concluded the minimum wage absolutely increases unemployment. To really do a valid study, you need to do an A/B comparison. Two different areas (preferably nearby or mixed with each other) with similar economies at the same time, but with different minimum wages.
Even then the results of all these studies tend to be mixed. I suspect what's going on is that a high minimum wage kills off low-income jobs (those where a worker would not be productive enough to pay for their own price of labor). So no neighborhood kid mowing your lawn, no people loading boxes from the warehouse onto the truck, fewer cleaning staff for the office, etc. But businesses don't give up and file for bankruptcy because of a small change like this. They adapt - they buy a forklift and hire a skilled forklift operator paid a lot more than the minimum wage, but who moves more boxes per hour per $ of wage than the workers he replaced. The company just hadn't wanted to pay the up-front money to buy a forklift, until the increased minimum wage tilted the scales against hiring multiple unskilled workers. Or they buy a robot vacuum which handles the nightly cleaning, so they only need the cleaning staff to come in once a week instead of every night. etc.
And so the net result of a too-high minimum wage isn't higher overall unemployment. It's a decrease in the number of jobs which require little or no skill or experience. Which is exactly the complaint you hear about from millenials (the neighborhood kid mowing lawns is out of luck - people will just do it themselves rather than overpay someone else to do it).
This is an impossible task. Copyright doesn't need to be registered. You simply have to create a work and it is automatically copyrighted (registering it just allows you to sue for greater damages if there's commercial infringement).
So in many cases, the only person on the planet who knows a work is copyrighted is the content creator. If someone else then steals his work and uploads it to YouTube, how the hell is YouTube supposed to know it's supposed to block it because it belongs to someone else?
Your clean floor analogy doesn't work because there's a distinct difference between a floor being clean or dirty. This is more like YouTube owning a fountain, and millions of people throw coins into that fountain each year. This court ruling is basically saying YouTube can't allow people to throw coins into the fountain unless YouTube first makes sure none of those coins were stolen. How is YouTube supposed to do that? In some cases, the *only* person on the planet who knows the coin was stolen is the thief - the original owner might not even realize yet that it's been stolen.
So a student is more likely to be killed by a deer than from a school shooting. Where are all the walk-outs and protests advocating deer population control?
For some perspective on the scope of the school shooting problem, look at the stats the CDC puts out. For 2015, the leading causes of death among the 15-19 year old demographic were:
3,919 deaths - Accidents (mostly automobile accidents and drug overdoses). 282x more than school shootings.
2.061 deaths - Suicide. 148x more.
1,587 deaths - Homicide (mostly outside school, and gang related). 114x more.
583 deaths - Malignant neoplasms (cancer). 42x more.
306 deaths - Heart disease. 22x more.
195 deaths - Birth defects. 14x more.
72 deaths - Influenza (the flu). 5.2x more.
63 deaths - Chronic lower respiratory diseases. 4.5x more.
61 deaths - Cerebrovascular diseases. 4.4x more.
52 deaths - Diabetes. 3.7x more.
41 deaths - Complications from pregnancy and childbirth. 3x more.
A protest over excessive rates of teen pregnancy could potentially save 3x more lives than a protest over school shootings. Likewise, teaching kids not to each too many sweets, to exercise, not to smoke, get the flu shot, use sunscreen, not to join gangs, to buckle their seat belt, not to use drugs, and offering them counseling for depression, would all be much more productive uses of our time and effort than worrying about or debating school shootings. For that matter, controlling deer populations to reduce the number of fatalities from striking deer could potentially save 1.35x as many students' lives as lost to school shootings.
If you want to tackle a life-threatening issue that students face, probably the best choice is suicide. It results in more than a hundred times as many student deaths as school shootings. But when's the last time you saw the media run a story about teen suicide? The only reason school shootings are even on the radar is because of the media using them to play the "think of the children!" card against guns.
One of the deadliest hurricanes on record was Hurricane Mitch, which reached category 5 at sea. But most of its devastation happened after dropped below hurricane status. It stalled over Honduras as a tropical storm and basically flooded the country back into the stone age (75 inches of rain over roughly 2 weeks). You can't evacuate when the storm spans all the way from the Atlantic to the Pacific coast.
The only reason it isn't the costliest Atlantic storm on record is because it hit relatively poor countries. Hurricane Harvey was a cake walk by comparison ("only" 40 inches of rain over 1 week). But Mitch was 20 years ago and doesn't quite fit the desired narrative of modern storms moving slower. So Harvey is mentioned instead.
The "core software" is Linux. Pretty much all routers (most standalone devices really) run Linux under the hood. There are a few oddball routers which use a RTOS like QNX or VxWorks (these are the ones you want to avoid if you're hoping to flash a third party firmware). But the vast majority run Linux because it's free. This malware probably wormed its way in via a universal Linux exploit which was patched in the production releases of Linux distros, but not in the much-slower-to-update router firmware releases.
Is this an official sign that MIcrosoft is finally listening to developers (albeit with a Slashdot-level of negative attention), or are they simply trying to appease the crowd while they are still in the news due to their acquisition of GitHub?
No, it's a sign that the GNOME brand is toxic. So toxic that even Microsoft doesn't want their products to be associated with it in any way. Even by mistake.
If you read Microsoft's news article on this, the point of this project isn't to cool a datacenter using the ocean as a heat sink. It's to build a self-contained datacenter which can be customized to order, is easily shipped (fits on the back of a truck), then deployed at the bottom of the ocean where it operates for years without human intervention. All you need to do is plug in the power and data cables to something onshore.
I suspect putting it at the bottom of the ocean is more about preventing people from breaking in and stealing the equipment inside, rather than cooling it. It's got a built-in self-destruct mechanism which triggers if anyone tries to break in while it's on the seabed, thus eliminating any incentive for thieves. This goes along with the self-contained theme (no need to hire security guards).
Microsoft realizes they're not beholden to Intel. Intel needs Microsoft more than Microsoft needs Intel. While it's been a profitable relationship for nearly 40 years, Intel is facing serious competition at the lower end from ARM ever since computers became "fast enough" that most people can get most of their computing needs done with a low-end computer. Microsoft is making sure they have a finger in every pie. If the Intel ship sinks, they don't want it to take Windows down with it. So they're doing what they can to make sure Windows and its API is hardware-agnostic and can run on both Intel/AMD and ARM.
Whether Windows 10 for the Snapdragon 850 sells well or not is immaterial to Microsoft. They are simply hedging their bets to insure their Windows cash cow survives regardless of whether the winner of the processor war ends up being Intel, AMD, or ARM.
Prior to Coffee Lake, the mobile i5 was just a mobile i3 with turbo boost. i.e. A faster clocked i3. That's it. In no way was the obscene markups for an i5 laptop over an i3 laptop justified. The fact that Macbooks were available with an i5 but not an i3 should've been a huge tip-off that there was an obscene profit margin for little performance gain there. Intel really milked that cash cow for roughly a decade. (It was probably driven by people wrongly assuming that what they knew about desktop processors also held for mobile processors. The desktop i5 was a quad core vs. the desktop i3 being a dual core, so was a worthy upgrade. But both the mobile i3 and i5 were dual cores.)
With Coffee Lake and the Kaby Lake refresh (i5-8xxx), most of the mobile i5s are now quad cores. So they're now a worthwhile upgrade over a mobile i3.
Try to stop these tracking methods. Which just results in the people doing the tracking coming up with new tracking methods. That kicks off an endless arms race where each side keeps countering the move the other side makes.
Pollute the data. Let them collect the data, but the browser should surreptitiously add fake data. Either generated by randomly crawling linked pages in the background, or by sharing anonymized sites other users have browsed. The moment the "user's" browsing data is no longer an accurate representation of the sites the user is actually browsing, that data loses its advertising value. And advertisers will be forced to place ads based on the type of people who like to visit a certain site (e.g. GPU ads on a gaming site), rather than trying to display ads targeted at the person browsing regardless of what site they visit.
The first method is a never-ending game of leapfrog. The second method favors users because there are a lot more of them than companies tracking this data. They can generate fake browsing data faster (up to the limit of their Internet bandwidth) than these companies can filter it out.
The vast majority of ratings are given by people even more clueless than the developers. It's basically the same problem as elections in a democracy - they've devolved into a popularity contest, where the winner is simply who can manipulate their public image via the media and advertisements, to most appeal to enough voters to get elected.
What's needed is some unbiased pool of experts who can evaluate software, and give it their stamp of approval that it's passed attacks along known vectors like SQL query buffer overruns. Sort of a UL label for software packages.
The alternative is licensing programmers so only ones who've received a certain education and passed a test of competency (e.g. they know about SQL query buffer overruns, and how to avoid them) are allowed to write software. This isn't as robust though because even competent programmers can have a bad day and make a mistake. In structural/mechanical engineering, this is offset by having multiple engineers inspect each others' work and signing off on it. Only when all have signed off is the design ready for production. That could work for software too, but as you point out the scope of possibilities is much larger with software, increasing the likelihood that all the programmers assigned to review some code may miss the same problem.
With software, you have a building. The earth may suddenly turn to quicksand, your building may be attacked by dinosaurs, the people in the building are usually trying to set it on fire, and the people who own it will never spend one cent on fixing any problems that appear. That and meteors.
Nobody is denying that software is different from physical design. But look at DNA. Babies heads don't explode just because the pregnant mother ate something that gave her a stomachache. Software can be designed robustly. The reason the "earth may suddenly turn to quicksand" in software is because the code you've written for "earth" is not robust and can be cajoled into doing things you don't expect (like turn to quicksand). So it's still a programmer's fault - the problem is just in the "earth" code rather than in the "building" code.
The "people trying to set the building on fire" problem is a great example of a way software could be improved. People do try to set buildings on fire. As a result, we've got a bunch of fire codes added to building design. The architect and structural engineer aren't allowed to just design the building any way they want. They're forced to contemplate the situation - "what if a fire breaks out?" And design the building to withstand that scenario. Likewise, if programmers were forced to contemplate the situation - "what if we suffered a buffer overrun exploit?" perhaps they would design their software to be more resistant to damage such an exploit could cause. Instead of all your code simply accepting whatever value for command messages passed between them, maybe they'll also communicate the maximum length they expect such strings to be. And when one piece of code sees another subroutine passing it a string which is longer than that subroutine says should be the maximum length, it knows something isn't right and can refuse to execute that string as a command. If all code did this, then buffer overruns would be impossible because anyone writing code would be forced to check for buffer overruns in their new code, if they wanted that code to work reliably with other code.
The biggest challenge to writing robust software isn't sloppy code. That can be countered by good education, good coding practices, and rigorous testing. The biggest challenge is the possibility of previously unknown exploits coming to light. When we build a building, we're relying on knowledge and experience gained over thousands of years as to how the ground and building materials behave. Nearly all the weird corner cases have already been found.
Software OTOH has only been around for a few decades (if you exclude programmable sewing design machines). We are still finding weird corner cases, whose discovery can turn a previously accepted-as-safe coding practice into a dangerous one.
Slashdot Mirror
"Retirement" has been a part of most cultures for over a thousand years. I wouldn't be surprised if it's been around for as long as humans have stuck together as a family. In most cultures, when your parents/grandparents get too old and frail to do physical labor, they move into your house and you take care of them until they die. (Though a few cultures took the opposite approach and simply killed the elderly.)
Social Security is the exact same thing, except it expands it from direct descendants taking care of their direct ancestors, to all descendants as a whole (everyone of working age in the country) taking care of all their ancestors as a whole. Young working people pay into Social Security, and the money is given to retired people as a monthly stipend. (No, the government does not put it into a savings account where it waits for you until you retire.)
I've been saying this for over a decade: Put a physical read-only switch on storage drives (and motherboard BIOSes). Then design OSes to boot off a read-only device, with things that need to be written (like logfiles) going to a different drive. Same for programs - the OS should only allow programs on the boot device to run. Double-clicking an executable on another drive should pop up an error (unless the read-only switch of the boot device is off).
Then, once you have the computer set up as you want it with the OS and and desired programs running, you can flip the switch and lock down the system. Anyone who uses the computer, whether remotely or locally cannot change the OS or programs without first physically opening it up to flip the switch. A hack might open up a crack to let a hacker's foot in the door, but they cannot then leverage it to root the entire system. If they got in via a memory overrun exploit, then all the modifications they try to make to the system have to be done through that memory overrun exploit. Malware might be able to take hold, but it cannot write itself to automatically start next time the computer reboots. Malware wouldn't be able to cause computers to fail to boot. In fact a reboot would clear out any such malware, though it might still be attached to a data file if a program is vulnerable to it when the data file is read. (Ransomware wouldn't change since it already leaves the OS and program files alone - it just wouldn't be able to set itself to load and run every time the computer boots - it would need to finish encrypting your data before you rebooted your computer.)
Yes it would make updates a pain. But the need for regular updates would be substantially diminished since it'd be much harder for malware to exploit a known vulnerability. You could make updates a once a month or once every few months thing, instead of needing daily updates like we do today. And the need to shutdown the computer before you opened it up to flip the read-only switch would clear out any malware laying in wait for update day. You'd just have to make sure the update was the first (and only) thing you ran when you turned the computer back on.
There is no justification for making these extensions retroactive. The purpose of copyright is to promote the creation of new works. You cannot encourage the creation of something which has already been created. So there is absolutely zero benefit to extending the copyright term of existing works. Since the only rationale given in the Constitution for copyright is the promotion of the creation of new works, extending the copyright term for existing works is unconstitutional.
If Congress wants to make copyright last 140 years, fine. But the lengthened term should only apply to works created after the term is changed. Those are the only works whose creation could have been encouraged by a lengthened copyright term, and thus fall under the justification outlined in the Constitution.
In gambling, the game is rigged by the house to statistically always be in their favor. The stock market is different because there is no rigging (outside of insider trading).
In cases where a company is obviously doing something right or wrong, this helps it rise or fall in the market, thus encouraging economic optimizations which increase our standard of living for the same amount of effort expended.
It only starts to look like gambling when all the obvious low-lying optimization fruit has already been picked. Then you start to get more into too-small-to-measure and luck-based optimizations, resulting in more random outcomes that look like gambling.
And pretty much all political proposals to penalize "bad" behavior is that they're not designed to respond to feedback. If you're going to do this, the correct way to do it is to designate some desired level of atmospheric carbon. The higher we go above that level, the greater the tax per ton of CO2 becomes. Likewise, the closer the actual CO2 level gets to the desired level, the smaller the tax becomes. (Conceivably it could go negative if it's deemed that too little atmospheric CO2 is also a danger.)
When you don't use a simple feedback mechanism like this to make incentives auto-adjusting, you get the screwed-up situation we currently have with gender "equality" in education. Girls are outperforming boys at every education level. We've reached the goal of gender equality in education, then blown right past it to where girls are now advantaged relative to boys. But the incentives (created when girls lagged behind boys) are still set to favor girls.
The electronic voting machine i used in this past Tuesday's election used a monitor, scroll wheel, and buttons to let you select which candidates you wanted to vote for. After you finished voting, it showed you your votes on the screen and asked you to make sure those votes were correct. Then it printed your votes onto a piece of paper, and asked you to confirm that the printed ballot accurately reflected your choices. (Presumably if you clicked no, it would shred the printed ballot.)
That seemed to me a sensible method of electronic voting. Use a computer to reduce the expense and confusion of custom-printed ballots for every election. But retain the paper ballot for counting and auditing purposes.
The problem is Amtrak competes against a service which is heavily subsidized by the government - cars and trucks run on freeways constructed with tax dollars (fuel taxes only pay for maintenance). If the roads weren't so heavily subsidized, the U.S. would be more like Europe and would make greater use of rail transport for both cargo and people.*
So expecting Amtrak to be profitable on its own is unrealistic. You need to subsidize it to the same degree you're subsidizing freeways just to level the playing field. And when you subsidize a small service to that degree, politicians start to play around with how the money should be spent.
* The idea back when freeways were first made was that trucks could transport goods from endpoint to endpoint, eliminating the need for expensive labor-intensive loading and unloading stages, where people at the railyard have to move cargo from the train onto a truck to make it to its final destination, or vice versa. The labor of the loading/unloading stages was the predominant cost to cargo transport at the time, so eliminating it was an economically sound idea. But since then, fuel costs have increased substantially, and the advent of container transport has reduced loading/unloading costs. But we're still stuck with a cargo transport system built based on the old cost structure, which is artificially keeping trucks competitive with trains for long-distance transport.
The long-term suicide rate for the U.S. has fluctuated between about 10-14 per 100,000 over the last 40 years. 1999 was just the minimum, which the reporter cherry picked to try to turn this into a story. It's been higher in the past.
Also, suicide rates have been climbing up throughout the world, not just in the U.S. Curiously, homicide rates also stopped the long-term decline around 2010 and started creeping up again. So average global temperature may actually be linked (violence and higher temperatures have been correlated in the past). But nobody really knows for sure why, and anyone who claims they do is just trying to spin the story to fit their preconceptions.
The link for CO2 emissions by fuel source didn't come through in that post.
Gasoline generates about 8.89 kg of CO2 per gallon. So producing 1 ton of CO2 would require burning (1000 kg)/(8.89 kg/gallon) = 112.5 gallons of gasoline. The current average price of gasoline is $2.934/gallon, so 112.5 gallons of gas would cost (112.5 gal)*($2.934/gal) = $330.
A $100-$200 surcharge per ton of CO2 would thus raise the price of gasoline by just 30%-61%.
Using the same EIA chart, coal generates roughly 2 tons of CO2 per ton of coal. One ton of coal contains roughly 24 Gigajoules of thermal energy, which is 6.67 MWh. If the coal plant is 40% efficient, that means that one ton of coal generates 2.67 MWh of electricity. Since that one ton of coal also emits 2 tons of CO2, we end up with (2 tons CO2) / (2.67 MWh) = 0.75 tons per MWh.
Natural gas generates roughly 53.12 kg of CO2 per thousand cubic feet. A thousand cubic feet of methane contains 1.037 million BTUs of thermal energy = 303.9 kWh. If the gas plant is 60% efficient, this means 53.12 kg of CO2 are emitted per 182.3 kWh, or (0.053 tons CO2) / (0.1823 MWh) = 0.29 tons per MWh.
Coal accounts for 30.1% of U.S. electricity. Natural gas accounts for 31.7%. So the fractional CO2 contribution of these fossil fuels to electricity is (0.75 tons/MWh)*(0.301)+(0.29 tons/MWh)*(0.317) = 0.318 tons of CO2 per MWh. A $100-$200 surcharge per ton of CO2 then ends up costing $31.80-$63.60 per MWh, or 3.2 cents - 6.4 cents per kWh.
Average electricity price in the U.S. is 12 cents/kWh. So a $100-$200 surcharge per ton of CO2 would raise the price of electricity by 27%-53%. Almost exactly the same percentage as gasoline.
Like I keep trying to explain to people: Electric vehicles aren't cheap to operate because they're more energy efficient. They use nearly as much energy as ICE vehicles. They're just cheaper to operate because the coal and natural gas used to generate electricity are roughly an order of magnitude cheaper per MJ than gasoline. If you want to reduce CO2 emissions, buying an EV presently doesn't help. When you replace an ICE vehicleswith an EV without changing the makeup of your electricity sources, all you've done is shift your CO2 emissions from the car's tailpipe to a fossil fuel power plant's smokestack. That's why the claim that EVs are "zero emissions" is BS at present. You need to replace fossil fuel power plants with nuclear and renewable plants to cause a reduction in CO2 emissions.
That is essentially the problem. The way things like this are supposed to happen is that a few local governments get an idea so they pass a law mandating net neutrality. After some time to gauge the results of the law, a few states take notice and say "this looks like a good idea" and pass their own laws. And after more time to gauge how this affects the states, the Federal government takes notice and says "maybe we should make this a national law." These incremental steps, gradually expanding the reach of a law, allow us to properly gauge the law's effect and make modifications to it if problems should arise, before it affects the entire nation.
Instead, Tom Wheeler short-circuited the entire process and unilaterally declared that net neutrality must be the law of the land. Net neutrality isn't the only possible solution to this problem. The problem is actually government-created - local governments granted service monopolies to cable companies. These companies, assured that their customers cannot flee to another ISP, then intentionally degrade online services like Netflix to extort payments from Netflix to restore service.
The way it should work is some areas try net neutrality, some areas try rescinding these government-granted monopolies and allow multiple ISPs to compete, some areas try some different solution that we haven't yet thought of. Let these different solutions play out for a few years. Then we can study actual empirical data, and decide what's best for the entire country. And only then do we pass a national law with a solution to the problem.
The knee-jerk reaction method used to get net neutrality implemented via the FCC is totally the wrong way for government to operate. Heavy-handed decisions like this without first exploring possible solutions is what nearly saddled us with GSM. The original GSM spec was built on TDMA - each phone takes turns talking to the tower. Europe mandated GSM, and most of the rest of the world followed. The U.S. refused to require it, which allowed a competing service based on CDMA to develop. When phones started being used more for data than talking, suddenly the achilles heel of TDMA reared up. TDMA requires each phone to get a full timeslice even if it has little or no data to transmit. This wastes a huge amount of bandwidth. CDMA on the other hand allows all phones to transmit at the same time (they see each others' transmissions as noise, thus reducing the signal-to-noise ratio), and bandwidth is automatically allocated in proportion to how much each phone is transmitting. No wasted bandwidth. Within a year GSM threw in the towel, licensed CDMA, and added wideband CDMA to the GSM spec for data services. If the U.S. had gone along with the "sensible" decision by bureaucrats to impose GSM, then CDMA wouldn't have happened, and our cellular data speeds today would probably down around 1 Mbps. And many of the services we enjoy on our phones today wouldn't yet exist.
Tablets more popular than desktop computers for browsing. The vast majority of desktop computers are used in businesses (where web browsing is usually prohibited) and by gamers. Most home users switched to laptops long ago, and most gamers also own a laptop for their non-gaming computer use.
And tablets are very popular. They just have a really long usable lifespan, which caused sales per year to plummet once the market was saturated. Mine is going on 4 years old and I have no intention of replacing it, whereas I get the itch to replace my laptop after 2 years and usually replace it by 3 years. I'd still be using my 8 year old tablet if the battery hadn't died.
Here's the study you requested.
The problem with using datasets over long times (like 19 years) is a myriad of other factors changing over time can influence the economic outcome. I suspect if you'd set the endpoint at the middle of the 2008/2009 recession, you would've concluded the minimum wage absolutely increases unemployment. To really do a valid study, you need to do an A/B comparison. Two different areas (preferably nearby or mixed with each other) with similar economies at the same time, but with different minimum wages.
Even then the results of all these studies tend to be mixed. I suspect what's going on is that a high minimum wage kills off low-income jobs (those where a worker would not be productive enough to pay for their own price of labor). So no neighborhood kid mowing your lawn, no people loading boxes from the warehouse onto the truck, fewer cleaning staff for the office, etc. But businesses don't give up and file for bankruptcy because of a small change like this. They adapt - they buy a forklift and hire a skilled forklift operator paid a lot more than the minimum wage, but who moves more boxes per hour per $ of wage than the workers he replaced. The company just hadn't wanted to pay the up-front money to buy a forklift, until the increased minimum wage tilted the scales against hiring multiple unskilled workers. Or they buy a robot vacuum which handles the nightly cleaning, so they only need the cleaning staff to come in once a week instead of every night. etc.
And so the net result of a too-high minimum wage isn't higher overall unemployment. It's a decrease in the number of jobs which require little or no skill or experience. Which is exactly the complaint you hear about from millenials (the neighborhood kid mowing lawns is out of luck - people will just do it themselves rather than overpay someone else to do it).
This is an impossible task. Copyright doesn't need to be registered. You simply have to create a work and it is automatically copyrighted (registering it just allows you to sue for greater damages if there's commercial infringement).
So in many cases, the only person on the planet who knows a work is copyrighted is the content creator. If someone else then steals his work and uploads it to YouTube, how the hell is YouTube supposed to know it's supposed to block it because it belongs to someone else?
Your clean floor analogy doesn't work because there's a distinct difference between a floor being clean or dirty. This is more like YouTube owning a fountain, and millions of people throw coins into that fountain each year. This court ruling is basically saying YouTube can't allow people to throw coins into the fountain unless YouTube first makes sure none of those coins were stolen. How is YouTube supposed to do that? In some cases, the *only* person on the planet who knows the coin was stolen is the thief - the original owner might not even realize yet that it's been stolen.
So a student is more likely to be killed by a deer than from a school shooting. Where are all the walk-outs and protests advocating deer population control?
For some perspective on the scope of the school shooting problem, look at the stats the CDC puts out. For 2015, the leading causes of death among the 15-19 year old demographic were:
3,919 deaths - Accidents (mostly automobile accidents and drug overdoses). 282x more than school shootings.
2.061 deaths - Suicide. 148x more.
1,587 deaths - Homicide (mostly outside school, and gang related). 114x more.
583 deaths - Malignant neoplasms (cancer). 42x more.
306 deaths - Heart disease. 22x more.
195 deaths - Birth defects. 14x more.
72 deaths - Influenza (the flu). 5.2x more.
63 deaths - Chronic lower respiratory diseases. 4.5x more.
61 deaths - Cerebrovascular diseases. 4.4x more.
52 deaths - Diabetes. 3.7x more.
41 deaths - Complications from pregnancy and childbirth. 3x more.
A protest over excessive rates of teen pregnancy could potentially save 3x more lives than a protest over school shootings. Likewise, teaching kids not to each too many sweets, to exercise, not to smoke, get the flu shot, use sunscreen, not to join gangs, to buckle their seat belt, not to use drugs, and offering them counseling for depression, would all be much more productive uses of our time and effort than worrying about or debating school shootings. For that matter, controlling deer populations to reduce the number of fatalities from striking deer could potentially save 1.35x as many students' lives as lost to school shootings.
If you want to tackle a life-threatening issue that students face, probably the best choice is suicide. It results in more than a hundred times as many student deaths as school shootings. But when's the last time you saw the media run a story about teen suicide? The only reason school shootings are even on the radar is because of the media using them to play the "think of the children!" card against guns.
One of the deadliest hurricanes on record was Hurricane Mitch, which reached category 5 at sea. But most of its devastation happened after dropped below hurricane status. It stalled over Honduras as a tropical storm and basically flooded the country back into the stone age (75 inches of rain over roughly 2 weeks). You can't evacuate when the storm spans all the way from the Atlantic to the Pacific coast.
The only reason it isn't the costliest Atlantic storm on record is because it hit relatively poor countries. Hurricane Harvey was a cake walk by comparison ("only" 40 inches of rain over 1 week). But Mitch was 20 years ago and doesn't quite fit the desired narrative of modern storms moving slower. So Harvey is mentioned instead.
The "core software" is Linux. Pretty much all routers (most standalone devices really) run Linux under the hood. There are a few oddball routers which use a RTOS like QNX or VxWorks (these are the ones you want to avoid if you're hoping to flash a third party firmware). But the vast majority run Linux because it's free. This malware probably wormed its way in via a universal Linux exploit which was patched in the production releases of Linux distros, but not in the much-slower-to-update router firmware releases.
No, it's a sign that the GNOME brand is toxic. So toxic that even Microsoft doesn't want their products to be associated with it in any way. Even by mistake.
If you read Microsoft's news article on this, the point of this project isn't to cool a datacenter using the ocean as a heat sink. It's to build a self-contained datacenter which can be customized to order, is easily shipped (fits on the back of a truck), then deployed at the bottom of the ocean where it operates for years without human intervention. All you need to do is plug in the power and data cables to something onshore.
I suspect putting it at the bottom of the ocean is more about preventing people from breaking in and stealing the equipment inside, rather than cooling it. It's got a built-in self-destruct mechanism which triggers if anyone tries to break in while it's on the seabed, thus eliminating any incentive for thieves. This goes along with the self-contained theme (no need to hire security guards).
Microsoft realizes they're not beholden to Intel. Intel needs Microsoft more than Microsoft needs Intel. While it's been a profitable relationship for nearly 40 years, Intel is facing serious competition at the lower end from ARM ever since computers became "fast enough" that most people can get most of their computing needs done with a low-end computer. Microsoft is making sure they have a finger in every pie. If the Intel ship sinks, they don't want it to take Windows down with it. So they're doing what they can to make sure Windows and its API is hardware-agnostic and can run on both Intel/AMD and ARM.
Whether Windows 10 for the Snapdragon 850 sells well or not is immaterial to Microsoft. They are simply hedging their bets to insure their Windows cash cow survives regardless of whether the winner of the processor war ends up being Intel, AMD, or ARM.
Prior to Coffee Lake, the mobile i5 was just a mobile i3 with turbo boost. i.e. A faster clocked i3. That's it. In no way was the obscene markups for an i5 laptop over an i3 laptop justified. The fact that Macbooks were available with an i5 but not an i3 should've been a huge tip-off that there was an obscene profit margin for little performance gain there. Intel really milked that cash cow for roughly a decade. (It was probably driven by people wrongly assuming that what they knew about desktop processors also held for mobile processors. The desktop i5 was a quad core vs. the desktop i3 being a dual core, so was a worthy upgrade. But both the mobile i3 and i5 were dual cores.)
With Coffee Lake and the Kaby Lake refresh (i5-8xxx), most of the mobile i5s are now quad cores. So they're now a worthwhile upgrade over a mobile i3.
Carbon neutral by free market works. Carbon neutral by free market + lawsuits against constructing nuclear plants doesn't work.
The first method is a never-ending game of leapfrog. The second method favors users because there are a lot more of them than companies tracking this data. They can generate fake browsing data faster (up to the limit of their Internet bandwidth) than these companies can filter it out.
What's needed is some unbiased pool of experts who can evaluate software, and give it their stamp of approval that it's passed attacks along known vectors like SQL query buffer overruns. Sort of a UL label for software packages.
The alternative is licensing programmers so only ones who've received a certain education and passed a test of competency (e.g. they know about SQL query buffer overruns, and how to avoid them) are allowed to write software. This isn't as robust though because even competent programmers can have a bad day and make a mistake. In structural/mechanical engineering, this is offset by having multiple engineers inspect each others' work and signing off on it. Only when all have signed off is the design ready for production. That could work for software too, but as you point out the scope of possibilities is much larger with software, increasing the likelihood that all the programmers assigned to review some code may miss the same problem.
Nobody is denying that software is different from physical design. But look at DNA. Babies heads don't explode just because the pregnant mother ate something that gave her a stomachache. Software can be designed robustly. The reason the "earth may suddenly turn to quicksand" in software is because the code you've written for "earth" is not robust and can be cajoled into doing things you don't expect (like turn to quicksand). So it's still a programmer's fault - the problem is just in the "earth" code rather than in the "building" code.
The "people trying to set the building on fire" problem is a great example of a way software could be improved. People do try to set buildings on fire. As a result, we've got a bunch of fire codes added to building design. The architect and structural engineer aren't allowed to just design the building any way they want. They're forced to contemplate the situation - "what if a fire breaks out?" And design the building to withstand that scenario. Likewise, if programmers were forced to contemplate the situation - "what if we suffered a buffer overrun exploit?" perhaps they would design their software to be more resistant to damage such an exploit could cause. Instead of all your code simply accepting whatever value for command messages passed between them, maybe they'll also communicate the maximum length they expect such strings to be. And when one piece of code sees another subroutine passing it a string which is longer than that subroutine says should be the maximum length, it knows something isn't right and can refuse to execute that string as a command. If all code did this, then buffer overruns would be impossible because anyone writing code would be forced to check for buffer overruns in their new code, if they wanted that code to work reliably with other code.
The biggest challenge to writing robust software isn't sloppy code. That can be countered by good education, good coding practices, and rigorous testing. The biggest challenge is the possibility of previously unknown exploits coming to light. When we build a building, we're relying on knowledge and experience gained over thousands of years as to how the ground and building materials behave. Nearly all the weird corner cases have already been found.
Software OTOH has only been around for a few decades (if you exclude programmable sewing design machines). We are still finding weird corner cases, whose discovery can turn a previously accepted-as-safe coding practice into a dangerous one.
Cost only a few hundred dollars to register. Any project of any reasonable size should be filing them.