But even when designing for 5 million users, it's very difficult to test for that load without having 5 million actual users. There will inevitably be a bottleneck you didn't think of that fails under real loads, that you didn't or couldn't test for. A bottleneck that can't simply be fixed with more servers.
Case in point; Steve Keen's Debunking Economics contains many examples of economic theories that are either provably false, or run counter to the empirical evidence, or both.
Building a web site for a hobby is very different to the approaches you need for massive scale. If you took a look at how the largest web sites scale, they all do things slightly differently, and they all had to fix their scalability problems gradually as their popularity increased. Throwing more CPU and RAM at the problem may be unable to fix anything if their current software design doesn't allow for it.
I was mainly thinking of his research into government responses to a credit crunch. Though he has also talked about the events of 1937, where the US government believed the crisis was over and they cut spending, leading to higher unemployment, which has more in common with the events of today.
A perspective on Australia's government debt and spending based on average trends.
For some more depth on the dynamics of the economy and how governments should react to crisis, I'd suggest reading some of Steve Keen's research in this area.
Invest in bridges, dams, highways, internet backbones, schools, housing... Invest in things with a long term payoff to the rest of society that the private sector can't afford to build. Then you should be improving our economic potential. But the important thing is to spend now and get that money into the hands of those people who are suffering the most due to the current economic conditions. The poor wont hold onto that money, they can't afford to.
The recent differences in measured GDP might be slight, but the recent changes in private debt levels & employment have been much larger, faster, strongly correlated and easy to measure. If you only look at government spending you aren't seeing the whole picture. You have to look at spending across the whole economy to quantify the dynamics of the system.
The US's current QE program is useless. Giving money to banks to hold in reserve may help keep the banks afloat, but it does nothing to help the common man. Also the scale of the current program, while large, is smaller than the reduction of consumer spending.
Austerity programs in other countries have not helped to reduce their level of government debt, the level of government debt has still grown during this period.
Why? Because the population is spending less from borrowed money. The scale of that change in consumer spending easily dwarfs the change in government spending.
If you want to boost the economy, the government needs to drastically *increase* their spending, with the majority of that money spent in ways that will flow through to the majority of citizens. Don't just throw cash at bank managers, that doesn't help anyone.
The fight to limit spending is a fight for the economy
No, no, no, no, no! Government spending reduces when then the economy is booming. But attempting to reduce spending does not trigger a boom!. Correlation does not imply causation.
That's not to say that the level of debt in the economy isn't important. Understanding the role of debt is critical to understanding the dynamics of the economy. But you can't just focus on government debt, you have to look at *all* of the debt in the economy to understand what is really going on. Once the crisis hit in 2008, the private sector started reducing their debts. Spending from credit essentially stopped. Government spending since 2008 has helped to cushion the impact of that reduction in credit. See this graph to get a feel for the scale of the problem.
If anything the role of government is to save for the future during a boom, and pay from those savings during a slump. Or borrow during a slump and pay it back in the next boom, which amounts to the same thing. Of course before this crisis started the government wasn't saving at all, but forcing the government to save now will have a drastic impact on the welfare of the country. Look how well austerity programs have been working in the rest of the world.
Source code seems to be available online here. A quick look at the User Manual indicates that all communication is routed via tor which raises the bar for tracing connections significantly.
Sure, track some limited data for later investigation, but encrypt everything that might be identifiable asymmetrically with a public key at the point of data capture. The private key being generated on a server controlled and accessed only by the judicial branch, only retrievable with a valid and very specific warrant.
So it looks like this was a deliberate addition so that the router's internal tools could use http requests to change config. Why didn't they just check for incoming requests from localhost? Surely that would have been simple and safe enough? So instead they create something that they *know* is a backdoor.
Browsers already have standards for embedding 3rd party plugins into pages. This new DRM scheme neither adds nor removes anything from that capability. Browsers will still need to hand over control of the decoding to 3rd party code. Perhaps the interface between browser and external code, or the method for specifying the content in html will be simpler. But this addition doesn't really change anything.
Because Ctrl-Alt-Del was already a low level hardware interrupt, and the NT kernel intercepted it, it was re-purposed for the login process. Mainly so the user could be certain that they weren't looking at a fake login prompt. It's a decision that made sense at the time.
Buyers and Sellers create orders at various prices and lock them in, but the order is initially hidden from everyone. After the order has existed in the system for 2x interval it is published. Attempting to cancel an order has a 1 interval delay. Every interval, if buy and sell orders overlap, match up trades. Resolve the imbalance between buyers and sellers by picking the winning trades randomly. Add a bias to the randomisation process to prefer smaller quantity orders.
Since there's a delay between publishing and finalising trades, a stock with a high volume of trades is likely to have overlapping buy and sell orders almost all the time.
Want to understand the dynamics of the economy? Track the level, velocity and acceleration of debt.
This might surprise you, but mainstream economists completely ignore the role of banks in issuing credit and the impact that credit has on the economy.
Now, I'm all for including the banking sector in stories where it's relevant; but why is it so crucial to a story about debt and leverage?
Paul Krugman
But there is an enormous correlation between debt growth and employment, margin lending and stock prices, mortgage debt and house prices. Changes in debt levels and asset prices create a feedback loop where each causes the other to change. Debt induces a euphoric blindness to it's potential downsides.
If you look at the historical trend in the level of debt compared to income, there was a big build up of debt leading up to the 1929 crash and ensuing depression. Plus the ratio of debt to income got even worse during the depression as income levels fell.
Leading up to the 2008 collapse of Lehman Bros. & AIG, we had far more debt than the 1920's. More debt than the highest peak during the 30's depression.
Most of that debt is not going to be repaid, it's simply impossible. So the only question that remains is, how we are going to manage not paying these debts. Are we going to continue waiting until everyone has gone bankrupt? Depressing the economy until debts are back to reasonable levels, while the population becomes increasingly desperate for a solution. Or inflate our way out by giving everyone free money to pay their debts with? Levelling the wealth distribution a little, which will of course annoy those who have been responsibly saving for the future.
Both solutions have morality questions, both solutions cause a massive disruption to the economy and fabric of society.
The recorded history of the domestic dog demonstrates an explosion of variations in a relatively short time frame, driven by both natural and artificial selection criteria. Selective breeding can result in quite rapid and drastic changes to the form of an organism. Almost as if this rapid adaptation was a deliberately designed feature of life. There's no need for a hypothetical Noah to have carried more than one pair of dogs / wolves and still account for all of the variations we see today. Similarly, other kinds of animals, living in different ecosystems, with different selection pressure could exhibit the same rapid variations without requiring aeons of time elapsing.
On the other hand, in a stable environment, natural selection pressure can also serve to keep the form of organisms constant. eg Predators eating the weak, females choosing the mates that match their ideal image of the species. If you assume there was a period of massive disruption, where each organism must adapt to the new ecosystem they find themselves in, change can be very rapid.
You're still assuming that the alps existed before the flood. The young earth creationist would assume that the flood *levelled everything*, deposited the fossils, then tectonic plate collisions / meteor impact / massive earthquakes rapidly pushed the alps above the water line followed by further erosion.
Another point from the creationist / young earth / Intelligent Design side, ignoring any argument based on the story of creation or Adam. The young earth creationist who takes the story of Noah literally, doesn't agree with your interpretation of the fossil record and evolutionary history at all. Picture Japan's violent tsunami multiplied to a global scale, eroding away practically everything. The majority of the fossils and layered geological records then deposited as the turbulent ocean calmed down and the water receded from the land. The large flow of receding waters carved out river basins and canyons quite quickly from the soft sediments.
For someone with this world view, the "Facts" of evolution are not incontrovertible. The story of evolution, as derived from the fossil record, is based on assumptions that the creationist doesn't agree with.
That's not to say that the creationist disagrees with the facts of biology, as derived from examining living animals and how they change over time. It's the extrapolation of currently observed processes into the unknowable past that they disagree with.
Exactly, which would force you to store every datetime with the relevant timezone instead of assuming all of your datetime values are in the same zone. Which is essentially what I implied.
They should just stick to one standard frame of reference, similar to GPS time. The translation to local timezones could include leap seconds if you want it to. Conversions to local timezones already has some ambiguity as you mentioned due to daylight saving time. Software developers need to abandon the idea that datetime values in your local timezone can be stored, compared and used in calculations as only a single number.
Multi-Monitor setups typically assume that all of your monitors are arranged in a single plane, with the scene rendered based on your nose being some distance away from the exact center of your primary display. Instead each screen should be rendered based on a separate camera, from a POV that is off center.
Slashdot Mirror
But even when designing for 5 million users, it's very difficult to test for that load without having 5 million actual users. There will inevitably be a bottleneck you didn't think of that fails under real loads, that you didn't or couldn't test for. A bottleneck that can't simply be fixed with more servers.
Case in point; Steve Keen's Debunking Economics contains many examples of economic theories that are either provably false, or run counter to the empirical evidence, or both.
Building a web site for a hobby is very different to the approaches you need for massive scale. If you took a look at how the largest web sites scale, they all do things slightly differently, and they all had to fix their scalability problems gradually as their popularity increased. Throwing more CPU and RAM at the problem may be unable to fix anything if their current software design doesn't allow for it.
Here's one.
I was mainly thinking of his research into government responses to a credit crunch. Though he has also talked about the events of 1937, where the US government believed the crisis was over and they cut spending, leading to higher unemployment, which has more in common with the events of today.
A perspective on Australia's government debt and spending based on average trends.
For some more depth on the dynamics of the economy and how governments should react to crisis, I'd suggest reading some of Steve Keen's research in this area.
Invest in bridges, dams, highways, internet backbones, schools, housing... Invest in things with a long term payoff to the rest of society that the private sector can't afford to build. Then you should be improving our economic potential. But the important thing is to spend now and get that money into the hands of those people who are suffering the most due to the current economic conditions. The poor wont hold onto that money, they can't afford to.
The recent differences in measured GDP might be slight, but the recent changes in private debt levels & employment have been much larger, faster, strongly correlated and easy to measure. If you only look at government spending you aren't seeing the whole picture. You have to look at spending across the whole economy to quantify the dynamics of the system.
The US's current QE program is useless. Giving money to banks to hold in reserve may help keep the banks afloat, but it does nothing to help the common man. Also the scale of the current program, while large, is smaller than the reduction of consumer spending.
Austerity programs in other countries have not helped to reduce their level of government debt, the level of government debt has still grown during this period.
Why? Because the population is spending less from borrowed money. The scale of that change in consumer spending easily dwarfs the change in government spending.
If you want to boost the economy, the government needs to drastically *increase* their spending, with the majority of that money spent in ways that will flow through to the majority of citizens. Don't just throw cash at bank managers, that doesn't help anyone.
The fight to limit spending is a fight for the economy
No, no, no, no, no! Government spending reduces when then the economy is booming. But attempting to reduce spending does not trigger a boom!. Correlation does not imply causation.
That's not to say that the level of debt in the economy isn't important. Understanding the role of debt is critical to understanding the dynamics of the economy. But you can't just focus on government debt, you have to look at *all* of the debt in the economy to understand what is really going on. Once the crisis hit in 2008, the private sector started reducing their debts. Spending from credit essentially stopped. Government spending since 2008 has helped to cushion the impact of that reduction in credit. See this graph to get a feel for the scale of the problem.
If anything the role of government is to save for the future during a boom, and pay from those savings during a slump. Or borrow during a slump and pay it back in the next boom, which amounts to the same thing. Of course before this crisis started the government wasn't saving at all, but forcing the government to save now will have a drastic impact on the welfare of the country. Look how well austerity programs have been working in the rest of the world.
Source code seems to be available online here. A quick look at the User Manual indicates that all communication is routed via tor which raises the bar for tracing connections significantly.
I hope that future transactions still require some kind of authorisation. On a completely unrelated subject, what's your email address?
Sure, track some limited data for later investigation, but encrypt everything that might be identifiable asymmetrically with a public key at the point of data capture. The private key being generated on a server controlled and accessed only by the judicial branch, only retrievable with a valid and very specific warrant.
... is so cheap already?
So it looks like this was a deliberate addition so that the router's internal tools could use http requests to change config. Why didn't they just check for incoming requests from localhost? Surely that would have been simple and safe enough? So instead they create something that they *know* is a backdoor.
Browsers already have standards for embedding 3rd party plugins into pages. This new DRM scheme neither adds nor removes anything from that capability. Browsers will still need to hand over control of the decoding to 3rd party code. Perhaps the interface between browser and external code, or the method for specifying the content in html will be simpler. But this addition doesn't really change anything.
Because Ctrl-Alt-Del was already a low level hardware interrupt, and the NT kernel intercepted it, it was re-purposed for the login process. Mainly so the user could be certain that they weren't looking at a fake login prompt. It's a decision that made sense at the time.
Buyers and Sellers create orders at various prices and lock them in, but the order is initially hidden from everyone. After the order has existed in the system for 2x interval it is published. Attempting to cancel an order has a 1 interval delay. Every interval, if buy and sell orders overlap, match up trades. Resolve the imbalance between buyers and sellers by picking the winning trades randomly. Add a bias to the randomisation process to prefer smaller quantity orders.
Since there's a delay between publishing and finalising trades, a stock with a high volume of trades is likely to have overlapping buy and sell orders almost all the time.
Want to understand the dynamics of the economy? Track the level, velocity and acceleration of debt.
This might surprise you, but mainstream economists completely ignore the role of banks in issuing credit and the impact that credit has on the economy.
Now, I'm all for including the banking sector in stories where it's relevant; but why is it so crucial to a story about debt and leverage?
Paul Krugman
But there is an enormous correlation between debt growth and employment, margin lending and stock prices, mortgage debt and house prices. Changes in debt levels and asset prices create a feedback loop where each causes the other to change. Debt induces a euphoric blindness to it's potential downsides.
If you look at the historical trend in the level of debt compared to income, there was a big build up of debt leading up to the 1929 crash and ensuing depression. Plus the ratio of debt to income got even worse during the depression as income levels fell.
Leading up to the 2008 collapse of Lehman Bros. & AIG, we had far more debt than the 1920's. More debt than the highest peak during the 30's depression.
Most of that debt is not going to be repaid, it's simply impossible. So the only question that remains is, how we are going to manage not paying these debts. Are we going to continue waiting until everyone has gone bankrupt? Depressing the economy until debts are back to reasonable levels, while the population becomes increasingly desperate for a solution. Or inflate our way out by giving everyone free money to pay their debts with? Levelling the wealth distribution a little, which will of course annoy those who have been responsibly saving for the future.
Both solutions have morality questions, both solutions cause a massive disruption to the economy and fabric of society.
The recorded history of the domestic dog demonstrates an explosion of variations in a relatively short time frame, driven by both natural and artificial selection criteria. Selective breeding can result in quite rapid and drastic changes to the form of an organism. Almost as if this rapid adaptation was a deliberately designed feature of life. There's no need for a hypothetical Noah to have carried more than one pair of dogs / wolves and still account for all of the variations we see today. Similarly, other kinds of animals, living in different ecosystems, with different selection pressure could exhibit the same rapid variations without requiring aeons of time elapsing.
On the other hand, in a stable environment, natural selection pressure can also serve to keep the form of organisms constant. eg Predators eating the weak, females choosing the mates that match their ideal image of the species. If you assume there was a period of massive disruption, where each organism must adapt to the new ecosystem they find themselves in, change can be very rapid.
You're still assuming that the alps existed before the flood. The young earth creationist would assume that the flood *levelled everything*, deposited the fossils, then tectonic plate collisions / meteor impact / massive earthquakes rapidly pushed the alps above the water line followed by further erosion.
Another point from the creationist / young earth / Intelligent Design side, ignoring any argument based on the story of creation or Adam. The young earth creationist who takes the story of Noah literally, doesn't agree with your interpretation of the fossil record and evolutionary history at all. Picture Japan's violent tsunami multiplied to a global scale, eroding away practically everything. The majority of the fossils and layered geological records then deposited as the turbulent ocean calmed down and the water receded from the land. The large flow of receding waters carved out river basins and canyons quite quickly from the soft sediments.
For someone with this world view, the "Facts" of evolution are not incontrovertible. The story of evolution, as derived from the fossil record, is based on assumptions that the creationist doesn't agree with.
That's not to say that the creationist disagrees with the facts of biology, as derived from examining living animals and how they change over time. It's the extrapolation of currently observed processes into the unknowable past that they disagree with.
Exactly, which would force you to store every datetime with the relevant timezone instead of assuming all of your datetime values are in the same zone. Which is essentially what I implied.
They should just stick to one standard frame of reference, similar to GPS time. The translation to local timezones could include leap seconds if you want it to. Conversions to local timezones already has some ambiguity as you mentioned due to daylight saving time. Software developers need to abandon the idea that datetime values in your local timezone can be stored, compared and used in calculations as only a single number.
Multi-Monitor setups typically assume that all of your monitors are arranged in a single plane, with the scene rendered based on your nose being some distance away from the exact center of your primary display. Instead each screen should be rendered based on a separate camera, from a POV that is off center.