I think CS has (or should have) grown out of applied mathematics (theory) a bit and should realize that computers are there in the real world (empirics). Computers are not (no longer??) cleanly analyzable. Do you actually think that with three ethernet cards, a network that has widely varying connection statistics, a CPU that has an interpreter to go from CISC to RISC, L1, L2 etc. caches, branch prediction, pipelining, RAM access, magnetic disk and SSD, you can compute cleanly how long your algorithm will take in a concurrent environment? At best, your 'model' of the idealized computer might predict a particular runtime, you will actually have to measure it to see if it is true. CS does not produce better models of computers, they're still working on the notion that there's a universal computer, and that constants don't matter.
This is where the disconnect comes from. CS is still viewed as applied discrete mathematics, while it should be more akin to applied discrete physics, with a theoretical branch and an empirical branch. Unfortunately, in CS the empirical branch is practically non-existent, and only studied in the field. This leads to the immense gap between CS and industry.
We can add 2 apples + 2 apples, but this is something WE do - it isn't part of the natural universe - we would see it as 4 apples, but the apples don't suddenly become an entity of "4 apples", they are still individual apples.
But now you also claim that even individual apples are not part of the natural universe, but are also constructs of our way of dealing with the universe (here I agree with you). How can you critique mathematics for not being natural, while at the same time you celebrate that our entire way of dealing with the natural world is in the same way not part of the natural universe? Why is the concept of four things constructed by the human mind less natural than one thing constructed by the human mind? Why is 'none' even more abstract? What is your point?
Yes. I myself have always been baffled by the fact that statisticians, no matter how much you push them, cannot create a syllabus/book that explains statistics from the ground up. From day one they throw models, parameters, assumptions, short-cuts, and methods at you without a thorough explaining of why you would ever need that. Statisticians will shake their heads, but often can't explain what you should have done otherwise, simply because they in their turn, don't have a clue what experimental evidence gathering entails. Statistics is one fucked up branch of mathematics, to the point that many serious mathematicians (who do their utmost to teach the concepts) don't consider it to be part of mathematics.
I don't think that Numerical Analysis is that central in a CS education. It usually is there to convince the students that all the assumptions that they have learned in calculus (e.g., that the real numbers exist) is false when dealing with computers. So why first teach them something they don't need, then teach them that this doesn't work in practice, and then ways to get around it, finally ending up with methods to solve problems they will not face in practice? Can't we just teach them directly what they need to know: floating point, fixed point, rational numbers, arbitrary precision, and when to use what? Where does calculus come in?
This holds for an idealized computer that works in pure serial mode. When the computer is connected to the network, works together with other machines that it is not synchronized with, has interrupts, shares memory, has caches, uses a disk, uses various forms of memory with various speed signatures, and interprets 'machine code' in a microcode interpreter to create a reduced instruction set which uses pipelining, then your mathematical analysis of the performance of this computer might fall seriously short. Constants ignored in computational analysis become the object of study.
Often in current day systems, the shortest description of the system is the system itself. When this is the case, studying computers becomes an empirical science, where mathematics forms the method of creating models of what happens, not being the object of study itself.
Nuclear plants cannot be insured fully because the price of maximum disaster is so big that the insurance company would be bankrupt instantly. In other words, no insurance company in the world can insure something that can create 250 billion dollars of damage, no matter how unlikely the event is.
Even if the risk (probability * cost) is easily affordable for a single insurance company, they will not be able to insure themselves for when they have to pay 'cost'. The claimants would have to be bailed out, so you would foot the bill anyway.
In response to your question, which came first, reduced training or job hopping, I'm pretty confident to say that 'firing on the spot', and 'downsizing in response to quarterly results' came first. When your corporate culture is based upon getting rid of parts of your workforce as easily as possible, the same workforce will soon adapt and refuse to be trained in corporate specific technology. They will go for employability, as the corporation has clearly indicated that you're an replaceable resource.
Oh, that's easy. The process of attaching irrelevant legislation to further the agenda of a particular congressman is in itself a corrupt activity. Outside of congress, this form of corruption is called bribery, within congress it's called pork. The congressman is voting for a bill because he's bribed, not because he thinks the bill has merit for his constituents. That this is legal does not make this less corrupt. So, each and every politician that has worked on extending a bill to attach pork to it, can be considered corrupt. That's most of them I would guess.
Given the definition of the reasons for getting the prize:
...shall have done the most or the best work for fraternity between nations, for the abolition or reduction of standing armies and for the holding and promotion of peace congresses
Obama certainly qualifies for 'the best work for fraternity between nations', by not being Bush. I know, I know, Bush could not have been re-elected simply for not being eligible anymore, but Obama also qualifies for not being McCain. McCain would've been a signal to the world that the US is Bush, and that this would have utterly destroyed any 'fraternity between nations', namely between the US and the rest of world. So, on two counts, not being Bush, and not being McCain, Obama has done the best work for fraternity between nations in 2009. True, it's actually the people of the United States that voted for Obama that deserve the Nobel Peace Prize, but given the last congress elections, they don't deserve that honor anyway (and those Norwegians probably figured that out). Obama has not voted Republican (though many democrats did). He deserves the prize for not being Bush, nor McCain, nor a Republican. Yes, that's it, he's not a Republican. That's worth a prize, right?
Yes, we in the rest of the world truly detest Republicans and don't fraternize with them. Reagan is still considered a complete asswipe.
So... how do you propose that AT&T distinguish user-initiated data from other data?
I propose that they measure the 90 percentile of data usage of an iphone in rest and don't bill for that amount. So in this case, they simply don't bill for the first 3Mb, thereby avoiding a class action lawsuit. Easy, isn't it?
How much less is quite a lot less. Is it a mere 127 bits, 126, maybe only 100 bits? given 6 * 10^9 people and 10^12 cells per person, we just need 70 bits to be addressable to cover all cells.
In my work as a data miner using lots of socio-economical data for a diverse range of purposes in many industries, I have yet to see the predictor 'month of birth' have any predictive value whatsoever. I did run a test on data I had a few years back, just out of curiosity. Nada.If you want randomness, you might just as well use data enriched by Experian (rimshot)
How many things have supposed "scientists" gotten wrong over the years?
That's the wrong question. The right question is: How many things have supposed "scientists" gotten wrong over the years, and still believe in? You might want to go with the following definition of the scientific method: stop believing things that have been proven false!
Primaries are a marketing fest funded by the government with the sole purpose of making sure that nobody but Rebs or Dems will have a shot at becoming president. As such it makes a joke of democracy and everything that shows that the emperor has no clothes should be applauded.
The parties should hold their internal elections (what primaries really are) on their dollar, not based on tax money.
Actually, the Alaska-Russia comment, together with an embarrassingly large number of other quotes were repeated verbatim by Tiny Fey in her SNL skits. Palin said these words. She's still a political factor. This says a lot about where we are.
If they're known, they should be leaked, otherwise only criminals will have access. If they're leaked, they will be changed, if they are stolen, they will not.
Excellent. You just found an argument to shut down any newspaper that reported on this. Now for extra points see if you can prove that money is speech.
That worked just now, with Bernanke printing 600 billion dollars to pay off debt. However, future debt will take into account the risk of inflation, and the interest rate on that debt will increase. So, can the US inflate itself out of debt when interest rates on that debt are 10000%? Of course not. That is hyper-inflation, to pay off the interest on the debt you have to increase inflation, which increases the interest on the debt, etc. etc. The only way out of this spiral is to default and start over again.
Wow, brilliant reasoning. So everytime the government is involved, it's socialism? Feudalism, monarchies, tyrannies, right-wing dictatorships like Pinochet's, all a form of socialism? You're painting with a very broad brush here.
As others pointed out, both the shoe bomber and captain underpants where not young Arab men, and given the mental capacity of the TSA and of people like Dolphinzilla, I wouldn't be surprised that these guys got so far as they did simply because everybody is really investigating young Arab men, instead of being on the lookout for potential terrorists.
Slashdot Mirror
This is where the disconnect comes from. CS is still viewed as applied discrete mathematics, while it should be more akin to applied discrete physics, with a theoretical branch and an empirical branch. Unfortunately, in CS the empirical branch is practically non-existent, and only studied in the field. This leads to the immense gap between CS and industry.
So, no, CS is not just mathematics.
But now you also claim that even individual apples are not part of the natural universe, but are also constructs of our way of dealing with the universe (here I agree with you). How can you critique mathematics for not being natural, while at the same time you celebrate that our entire way of dealing with the natural world is in the same way not part of the natural universe? Why is the concept of four things constructed by the human mind less natural than one thing constructed by the human mind? Why is 'none' even more abstract? What is your point?
Yes. I myself have always been baffled by the fact that statisticians, no matter how much you push them, cannot create a syllabus/book that explains statistics from the ground up. From day one they throw models, parameters, assumptions, short-cuts, and methods at you without a thorough explaining of why you would ever need that. Statisticians will shake their heads, but often can't explain what you should have done otherwise, simply because they in their turn, don't have a clue what experimental evidence gathering entails. Statistics is one fucked up branch of mathematics, to the point that many serious mathematicians (who do their utmost to teach the concepts) don't consider it to be part of mathematics.
I don't think that Numerical Analysis is that central in a CS education. It usually is there to convince the students that all the assumptions that they have learned in calculus (e.g., that the real numbers exist) is false when dealing with computers. So why first teach them something they don't need, then teach them that this doesn't work in practice, and then ways to get around it, finally ending up with methods to solve problems they will not face in practice? Can't we just teach them directly what they need to know: floating point, fixed point, rational numbers, arbitrary precision, and when to use what? Where does calculus come in?
This holds for an idealized computer that works in pure serial mode. When the computer is connected to the network, works together with other machines that it is not synchronized with, has interrupts, shares memory, has caches, uses a disk, uses various forms of memory with various speed signatures, and interprets 'machine code' in a microcode interpreter to create a reduced instruction set which uses pipelining, then your mathematical analysis of the performance of this computer might fall seriously short. Constants ignored in computational analysis become the object of study. Often in current day systems, the shortest description of the system is the system itself. When this is the case, studying computers becomes an empirical science, where mathematics forms the method of creating models of what happens, not being the object of study itself.
Nuclear plants cannot be insured fully because the price of maximum disaster is so big that the insurance company would be bankrupt instantly. In other words, no insurance company in the world can insure something that can create 250 billion dollars of damage, no matter how unlikely the event is. Even if the risk (probability * cost) is easily affordable for a single insurance company, they will not be able to insure themselves for when they have to pay 'cost'. The claimants would have to be bailed out, so you would foot the bill anyway.
In response to your question, which came first, reduced training or job hopping, I'm pretty confident to say that 'firing on the spot', and 'downsizing in response to quarterly results' came first. When your corporate culture is based upon getting rid of parts of your workforce as easily as possible, the same workforce will soon adapt and refuse to be trained in corporate specific technology. They will go for employability, as the corporation has clearly indicated that you're an replaceable resource.
Oh, that's easy. The process of attaching irrelevant legislation to further the agenda of a particular congressman is in itself a corrupt activity. Outside of congress, this form of corruption is called bribery, within congress it's called pork. The congressman is voting for a bill because he's bribed, not because he thinks the bill has merit for his constituents. That this is legal does not make this less corrupt. So, each and every politician that has worked on extending a bill to attach pork to it, can be considered corrupt. That's most of them I would guess.
Your freedom to swing your arms stops at the tip of my nose.
Obama certainly qualifies for 'the best work for fraternity between nations', by not being Bush. I know, I know, Bush could not have been re-elected simply for not being eligible anymore, but Obama also qualifies for not being McCain. McCain would've been a signal to the world that the US is Bush, and that this would have utterly destroyed any 'fraternity between nations', namely between the US and the rest of world. So, on two counts, not being Bush, and not being McCain, Obama has done the best work for fraternity between nations in 2009. True, it's actually the people of the United States that voted for Obama that deserve the Nobel Peace Prize, but given the last congress elections, they don't deserve that honor anyway (and those Norwegians probably figured that out). Obama has not voted Republican (though many democrats did). He deserves the prize for not being Bush, nor McCain, nor a Republican. Yes, that's it, he's not a Republican. That's worth a prize, right?
Yes, we in the rest of the world truly detest Republicans and don't fraternize with them. Reagan is still considered a complete asswipe.
I propose that they measure the 90 percentile of data usage of an iphone in rest and don't bill for that amount. So in this case, they simply don't bill for the first 3Mb, thereby avoiding a class action lawsuit. Easy, isn't it?
How much less is quite a lot less. Is it a mere 127 bits, 126, maybe only 100 bits? given 6 * 10^9 people and 10^12 cells per person, we just need 70 bits to be addressable to cover all cells.
In my work as a data miner using lots of socio-economical data for a diverse range of purposes in many industries, I have yet to see the predictor 'month of birth' have any predictive value whatsoever. I did run a test on data I had a few years back, just out of curiosity. Nada.If you want randomness, you might just as well use data enriched by Experian (rimshot)
That's the wrong question. The right question is: How many things have supposed "scientists" gotten wrong over the years, and still believe in? You might want to go with the following definition of the scientific method: stop believing things that have been proven false!
The parties should hold their internal elections (what primaries really are) on their dollar, not based on tax money.
Actually, the Alaska-Russia comment, together with an embarrassingly large number of other quotes were repeated verbatim by Tiny Fey in her SNL skits. Palin said these words. She's still a political factor. This says a lot about where we are.
If they're known, they should be leaked, otherwise only criminals will have access. If they're leaked, they will be changed, if they are stolen, they will not.
Yes, under Socialism man suppresses man. Under Capitalism it's the other way around.
Excellent. You just found an argument to shut down any newspaper that reported on this. Now for extra points see if you can prove that money is speech.
So, they're closing down the New York Times & the Washington Post? They've committed exactly the same 'crime' that WikiLeaks did.
Wikipedia thinks that the contribution of small (500) companies is around 50% of the GDP. So yes, I'd say that's a lucrative market.
That worked just now, with Bernanke printing 600 billion dollars to pay off debt. However, future debt will take into account the risk of inflation, and the interest rate on that debt will increase. So, can the US inflate itself out of debt when interest rates on that debt are 10000%? Of course not. That is hyper-inflation, to pay off the interest on the debt you have to increase inflation, which increases the interest on the debt, etc. etc. The only way out of this spiral is to default and start over again.
Wow, brilliant reasoning. So everytime the government is involved, it's socialism? Feudalism, monarchies, tyrannies, right-wing dictatorships like Pinochet's, all a form of socialism? You're painting with a very broad brush here.
It shouldn't even be too difficult to use bittorrent to distribute the correct database
As others pointed out, both the shoe bomber and captain underpants where not young Arab men, and given the mental capacity of the TSA and of people like Dolphinzilla, I wouldn't be surprised that these guys got so far as they did simply because everybody is really investigating young Arab men, instead of being on the lookout for potential terrorists.