You can't set constraints to these values, they're determined by the relative positions between the satellites and the GPS receiver.
You will only have t1=t2 when the distance between the satellite 1 and the GPS receiver is the same as the distance between the satellite 2 and the GPS receiver.
There is NO geometric problem here... you are getting transmissions from 3 satellites that know their own positions in time and space... therefore you know your position. In full 3D. With only 3 satellites.
That's not how I learned it, but I want to change my mind on this if you're really right. But if what you're saying is true, then what exactly is wrong with the following reasoning:
The GPS receiver gets signals from 3 satellites telling the exact time and position of each satellite at the exact moment the signal was sent. Let's call (xi,yi,zi,ti) the position ant time of satellite i (i=1,2,3), and (x,y,z,t) the position and time of the GPS receiver when the message was received. We want to find out the position of the GPS receiver, that is, x, y and z (we don't care about t).
From the data we get from the satellites, we can make an equation like this for each satellite:
sqrt((xi-x)^2 + (yi-y)^2 + (yi-y)^2) = c*(ti-t)
Where c is the speed of light, and i is the number of the satellite (i=1,2,3). This equations simply says that the time it will take for the signal to leave the satellite and reach the receiver, multiplied by the speed of light, is equal to the distance between the satellite and the receiver.
Now, we have 3 equations and 4 unknowns (x,y,z,t). Granted, we're not interested on the value of t if all we care about is position. Still, if you allow t to be free, there are an infinite number of possible values for (x,y,z) that satisfy these equations. Obviously we can discard a lot of silly values for t, like any value less than any of the ti (i.e., the time the signal was received must be later than the time it was sent), and also t can't be much more than any of the ti (i.e., we know the satellites aren't to too far from the receiver). Which means you can get a range of possible values for (x,y,z), but that's not very precise.
If you have 4 satellites, however, there's only one solution (x,y,z,t) for the system of 4 equations.
So, in the end: what's wrong with this reasoning? Or, alternatively, What extra information the GPS satellites could send that would change things?
(By the way, the method I described is -- possibly -- the easiest way to describe the problem, but not the way it's usually stated. See for example this paper for a more realistic approach)
About the wave/particle duality: the explanation you read is extremely confused.
We know light is made of particles because it's possible to make the light source so weak that only one particle is produced every minute or so. It's possible to confirm this by detecting each individual particle. There's no way to explain this behavior if you assume that light is a wave.
But then comes the strange part: even when the particles go one at a time, we still see the interference described in the double-slit experiment you mentioned -- so the particle must be interfering with itself (since there are no other particles close to it). There's no way to explain this behavior if you assume light is a particle.
In the end, it's clear that light is really neither a simple wave nor a simple particle (and why should it be?). Still, if you insist in understanding it in these terms (and it's very useful in practice), you're stuck with this "wave/particle" duality.
Also, the exact same "wave/particle" behavior has been observed countless times with lots of different things: photons (i.e., light), electrons, atoms and even molecules as large as buckyballs (see here).
Structure and Interpretation of Computer Programs - Abelson and Sussman.
It's available online (completely free) here: http://mitpress.mit.edu/sicp/full-text/book/book.html. I had been programming for 5 years when I read it for the first time, and it completely changed the way I think about programming.
That's interesting. Apparently, another important factor for the early movie industry was how the local community received the movie makers. Compare Hollywood:
While [in Los Angeles], the company decided to explore new territories, traveling several miles north to Hollywood, a little village that was friendly and enjoyed the movie company filming there.
Jacksonville's mostly conservative residents, however, objected to the hallmarks of the early movie industry, such as car chases in the streets, simulated bank robberies and fire alarms in public places, and even the occasional riot. In 1917, conservative Democrat John W. Martin was elected mayor on the platform of taming the city's movie industry.
Still, it's clear that they were avoiding New York, which already had a movie industry.
Patent disputes over motion picture films were a major influence on the development of Hollywood. From Wikipedia:
The film patents wars of the early 20th century led to the spread of film companies across the U.S. Many worked with equipment for which they did not own the rights, and thus filming in New York could be dangerous; it was close to Edison's Company headquarters, and to agents the company set out to seize cameras. By 1912, most major film companies had set up production facilities in Southern California near or in Los Angeles because of the location's proximity to Mexico, as well as the region's favorable year-round weather.
So, at least for the movie industry, it was possible to avoid or minimize patent harassment by moving elsewhere. In today's world that's obviously impossible.
First, there's a lot of stigma associated with smoking; smoking has declined from over 40% of the adult population (in the US) in the 1960's to 20% in the 2000's.
Second, the number of vegetarians in the US is very small (only about 5% of people in the US strictly don't eat meat), so anyone not hiring meat eaters would severely limit their options.
Third, while there are real health benefits associated with being vegetarian, some of the benefits that are usually associated with vegetarianism are actually due to other lifestyle factors, like not smoking and being wealthier, or even having a more balanced diet (that is, other than not eating meat: to be a strict vegetarian, you have to pay attention to what you eat, because it's not that easy to get some of the necessary nutrients you'd otherwise get from simply eating meat). That's all just to say that the health benefits are not as clear-cut as with smoking.
All that said, I agree with the general point that not hiring smokers is stupid, probably as stupid as not hiring meat eaters. My point is that there's much more bias against smokers than meat eaters.
Did you even READ the article you're quoting? Merkle-Damgard is known to have many undesirable properties. They're even mentioned in that same WIkipedia article...
And yet, NIST has no plans to phase-out of SHA-2, because SHA-2 is fine. There's a reason I wrote "If there's something really wrong with it": as far as we know, there's nothing really wrong with it. Cryptographic algorithms are all about trade-offs, and SHA-2 is certainly not perfect, but the same is true of almost everything else we use.
When Scheier said "we don't need SHA-3" last week, he was lambasted by the remaining cryptographic community for ignoring exacly that.
Don't be silly. If that was true, people would be recommending moving away from SHA-2, and nobody is saying that.
In fact it should be a given that SHA-X does not suffer from the same vulnerabilities as SHA-X-1.
No, it shouldn't. Both SHA-1 and SHA-2 are based on the Merkle–Damgard construction. If there's something really wrong with it (not that there's any reason to believe so, today), both SHA-1 and SHA-2 would be affected.
Keccak (SHA-3) has a completely different design based on the sponge construction.
often enough the scientist will have a pretty good idea of the expected nature of the data to be collected.
Sure. I was clarifying that you can't just change to a different method of analysis according to the data you got, just because your original analysis didn't give you the result you expected.
In other words, you simply can't change your plan once you've seen the data. In this lecture, Feynman gives a very clear and real example of what can go wrong even if you're trying to be completely honest (look for the part where he talks about Millikan). That's a neat example because it's a very controlled experiment measuring a completely objective thing. In other fields, like Medicine or Psychology, it's much harder to know if your own expectations are influencing the study, so it's very important that you follow a very clearly defined plan from the beginning.
then the way you expected your study to come out is a null hypothesis which can supposedly be rejected with some measurable degree of significance.
You have to be very careful here. In serious studies, you don't get to choose your null hypothesis or how you're going to analyse the data after collecting it. That's a textbook example of introducing confirmation bias.
failing to protect those invested in a market is the same as giving them the finger.
You can phrase it like that if you want to, but there's nothing sacred about protecting someone's investment in a market; the government should protect things that are good for the whole society. Copyright exists "To promote the Progress of Science and useful Arts", and this is done by giving an exclusive right to authors and inventors. This exclusive right is not the reason why we have copyright, it's just the means to a goal.
Just because someone worked hard or invested in something, it doesn't mean they should be protected by the government.
Nice story, but I really don't see the connection with reality.
Your original post got two answers: one saying that it wouldn't make sense to add the feature you want, and another saying that Intel should provide documentation.
You, on the other hand, imply that the reasons given by the ffmpeg authors for not implementing the feature you want are "Saint Stallman does not approve it" and "The Balance of The Force will be tipped in a wrong way", and you give no evidence of that at all.
You make it sound like the economy was shit before Lula took office in 2003, and he fixed it. But anyone who lives in Brazil knows this is not true.
Brazil's current economic situation started to take shape in as early as 1994 with Plano Real. With inflation under control and general stabilization, companies started investing in the country, and the economy started growing a lot. That was slowed down a lot by the economic crisis in Argentina around 2000 -- Argentina is a really important trade partner for Brazil (and even more at the time). Still, the economy kept growing and the situation kept improving way before Lula was elected.
It's true that Lula's and Dilma's government were mostly successful for the economy, but to say they're solely responsible for the relatively good current economic situation of Brazil is a flat-out lie.
Everything you wrote is true, but I think it misses the point.
The reason for the law has nothing to do with making a reliable estimation of the gestational age -- if it was, they would have made the change and then added two weeks to the amount of time given for a legal abortion (this would have kept it consistent with the amount of time given in other states). The way it was done, the only practical effect of this change is to shorten by about two weeks the amount of time available for legal abortions.
My guess is that they didn't want to release the code because, perhaps, they didn't have any, or perhaps because it was all chewing gum and bailing wire and they didn't even have it under source control.
If that had been the case, he could simply have tar'ed his whole tree and put it up in the same place he was distributing the installer. The GPL defines "source code" as:
The “source code” for a work means the preferred form of the work for making modifications to it.
I don't think anyone can argue that the exact tree that was used to develop the code is not the "preferred" form to make modifications to it -- it is the form he used to make his modifications.
I wouldn't see anything wrong with a plumbing company that made accountants have some plumbing experience nor the secretaries for the CEO of Ferrari/Lamborghini/etc probably should know a bit about swapping engines.
That's reasonable, but TFA says:
Each team is also responsible for development of a new coding project that it will present to the company later this year -- projects may involve creation of a new feature or improved functionality for a Web page within the FreeCause application.
This is not about teaching the company's core product, all employees are expected to produce it. Still, it looks like they're doing it only for 2 hours per week, and it's all supervised. So, maybe it's not a bad idea.
Again, no. Just because it's a single assembly instruction, it doesn't make it atomic -- x86 will do exactly the same as any RISC processor: it will load the memory contents, perform the add and then store the result. Nothing guarantees that another processor will not modify the memory content during the execution of the instruction.
If you want to do it atomically, you'd have to do something like this.
If you remove "The single-cycle rule", then there is no difference between RISC and CISC.
No. With RISC, the only way to access memory is via load/store instructions. The general data processing instructions (arithmetic, comparison, etc.) are always done on registers. That's why instructions are very simple.
On CISC, on the other hand, you can use data from memory without loading it into registers, for example in x86 you can do something like "add eax,dword ptr[ebx]" or "add dword ptr[ebx],eax".
Pauling is really great, I agree that he should be more well-known (I, myself, am partial to Dirac, who predicted the existence of anti-matter out of pure math). But you have to realize that Newton and (maybe to a lesser degree) Einstein contributions were of another class.
Newton took Galileo's ideas -- that things tend to keep moving if they're left alone -- and built a whole mathematical theory on top of that, inventing calculus in the process. In a sense, it was the beginning of what we today call Physics.
Einstein was the first to notice (and convinced everyone) that the Lorentz transformation is not just a mathematical trick, it's the very way the space and time works. This itself was not that impressive, he was just giving a "new spin" on what everyone had already observed. But then he went on to show that the "right way" to understand gravity is by noting that it's just a side-effect of mass bending the space and time -- this has lots of consequences that were unknown at the time, like gravity bending light, gravity making time pass at different rates, and a lot other stuff, all of which turned out to be right.
Quantum Mechanics and its implications (like the electron shell), on the other hand, were discovered bit by bit by a lot of different people. That's why no one is hugely famous for it (even though there are certainly big names like Planck, Einstein, Bohr, Heisenberg, Born, Schrodinger, Dirac, Pauli, etc.).
When Paul pointed out that Gentiles didn't have to become Jews before becoming Christians, it opened the floodgates to the spirit of the law, not the letter
On the other hand, Jesus said (Matthew 5:17-5:19)
[17] Do not think that I have come to abolish the Law or the Prophets; I have not come to abolish them but to fulfill them. [18] For truly I tell you, until heaven and earth disappear, not the smallest letter, not the least stroke of a pen, will by any means disappear from the Law until everything is accomplished. [19] Therefore anyone who sets aside one of the least of these commands and teaches others accordingly will be called least in the kingdom of heaven, but whoever practices and teaches these commands will be called great in the kingdom of heaven. [my ephasis]
But, I don't pretend to tell Christians what Christianity is about. And of course Jews also have their own reasons not to follow those barbaric rules. I think it's great. My whole point is that Islam should be treated the same way.
Slashdot Mirror
You can't set constraints to these values, they're determined by the relative positions between the satellites and the GPS receiver.
You will only have t1=t2 when the distance between the satellite 1 and the GPS receiver is the same as the distance between the satellite 2 and the GPS receiver.
Erm, I messed up the equation. It should be:
That doesn't change the reasoning, though.
There is NO geometric problem here... you are getting transmissions from 3 satellites that know their own positions in time and space... therefore you know your position. In full 3D. With only 3 satellites.
That's not how I learned it, but I want to change my mind on this if you're really right. But if what you're saying is true, then what exactly is wrong with the following reasoning:
The GPS receiver gets signals from 3 satellites telling the exact time and position of each satellite at the exact moment the signal was sent. Let's call (xi,yi,zi,ti) the position ant time of satellite i (i=1,2,3), and (x,y,z,t) the position and time of the GPS receiver when the message was received. We want to find out the position of the GPS receiver, that is, x, y and z (we don't care about t).
From the data we get from the satellites, we can make an equation like this for each satellite:
Where c is the speed of light, and i is the number of the satellite (i=1,2,3). This equations simply says that the time it will take for the signal to leave the satellite and reach the receiver, multiplied by the speed of light, is equal to the distance between the satellite and the receiver.
Now, we have 3 equations and 4 unknowns (x,y,z,t). Granted, we're not interested on the value of t if all we care about is position. Still, if you allow t to be free, there are an infinite number of possible values for (x,y,z) that satisfy these equations. Obviously we can discard a lot of silly values for t, like any value less than any of the ti (i.e., the time the signal was received must be later than the time it was sent), and also t can't be much more than any of the ti (i.e., we know the satellites aren't to too far from the receiver). Which means you can get a range of possible values for (x,y,z), but that's not very precise.
If you have 4 satellites, however, there's only one solution (x,y,z,t) for the system of 4 equations.
So, in the end: what's wrong with this reasoning? Or, alternatively, What extra information the GPS satellites could send that would change things?
(By the way, the method I described is -- possibly -- the easiest way to describe the problem, but not the way it's usually stated. See for example this paper for a more realistic approach)
About the wave/particle duality: the explanation you read is extremely confused.
We know light is made of particles because it's possible to make the light source so weak that only one particle is produced every minute or so. It's possible to confirm this by detecting each individual particle. There's no way to explain this behavior if you assume that light is a wave.
But then comes the strange part: even when the particles go one at a time, we still see the interference described in the double-slit experiment you mentioned -- so the particle must be interfering with itself (since there are no other particles close to it). There's no way to explain this behavior if you assume light is a particle.
In the end, it's clear that light is really neither a simple wave nor a simple particle (and why should it be?). Still, if you insist in understanding it in these terms (and it's very useful in practice), you're stuck with this "wave/particle" duality.
Also, the exact same "wave/particle" behavior has been observed countless times with lots of different things: photons (i.e., light), electrons, atoms and even molecules as large as buckyballs (see here).
May I add:
Structure and Interpretation of Computer Programs - Abelson and Sussman.
It's available online (completely free) here: http://mitpress.mit.edu/sicp/full-text/book/book.html. I had been programming for 5 years when I read it for the first time, and it completely changed the way I think about programming.
That's interesting. Apparently, another important factor for the early movie industry was how the local community received the movie makers. Compare Hollywood:
With Jacksonville:
Still, it's clear that they were avoiding New York, which already had a movie industry.
Patent disputes over motion picture films were a major influence on the development of Hollywood. From Wikipedia:
So, at least for the movie industry, it was possible to avoid or minimize patent harassment by moving elsewhere. In today's world that's obviously impossible.
I doubt that.
First, there's a lot of stigma associated with smoking; smoking has declined from over 40% of the adult population (in the US) in the 1960's to 20% in the 2000's.
Second, the number of vegetarians in the US is very small (only about 5% of people in the US strictly don't eat meat), so anyone not hiring meat eaters would severely limit their options.
Third, while there are real health benefits associated with being vegetarian, some of the benefits that are usually associated with vegetarianism are actually due to other lifestyle factors, like not smoking and being wealthier, or even having a more balanced diet (that is, other than not eating meat: to be a strict vegetarian, you have to pay attention to what you eat, because it's not that easy to get some of the necessary nutrients you'd otherwise get from simply eating meat). That's all just to say that the health benefits are not as clear-cut as with smoking.
All that said, I agree with the general point that not hiring smokers is stupid, probably as stupid as not hiring meat eaters. My point is that there's much more bias against smokers than meat eaters.
Did you even READ the article you're quoting? Merkle-Damgard is known to have many undesirable properties. They're even mentioned in that same WIkipedia article...
And yet, NIST has no plans to phase-out of SHA-2, because SHA-2 is fine. There's a reason I wrote "If there's something really wrong with it": as far as we know, there's nothing really wrong with it. Cryptographic algorithms are all about trade-offs, and SHA-2 is certainly not perfect, but the same is true of almost everything else we use.
When Scheier said "we don't need SHA-3" last week, he was lambasted by the remaining cryptographic community for ignoring exacly that.
Don't be silly. If that was true, people would be recommending moving away from SHA-2, and nobody is saying that.
In fact it should be a given that SHA-X does not suffer from the same vulnerabilities as SHA-X-1.
No, it shouldn't. Both SHA-1 and SHA-2 are based on the Merkle–Damgard construction. If there's something really wrong with it (not that there's any reason to believe so, today), both SHA-1 and SHA-2 would be affected.
Keccak (SHA-3) has a completely different design based on the sponge construction.
often enough the scientist will have a pretty good idea of the expected nature of the data to be collected.
Sure. I was clarifying that you can't just change to a different method of analysis according to the data you got, just because your original analysis didn't give you the result you expected.
In other words, you simply can't change your plan once you've seen the data. In this lecture, Feynman gives a very clear and real example of what can go wrong even if you're trying to be completely honest (look for the part where he talks about Millikan). That's a neat example because it's a very controlled experiment measuring a completely objective thing. In other fields, like Medicine or Psychology, it's much harder to know if your own expectations are influencing the study, so it's very important that you follow a very clearly defined plan from the beginning.
then the way you expected your study to come out is a null hypothesis which can supposedly be rejected with some measurable degree of significance.
You have to be very careful here. In serious studies, you don't get to choose your null hypothesis or how you're going to analyse the data after collecting it. That's a textbook example of introducing confirmation bias.
failing to protect those invested in a market is the same as giving them the finger.
You can phrase it like that if you want to, but there's nothing sacred about protecting someone's investment in a market; the government should protect things that are good for the whole society. Copyright exists "To promote the Progress of Science and useful Arts", and this is done by giving an exclusive right to authors and inventors. This exclusive right is not the reason why we have copyright, it's just the means to a goal.
Just because someone worked hard or invested in something, it doesn't mean they should be protected by the government.
That's nice, how about replying to the relevant post?
Nice story, but I really don't see the connection with reality.
Your original post got two answers: one saying that it wouldn't make sense to add the feature you want, and another saying that Intel should provide documentation.
You, on the other hand, imply that the reasons given by the ffmpeg authors for not implementing the feature you want are "Saint Stallman does not approve it" and "The Balance of The Force will be tipped in a wrong way", and you give no evidence of that at all.
Better luck on your future trollings!
You make it sound like the economy was shit before Lula took office in 2003, and he fixed it. But anyone who lives in Brazil knows this is not true.
Brazil's current economic situation started to take shape in as early as 1994 with Plano Real. With inflation under control and general stabilization, companies started investing in the country, and the economy started growing a lot. That was slowed down a lot by the economic crisis in Argentina around 2000 -- Argentina is a really important trade partner for Brazil (and even more at the time). Still, the economy kept growing and the situation kept improving way before Lula was elected.
It's true that Lula's and Dilma's government were mostly successful for the economy, but to say they're solely responsible for the relatively good current economic situation of Brazil is a flat-out lie.
Everything you wrote is true, but I think it misses the point.
The reason for the law has nothing to do with making a reliable estimation of the gestational age -- if it was, they would have made the change and then added two weeks to the amount of time given for a legal abortion (this would have kept it consistent with the amount of time given in other states). The way it was done, the only practical effect of this change is to shorten by about two weeks the amount of time available for legal abortions.
If that had been the case, he could simply have tar'ed his whole tree and put it up in the same place he was distributing the installer. The GPL defines "source code" as:
The “source code” for a work means the preferred form of the work for making modifications to it.
I don't think anyone can argue that the exact tree that was used to develop the code is not the "preferred" form to make modifications to it -- it is the form he used to make his modifications.
Linux was a very different beast when Raymond wrote that.
Actually, Raymond never wrote that. It was Jamie Zawinski in 1998.
Between that and writing "The Cathedral and the Bizarre" I think the AC is just trolling.
That's reasonable, but TFA says:
Each team is also responsible for development of a new coding project that it will present to the company later this year -- projects may involve creation of a new feature or improved functionality for a Web page within the FreeCause application.
This is not about teaching the company's core product, all employees are expected to produce it. Still, it looks like they're doing it only for 2 hours per week, and it's all supervised. So, maybe it's not a bad idea.
Again, no. Just because it's a single assembly instruction, it doesn't make it atomic -- x86 will do exactly the same as any RISC processor: it will load the memory contents, perform the add and then store the result. Nothing guarantees that another processor will not modify the memory content during the execution of the instruction.
If you want to do it atomically, you'd have to do something like this.
If you remove "The single-cycle rule", then there is no difference between RISC and CISC.
No. With RISC, the only way to access memory is via load/store instructions. The general data processing instructions (arithmetic, comparison, etc.) are always done on registers. That's why instructions are very simple.
On CISC, on the other hand, you can use data from memory without loading it into registers, for example in x86 you can do something like "add eax,dword ptr[ebx]" or "add dword ptr[ebx],eax".
True. I was thinking of the earlier developments, but you're absolutely right that Feynman belongs there.
I also forgot de Broglie.
Pauling is really great, I agree that he should be more well-known (I, myself, am partial to Dirac, who predicted the existence of anti-matter out of pure math). But you have to realize that Newton and (maybe to a lesser degree) Einstein contributions were of another class.
Newton took Galileo's ideas -- that things tend to keep moving if they're left alone -- and built a whole mathematical theory on top of that, inventing calculus in the process. In a sense, it was the beginning of what we today call Physics.
Einstein was the first to notice (and convinced everyone) that the Lorentz transformation is not just a mathematical trick, it's the very way the space and time works. This itself was not that impressive, he was just giving a "new spin" on what everyone had already observed. But then he went on to show that the "right way" to understand gravity is by noting that it's just a side-effect of mass bending the space and time -- this has lots of consequences that were unknown at the time, like gravity bending light, gravity making time pass at different rates, and a lot other stuff, all of which turned out to be right.
Quantum Mechanics and its implications (like the electron shell), on the other hand, were discovered bit by bit by a lot of different people. That's why no one is hugely famous for it (even though there are certainly big names like Planck, Einstein, Bohr, Heisenberg, Born, Schrodinger, Dirac, Pauli, etc.).
When Paul pointed out that Gentiles didn't have to become Jews before becoming Christians, it opened the floodgates to the spirit of the law, not the letter
On the other hand, Jesus said (Matthew 5:17-5:19)
But, I don't pretend to tell Christians what Christianity is about. And of course Jews also have their own reasons not to follow those barbaric rules. I think it's great. My whole point is that Islam should be treated the same way.