Heh. Magic casting characters were weak? Adding an area-effect to touch spells was amazingly cheap, so killing everyone in a 50 foot radius simply required finding one puny guy. And an "open lock 100" with a high fizzle chance was an easy way for a very low level character to steal nearly anything...
OT: Depends on the wireless arena... Chipcon has all of their datasheets available on their website for their 802.15.4 radios. Then again, 802.15.4 is a little more hobbyist friendly than the MESH stuff, looks like.
That's a good point: state machines are often better expressed as diagrams than as code, and automatic translation of them into code is not terribly difficult. If your problem fits well into a state machine framework (some embedded work, network stack, etc) it can be very useful.
While most people think of class hierarchy diagrams when they think of UML, but there are a bunch of other (unrelated, really) types of diagrams that are under the UML umbrella as well.
Still, UPSes, Mobos, and drive controllers can fail in ways to fry all of your RAID drives (well, at least the electronics) at once. Having better quality stuff helps minimize this risk, of course.
And ok, then it is theoretically possible to repair the drive electronics, if it was an electrical failure. But that'll cost you way more than the backup solution would have.
In order to break the laws of a another land, you have to be there at the time. Otherwise, their laws don't apply to you.
Nice thought, but that's not true. There are actions that are against US law no matter where you were when you did them. The US prosecutes "Drug Kingpins" living in foreign countries all the time (think Noriega). I'm pretty sure that this is also true of murder: murdering anyone, anywhere, is against US law. US courts will generally let other countries take jurisdiction, unless there is a clear reason why they shouldn't (i.e. there are no functioning courts in that country).
Most countries will not extradite someone for something that is not a criminal offense in their own country (see Salmon Rushdie not being sent to Iran after they sentenced him to death in absentia). Unless the laws are also made uniform, requiring such extradition for computer crime/non-crimes seems a little scary.
Lessee: there are 1 Billion (1 thousand million for our English friends) possible SSNs. If you give me the MD5 hash of an SSN, I'll have to try 500M possibilities on average to figure out which SSN. I'm guessing that you could do 100K hashes per second, so that's 5K seconds to produce the SSN. An hour or two.
Or, just generate a 16 gigabyte table of all possible SSN hashes....
Still, I suppose, you'd be making yourself a slightly harder target.
Now, if you salted the hash with the name, or something, then you'd be all set. But then you're stuck with the folks who are listed as Joe Shmoe Jr. in one listing and Joseph Shmoe in another.
Actually, no. You can make some fairly simple arguments that limit the density of neutrinos in the universe, and thus show that neutrinos can't be the dark matter used to explain galactic rotation curves or galactic cluster dynamics. (If you've had some Quantum Mechanics: treat them as fermions in a box, ie a Fermi gas. The density requrired leads to a very high temperature for those on the surface of the Fermi sphere.)
Another problem with neutrinos being the dark matter is that they are so light that they would be "hot", ie moving at very nearly the speed of light. As such, they wouldn't be bound to a given mass center (ie a galaxy), and they would tend to wash out density fluctuations rather than causing them (which is another reason why theorists like dark matter: explaining structure formation in the early universe).
The results at Kamiokande that show neutrino mixing (and thus that they have mass) are very interesting, and I expect that someone will eventually get the Nobel prize for that. But, it still doesn't find all of the missing mass.
Oh, and "esoteric, non-euclidian form of matter"? Dark matter is just stuff that interacts via gravity, but not via the electromagnetic and strong forces. Much like neutrinos, in fact (which interact with "normal" matter via the weak force). I'm not sure how this makes anything "non-euclidian".
Yes, the book is mainly about implementing HSMs. One of his classes is called "QHsm", for example. The distinction is that a Statechart is the drawing of the HSM (Samek just uses Visio for that).
I'm a little surprised that the poster didn't use the term, it's not like Samek doesn't use it everywhere in his book.
The first half of the book is just exploring different ways of implementing FSMs/HSMs in C and C++. It doesn't seem like a hard thing to do, but a simplistic version with nested switch/case statements is really a recipe for hard to maintain code.
The sencond half of the book though, explains Samek's "Quantum Framework", of interacting HSMs. I don't think that this is earth-shattering, but it's not the way most people are taught to code in C/C++, and I think it is a neat framework to use, especially in embedded/safety-critical systems.
The article says they're going to stop production of these machines in June. With a month or so for the production pipeline to empty out, that means that they'll be selling these things for another six months at least.
The article says that they were introduced last January. So that's a 1 1/2 year shelf life for a computer line. So what's the big deal? If they don't come out with a new model to replace it, now that would be a big deal.
It's actually not irrelevant at all. At the end of the day, the scientific method is practiced by the "people" (and "experts") that you say don't matter.
If no reputable member of the scientific community (ie expert) believes that this can possibly be true, then none of them will bother trying to replicate the experiment. And, as the advisors to the folks with the money, it probably won't be funded nearly as much as if the experts did believe in it. If it does happen to be true, then it does indeed matter to all of us that these experts take it seriously, since that's the difference between having this become reality in a year or in a few decades.
That said, it doesn't really matter what the average slashdot reader's opinion is, since he/she is not going to replicate the experiment in any event.
Also, I don't mean to imply that I think that the process is bad- it sure beats wasting a lot of everyone's time and money chasing down every crackpot perpetual-motion/free-energy theory that comes along. But, it does lead to situations where, as Pauli said, you need to wait for most of the current scientists to die off before your new really revolutionary theory is accepted by a majority of the scientific community.
That salon article is pretty interesting, but you have to read more than just the first page to get the whole story.
There are a lot of GR corrections that one must make to GPS to get decent accuracy out of it. In fact, there is a bit of a scandal because none of the GPS receiver makers have made public their correction formulae (standard "if we publish then we'll just be helping our competitors" philosophy).
There was a good article in the May edition of Physics Today about relativistic corrections to GPS, unfortunately the web version isn't free. If you're near a university, they should have copies of PT in their physics library.
There are actually two different effects that can cause one observer's time to move faster than anothers:
1) From special relativity, if the observers are moving relative to each other, _each_ will see the other's clock as moving slower than theirs. This causes the twin paradox, which ends up being resolved by...
2) From general relativity, if an observer is "accelerating", their clock moves slower relative to a non-accelerating observer. Note that in GR, staying stationary in a gravitational field is actually "accelerating": You can't distinguish between sitting in an elevator that is in free-fall and an elevator that is floating in space, or between an elevator that is sitting still in the earth's gravity and one that is accelerating upward while in space.
The effect that causes the atomic clocks to show different times is the latter: one plane is 30K feet higher than the other, so assuming that it doesn't do a lot of stopping and starting or high-G turning, its average acceleration is less.
Re:"Up" quarks and "down" quarks.
on
Quark Stars
·
· Score: 1
Although there's a possibility other properties like color may be discovered to make 2 "up" quarks differ
Actually, the Pauli exclusion principle tells you that any two Up, Red, spin up quarks are really identical. In quantum mechanics, they're both described by the same "wave", and the predictions of the theory would be different if they were described by different waves (the way Up and Down quarks are, say). So either something very basic in QM is wrong, or they really are the same.
The same is true of electrons, or even of the atomic nuclei that made up of protons & neutrons (which are in turn made up of quarks).
I don't know of a good Linux C++ book... C++ really isn't the default language for Linux, and it shows, especially in gcc and the standard library. It's gotten alot better in the past few years, but it still has a ways to go. Not to say that Windows is any better... VC++ is even worse (although the APIs are all there).
My advice:
Use the STLPort, not the GNU Standard C++ Library. It actually conforms to the spec, and has fewer bugs. This is huge if you're going to be doing multithreaded programming.
Use the Boost libraries. They're gonna be put into the next generation standard, and it is good stuff.
Databases: check out the OTL (Oracle Template Library) for interfacing with Oracle. I haven't done too much with other DBs. The OTL web page makes it look like a "make money fast" scheme, but it works, and the mainainer apparently has no life, and has nothing better to do than patch it whenever a user whines or Oracle changes versions:)
Makefiles: look at the man/info pages for "make", or look at the (old but still good) O'Reilly book on Make. There's really no difference between C and C++ there.
I haven't done much with GUI dev, so I don't know of a good ref for QT or gtk--. I can say that FLTK has good online docs, and is quite easy to use.
As many others have said elsewhere: the best documentation is usually on the project page for the tool you're interested in. Finding the right tool, though, can be hard.
The point of a programming class is to learn how to program, not to learn how to copy code. If I were hiring somebody fresh out of school who got an A in their C++ programming class, I'd expect that they can actually _code_ in C++, not just _copy_ C++ code.
First year physics
on
This is IT?
·
· Score: 2, Insightful
Yes, heavier objects have more momentum.
They also have more friction stopping them, as the force of friction equals the force down (gravity * mass) times the coefficient of friction.
Since both the momentum and the stopping force are proportional to mass, it cancels out.
Of course, real rubber tires don't act quite like what you learned in high-school physics, but we'll leave that for another day.
If you're gonna flame somebody, try to get a bit of a clue first.
Good explanation, but there's a couple things that are missing. For a somewhat longer explanatory article, people can look at one from Scientific American a few years back.
The importance of the 7 qubits is mainly just the same as the importance of bits in a classical computer: this is essentially the memory capacity of the quantum computer. It is also very important to have extra bits since, as you say, you can use redundancies to keep your computer from decohering (becoming non-quantum). But I remember a talk by a guy using chloroform as the molecule (CHCl3, for 2 qubits), and talking about using Shur's factoring algorithm (the quantum algorithm to factor numbers very quickly) to factor a number like 4. Not very exciting from a mathematical standpoint.
The cute thing about using NMR with organic molecules this way is that different frequencies "talk" to different atoms in the molecule. So you can set or manipulate different qubits using different frequencies. The qubits talk to neighboring qubits as well, through the atomic interactions: this is actually a good thing- it's most of how your calculation gets done (the NMR is essentially I/O). Unfortunately, there's a problem with this approach: there's a physical limit to the size of these molecules (and thus the number of qubits). I think this is where the 15 bit limit came from. It is hard to imagine how you can find arbitrarily large organic molecules where each nuclear spin has an appreciably different resonance frequency, and where the atoms are spaced so that they only talk with
There may be some breakthrough with this latest announcement, but it looks to me like a straightforward extension of previous results.
Luckily there are some other approaches to quantum computers that (if they work) should scale up much more easily.
The present invention has discovered the apparent existence of a new dimension capable of acting as a medium for RF signals. Initial benefits of penetrating this new dimension include sending RF signals faster than the speed of light, extending the effective distance of RF transmitters at the same power radiated, penetrating known RF shielding devices, and accelerating plant growth exposed to the by-product energy of the RF transmissions.
I love how they tack on "accelerating plant growth" right on the end. Classic Kook stuff. Reading more about the plants:
It has been observed by the inventor and witnesses that accelerated plant growth can occur using the present invention.
For accelerated plant growth, first, you need to create a hot surface that is more than 1000 degrees Fahrenheit. Next, you need a strong magnetic field. Only one device is needed for this function. This allows energy from another dimension to influence plant growth.
Later he goes on to talk about how to make this thing with a halogen lamp (honest! That's his heat source) and some wires wrapped around it (the accelerator). So I guess he really could send a "working model" to the USPTO.
I bet you could influence the plant growth even more if you use a grow light in place of the halogen.
Wavelets CAN be theoretically best, depending on the type of the information. Each basis is best at encoding things that look like the basis vectors: thus the Fourier basis is best for encoding sine waves. And for pictures that have localized bumps (like most normal pictures), wavelets that have localized bumps will do best.
Oh, and best here can be rigorously defined: just plot mean squared error vs number of components. The lower, the better.
Also, there are fast wavelet transforms. In fact, some of them are O(N), rather than O(N ln N) as for the FFT. Again though, depends on the wavelets that you're using.
Slashdot Mirror
Heh. Magic casting characters were weak? Adding an area-effect to touch spells was amazingly cheap, so killing everyone in a 50 foot radius simply required finding one puny guy. And an "open lock 100" with a high fizzle chance was an easy way for a very low level character to steal nearly anything...
OT: Depends on the wireless arena... Chipcon has all of their datasheets available on their website for their 802.15.4 radios. Then again, 802.15.4 is a little more hobbyist friendly than the MESH stuff, looks like.
That's a good point: state machines are often better expressed as diagrams than as code, and automatic translation of them into code is not terribly difficult. If your problem fits well into a state machine framework (some embedded work, network stack, etc) it can be very useful.
While most people think of class hierarchy diagrams when they think of UML, but there are a bunch of other (unrelated, really) types of diagrams that are under the UML umbrella as well.
Still, UPSes, Mobos, and drive controllers can fail in ways to fry all of your RAID drives (well, at least the electronics) at once. Having better quality stuff helps minimize this risk, of course.
And ok, then it is theoretically possible to repair the drive electronics, if it was an electrical failure. But that'll cost you way more than the backup solution would have.
In order to break the laws of a another land, you have to be there at the time. Otherwise, their laws don't apply to you.
Nice thought, but that's not true. There are actions that are against US law no matter where you were when you did them. The US prosecutes "Drug Kingpins" living in foreign countries all the time (think Noriega). I'm pretty sure that this is also true of murder: murdering anyone, anywhere, is against US law. US courts will generally let other countries take jurisdiction, unless there is a clear reason why they shouldn't (i.e. there are no functioning courts in that country).
Most countries will not extradite someone for something that is not a criminal offense in their own country (see Salmon Rushdie not being sent to Iran after they sentenced him to death in absentia). Unless the laws are also made uniform, requiring such extradition for computer crime/non-crimes seems a little scary.
One company I heard of was pretty sure they could foam diamond, but were looking for a customer to foot the bill.
Well, if you like RPN, just use "dc" (desk calculator).
You might want to look at the man page before you run it, though.
Lessee: there are 1 Billion (1 thousand million for our English friends) possible SSNs. If you give me the MD5 hash of an SSN, I'll have to try 500M possibilities on average to figure out which SSN. I'm guessing that you could do 100K hashes per second, so that's 5K seconds to produce the SSN. An hour or two.
Or, just generate a 16 gigabyte table of all possible SSN hashes....
Still, I suppose, you'd be making yourself a slightly harder target.
Now, if you salted the hash with the name, or something, then you'd be all set. But then you're stuck with the folks who are listed as Joe Shmoe Jr. in one listing and Joseph Shmoe in another.
Another problem with neutrinos being the dark matter is that they are so light that they would be "hot", ie moving at very nearly the speed of light. As such, they wouldn't be bound to a given mass center (ie a galaxy), and they would tend to wash out density fluctuations rather than causing them (which is another reason why theorists like dark matter: explaining structure formation in the early universe).
The results at Kamiokande that show neutrino mixing (and thus that they have mass) are very interesting, and I expect that someone will eventually get the Nobel prize for that. But, it still doesn't find all of the missing mass.
Oh, and "esoteric, non-euclidian form of matter"? Dark matter is just stuff that interacts via gravity, but not via the electromagnetic and strong forces. Much like neutrinos, in fact (which interact with "normal" matter via the weak force). I'm not sure how this makes anything "non-euclidian".
I'm a little surprised that the poster didn't use the term, it's not like Samek doesn't use it everywhere in his book.
The first half of the book is just exploring different ways of implementing FSMs/HSMs in C and C++. It doesn't seem like a hard thing to do, but a simplistic version with nested switch/case statements is really a recipe for hard to maintain code.
The sencond half of the book though, explains Samek's "Quantum Framework", of interacting HSMs. I don't think that this is earth-shattering, but it's not the way most people are taught to code in C/C++, and I think it is a neat framework to use, especially in embedded/safety-critical systems.
The article says that they were introduced last January. So that's a 1 1/2 year shelf life for a computer line. So what's the big deal? If they don't come out with a new model to replace it, now that would be a big deal.
If no reputable member of the scientific community (ie expert) believes that this can possibly be true, then none of them will bother trying to replicate the experiment. And, as the advisors to the folks with the money, it probably won't be funded nearly as much as if the experts did believe in it. If it does happen to be true, then it does indeed matter to all of us that these experts take it seriously, since that's the difference between having this become reality in a year or in a few decades.
That said, it doesn't really matter what the average slashdot reader's opinion is, since he/she is not going to replicate the experiment in any event.
Also, I don't mean to imply that I think that the process is bad- it sure beats wasting a lot of everyone's time and money chasing down every crackpot perpetual-motion/free-energy theory that comes along. But, it does lead to situations where, as Pauli said, you need to wait for most of the current scientists to die off before your new really revolutionary theory is accepted by a majority of the scientific community.
There are a lot of GR corrections that one must make to GPS to get decent accuracy out of it. In fact, there is a bit of a scandal because none of the GPS receiver makers have made public their correction formulae (standard "if we publish then we'll just be helping our competitors" philosophy).
There was a good article in the May edition of Physics Today about relativistic corrections to GPS, unfortunately the web version isn't free. If you're near a university, they should have copies of PT in their physics library.
There are actually two different effects that can cause one observer's time to move faster than anothers:
1) From special relativity, if the observers are moving relative to each other, _each_ will see the other's clock as moving slower than theirs. This causes the twin paradox, which ends up being resolved by...
2) From general relativity, if an observer is "accelerating", their clock moves slower relative to a non-accelerating observer. Note that in GR, staying stationary in a gravitational field is actually "accelerating": You can't distinguish between sitting in an elevator that is in free-fall and an elevator that is floating in space, or between an elevator that is sitting still in the earth's gravity and one that is accelerating upward while in space.
The effect that causes the atomic clocks to show different times is the latter: one plane is 30K feet higher than the other, so assuming that it doesn't do a lot of stopping and starting or high-G turning, its average acceleration is less.
Actually, the Pauli exclusion principle tells you that any two Up, Red, spin up quarks are really identical. In quantum mechanics, they're both described by the same "wave", and the predictions of the theory would be different if they were described by different waves (the way Up and Down quarks are, say). So either something very basic in QM is wrong, or they really are the same.
The same is true of electrons, or even of the atomic nuclei that made up of protons & neutrons (which are in turn made up of quarks).
My advice:
Use the STLPort, not the GNU Standard C++ Library. It actually conforms to the spec, and has fewer bugs. This is huge if you're going to be doing multithreaded programming.
Use the Boost libraries. They're gonna be put into the next generation standard, and it is good stuff.
Databases: check out the OTL (Oracle Template Library) for interfacing with Oracle. I haven't done too much with other DBs. The OTL web page makes it look like a "make money fast" scheme, but it works, and the mainainer apparently has no life, and has nothing better to do than patch it whenever a user whines or Oracle changes versions :)
Makefiles: look at the man/info pages for "make", or look at the (old but still good) O'Reilly book on Make. There's really no difference between C and C++ there.
I haven't done much with GUI dev, so I don't know of a good ref for QT or gtk--. I can say that FLTK has good online docs, and is quite easy to use.
As many others have said elsewhere: the best documentation is usually on the project page for the tool you're interested in. Finding the right tool, though, can be hard.
Nice troll, so I'll respond :)
The point of a programming class is to learn how to program, not to learn how to copy code. If I were hiring somebody fresh out of school who got an A in their C++ programming class, I'd expect that they can actually _code_ in C++, not just _copy_ C++ code.
Yes, heavier objects have more momentum.
They also have more friction stopping them, as the force of friction equals the force down (gravity * mass) times the coefficient of friction.
Since both the momentum and the stopping force are proportional to mass, it cancels out.
Of course, real rubber tires don't act quite like what you learned in high-school physics, but we'll leave that for another day.
If you're gonna flame somebody, try to get a bit of a clue first.
Yet another occasional BSDer.
Good explanation, but there's a couple things that are missing. For a somewhat longer explanatory article, people can look at one from Scientific American a few years back.
The importance of the 7 qubits is mainly just the same as the importance of bits in a classical computer: this is essentially the memory capacity of the quantum computer. It is also very important to have extra bits since, as you say, you can use redundancies to keep your computer from decohering (becoming non-quantum). But I remember a talk by a guy using chloroform as the molecule (CHCl3, for 2 qubits), and talking about using Shur's factoring algorithm (the quantum algorithm to factor numbers very quickly) to factor a number like 4. Not very exciting from a mathematical standpoint.
The cute thing about using NMR with organic molecules this way is that different frequencies "talk" to different atoms in the molecule. So you can set or manipulate different qubits using different frequencies. The qubits talk to neighboring qubits as well, through the atomic interactions: this is actually a good thing- it's most of how your calculation gets done (the NMR is essentially I/O). Unfortunately, there's a problem with this approach: there's a physical limit to the size of these molecules (and thus the number of qubits). I think this is where the 15 bit limit came from. It is hard to imagine how you can find arbitrarily large organic molecules where each nuclear spin has an appreciably different resonance frequency, and where the atoms are spaced so that they only talk with
There may be some breakthrough with this latest announcement, but it looks to me like a straightforward extension of previous results.
Luckily there are some other approaches to quantum computers that (if they work) should scale up much more easily.
From the application:
The present invention has discovered the apparent existence of a new dimension capable of acting as a medium for RF signals. Initial benefits of penetrating this new dimension include sending RF signals faster than the speed of light, extending the effective distance of RF transmitters at the same power radiated, penetrating known RF shielding devices, and accelerating plant growth exposed to the by-product energy of the RF transmissions.
I love how they tack on "accelerating plant growth" right on the end. Classic Kook stuff.
Reading more about the plants:
It has been observed by the inventor and witnesses that accelerated plant growth can occur using the present invention.
For accelerated plant growth, first, you need to create a hot surface that is more than 1000 degrees Fahrenheit. Next, you need a strong magnetic field. Only one device is needed for this function. This allows energy from another dimension to influence plant growth.
Later he goes on to talk about how to make this thing with a halogen lamp (honest! That's his heat source) and some wires wrapped around it (the accelerator). So I guess he really could send a "working model" to the USPTO.
I bet you could influence the plant growth even more if you use a grow light in place of the halogen.
Wavelets CAN be theoretically best, depending on the type of the information. Each basis is best at encoding things that look like the basis vectors: thus the Fourier basis is best for encoding sine waves. And for pictures that have localized bumps (like most normal pictures), wavelets that have localized bumps will do best.
Oh, and best here can be rigorously defined: just plot mean squared error vs number of components. The lower, the better.
Also, there are fast wavelet transforms. In fact, some of them are O(N), rather than O(N ln N) as for the FFT. Again though, depends on the wavelets that you're using.
Papa Smurf aside, I always figured that AFS was developed at CMU, and named after Andrew Carnegie.
Just a guess, though.