That would be relevant if the unused code was very tightly mixed with the used code.
In this case, if you look at the way they detect unused code, you'll see that all the code that was removed consists of whole methods and sometimes whole classes. This means that the vast majority of the code that was removed would probably never be loaded in the instruction cache anyway -- very small methods tend to be inlined by the compiler and so don't cause surrounding (and potentially unused) code to be loaded to the cache.
There's a big difference between forcing everything to go through to committee, and having a strong leadership but with a "release early, release often" approach and seeking input.
I don't understand your position. Are you suggesting no one should develop anything except in the open and accepting suggestions from anyone, or merely that it might be a good idea to be considered?
Javascript surely would have benefited from input from other people. OR it would never have existed, because Netscape and others would be too busy deciding the features of the language to ship it with Netscape 2.0, and then someone else would do something else first. And maybe the CERN people, after receiving input from other parties, would think better and decide that a text-only protocol really was too simple to work for any serious data transfer, what were they thinking?
In the end, it doesn't seem productive to try to dictate how innovation should happen. And we certainly shouldn't prevent people from doing new things a certain way because they *might* do it wrong.
Much like the US Constitution provides us the concept of "Innocent until proven guilty", I assume all works are Copyable unless proven protected.
Your assumption is wrong. From Wikipedia (my emphasis):
In all countries where the Berne Convention standards apply, copyright is automatic, and need not be obtained through official registration with any government office.
[...]
In 1989, the U.S. enacted the Berne Convention Implementation Act. [...] As a result, the use of copyright notices has become optional to claim copyright, because the Berne Convention makes copyright automatic.
Re:Is Google trying to fragment web?
on
MAME Running In Chrome
·
· Score: 3, Informative
A standard should be discussed and developed in partnership with the other parties [...]
A lot of standards we have today (Ethernet, HTTP, Javascript, to name a few) were born out of small groups of individuals with no input from other parties, at first. In each case, the new technology became a standard only after it was implemented and proven really useful.
Demanding that every new technology comes out of a committee is insane: it's a great way to stifle innovation.
Now, there is a danger about non-standardized technology: it can be very bad if the creator tries to use it to prevent everyone else from competing in the same area, like was the case with ActiveX. However, I don't see how that would be the case with NaCl: anyone can see how it works, take the code, make it work in any browser (or anything else), and change it at will with no strings attached -- it has a BSD license. It could be a problem if Google refuses to participate in an effort to make a standard out of it, when (and if) it becomes widely used and has competing implementations. But it's way too early to be worried about it -- currently, NaCl is a mere curiosity.
He said, in response to your mention of Stalin and Mao (paraphrased): "Atheists committing atrocities doesn't tell anything about atheism; similar atrocities were also committed by religious people (e.g. Hitler)."
And your answer was (paraphrased) "I don't accept that; Hitler being religious shouldn't be used to blame religion".
You used the same exact form of reasoning, yet you don't accept it?
Further, the main point of his phrase which you seem to disagree so much (since you called it a lie):
I think we can, however, all agree that when religion is given the only say, that the results are uniformly horrifying.
Is that no one should be forced to accept religion ("religion [shouldn't be] given the only say"), because anything that forces everyone accept one doctrine of thought is eventually abused pretty badly, as we have seen many times. What's so controversial about that? Do you really think religion should be given the only say?
Your points 1-4 seem valid to some extent (even though, for example, I have never seen any difference in google search or gmail between Firefox and Chrome, and I use all combinations), but I think saying that's even remotely similar to what MS did is blowing things out of proportion. I think (2) and (3) are somewhat natural for anyone developing a web browser. For example, Mozilla did essentially (3) and allowed people to do (2) with Mozilla (now Seamonkey) and then Firefox for the longest time, in the form of easy-to-write extensions -- for instance, there are a *lot* of site-specific extensions for Firefox that completely change the way a site appears and behaves. Maybe no one cares because the amount of users of these extensions is minimal? And (4) is indeed worrying; this is the first I hear of it (I don't come even close to using any of the software that bundles Chrome). I looked around, and this actually changes my mind about some things; I'll have start following these kinds of thing and see what comes out of it.
To the next point, it would indeed be nicer if Google allowed more participation from the community in V8, but I don't think too much of it. As long as there's a workable option to fork the code *and* keep receiving updates, I don't think anyone who is serious about developing a web browser should have too much difficulty in merging updates from Google or other sources. Things like these happen a lot in other contexts -- for example, most Linux distributions do that in some form or another in a much larger scale (but I can see that it takes a lot of people to do it).
The link you posted seems related specifically to Android, not Chrome, so I'll take that with a grain of salt. It *might* give an insight into Google's general thinking, or just their strategy regarding Android.
With all that said, in the end, I can't shake the feeling that most of these arguments are a tad paranoid, and there's a lot of speculation involved. Maybe the writing is on the wall and I simply can not see it. I guess I'll find out in the next few years.
And I agree that a browser monopoly in the hands of Google (or anyone else) would be bad. I just wish the Mozilla people would stop making Firefox increasingly unlikable:)
So, the fact that V8 runs better some javascript in the wild has nothing to do with Google being evil. And "self-interested" seems a little uncharitable, seeing that V8 has a BSD license -- in other words, they're going out of their way to make the javascript engine they develop and use available to anyone interested, not even requiring modifications to be given back to them.
Google might be evil in other ways, but can we agree that the original point of this thread -- comparing what Google is doing with Chrome (which, by the way, ALSO has an open-source version) with what MS did with IE6 -- is ludicrous?
That would be bad, but I find it hard to believe. Do you have an example of an optimization in V8 that makes some code run *slower* than it would without that optimization? It seems more likely that other engines are just better than V8 for some types of code; it's hardly surprising that V8's optimizations are focused on the type of code that appears in Google's libraries.
I didn't say anything about preventing Google from doing whatever they want. I'm not sure where you got that.
I'm also not sure where you got your "evil" bit from. I made a factual statement, with no value judgments attached to it.
Sorry, I misunderstood you. I thought you were supporting the argument that Google is trying to do something similar to what Microsoft did back in the days of Netscape vs. IE, like the thread's parent comment (the "evil" bit came from there).
Google's JavaScript libraries are purposefully written to run faster on V8 specifically, often at the expense of performance in other browsers. And at the same time, V8 is written to run the code patterns those libraries use faster, often at the expense of other code patterns used elsewhere.
That might be a problem if V8 wasn't open source, or if Google was preventing anyone from seeing how it works. As it stands, your argument makes about as much sense as saying that Linux and GCC were evil because earlier versions of Linux couldn't be compiled on anything but GCC. The thing is, nothing prevented anyone from changing Linux to compile in other compilers, or studying GCC to learn how to make other compilers compile Linux (and, indeed, today it's possible to compile Linux with other compilers).
So, if Google's libraries do indeed run that much faster in V8, what's preventing anyone from implementing these optimizations in other javascript engines, or writing their libraries to run fast on V8? Is it reasonable to prevent Google from optimizing anything just because other browsers and libraries would then be slower?
You're right in saying that the system is hopelessly entangled with the environment, but there is no way of telling how it is entangled without observing either the system or the environment.
We have to thread carefully, here. There's no way for person X to tell if it's entangled before person X observes the system, but that has nothing to do with quantum mechanics: it's true in classical mechanics, so I don't think that's what you mean. I assume, then, that you mean that the physics of the system itself doesn't change at all when it becomes entangled with the environment (but before anyone observes it). That's not true. The change is exactly that, after becoming entangled with the environment, the branches of the wavefunction can't interfere with each other.
Let me give a more concrete example: think of a photon in the double-slit experiment. We have one branch of the wavefunction for the photon going through each slit. If the photon is completely isolated throughout the experiment, the branches interfere, causing the characteristic interference pattern.
Now change things so that there's an electron close to the slit in the left. Further, the electron is placed in such a way that, if the photon goes through that slit, it flips the electron's spin (becoming entangled with it). In this case, the branches of the wavefunction of the photon can't interfere anymore, because the branches became orthogonal: one of them has the electron with the original spin, and the other has the electron with the spin flipped.
The whole point is that, in the second experiment, the interference is lost without anyone observing anything. With decoherence, the same thing happens: interference is lost before regardless of anyone observing anything. (You could argue that the electron observed the photon, thus causing the collapse of the wavefunction; and that's a valid interpretation. But then, the same argument can be done with decoherence: the environment observed the system, causing the collapse of the wavefunction, and so no further observation is required).
There's no reason why you should consider the state to have de-cohered in the absence of your conscious perception of it. None whatsoever!
Except in everyone's definition of "decoherence". For example, Wikipedia says:
Decoherence occurs when a system interacts with its environment in a thermodynamically irreversible way.
"Thermodynamically irreversible" means that it's practically impossible to make the environment go back to the exact state it was when it interacted with the system, much like it's practically impossible to unscramble an egg. This means that the system is hopelessly entangled with the environment and can never go back to the way it was before, when the branches of its wavefunction could interfere with each other. This has a real physical effect that happens regardless of whether or not someone is watching.
The way people often describe quantum decoherence is that an "observation" occurs that "collapses the wave function" and causes a superposition to converge to a single classical state. But I really think that's a misleading explanation. For one thing, surely the same phenomena occurred long before there were any intelligent observers [...]
I can't say if it's misleading or not (it might mislead someone...), but it doesn't sound misleading to me. But an important point is that most people nowadays accept that "observation" doesn't require (and has nothing to do with) an intelligent observer. A photon can "observe" a system just as well as a person.
Also, there are many interpretations of quantum mechanics where the collapse of the wavefunction is not as fundamental as in Copenhagen. For example, here's an account of an extension of the Schrödinger's cat experiment according to one such interpretation: Schrödinger is locked in his lab looking at the box that houses the famous cat that is in a superposition of "dead" and "alive". When he opens the box, he observes the cat, and collapses the cat's wavefunction to either dead or alive. But Dirac, who is outside the lab, describes the situation inside the lab as a superposition of "dead cat + sad Schrödinger" and "live cat + happy Schrödinger". When Dirac opens the door of the lab, he observes the situation and collapses (cat + Schrödinger)'s wavefunction to either "dead+sad" or "alive+happy". These "observations" are completely arbitrary, and in fact, you could imagine one such "observation" for each photon that interacts with the system for the first time.
[...] and secondly, scientists have observed things in states of quantum superposition WITHOUT causing decoherence.
I think there's some confusion here. An observation necessarily causes decoherence, because after the observation, a description of the observed system must necessarily include the different states of the observer in each branch of the wavefunction, so the branches become orthogonal. (If you believe in the collapse of the wavefunction, you just throw away all but one of the branches, but wither way, there's decoherence.) Well, technically, you could have the observer with the same state in some of the branches, but that would mean that there would be no way for the observer to distinguish between these branches, so there would be no "observation" for these branches. I think this is getting too technical, so I'll stop here, but the point is: if the observer can distinguish between two states of the observed system, then there's decoherence.
The way think of it (as a total amateur in the area) is that rather than the wave function representing probabilities of states, it represents the degrees to which something is in all of those states. An "observation" is just like many other interactions with the environment that change those probabilities (or degrees of state).
That's (part of) another perfectly fine interpretation of quantum mechanics.
Then there's the question of why subatomic particles (and some larger things) can be in states of quantum superposition, while larger things cannot. Penrose had a suggestion here. It's gravity. The more massive you are, the less your superimposed states can diverge from one another. Even a planet is in a state of superposition, but all of those states overlap so much relative to the dimensions of the object that you cannot distinguish them.
Well, let's be fair here: first of all, there's no agreement that larger things cannot be in a superposition of states -- many people believe it's just a matter of isolating the system well enough. Secondly, Pen
OK, I should have said "as far as we know, it's not impossible, in principle, to simulate [...]". The point I was trying to make is this: a qubit (which exactly encodes the state of a spin of an electron) is fundamentally different than a bit, but it's still possible to simulate the evolution of the state of an electron spin in a classical computer.
And trying to distinguish "simulating" and "creating" in this context is begging the question. This whole discussion is about whether the evolution of universe can be understood as running a computer program.
Another example: take a random vector V of 300 values and consider the subset-sum problem: does it exist a partition of V into two subsets A and B such that sum(A) = sum(B). This is a known NP-hard problem. Solving this problem for a given vector V only once would require much more energy than exists in the entire visible universe, for any physical computer. Do the math yourself as an exercise...
I don't see what's the point of this example. Where in the universe do you see hard instances of the subset-sum problem (or any other NP-hard problem) of that size being solved?
If there is a smallest fundamental particle in the universe, it is binary.
I agree with the rest of what you said, but this is not really true. As far as we know, the "logic" of the universe is not classical: there are some fundamental properties of particles that can't be reduced to a single bit. See, for instance the spin of an electron.
Despite that, as far as we know it's still possible, in principle, to simulate anything in the universe (including quantum mechanics, which includes the electron spin) in a classical computer (to any precision you'd like). So your main point still stands.
"Random" and "entropy" are already used in computer science with the meanings you seem to not want them to have (likethis). Maybe someone should complain to the president of computer science. (I'm sorry, I couldn't resist. This discussion is too silly.)
It's not my theory; maybe you heard about a guy named Kolmogorov that lived in the last century? I bet the great mind of Robert Coveyou studied a lot of his theory:).
But, more seriously, of course a random source will output compressible data sometimes. What happens is this: as you collect more output from a truly random source, the probability of it being compressible goes to zero very fast.
But the point is that it *is* useful to distinguish between "false" and "true" randomness, otherwise it wouldn't be true that "the generation of random numbers is too important to be left to chance".
Distinguishing between "false" and "true" randomness is pointless.
Not really, it's done all the time for many different purposes.
Take, for example, how computer scientists define it: roughly, a sequence is random if it can't be compressed, that is, any (program+data) that generates it must be at least as large than the sequence itself. It distinguishes between "random" and "not having enough information to predict it": it doesn't matter if it looks random to YOU; if it could in principle be compressed, it's not random.
That's not pointless hair splitting, it has real consequences for many areas of computer science, some very practical (cryptography, for example).
Do you want to claim that breaking entanglement is not instantly? Sorry, but then you are a bit out of the loop. It was never assumed that it took any time to be observed. It was assumed "it could not be observed". And this later point has changed in recent history. It was always clear that breaking entanglement has its "effect" transported instanly regardless of distance. That is not my "interpretration" but what is tought in school since ever.
"Breaking the entanglement is instantaneous" is a question that, currently, nobody knows the answer, but that's because this question is about the interpretation of quantum mechanics, and can't be settled, as far as we know, with experiments. What people do believe, though, is that even if breaking the entanglement is "instantaneous" (I write it in quotes because that can't even be made precise with current physics, as I've explained), no information can be sent "instantaneously". I'm curious to see this article on Scientific American that gave you the impression that this is not so.
Some interpretations of quantum mechanics try to explain things in terms of faster than light effects (that's what Einstein derisively called "spooky action"), but these are just interpretations of the math and the outcomes of the experiments. There's no consensus of whether it actually happens or not. And, more importantly, not even people who use these interpretations think that these "effects" can be used to send information faster than light. It should be clear how this works once you understand the math and experiments.
I've resisted doing this so far, because I didn't want to sound disrespectful, but it seems that you have a wrong impression about entanglement. Do you know how the mathematics of entanglement works, or just things you read in science popularization articles? There's an excellent series of introductory lectures from Stanford University in Youtube that teaches the basics of quantum entanglements starting here. The lecturer is a very famous physicist, Leonard Susskind. If you watch the whole thing (and follow the math explanations carefully), you'll notice that he never talks about "instantaneous" anything, and it's there's no reason to do so. And the math isn't too hard to follow, you only need to know basic algebra, complex numbers and matrices.
Regarding your strange attempt to citate me: 'but "it's ok because you had to send the particles beforehand"'... I did not say that. Sigh. How hard is it to read? I said: some phyics say this! And the "it is ok" is not making the point.
Sorry, I misunderstood you. I had never seen it before, and I thought you believed this.
I don't know if it is "confirmed" but I assumed that quantum secured communication where wiretapping can be detected as certain properties of entangeld photons change supports this.
I't true that it's possible to detect eavesdropping in quantum secure channels, but that's due to the uncertainty principle and the no cloning theorem, not entanglement. It's almost impossible to really understand these intricate things, though, without understanding the underlying math.
Sigh... I don't think we're going to make any progress.
I've explained to you that in current physics, it doesn't make any sense to talk about "instant transmission" in the sense you're talking about. I've explained that in the current model for entanglement, there's no need or place for any such "instant transmission". Yet you keep insisting on a non-standard model for entanglement where information is sent faster than light, but "it's ok because you had to send the particles beforehand" (which, I have mentioned, is not only inconsistent with current physics but seems silly).
As I said, Scientific American is a magazine that does science popularization. While sometimes some articles are written by researchers, it's mostly written by journalists. If you're looking for papers with the current research on the foundations of quantum mechanics, you should check peer-reviewed physics journals; there's a nice compilation in Wikipedia. There's also Arxiv, but the papers there are not necessarily peer-reviewed.
Quantum encrypted channels work on the exact same principle and are already in general use. I don't see your point.
Quantum encryption, as all other established results that use entanglement (like quantum teleportation, etc.), don't presuppose or rely on entanglement being able to send information faster than light.
Also I assumed Scientific American was a respected peer reviewed physics magazine. Am I mistaken here?
Scientific American is a magazine that does science popularization, its articles are nowhere near the level of rigor necessary for a peer-reviewed physics journal.
I have read that in Scientific American, quite a few years ago.
Well, as great as Scientific American is, can we at least agree that it publishes a lot of stories about non-established and speculative theories? I'm talking about papers by real physicists that have been peer-reviewed and referenced by other physicists.
It does. As most physics accept that such watching on entanglements is/might be possible. However the general argument is: the first movement counts. Regarding "causality" at least.
I doubt that "most physics accept" that, at least in the terms you describe. Within one of the most basic foundations of modern physics (Special Relativity), it doesn't even make sense to talk about which came first between A and B if B is outside the light cone of A (that is, if light from A can't reach B), so your argument can't even be made precise within current physics. That's not a pointless objection: in your scenario, as far as Special Relativity goes, it would be possible to have one observer concluding that Mars measured first, and another one concluding that Earth measured first.
It is possible to propose new alternative theories, of course. But it's usually best to suspend judgement until they have been made precise and published for general scrutiny. That's why I wanted references to papers; I wanted see the reactions from real physicists.
[...] it was believed you can not watch one particel and notice when the entanglement is broken by the other one. This is disproved by experiments last century.
Can you post (established) references, please? The only thing I heard about that is a paper from B. Dopfer in 1999 (I can't find a link right now) where they supposedly were able to stop interference by measuring entangled particles. I have seen a few critics, but mostly the lack of references from other papers is what makes me suspicious about it: surely this would be huge news, and a lot more people would be talking about it, no?
And in you scenario, nothing depends on anyone's definition of "faster than light". If you can make information from Earth (at point A) available in Mars (at point B) before light has time to travel from A to B, there's no discussion: there is faster than light communication, regardless of how you prepared the whole stuff. Claiming otherwise would be as silly as claiming that communication via radio is no faster than a train because you had to send the radio via train in the first place.
Yes, and my point is: it's no sillier than suggesting that quantum entanglement can be used to transmit information faster than light.
Really. In the current model of quantum entanglement, which is part of the foundations of quantum mechanics and exists more or less unchanged since (at least) the 1950s, it's very clear that information (or anything else) can't travel faster than light. Beyond that, there have been lots of experiments testing and exploring all subtleties of entanglement, since the 1970s. All the results conform in all aspects to the "standard" theory for quantum mechanics. All phenomena that we have been able to observe fit the current theories.
You could say "that doesn't mean that it's impossible, maybe some other experiments will find something someday", and that's true. But that's also true of conservation of energy, or the ontological status of the Easter Bunny (there's a phrase you don't head often:)).
There's a misconception that I've noticed in some people (I don't know if it includes you): the impression that people who think that quantum computing is an attainable goal also believe that entanglement must have some power that breaks the known laws of physics (for example, the speed of light limit). Usually, though, people with this misconception come to believe that quantum computing must be impossible (which, ironically, is the opposite of what you suggest -- that not only quantum computing is possible, but that entanglement does have the power to send information faster than light)
As far as our current understanding goes, though, none of that is true. Quantum computing is firmly founded in the current theory of quantum mechanics, respecting special relativity and all (i.e., the limit of the speed of light).
Slashdot Mirror
That would be relevant if the unused code was very tightly mixed with the used code.
In this case, if you look at the way they detect unused code, you'll see that all the code that was removed consists of whole methods and sometimes whole classes. This means that the vast majority of the code that was removed would probably never be loaded in the instruction cache anyway -- very small methods tend to be inlined by the compiler and so don't cause surrounding (and potentially unused) code to be loaded to the cache.
There's a big difference between forcing everything to go through to committee, and having a strong leadership but with a "release early, release often" approach and seeking input.
I don't understand your position. Are you suggesting no one should develop anything except in the open and accepting suggestions from anyone, or merely that it might be a good idea to be considered?
Javascript surely would have benefited from input from other people. OR it would never have existed, because Netscape and others would be too busy deciding the features of the language to ship it with Netscape 2.0, and then someone else would do something else first. And maybe the CERN people, after receiving input from other parties, would think better and decide that a text-only protocol really was too simple to work for any serious data transfer, what were they thinking?
In the end, it doesn't seem productive to try to dictate how innovation should happen. And we certainly shouldn't prevent people from doing new things a certain way because they *might* do it wrong.
Much like the US Constitution provides us the concept of "Innocent until proven guilty", I assume all works are Copyable unless proven protected.
Your assumption is wrong. From Wikipedia (my emphasis):
In all countries where the Berne Convention standards apply, copyright is automatic, and need not be obtained through official registration with any government office.
[...]
In 1989, the U.S. enacted the Berne Convention Implementation Act. [...] As a result, the use of copyright notices has become optional to claim copyright, because the Berne Convention makes copyright automatic.
A standard should be discussed and developed in partnership with the other parties [...]
A lot of standards we have today (Ethernet, HTTP, Javascript, to name a few) were born out of small groups of individuals with no input from other parties, at first. In each case, the new technology became a standard only after it was implemented and proven really useful.
Demanding that every new technology comes out of a committee is insane: it's a great way to stifle innovation.
Now, there is a danger about non-standardized technology: it can be very bad if the creator tries to use it to prevent everyone else from competing in the same area, like was the case with ActiveX. However, I don't see how that would be the case with NaCl: anyone can see how it works, take the code, make it work in any browser (or anything else), and change it at will with no strings attached -- it has a BSD license. It could be a problem if Google refuses to participate in an effort to make a standard out of it, when (and if) it becomes widely used and has competing implementations. But it's way too early to be worried about it -- currently, NaCl is a mere curiosity.
For those interested, the last draft before the official version is available for free here: http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1570.pdf
And Hitler was a Catholic -- so what?
No. I don't accept this line of reasoning at all.
I don't understand your argument.
He said, in response to your mention of Stalin and Mao (paraphrased): "Atheists committing atrocities doesn't tell anything about atheism; similar atrocities were also committed by religious people (e.g. Hitler)."
And your answer was (paraphrased) "I don't accept that; Hitler being religious shouldn't be used to blame religion".
You used the same exact form of reasoning, yet you don't accept it?
Further, the main point of his phrase which you seem to disagree so much (since you called it a lie):
I think we can, however, all agree that when religion is given the only say, that the results are uniformly horrifying.
Is that no one should be forced to accept religion ("religion [shouldn't be] given the only say"), because anything that forces everyone accept one doctrine of thought is eventually abused pretty badly, as we have seen many times. What's so controversial about that? Do you really think religion should be given the only say?
Your points 1-4 seem valid to some extent (even though, for example, I have never seen any difference in google search or gmail between Firefox and Chrome, and I use all combinations), but I think saying that's even remotely similar to what MS did is blowing things out of proportion. I think (2) and (3) are somewhat natural for anyone developing a web browser. For example, Mozilla did essentially (3) and allowed people to do (2) with Mozilla (now Seamonkey) and then Firefox for the longest time, in the form of easy-to-write extensions -- for instance, there are a *lot* of site-specific extensions for Firefox that completely change the way a site appears and behaves. Maybe no one cares because the amount of users of these extensions is minimal? And (4) is indeed worrying; this is the first I hear of it (I don't come even close to using any of the software that bundles Chrome). I looked around, and this actually changes my mind about some things; I'll have start following these kinds of thing and see what comes out of it.
To the next point, it would indeed be nicer if Google allowed more participation from the community in V8, but I don't think too much of it. As long as there's a workable option to fork the code *and* keep receiving updates, I don't think anyone who is serious about developing a web browser should have too much difficulty in merging updates from Google or other sources. Things like these happen a lot in other contexts -- for example, most Linux distributions do that in some form or another in a much larger scale (but I can see that it takes a lot of people to do it).
The link you posted seems related specifically to Android, not Chrome, so I'll take that with a grain of salt. It *might* give an insight into Google's general thinking, or just their strategy regarding Android.
With all that said, in the end, I can't shake the feeling that most of these arguments are a tad paranoid, and there's a lot of speculation involved. Maybe the writing is on the wall and I simply can not see it. I guess I'll find out in the next few years.
And I agree that a browser monopoly in the hands of Google (or anyone else) would be bad. I just wish the Mozilla people would stop making Firefox increasingly unlikable :)
So, the fact that V8 runs better some javascript in the wild has nothing to do with Google being evil. And "self-interested" seems a little uncharitable, seeing that V8 has a BSD license -- in other words, they're going out of their way to make the javascript engine they develop and use available to anyone interested, not even requiring modifications to be given back to them.
Google might be evil in other ways, but can we agree that the original point of this thread -- comparing what Google is doing with Chrome (which, by the way, ALSO has an open-source version) with what MS did with IE6 -- is ludicrous?
They make other (non-Google) sites run slower?
That would be bad, but I find it hard to believe. Do you have an example of an optimization in V8 that makes some code run *slower* than it would without that optimization? It seems more likely that other engines are just better than V8 for some types of code; it's hardly surprising that V8's optimizations are focused on the type of code that appears in Google's libraries.
I didn't say anything about preventing Google from doing whatever they want. I'm not sure where you got that.
I'm also not sure where you got your "evil" bit from. I made a factual statement, with no value judgments attached to it.
Sorry, I misunderstood you. I thought you were supporting the argument that Google is trying to do something similar to what Microsoft did back in the days of Netscape vs. IE, like the thread's parent comment (the "evil" bit came from there).
Google's JavaScript libraries are purposefully written to run faster on V8 specifically, often at the expense of performance in other browsers. And at the same time, V8 is written to run the code patterns those libraries use faster, often at the expense of other code patterns used elsewhere.
That might be a problem if V8 wasn't open source, or if Google was preventing anyone from seeing how it works. As it stands, your argument makes about as much sense as saying that Linux and GCC were evil because earlier versions of Linux couldn't be compiled on anything but GCC. The thing is, nothing prevented anyone from changing Linux to compile in other compilers, or studying GCC to learn how to make other compilers compile Linux (and, indeed, today it's possible to compile Linux with other compilers).
So, if Google's libraries do indeed run that much faster in V8, what's preventing anyone from implementing these optimizations in other javascript engines, or writing their libraries to run fast on V8? Is it reasonable to prevent Google from optimizing anything just because other browsers and libraries would then be slower?
You're right in saying that the system is hopelessly entangled with the environment, but there is no way of telling how it is entangled without observing either the system or the environment.
We have to thread carefully, here. There's no way for person X to tell if it's entangled before person X observes the system, but that has nothing to do with quantum mechanics: it's true in classical mechanics, so I don't think that's what you mean. I assume, then, that you mean that the physics of the system itself doesn't change at all when it becomes entangled with the environment (but before anyone observes it). That's not true. The change is exactly that, after becoming entangled with the environment, the branches of the wavefunction can't interfere with each other.
Let me give a more concrete example: think of a photon in the double-slit experiment. We have one branch of the wavefunction for the photon going through each slit. If the photon is completely isolated throughout the experiment, the branches interfere, causing the characteristic interference pattern.
Now change things so that there's an electron close to the slit in the left. Further, the electron is placed in such a way that, if the photon goes through that slit, it flips the electron's spin (becoming entangled with it). In this case, the branches of the wavefunction of the photon can't interfere anymore, because the branches became orthogonal: one of them has the electron with the original spin, and the other has the electron with the spin flipped.
The whole point is that, in the second experiment, the interference is lost without anyone observing anything. With decoherence, the same thing happens: interference is lost before regardless of anyone observing anything. (You could argue that the electron observed the photon, thus causing the collapse of the wavefunction; and that's a valid interpretation. But then, the same argument can be done with decoherence: the environment observed the system, causing the collapse of the wavefunction, and so no further observation is required).
There's no reason why you should consider the state to have de-cohered in the absence of your conscious perception of it. None whatsoever!
Except in everyone's definition of "decoherence". For example, Wikipedia says:
Decoherence occurs when a system interacts with its environment in a thermodynamically irreversible way.
"Thermodynamically irreversible" means that it's practically impossible to make the environment go back to the exact state it was when it interacted with the system, much like it's practically impossible to unscramble an egg. This means that the system is hopelessly entangled with the environment and can never go back to the way it was before, when the branches of its wavefunction could interfere with each other. This has a real physical effect that happens regardless of whether or not someone is watching.
The way people often describe quantum decoherence is that an "observation" occurs that "collapses the wave function" and causes a superposition to converge to a single classical state. But I really think that's a misleading explanation. For one thing, surely the same phenomena occurred long before there were any intelligent observers [...]
I can't say if it's misleading or not (it might mislead someone...), but it doesn't sound misleading to me. But an important point is that most people nowadays accept that "observation" doesn't require (and has nothing to do with) an intelligent observer. A photon can "observe" a system just as well as a person.
Also, there are many interpretations of quantum mechanics where the collapse of the wavefunction is not as fundamental as in Copenhagen. For example, here's an account of an extension of the Schrödinger's cat experiment according to one such interpretation: Schrödinger is locked in his lab looking at the box that houses the famous cat that is in a superposition of "dead" and "alive". When he opens the box, he observes the cat, and collapses the cat's wavefunction to either dead or alive. But Dirac, who is outside the lab, describes the situation inside the lab as a superposition of "dead cat + sad Schrödinger" and "live cat + happy Schrödinger". When Dirac opens the door of the lab, he observes the situation and collapses (cat + Schrödinger)'s wavefunction to either "dead+sad" or "alive+happy". These "observations" are completely arbitrary, and in fact, you could imagine one such "observation" for each photon that interacts with the system for the first time.
[...] and secondly, scientists have observed things in states of quantum superposition WITHOUT causing decoherence.
I think there's some confusion here. An observation necessarily causes decoherence, because after the observation, a description of the observed system must necessarily include the different states of the observer in each branch of the wavefunction, so the branches become orthogonal. (If you believe in the collapse of the wavefunction, you just throw away all but one of the branches, but wither way, there's decoherence.) Well, technically, you could have the observer with the same state in some of the branches, but that would mean that there would be no way for the observer to distinguish between these branches, so there would be no "observation" for these branches. I think this is getting too technical, so I'll stop here, but the point is: if the observer can distinguish between two states of the observed system, then there's decoherence.
The way think of it (as a total amateur in the area) is that rather than the wave function representing probabilities of states, it represents the degrees to which something is in all of those states. An "observation" is just like many other interactions with the environment that change those probabilities (or degrees of state).
That's (part of) another perfectly fine interpretation of quantum mechanics.
Then there's the question of why subatomic particles (and some larger things) can be in states of quantum superposition, while larger things cannot. Penrose had a suggestion here. It's gravity. The more massive you are, the less your superimposed states can diverge from one another. Even a planet is in a state of superposition, but all of those states overlap so much relative to the dimensions of the object that you cannot distinguish them.
Well, let's be fair here: first of all, there's no agreement that larger things cannot be in a superposition of states -- many people believe it's just a matter of isolating the system well enough. Secondly, Pen
OK, I should have said "as far as we know, it's not impossible, in principle, to simulate [...]". The point I was trying to make is this: a qubit (which exactly encodes the state of a spin of an electron) is fundamentally different than a bit, but it's still possible to simulate the evolution of the state of an electron spin in a classical computer.
And trying to distinguish "simulating" and "creating" in this context is begging the question. This whole discussion is about whether the evolution of universe can be understood as running a computer program.
Another example: take a random vector V of 300 values and consider the subset-sum problem: does it exist a partition of V into two subsets A and B such that sum(A) = sum(B). This is a known NP-hard problem. Solving this problem for a given vector V only once would require much more energy than exists in the entire visible universe, for any physical computer. Do the math yourself as an exercise...
I don't see what's the point of this example. Where in the universe do you see hard instances of the subset-sum problem (or any other NP-hard problem) of that size being solved?
If there is a smallest fundamental particle in the universe, it is binary.
I agree with the rest of what you said, but this is not really true. As far as we know, the "logic" of the universe is not classical: there are some fundamental properties of particles that can't be reduced to a single bit. See, for instance the spin of an electron.
Despite that, as far as we know it's still possible, in principle, to simulate anything in the universe (including quantum mechanics, which includes the electron spin) in a classical computer (to any precision you'd like). So your main point still stands.
"Random" and "entropy" are already used in computer science with the meanings you seem to not want them to have (like this). Maybe someone should complain to the president of computer science. (I'm sorry, I couldn't resist. This discussion is too silly.)
It's not my theory; maybe you heard about a guy named Kolmogorov that lived in the last century? I bet the great mind of Robert Coveyou studied a lot of his theory :).
But, more seriously, of course a random source will output compressible data sometimes. What happens is this: as you collect more output from a truly random source, the probability of it being compressible goes to zero very fast.
But the point is that it *is* useful to distinguish between "false" and "true" randomness, otherwise it wouldn't be true that "the generation of random numbers is too important to be left to chance".
Distinguishing between "false" and "true" randomness is pointless.
Not really, it's done all the time for many different purposes.
Take, for example, how computer scientists define it: roughly, a sequence is random if it can't be compressed, that is, any (program+data) that generates it must be at least as large than the sequence itself. It distinguishes between "random" and "not having enough information to predict it": it doesn't matter if it looks random to YOU; if it could in principle be compressed, it's not random.
That's not pointless hair splitting, it has real consequences for many areas of computer science, some very practical (cryptography, for example).
Do you want to claim that breaking entanglement is not instantly? Sorry, but then you are a bit out of the loop. It was never assumed that it took any time to be observed. It was assumed "it could not be observed". And this later point has changed in recent history. It was always clear that breaking entanglement has its "effect" transported instanly regardless of distance. That is not my "interpretration" but what is tought in school since ever.
"Breaking the entanglement is instantaneous" is a question that, currently, nobody knows the answer, but that's because this question is about the interpretation of quantum mechanics, and can't be settled, as far as we know, with experiments. What people do believe, though, is that even if breaking the entanglement is "instantaneous" (I write it in quotes because that can't even be made precise with current physics, as I've explained), no information can be sent "instantaneously". I'm curious to see this article on Scientific American that gave you the impression that this is not so.
Some interpretations of quantum mechanics try to explain things in terms of faster than light effects ( that 's what Einstein derisively called "spooky action"), but these are just interpretations of the math and the outcomes of the experiments. There's no consensus of whether it actually happens or not. And, more importantly, not even people who use these interpretations think that these "effects" can be used to send information faster than light. It should be clear how this works once you understand the math and experiments.
I've resisted doing this so far, because I didn't want to sound disrespectful, but it seems that you have a wrong impression about entanglement. Do you know how the mathematics of entanglement works, or just things you read in science popularization articles? There's an excellent series of introductory lectures from Stanford University in Youtube that teaches the basics of quantum entanglements starting here. The lecturer is a very famous physicist, Leonard Susskind. If you watch the whole thing (and follow the math explanations carefully), you'll notice that he never talks about "instantaneous" anything, and it's there's no reason to do so. And the math isn't too hard to follow, you only need to know basic algebra, complex numbers and matrices.
Regarding your strange attempt to citate me: 'but "it's ok because you had to send the particles beforehand"' ... I did not say that. Sigh. How hard is it to read? I said: some phyics say this! And the "it is ok" is not making the point.
Sorry, I misunderstood you. I had never seen it before, and I thought you believed this.
I don't know if it is "confirmed" but I assumed that quantum secured communication where wiretapping can be detected as certain properties of entangeld photons change supports this.
I't true that it's possible to detect eavesdropping in quantum secure channels, but that's due to the uncertainty principle and the no cloning theorem, not entanglement. It's almost impossible to really understand these intricate things, though, without understanding the underlying math.
Sigh... I don't think we're going to make any progress.
I've explained to you that in current physics, it doesn't make any sense to talk about "instant transmission" in the sense you're talking about. I've explained that in the current model for entanglement, there's no need or place for any such "instant transmission". Yet you keep insisting on a non-standard model for entanglement where information is sent faster than light, but "it's ok because you had to send the particles beforehand" (which, I have mentioned, is not only inconsistent with current physics but seems silly).
As I said, Scientific American is a magazine that does science popularization. While sometimes some articles are written by researchers, it's mostly written by journalists. If you're looking for papers with the current research on the foundations of quantum mechanics, you should check peer-reviewed physics journals; there's a nice compilation in Wikipedia. There's also Arxiv, but the papers there are not necessarily peer-reviewed.
Quantum encrypted channels work on the exact same principle and are already in general use. I don't see your point.
Quantum encryption, as all other established results that use entanglement (like quantum teleportation, etc.), don't presuppose or rely on entanglement being able to send information faster than light.
Also I assumed Scientific American was a respected peer reviewed physics magazine. Am I mistaken here?
Scientific American is a magazine that does science popularization, its articles are nowhere near the level of rigor necessary for a peer-reviewed physics journal.
I have read that in Scientific American, quite a few years ago.
Well, as great as Scientific American is, can we at least agree that it publishes a lot of stories about non-established and speculative theories? I'm talking about papers by real physicists that have been peer-reviewed and referenced by other physicists.
It does. As most physics accept that such watching on entanglements is/might be possible. However the general argument is: the first movement counts. Regarding "causality" at least.
I doubt that "most physics accept" that, at least in the terms you describe. Within one of the most basic foundations of modern physics (Special Relativity), it doesn't even make sense to talk about which came first between A and B if B is outside the light cone of A (that is, if light from A can't reach B), so your argument can't even be made precise within current physics. That's not a pointless objection: in your scenario, as far as Special Relativity goes, it would be possible to have one observer concluding that Mars measured first, and another one concluding that Earth measured first.
It is possible to propose new alternative theories, of course. But it's usually best to suspend judgement until they have been made precise and published for general scrutiny. That's why I wanted references to papers; I wanted see the reactions from real physicists.
Can you post (established) references, please? The only thing I heard about that is a paper from B. Dopfer in 1999 (I can't find a link right now) where they supposedly were able to stop interference by measuring entangled particles. I have seen a few critics, but mostly the lack of references from other papers is what makes me suspicious about it: surely this would be huge news, and a lot more people would be talking about it, no?
And in you scenario, nothing depends on anyone's definition of "faster than light". If you can make information from Earth (at point A) available in Mars (at point B) before light has time to travel from A to B, there's no discussion: there is faster than light communication, regardless of how you prepared the whole stuff. Claiming otherwise would be as silly as claiming that communication via radio is no faster than a train because you had to send the radio via train in the first place.
Yes, and my point is: it's no sillier than suggesting that quantum entanglement can be used to transmit information faster than light.
Really. In the current model of quantum entanglement, which is part of the foundations of quantum mechanics and exists more or less unchanged since (at least) the 1950s, it's very clear that information (or anything else) can't travel faster than light. Beyond that, there have been lots of experiments testing and exploring all subtleties of entanglement, since the 1970s. All the results conform in all aspects to the "standard" theory for quantum mechanics. All phenomena that we have been able to observe fit the current theories.
You could say "that doesn't mean that it's impossible, maybe some other experiments will find something someday", and that's true. But that's also true of conservation of energy, or the ontological status of the Easter Bunny (there's a phrase you don't head often :)).
There's a misconception that I've noticed in some people (I don't know if it includes you): the impression that people who think that quantum computing is an attainable goal also believe that entanglement must have some power that breaks the known laws of physics (for example, the speed of light limit). Usually, though, people with this misconception come to believe that quantum computing must be impossible (which, ironically, is the opposite of what you suggest -- that not only quantum computing is possible, but that entanglement does have the power to send information faster than light)
As far as our current understanding goes, though, none of that is true. Quantum computing is firmly founded in the current theory of quantum mechanics, respecting special relativity and all (i.e., the limit of the speed of light).