hubie wrote:
>The effect of apparent superluminal speeds, which
>has to do with propagation of waves through a
>medium with a large changing index of
>refraction, was well known for almost a century.
I'm aware of this, but this article shouldn't be dismissed as "old news." The researchers who published the paper in Nature recently claimed to have not only had apparent superluminal speeds, but to actually have a negative group velocity, and apparently there was some speculation that this process could actually have applications. The paper that was submitted to arXiv.org recently argues that their group velocity calculation was incorrect and the pulse did not have a superluminal group velocity.
So I don't think this article should be dismissed. I was hoping someone with more physics knowledge than I have could read through the paper; I read it and it seems correct but I don't have enough knowledge to be entirely sure their approximations are valid. The basic question is not whether it is possible to have apparent superluminal speeds - it certainly is - but whether the researchers actually achieved this.
Anonymous Coward wrote:
>Hasn't this already been discussed? I seem to
>remember reading about some scientists supposedly
>speeding up light, then later saying that the
>media had completely misinterpreted the
>experiment and they didn't speed up light. Or am
>I remembering a completely different incident?
Yes, the experiment was discussed; no, this paper was not. Here is what happened: the original researchers claimed to have created a laser pulse with superluminal group velocity. Still, no claim was made that information was transferred faster than light. Some of the media misinterpreted, thinking information had been transferred and this conflicted with relativity. THe original researchers made no claim that their results conflicted with relativity. The first Slashdot article came at this point, and there was a lot of discussion about how the media misinterpreted the findings.
The new paper posted to arXiv.org by the researchers from the Plasma Physics lab, however, disputes the claim made by the researchers that the pulse had superluminal group velocity. It is not disputing any media interpretation, but the actual paper in Nature that first presented the results.
So, in short, it may be that both the original research and the media interpretations were incorrect.
karma_hax0r wrote:
A mathematical equation such as a line, curve, or a plan can't really be imagined without a model using the Cartesian co-ordinate system. After hundreds of years, it's still the best thing that we have. Models are also important in science, which is replete with things that cannot be directly observed - and therefore require modeling. The atom. The quark. String theory.
I think that to some extent you are correct, but we also have to be careful about our models. We have fairly accurate equations for things like atoms and quarks - and then we have extremely simplified models. We have to be careful that if we draw conclusions from a model, those conclusions are still valid in the most sophisticated understanding we have of a subject, and we have to test our model against actual data. Younger students learn to think of electrons moving in perfect circles around a nucleus, in discrete shells, in one region of space. That isn't valid, and any ideas drawn from the "electron-shell" model must be checked against the full complexity of quantum theory and relativity. Even those models are only approximations of our world. String theory is probably the best model we have currently, but it is still a work in progress at best. Even when (if?) it is completed as a valid basis for physics, more "coarse" chemical and quantum equations will be vastly more useful for most purposes. So I would agree that models are important, but they are also dangerous. We must always check the results of our models against the real world, and work to refine our models and make them more accurate. Of course metaphors and models are the best tools we have for understanding our world. Everything we see is filtered through our thought processes. Information must be encoded in our brain in a certain way, and however it is encoded, it will never approach the full complexity of the vast array of information in the "real world." Language, mathematics, images - they are all approximations, ways of encoding information for efficient understanding and communication. But we must be careful that we do not distort information through our encoding of it.
Some models are better than others. None are perfect.
Re:Information is critical in war, class war inclu
on
Focusing Audio
·
· Score: 1
>Information is crucial to waging war; the class war is no exception.
...
>Their objective, again, is to manipulate us, to control us. My interest in preserving my privacy in this case is the same interest a country at war has in keeping its sensitive information
>secure.
Let's not exaggerate this.... You're not at war with these grocery stores. No one can make you have any dealings with them whatsoever. The only reason they exist is to provide a service to consumers. Yes, they might be ripping people off, but obviously people feel they are useful or no one would shop at them. If they were really so dangerous as you say, their entire reason for existence would vanish because no one would actually shop in them. There is no way a store is going to "manipulate" you.... You always have the option of avoiding them. But stores depend upon consumers for their survival; they certainly are not at "war" with them. It is in their best interests to keep customers happy.
It would seem to me that, while a guide to installation and basic command line info would be useful, complete newbies to Linux would probably begin by wondering how they can do the same things under Linux that they do under Windows. And for most people, this does not mean using an editor like vi or writing shell scripts. It means connecting to the Internet, using e-mail, web browsing, and word processing. From this review, it appears that these books do not explain how to do all of this under Linux. If not, I would thinkk that the average newbie Linux user would find these books disappointing, or, even worse, they may decide that these tasks cannot be done in Linux without resorting to the CLI. A good introduction to Linux, in my opinion, should cover installation and the use of a user-friendly GUI (like KDE or Gnome). It should explain how to set up an Internet connection, how to run Mozilla or Netscape, and the use of StarOffice or KOffice (when it becomes stable, that is) or some other office suite. Someone who buys "Linux for dummies" probably won't be using the command line very soon. Leave that for a later book, once they know how to do what they need to and they're interested in learning more.
So now we have the N-Cube and the X-Box. How boring. In the form of cubes and boxes, these consoles are useful for little other than being consoles. Why not more fun and interesting shapes? Maybe Sega or Sony can be more creative. The S-frisbee would be great. Power outage leaves you unable to play video games? Try playing with the frisbee outside! (Gasp!) Just imagine the marketing possibilities of such a device.
No, I don't think scientists were "covering up" or hiding results. I do think that someone - either the press, or NEC's PR department - decided to try to generate more interest in the story through reporting that was basically factually correct, but still misleading. The headlines all said things like "Speed of light broken," and the articles generally prominently mentioned that special relativity prohibits anything moving faster than the speed of light, and then said that scientists made light go faster than the lightspeed barrier. Most readers, not reading very carefully and not having much knowledge of physics, would get the impression that somehow Einstein's famous theory had been proven wrong. Generally, the news articles about the research contained a brief quote saying that the results do not violate relativity, but this quote was not explained in any detail. The overall spirit of the articles seemed to be "Law of physics overturned!"
I don't think there was any cover-up, but I do think the reporting could have been more careful.
I think the idea is that tunneling takes no time (or at least less time than it takes for a photon to move that far), though the particle is translated in space, resulting in FTL movement of the particle.
This is very interesting. I wasn't aware of the debate over tunneling time, but I have just been reading through various preprints and web sites and it appears that there are currently several different approaches, with some researchers claiming superluminal and some subluminal tunneling times. From what I have read, it looks as though tunneling through a thin enough barrier might actually be instantaneous.
However, the papers that I have skimmed so far mostly say that while the group velocity is superluminal, the frontal velocity is not. Apparently it is the frontal velocity that is involved in causality and information transfer, so again, the researchers do not seem to claim to have violated relativity.
Interestingly, the Nature article by Wang et. al. (from NEC) about superluminal propagation that started this discussion also indicates a group velocity faster than c. Everything I have been reading lately indicates that the "frontal velocity," not the group velocity, is what cannot exceed c in information transfer. I wonder if this concept is a relatively recent one, as I was able to talk to a rather famous (Nobel laureate) physicist a few months ago and he told me that the group velocity of any wave could never exceed c, as this would violate relativity. Has the understanding of which velocity determines information transfer been called into question recently?
It seems that there are many open questions and debates in this field, even though almost everyone agrees that FTL information transfer is not possible.
I'm not familiar with any way in which quantum tunneling can be used to send signals, but I don't know enough to say for sure whether it is possible or not.... I thought the point was simply that quantum particles had a nonzero probability of crossing potential barriers that classical particles could not. Is there more to tunneling?
Note: Quantum physics and a bit of math follow. I've highlighted the important stuff in bold in case you don't want to read the boring details. Or, you can just skip to the last paragraph.
However, I am somewhat more familiar with the "quantum teleportation" of a photon, and my understanding of this is that the actual signal transfer happens by classical means, and is not superluminal.
The quantum teleportation depends on an "EPR" device (named after the famous Einstein-Podolsky-Rosen thought experiment). This device can produce a pair of entangled photons, so that when a measurement is performed on one, the state of the other is determined. That is, the particles start out in some entangled state, but when particle A is measured, it will become either a |0> or a |1>. If A is |0>, B is |1>, and if A is |1>, B is |0>. They are said to be "orthogonal."
The basic idea is that A and the signal are entangled during a measurement, and this affects B, so that B becomes similar to the signal. Now, in order to determine the signal from B, one has to know not only the state of B, but the result of the measurement that entangled A and the signal. Thus, that result has to be transmitted by a classical method, which can't be faster than light. The advantage of quantum teleportation, as I understand it, is not faster transfer, but more accurate transfer.
Here's a more detailed explanation (let's hope I get the details right): the quantum teleportation method has a signal. Let's call it S. Now, Alice wants to send Bob the signal. Alice has a particle "A" (from the EPR device) and a particle "S." Bob also has a particle from the EPR device, particle "B". Particles A and B are entangled, in the state k(|0A>|1B> + |1A>|0B>). (k is a normalization constant = 1/sqrt(2)). When measured, either they will become |0A>|1B>, or |1A>|0B>. Particle S is in some unknown superposition of states, a|0S> + b|1S>. So Alice starts out with the overall state k(|0A>|1B> + |1A>|0B>)(a|0S> + b|1S>). She performs a Bell measurement on A and S. The Bell measurement entangles the photons, thus producing one of the four eigenstates:
PSI(+/-) = k(|0A>|1C> +/- |1A>|0C>)
PHI(+/-) = k(|0A>|0C> +/- |1A>|1C>)
Now, the combination of states (of A and S) that Alice begins with can be rewritten as:
(1/2)[ |PSI+>(a|1B>+b|0B>) + |PSI->(a|1B>-b|0B>) + |PHI+>(a|0B>+b|1B>) + |PHI->(a|0B>-b|1B>) ],
so that entangling the photons causes the state of particle B to become one of the four terms in parentheses. There is a nonzero probability (25%, to be specific) that any of these four states will occur.
Due to the way the beam-splitter technology used in the device works, the teleportation is only successful if the |PSI+> or |PSI-> state results; otherwise, there is a probability that either the |PHI+> or |PHI-> state caused the observed result. As far as I know, 50% efficiency is the best possible for quantum teleportation.
Still, the key point is that, in order for Bob to know if he has the exact state S that was intended to be teleported (that is, the |PSI+> eigenstate was measured so that a|1C>+b|0C> is the current state), or if he needs to apply a phase shift (that is, the |PSI-> eigenstate was measured so that a|1C>-b|1C> is selected), Alice has to tell Bob which Bell eigenstate her detectors measured. And that information must be transferred classically. And so, quantum teleportation does not transfer information faster than light.
By the way, I think a lot of these quantum teleportation experiments happened at IBM. There's a web site at http://www.research.ibm.com/ quantuminfo/teleportation/ that gives some information about the researchers who have done this. Also, the Los Alamos (xxx.lanl.gov or arXiv.org) preprint archive has a nice paper at quant-ph/0007106, written by Hai-Woong Lee and Jaewan Kim, if you want more detail.
In short, it seems that, no matter how clever scientists become, nature always leaves some sort of catch that keeps us from sending information faster than light and violating causality.
It does seem that someone... whether NEC or reporters, I'm not sure... was a little irresponsible with the "faster-than-light" story. Even the title of the Nature article, "Gain-assisted superluminal light propagation," seems a little misleading. Although all the articles clearly indicated that this experiment was not at odds with Einsteinian relativity, none of them really explained it.
I don't know that much about physics, but I knew something weird was happening here, and I found a little bit of explanation in the Feynman Lectures on Physics (Volume 1, Chapter 31).
For light of frequency omega, in a material with electrons having resonant frequency omega0, the index of refraction is:
n = 1 + (Ne^2)/(2 epsilon0 m (omega0^2 - omega^2))
The dependence on omega shows that a material transmits light at different speeds, depending on the frequency (or, from a different point of view, the wavelength) of the light. This phenomenon is called "dispersion." Now, for some frequencies, (omega0^2 - omega^2) will be negative and n can be less than one, implying "superluminal" propagation in the sense that light of that frequency may be transmitted faster than "c", the speed of light in vacuum.
Feynman notes that the difference in index of refraction indicates a "that the phase shift which is produced by the scattered light can be either positive or negative." However, he is careful to point out that signals themselves are not transmitted faster than c, because transmission of a signal depends on the index of refraction at multiple frequencies. The index tells the speed at which the node of the wave travels, but the node in itself can carry no information. In order to transmit information, the frequency of the wave must be varied.
So, it appears that the idea of sending light at "faster-than-light" speeds is an old one, well understood by physicists. The theory of relativity has not been violated, and this has been known for some time. Feynman, apparently, taught it to beginning physics students at Caltech in the 1960s. News sources must have simply been attempting to make the story into something more appealing to the public. "Laws of physics break down!" But in reality, no laws have been violated, physics is fundamentally unchanged, and the net result seems to have been a confused public.
nerds, geeks, chickens, and local definitions
on
Geeks vs. Nerds
·
· Score: 1
This thread seems particularly interesting, given that "Gunther," myself, and several other people living in Louisville, Kentucky [but not for much longer! yay, out-of-state-universities! um, anyway] appear to hold practically the opposite opinion from that held by most Slashdotters.
I think I've always more or less agreed with Ryan's definition of nerd and geek. That is, nerds like computers, math, physics, and generally knowing how things work and fit together; geeks like role-playing, anime, and science fiction. I would consider myself a nerd; I would probably be vaguely offended if someone called me a geek, although there doesn't seem to be an agreed-upon definition... It would be much simpler if we would all agree that "geek" means "one who bites the heads off chickens."
I wonder if the definitions of "nerd" and "geek" vary from one locality to another; if not, then there must be some isolated pocket of weirdness surrounding Louisville - or certain parts thereof - in which the definitions are generally assumed to be reversed. Since it is not unknown for things in Kentucky to be backward, I suppose that is a distinct possibility;-).
It appears that many people agree that a nerd is singlemindedly focused on one thing. That is almost exactly the opposite of what I would have said. I usually think of a nerd as someone with a wide variety of interests.
There also seems to be some debate over whether it is more socially acceptable to be a nerd or a geek, although the general consensus on Slashdot appears to be that geeks are more socially acceptable. I would have thought nerds are more socially acceptable, although geeks may be very acceptable within their own circles... On the other hand, perhaps being a geek (by Ryan's definition) does involve more inherently social actions (role-playing, watching anime) than being a nerd. That is one thing that comes to mind.... nerds do not necessarily do nerdy things when they're together. Writing code for some numerical or scientific task, for instance, is not a thing that many people would consider a fun group activity. Still, I don't think nerds are socially limited as some of the quotes indicate. I consider myself a nerd, and I have some vague semblance of a social life.... At least, I have a girlfriend.
Also, I wouldn't limit "nerdness" ("nerddom?") to interest in math, physics, or technology. I think there can be history nerds, literature nerds... although I think other sorts of nerds seem to be less willing to have the term "nerd" applied to them. In general, though, I would characterize most nerds as having an interest in more than one field... It seems like the Jargon file has some sort of description of hackers as being unusually knowledgeable about subjects outside of computing. That would seem to be a general characteristic of nerds, to me.
In short, I would consider myself and most of my friends nerds but not geeks. And I don't consider "nerd" an insulting term... Given that Slashdot is "News for Nerds," it would appear that at least a few people agree with me on that point.
>"given the state of the universe at time t, you could determine what it was at time t-1. But black holes screw that up"
Unfortunately, the universe doesn't work that way. Classical mechanics is deterministic, but quantum mechanics is not. We can't trace back to an earlier state of the universe, because we can only find the probability of one state evolving into another.
I'm not at all an expert, but I think this question is a lot more complex than most posts are making it sound. If problem #8 could be solved so easily by saying "the information is lost; the entropy difference is made up for by gravitational potential energy or Hawking radiation or some combination thereof," then the physicists who made this list would have said that.
It would be nice if someone who knows more about the underlying problems than I do could explain. However, this looks vaguely related to something known as the holographic principle, which seems to be important in string theory. My (very limited) understanding of it is that the equations of string theory could work on a smaller-dimensional subset... in other words, everything inside the universe might just be some sort of "reflection" of something on the "surface" of the universe. The theory that information inside black holes is reflected on the surface seems to be very similar. I'm fairly sure - correct me if I'm wrong - that the current theory is that a black hole's entropy is related to its surface area; i.e., when something falls into a black hole, the surface area increases by the same amount that the entropy of the rest of the universe decreases.
As I said, my understanding of this physics is limited, but one thing I do know is that quantum states can not be copied or destroyed. This is the source of some interesting differences between classical computers and quantum computers. Black holes seem to violate this.
By the way, if anyone reading this thread is unfamiliar with Hawking radiation, that's one thing I can explain... Black holes have a tendency to "evaporate." They give off particle/antiparticle pairs, so that they aren't entirely "black." On the other hand, these particles don't contain any of the information that actually goes into a black hole. They're more or less empty of information; black-hole radiation follows a statistical distribution that is roughly the same as the distribution of energy emitted from the sun (i.e., they work like blackbodies). As black holes emit this energy, they lose mass (remember E=mc^2?) and, correspondingly, "evaporate" over time. Because of this, every black hole can be said to have a temperature, which is determined from the Stefan-Boltzmann formula F=sigma*T^4, where F is the energy flux (watts per square meter) coming out of the black hole. So, a black hole shrinks over time. The question is: as it shrinks, does any information that was lost in it come back out?
If I get around to it, I might dig through some books later today and see if I can understand more about the issues involved here. I'm pretty sure it isn't a simple question, and there are very good reasons for it to be part of the list.
I think the reason "disappearing information" is considered so important is that, if information in a black hole is completely gone, then theoretically that would imply a decrease in entropy in the universe. The second law of thermodynamics says entropy always increases, so this causes a bit of a problem.
OK, now that I've read the opinions of so many others who have no clue, I feel like injecting my own clueless thoughts:-)
First of all, I'm sick of reading stuff written by those who know little about physics or math.
Secondly, I know very little about physics or math compared to the people doing this research.
Now that I've said that, I'll go do exactly what I hate everyone else doing and apply my high school math and physics knowledge (along with a college electricity & magnetism and course and some random knowledge I've picked up my own, none of which is particularly useful here).
Look at the paper, page 7. I don't know much about GR, and I know even less about the quantum theory involved in computing a term $T^{Q}_{ik}$. However, I do see something that looks very odd to me.
After eq. 5, we see the statement "where $\Omega$ and $K$ are smooth positive even functions." Is the *even* part important? If so, look at eq. 6. It says $\Omega ~ \Omega_{0} exp(Bx)$. Last I heard, the exponential function was not even.
Does anyone with enough knowledge of the physics of this problem care to explain why the paper contradicts itself here? It would be nice if the paper didn't omit most of the mathematical derivations...
Another thing: for those of you who complain about time paradoxes and the like... (please keep in mind I'm far from an expert here, but this is my understanding) notice that, according to the paper, these wormholes would only facilitate FTL travel if they could be moved. A stationary wormhole would allow faster travel (because the areas are essentially closer in space-time than was previously thought, if a wormhole connects them) but would not lead to time travel or other paradoxes. Since we have no technology for moving wormholes, FTL paradoxes don't seem to be a problem... it appears that the only way this theory would be useful for travel, in any reasonable time frame, would be if we find a way to discover pre-existing wormholes. Don't expect us to be building our own anytime soon...
That's just my $0.02... probably pretty worthless given my lack of background knowledge on this subject, but maybe someone who knows a bit more can explain why I'm right or wrong.
Yes, Intel is more or less just footing the bill. But that's what Westinghouse used to do, until two years ago. Science Service is the organization that arranges most of the activities, and has been (as far as I know) since the competition started in 1942.
The Westinghouse competition became the Intel competition two years ago. So perhaps it would be more correct to simply say "The STS is America's oldest...."
I found it very interesting that so many abstract mathematics projects placed in the top ten. Jayce Getz (from Montana) did the project extending one of Ramanujan's theorems, dealing with partitions. That placed 2nd.
In 4th place, Sasha Schwartz (from Pennsylvania) worked on coset partitions of Abelian groups.
And then in 9th place, there was Zach Cohn (from New York) with work on quadratic reciprocity in certain polynomial fields.
So, that's three abstract mathematics (i.e. number theory and group theory) projects. It surprised me. I thought that the Intel judges were looking for things that had immediate applications that the public could understand. Apparently they weren't. That was a pleasant surprise. I think the math projects were all at a very high level and I'm glad to see them recognized. Of course, all 40 finalists had very good research, and I'm glad I didn't have to try to choose the best among them...
Anyway, I just thought I would point out that there were several very interesting pure math projects at the competition and they all did very well.
Well, the descriptions may sound that way... All the bios of the finalists (I was one) are put together by Intel based on our applications. The application asks many questions about hobbies, interests, membership in clubs, and so on, and the Intel PR people put together the bios based on that. So yes, they don't always come out the way we would prefer to have ourselves described.
I did not do my work for the Intel STS (I was the 6th place winner) in a laboratory, but I have worked in a university lab in the past. There are many high school students who do research in well-equipped labs. It isn't an unfair advantage - these opportunities are available to most people who have enough initiative and intelligence to pursue them.
This year (if I remember correctly) there were 4 Intel finalists who participated in RSI (if I remember right, it was Feng Zhang, Viviana, Sasha Schwartz, and Elizabeth Williams). A few other finalists were in other summer research programs.
Also, I recall Viviana saying that she ordered the DNA from a lab somewhere that will manufacture DNA with a given base-pair sequence... Apparently it isn't too expensive.
I was one of the participants in this competition - I finished in 6th place with a project on adaptive wavelet methods for fluid dynamics problems. (I'm Matt Reece from Louisville, Kentucky).
First of all, I would like to say that if anyone reading this is a high school student considering entering this competition, do it. It is very much worth the time you spend on your research if you can become a finalist. All 40 finalists get $5000, a laptop (650 MHz Pentium III), and a trip to D.C. where Intel pays for everything - very nice expensive dinners, meetings with Nobel laureates... it's an incredible program. The best part was definitely meeting the other finalists, though. They were all wonderful people and I have had a great week... Don't think these people are just science nerds (not that that's a bad thing, mind you). They're very well-rounded. Many speak foreign languages, play musical instruments, sports, etc.
Also, the week is not all science - Intel provided a web center in the hotel with lots of nice computers equipped with Quake 3, so we could have big multiplayer deathmatches over the LAN. I also played cards more in the past week than I have in months, and generally just spent a lot of time hanging out with the other finalists.
Anyway, to get on to some of the comments the rest of you have made about Viviana's project. First, I will say that I'm not as familiar with her work as I am with some of the other projects.
She does attend a U.S. school - I think it's a public one but I'll have to look that up later. Personally, I attend a public magnet school (duPont Manual High School) and I know many of the other finalists do attend public schools.
It would probably be best if Viviana responded to your comments about DNA steganography, as I'm not an expert in the area. Still, the project did seem to be very well done and she did an excellent job of presenting it to the public.
As far as your comment about open source programmers... If an open source project involved a new algorithm or some other method that could be applied to science, then it would certainly stand a chance in the Intel competition. My wavelet code is open source, although at this point I haven't implemented enough features to make it very useful.
Also, you might be interested to know that the judging is not solely based on the research. The first stages are based on a research paper - out of about 1500 applicants, 300 semifinalists were chosen and then from those 300, forty were chosen as finalists.
The finalist judging is based on three 15-minute interviews in which judges ask questions related to science in general. Some questions are straightforward tests of scientific knowledge, others are more open-ended questions meant to see how well you can think. Some of the questions are things that no one knows...
These judging interviews took place on Thursday and Friday (the 9th and 10th). The next two days, March 11th and 12th, involved the public presentations, where we set up display boards at the National Academy of Science and talked about our research with judges, scientists, and anyone else who showed up. The judges talked to students on Saturday, and from what I understand had made all their decisions just before the dinner at Mr. K's (great Chinese restaurant) Saturday night. The winners were announced Monday evening.
So anyway, judging is based initially on the research, but the final awards are also based on general scientific knowledge and also ability to communicate that knowledge to others. The emphasis on communication is also evident in the Seaborg award, given to the student who best displays an excitement about science and a willingness to share that excitement - that award went to Eugene Simuni, who finished 5th. His work was all the more amazing because he's only lived in the U.S. for two years (he came here from Russia) and yet he's better at communicating science to the general public (in English, a language that he more or less taught himself) than most or maybe all of the rest of us who have been speaking English our whole lives.
Well, there is much more I could say, but I just wanted to give you a better idea of what this competition is all about. It's a great program, and I would recommend it to anyone. If you have any questions about the Intel STS, feel free to ask me.
Sorry for not making the "overkill" comment clear. No, the other stuff in Cygwin doesn't hurt. But for someone who simply wants to compile an open-source program for Windows, it isn't necessary. Someone who finds open-source C++ code for a useful program and wishes to run it under Windows, but who doesn't already have a C++ compiler for Windows, might find it to be too much trouble to use Cygwin.
I seem to disagree with most of you so here are my $0.02 on the issue.
Yes, Borland clearly wants people to buy their C++ Builder product. What's so evil about that? They're a business. They're trying to make money.
On the other hand, they have released a free compiler for Windows. Many of you imply this is useless. On the contrary, I think this is perhaps more useful than a Linux version of the compiler.
Why? Because Linux has gcc. gcc is a reasonably ANSI/ISO compliant C++ compiler. It includes a debugger, linker, etc. It's very useful. Borland C++, with no debugger, for Linux would not be particularly useful.
On the other hand, Windows has limited options for free C++ compilers. There is DJGPP, a DOS port of GNU tools. And then there's Cygwin. Cygwin is nice, but it's a large download and it's a complete set of tools. If all you want to do is compile C++ code, it's overkill. Especially since you have to distribute some large Cygwin DLLs.
Borland claims that this free compiler can support Open Source software. Most of you seem to disagree. I don't.
Imagine this scenario: someone writes open-source C++ software, using standard-compliant code and nothing operating-system specific. (It could even have a GUI, using something like GLUT or FLTK). They write this code on Linux. Compile it with gcc. Debug with gdb. Perhaps use one of the Linux IDEs.
Currently, this code would be difficult to use under Windows. You could compile with Cygwin, but this is overkill. Now, there's a new option. Compile the code with Borland's C++ compiler.
The lack of a debugger doesn't necessarily hurt this software. And this compiler does make it easier to write cross-platform, open-source software for Windows and Linux.
Go ahead and flame me. Say it's useless to write code that runs on Windows, or that it's supporting Micro$oft's evil empire. I disagree. Many people use Windows. Open-source software for these people isn't necessarily a bad thing. If they become interested enough in it, they can switch to Linux. And don't knock Borland for giving away free software without source. It's not the best thing we could have hoped for, true. But on the other hand, they are planning to develop their compiler software for Linux. Let's not change their minds by flaming them for trying to help out developers.
No, Visual C++ 6 is not yet ANSI compliant. Not if you're doing much of anything with templates. It doesn't support partial template specialization and member templates are still very screwed up. As are classes nested in templates. This code, for instance, won't compile:
// interface template<typename T> class A { public: class B { public: int do_something(); }; };
// implementation template<typename T> int A<T>::B::do_something() { return 1; };
On the other hand, inlining the code does work. But it's ugly and I don't like having to do it. It would be nice if Microsoft would produce a reasonably standard-compliant compiler for those of us who have to do Windows development...
My knowledge of this subject is limited (to what I've learned in AP psychology in high school) but I'll take a stab at this question.
The brain does need sleep. Furthermore, it needs dreams. Dreams happen in REM sleep and everyone dreams every night, if they sleep normally. Even if you don't remember it, you do.
(Another fact - sleepwalking does not happen in REM sleep and so sleepwalkers are not dreaming. Rather it happens in the deepest sleep stage. REM sleep happens in lighter sleep when the brain waves more closely resemble the alpha waves of awake but relaxed people. Sleepwalking involves deep sleep and delta waves.)
People who take sleeping pills will fail to have REM sleep, so this can be harmful. When taken off sleeping medication, people who fall asleep will spend much more time in REM sleep. Also, people who are sleep deprived will enter REM sleep very quickly. It is apparent that the brain needs REM sleep.
People tend to remember the preceding day's events better when they have more sleep, suggesting that REM sleep is involved in encoding short-term memories into long-term memory. Staying up all night to cram for a test might help but the information is not as likely to be remembered over a long period of time (what is forgotten will likely be forgotten very quickly - over time the amount of forgotten material levels off to a steady value - this is Ebbinghaus's forgetting curve).
The brain is probably not "exercising itself" through dreams. Rather, it is busily sorting through the day's events and storing the important material in long-term memory. Dreams appear to be the way that our mind superimposes a structure on this activity. This is the reason we see sudden changes in setting or events in a dream - our neurons are firing, and the brain tries to fit this somewhat random activity into a rational pattern through dreams.
Freud's ideas on latent and manifest content of dreams seem much less credible than the biological picture.
The events in dreams often echo the preceding day's events or things that have been on our minds. This lends support to the idea that the brain is encoding short-term memories into long-term storage.
Sleep deprivation may not hamper activity on some tasks - for instance, one research subject, a teenager who had been kept awake for several days, was able to beat the researcher in a game - pinball, if I remember correctly - even after several days without sleep.
From personal experience, it seems that sleep deprivation hurts mathematical ability but enhances creativity.
I don't have time to write more now, but hopefully this sheds a little light on the subject.
Slashdot Mirror
hubie wrote:
>The effect of apparent superluminal speeds, which
>has to do with propagation of waves through a
>medium with a large changing index of
>refraction, was well known for almost a century.
I'm aware of this, but this article shouldn't be dismissed as "old news." The researchers who published the paper in Nature recently claimed to have not only had apparent superluminal speeds, but to actually have a negative group velocity, and apparently there was some speculation that this process could actually have applications. The paper that was submitted to arXiv.org recently argues that their group velocity calculation was incorrect and the pulse did not have a superluminal group velocity.
So I don't think this article should be dismissed. I was hoping someone with more physics knowledge than I have could read through the paper; I read it and it seems correct but I don't have enough knowledge to be entirely sure their approximations are valid. The basic question is not whether it is possible to have apparent superluminal speeds - it certainly is - but whether the researchers actually achieved this.
Anonymous Coward wrote:
>Hasn't this already been discussed? I seem to
>remember reading about some scientists supposedly
>speeding up light, then later saying that the
>media had completely misinterpreted the
>experiment and they didn't speed up light. Or am
>I remembering a completely different incident?
Yes, the experiment was discussed; no, this paper was not. Here is what happened: the original researchers claimed to have created a laser pulse with superluminal group velocity. Still, no claim was made that information was transferred faster than light. Some of the media misinterpreted, thinking information had been transferred and this conflicted with relativity. THe original researchers made no claim that their results conflicted with relativity. The first Slashdot article came at this point, and there was a lot of discussion about how the media misinterpreted the findings.
The new paper posted to arXiv.org by the researchers from the Plasma Physics lab, however, disputes the claim made by the researchers that the pulse had superluminal group velocity. It is not disputing any media interpretation, but the actual paper in Nature that first presented the results.
So, in short, it may be that both the original research and the media interpretations were incorrect.
karma_hax0r wrote:
A mathematical equation such as a line, curve, or a plan can't really be imagined without a model using the Cartesian co-ordinate system. After hundreds of years, it's still the best thing that we have. Models are also important in science, which is replete with things that cannot be directly observed - and therefore require modeling. The atom. The quark. String theory.
I think that to some extent you are correct, but we also have to be careful about our models. We have fairly accurate equations for things like atoms and quarks - and then we have extremely simplified models. We have to be careful that if we draw conclusions from a model, those conclusions are still valid in the most sophisticated understanding we have of a subject, and we have to test our model against actual data. Younger students learn to think of electrons moving in perfect circles around a nucleus, in discrete shells, in one region of space. That isn't valid, and any ideas drawn from the "electron-shell" model must be checked against the full complexity of quantum theory and relativity. Even those models are only approximations of our world. String theory is probably the best model we have currently, but it is still a work in progress at best. Even when (if?) it is completed as a valid basis for physics, more "coarse" chemical and quantum equations will be vastly more useful for most purposes. So I would agree that models are important, but they are also dangerous. We must always check the results of our models against the real world, and work to refine our models and make them more accurate. Of course metaphors and models are the best tools we have for understanding our world. Everything we see is filtered through our thought processes. Information must be encoded in our brain in a certain way, and however it is encoded, it will never approach the full complexity of the vast array of information in the "real world." Language, mathematics, images - they are all approximations, ways of encoding information for efficient understanding and communication. But we must be careful that we do not distort information through our encoding of it.
Some models are better than others. None are perfect.
>Information is crucial to waging war; the class war is no exception.
...
>Their objective, again, is to manipulate us, to control us. My interest in preserving my privacy in this case is the same interest a country at war has in keeping its sensitive information
>secure.
Let's not exaggerate this.... You're not at war with these grocery stores. No one can make you have any dealings with them whatsoever. The only reason they exist is to provide a service to consumers. Yes, they might be ripping people off, but obviously people feel they are useful or no one would shop at them. If they were really so dangerous as you say, their entire reason for existence would vanish because no one would actually shop in them. There is no way a store is going to "manipulate" you.... You always have the option of avoiding them. But stores depend upon consumers for their survival; they certainly are not at "war" with them. It is in their best interests to keep customers happy.
It would seem to me that, while a guide to installation and basic command line info would be useful, complete newbies to Linux would probably begin by wondering how they can do the same things under Linux that they do under Windows. And for most people, this does not mean using an editor like vi or writing shell scripts. It means connecting to the Internet, using e-mail, web browsing, and word processing. From this review, it appears that these books do not explain how to do all of this under Linux. If not, I would thinkk that the average newbie Linux user would find these books disappointing, or, even worse, they may decide that these tasks cannot be done in Linux without resorting to the CLI. A good introduction to Linux, in my opinion, should cover installation and the use of a user-friendly GUI (like KDE or Gnome). It should explain how to set up an Internet connection, how to run Mozilla or Netscape, and the use of StarOffice or KOffice (when it becomes stable, that is) or some other office suite. Someone who buys "Linux for dummies" probably won't be using the command line very soon. Leave that for a later book, once they know how to do what they need to and they're interested in learning more.
So now we have the N-Cube and the X-Box. How boring. In the form of cubes and boxes, these consoles are useful for little other than being consoles. Why not more fun and interesting shapes? Maybe Sega or Sony can be more creative. The S-frisbee would be great. Power outage leaves you unable to play video games? Try playing with the frisbee outside! (Gasp!) Just imagine the marketing possibilities of such a device.
No, I don't think scientists were "covering up" or hiding results. I do think that someone - either the press, or NEC's PR department - decided to try to generate more interest in the story through reporting that was basically factually correct, but still misleading. The headlines all said things like "Speed of light broken," and the articles generally prominently mentioned that special relativity prohibits anything moving faster than the speed of light, and then said that scientists made light go faster than the lightspeed barrier. Most readers, not reading very carefully and not having much knowledge of physics, would get the impression that somehow Einstein's famous theory had been proven wrong. Generally, the news articles about the research contained a brief quote saying that the results do not violate relativity, but this quote was not explained in any detail. The overall spirit of the articles seemed to be "Law of physics overturned!"
I don't think there was any cover-up, but I do think the reporting could have been more careful.
I think the idea is that tunneling takes no time (or at least less time than it takes for a photon to move that far), though the particle is translated in space, resulting in FTL movement of the particle.
This is very interesting. I wasn't aware of the debate over tunneling time, but I have just been reading through various preprints and web sites and it appears that there are currently several different approaches, with some researchers claiming superluminal and some subluminal tunneling times. From what I have read, it looks as though tunneling through a thin enough barrier might actually be instantaneous.
However, the papers that I have skimmed so far mostly say that while the group velocity is superluminal, the frontal velocity is not. Apparently it is the frontal velocity that is involved in causality and information transfer, so again, the researchers do not seem to claim to have violated relativity.
Interestingly, the Nature article by Wang et. al. (from NEC) about superluminal propagation that started this discussion also indicates a group velocity faster than c. Everything I have been reading lately indicates that the "frontal velocity," not the group velocity, is what cannot exceed c in information transfer. I wonder if this concept is a relatively recent one, as I was able to talk to a rather famous (Nobel laureate) physicist a few months ago and he told me that the group velocity of any wave could never exceed c, as this would violate relativity. Has the understanding of which velocity determines information transfer been called into question recently?
It seems that there are many open questions and debates in this field, even though almost everyone agrees that FTL information transfer is not possible.
I'm not familiar with any way in which quantum tunneling can be used to send signals, but I don't know enough to say for sure whether it is possible or not.... I thought the point was simply that quantum particles had a nonzero probability of crossing potential barriers that classical particles could not. Is there more to tunneling?
Note: Quantum physics and a bit of math follow. I've highlighted the important stuff in bold in case you don't want to read the boring details. Or, you can just skip to the last paragraph.
However, I am somewhat more familiar with the "quantum teleportation" of a photon, and my understanding of this is that the actual signal transfer happens by classical means, and is not superluminal.
The quantum teleportation depends on an "EPR" device (named after the famous Einstein-Podolsky-Rosen thought experiment). This device can produce a pair of entangled photons, so that when a measurement is performed on one, the state of the other is determined. That is, the particles start out in some entangled state, but when particle A is measured, it will become either a |0> or a |1>. If A is |0>, B is |1>, and if A is |1>, B is |0>. They are said to be "orthogonal."
The basic idea is that A and the signal are entangled during a measurement, and this affects B, so that B becomes similar to the signal. Now, in order to determine the signal from B, one has to know not only the state of B, but the result of the measurement that entangled A and the signal. Thus, that result has to be transmitted by a classical method, which can't be faster than light. The advantage of quantum teleportation, as I understand it, is not faster transfer, but more accurate transfer.
Here's a more detailed explanation (let's hope I get the details right): the quantum teleportation method has a signal. Let's call it S. Now, Alice wants to send Bob the signal. Alice has a particle "A" (from the EPR device) and a particle "S." Bob also has a particle from the EPR device, particle "B". Particles A and B are entangled, in the state k(|0A>|1B> + |1A>|0B>). (k is a normalization constant = 1/sqrt(2)). When measured, either they will become |0A>|1B>, or |1A>|0B>. Particle S is in some unknown superposition of states, a|0S> + b|1S>. So Alice starts out with the overall state k(|0A>|1B> + |1A>|0B>)(a|0S> + b|1S>). She performs a Bell measurement on A and S. The Bell measurement entangles the photons, thus producing one of the four eigenstates:
PSI(+/-) = k(|0A>|1C> +/- |1A>|0C>)
PHI(+/-) = k(|0A>|0C> +/- |1A>|1C>)
Now, the combination of states (of A and S) that Alice begins with can be rewritten as:
(1/2)[ |PSI+>(a|1B>+b|0B>) + |PSI->(a|1B>-b|0B>) + |PHI+>(a|0B>+b|1B>) + |PHI->(a|0B>-b|1B>) ],
so that entangling the photons causes the state of particle B to become one of the four terms in parentheses. There is a nonzero probability (25%, to be specific) that any of these four states will occur.
Due to the way the beam-splitter technology used in the device works, the teleportation is only successful if the |PSI+> or |PSI-> state results; otherwise, there is a probability that either the |PHI+> or |PHI-> state caused the observed result. As far as I know, 50% efficiency is the best possible for quantum teleportation.
Still, the key point is that, in order for Bob to know if he has the exact state S that was intended to be teleported (that is, the |PSI+> eigenstate was measured so that a|1C>+b|0C> is the current state), or if he needs to apply a phase shift (that is, the |PSI-> eigenstate was measured so that a|1C>-b|1C> is selected), Alice has to tell Bob which Bell eigenstate her detectors measured. And that information must be transferred classically. And so, quantum teleportation does not transfer information faster than light.
By the way, I think a lot of these quantum teleportation experiments happened at IBM. There's a web site at http://www.research.ibm.com/ quantuminfo/teleportation/ that gives some information about the researchers who have done this. Also, the Los Alamos (xxx.lanl.gov or arXiv.org) preprint archive has a nice paper at quant-ph/0007106, written by Hai-Woong Lee and Jaewan Kim, if you want more detail.
In short, it seems that, no matter how clever scientists become, nature always leaves some sort of catch that keeps us from sending information faster than light and violating causality.
It does seem that someone... whether NEC or reporters, I'm not sure... was a little irresponsible with the "faster-than-light" story. Even the title of the Nature article, "Gain-assisted superluminal light propagation," seems a little misleading. Although all the articles clearly indicated that this experiment was not at odds with Einsteinian relativity, none of them really explained it.
I don't know that much about physics, but I knew something weird was happening here, and I found a little bit of explanation in the Feynman Lectures on Physics (Volume 1, Chapter 31).
For light of frequency omega, in a material with electrons having resonant frequency omega0, the index of refraction is:
n = 1 + (Ne^2)/(2 epsilon0 m (omega0^2 - omega^2))
The dependence on omega shows that a material transmits light at different speeds, depending on the frequency (or, from a different point of view, the wavelength) of the light. This phenomenon is called "dispersion." Now, for some frequencies, (omega0^2 - omega^2) will be negative and n can be less than one, implying "superluminal" propagation in the sense that light of that frequency may be transmitted faster than "c", the speed of light in vacuum.
Feynman notes that the difference in index of refraction indicates a "that the phase shift which is produced by the scattered light can be either positive or negative." However, he is careful to point out that signals themselves are not transmitted faster than c, because transmission of a signal depends on the index of refraction at multiple frequencies. The index tells the speed at which the node of the wave travels, but the node in itself can carry no information. In order to transmit information, the frequency of the wave must be varied.
So, it appears that the idea of sending light at "faster-than-light" speeds is an old one, well understood by physicists. The theory of relativity has not been violated, and this has been known for some time. Feynman, apparently, taught it to beginning physics students at Caltech in the 1960s. News sources must have simply been attempting to make the story into something more appealing to the public. "Laws of physics break down!" But in reality, no laws have been violated, physics is fundamentally unchanged, and the net result seems to have been a confused public.
This thread seems particularly interesting, given that "Gunther," myself, and several other people living in Louisville, Kentucky [but not for much longer! yay, out-of-state-universities! um, anyway] appear to hold practically the opposite opinion from that held by most Slashdotters.
;-).
I think I've always more or less agreed with Ryan's definition of nerd and geek. That is, nerds like computers, math, physics, and generally knowing how things work and fit together; geeks like role-playing, anime, and science fiction. I would consider myself a nerd; I would probably be vaguely offended if someone called me a geek, although there doesn't seem to be an agreed-upon definition... It would be much simpler if we would all agree that "geek" means "one who bites the heads off chickens."
I wonder if the definitions of "nerd" and "geek" vary from one locality to another; if not, then there must be some isolated pocket of weirdness surrounding Louisville - or certain parts thereof - in which the definitions are generally assumed to be reversed. Since it is not unknown for things in Kentucky to be backward, I suppose that is a distinct possibility
It appears that many people agree that a nerd is singlemindedly focused on one thing. That is almost exactly the opposite of what I would have said. I usually think of a nerd as someone with a wide variety of interests.
There also seems to be some debate over whether it is more socially acceptable to be a nerd or a geek, although the general consensus on Slashdot appears to be that geeks are more socially acceptable. I would have thought nerds are more socially acceptable, although geeks may be very acceptable within their own circles... On the other hand, perhaps being a geek (by Ryan's definition) does involve more inherently social actions (role-playing, watching anime) than being a nerd. That is one thing that comes to mind.... nerds do not necessarily do nerdy things when they're together. Writing code for some numerical or scientific task, for instance, is not a thing that many people would consider a fun group activity. Still, I don't think nerds are socially limited as some of the quotes indicate. I consider myself a nerd, and I have some vague semblance of a social life.... At least, I have a girlfriend.
Also, I wouldn't limit "nerdness" ("nerddom?") to interest in math, physics, or technology. I think there can be history nerds, literature nerds... although I think other sorts of nerds seem to be less willing to have the term "nerd" applied to them. In general, though, I would characterize most nerds as having an interest in more than one field... It seems like the Jargon file has some sort of description of hackers as being unusually knowledgeable about subjects outside of computing. That would seem to be a general characteristic of nerds, to me.
In short, I would consider myself and most of my friends nerds but not geeks. And I don't consider "nerd" an insulting term... Given that Slashdot is "News for Nerds," it would appear that at least a few people agree with me on that point.
>"given the state of the universe at time t, you could determine what it was at time t-1. But black holes screw that up"
Unfortunately, the universe doesn't work that way. Classical mechanics is deterministic, but quantum mechanics is not. We can't trace back to an earlier state of the universe, because we can only find the probability of one state evolving into another.
I'm not at all an expert, but I think this question is a lot more complex than most posts are making it sound. If problem #8 could be solved so easily by saying "the information is lost; the entropy difference is made up for by gravitational potential energy or Hawking radiation or some combination thereof," then the physicists who made this list would have said that.
It would be nice if someone who knows more about the underlying problems than I do could explain. However, this looks vaguely related to something known as the holographic principle, which seems to be important in string theory. My (very limited) understanding of it is that the equations of string theory could work on a smaller-dimensional subset... in other words, everything inside the universe might just be some sort of "reflection" of something on the "surface" of the universe. The theory that information inside black holes is reflected on the surface seems to be very similar. I'm fairly sure - correct me if I'm wrong - that the current theory is that a black hole's entropy is related to its surface area; i.e., when something falls into a black hole, the surface area increases by the same amount that the entropy of the rest of the universe decreases.
As I said, my understanding of this physics is limited, but one thing I do know is that quantum states can not be copied or destroyed. This is the source of some interesting differences between classical computers and quantum computers. Black holes seem to violate this.
By the way, if anyone reading this thread is unfamiliar with Hawking radiation, that's one thing I can explain... Black holes have a tendency to "evaporate." They give off particle/antiparticle pairs, so that they aren't entirely "black." On the other hand, these particles don't contain any of the information that actually goes into a black hole. They're more or less empty of information; black-hole radiation follows a statistical distribution that is roughly the same as the distribution of energy emitted from the sun (i.e., they work like blackbodies). As black holes emit this energy, they lose mass (remember E=mc^2?) and, correspondingly, "evaporate" over time. Because of this, every black hole can be said to have a temperature, which is determined from the Stefan-Boltzmann formula F=sigma*T^4, where F is the energy flux (watts per square meter) coming out of the black hole. So, a black hole shrinks over time. The question is: as it shrinks, does any information that was lost in it come back out?
If I get around to it, I might dig through some books later today and see if I can understand more about the issues involved here. I'm pretty sure it isn't a simple question, and there are very good reasons for it to be part of the list.
I think the reason "disappearing information" is considered so important is that, if information in a black hole is completely gone, then theoretically that would imply a decrease in entropy in the universe. The second law of thermodynamics says entropy always increases, so this causes a bit of a problem.
OK, now that I've read the opinions of so many others who have no clue, I feel like injecting my own clueless thoughts :-)
First of all, I'm sick of reading stuff written by those who know little about physics or math.
Secondly, I know very little about physics or math compared to the people doing this research.
Now that I've said that, I'll go do exactly what I hate everyone else doing and apply my high school math and physics knowledge (along with a college electricity & magnetism and course and some random knowledge I've picked up my own, none of which is particularly useful here).
Look at the paper, page 7. I don't know much about GR, and I know even less about the quantum theory involved in computing a term $T^{Q}_{ik}$. However, I do see something that looks very odd to me.
After eq. 5, we see the statement "where $\Omega$ and $K$ are smooth positive even functions." Is the *even* part important? If so, look at eq. 6. It says $\Omega ~ \Omega_{0} exp(Bx)$. Last I heard, the exponential function was not even.
Does anyone with enough knowledge of the physics of this problem care to explain why the paper contradicts itself here? It would be nice if the paper didn't omit most of the mathematical derivations...
Another thing: for those of you who complain about time paradoxes and the like... (please keep in mind I'm far from an expert here, but this is my understanding) notice that, according to the paper, these wormholes would only facilitate FTL travel if they could be moved. A stationary wormhole would allow faster travel (because the areas are essentially closer in space-time than was previously thought, if a wormhole connects them) but would not lead to time travel or other paradoxes. Since we have no technology for moving wormholes, FTL paradoxes don't seem to be a problem... it appears that the only way this theory would be useful for travel, in any reasonable time frame, would be if we find a way to discover pre-existing wormholes. Don't expect us to be building our own anytime soon...
That's just my $0.02... probably pretty worthless given my lack of background knowledge on this subject, but maybe someone who knows a bit more can explain why I'm right or wrong.
Yes, Intel is more or less just footing the bill. But that's what Westinghouse used to do, until two years ago. Science Service is the organization that arranges most of the activities, and has been (as far as I know) since the competition started in 1942.
The Westinghouse competition became the Intel competition two years ago. So perhaps it would be more correct to simply say "The STS is America's oldest...."
I found it very interesting that so many abstract mathematics projects placed in the top ten. Jayce Getz (from Montana) did the project extending one of Ramanujan's theorems, dealing with partitions. That placed 2nd.
In 4th place, Sasha Schwartz (from Pennsylvania) worked on coset partitions of Abelian groups.
And then in 9th place, there was Zach Cohn (from New York) with work on quadratic reciprocity in certain polynomial fields.
So, that's three abstract mathematics (i.e. number theory and group theory) projects. It surprised me. I thought that the Intel judges were looking for things that had immediate applications that the public could understand. Apparently they weren't. That was a pleasant surprise. I think the math projects were all at a very high level and I'm glad to see them recognized. Of course, all 40 finalists had very good research, and I'm glad I didn't have to try to choose the best among them...
Anyway, I just thought I would point out that there were several very interesting pure math projects at the competition and they all did very well.
Notice the second name in what you wrote:
Clelland, C.T., Risca, Viviana, and Bancroft, C. (1999). Hiding messages in DNA microdots. Nature 399, 533-534
"Risca, Viviana" would be none other than Viviana Risca, the winner of the Intel Science Talent Search. Same person, same research.
Well, the descriptions may sound that way... All the bios of the finalists (I was one) are put together by Intel based on our applications. The application asks many questions about hobbies, interests, membership in clubs, and so on, and the Intel PR people put together the bios based on that. So yes, they don't always come out the way we would prefer to have ourselves described.
Just to expand on the previous comment...
I did not do my work for the Intel STS (I was the 6th place winner) in a laboratory, but I have worked in a university lab in the past. There are many high school students who do research in well-equipped labs. It isn't an unfair advantage - these opportunities are available to most people who have enough initiative and intelligence to pursue them.
This year (if I remember correctly) there were 4 Intel finalists who participated in RSI (if I remember right, it was Feng Zhang, Viviana, Sasha Schwartz, and Elizabeth Williams). A few other finalists were in other summer research programs.
Also, I recall Viviana saying that she ordered the DNA from a lab somewhere that will manufacture DNA with a given base-pair sequence... Apparently it isn't too expensive.
I hope this clarifies some things...
I was one of the participants in this competition - I finished in 6th place with a project on adaptive wavelet methods for fluid dynamics problems. (I'm Matt Reece from Louisville, Kentucky).
First of all, I would like to say that if anyone reading this is a high school student considering entering this competition, do it. It is very much worth the time you spend on your research if you can become a finalist. All 40 finalists get $5000, a laptop (650 MHz Pentium III), and a trip to D.C. where Intel pays for everything - very nice expensive dinners, meetings with Nobel laureates... it's an incredible program. The best part was definitely meeting the other finalists, though. They were all wonderful people and I have had a great week... Don't think these people are just science nerds (not that that's a bad thing, mind you). They're very well-rounded. Many speak foreign languages, play musical instruments, sports, etc.
Also, the week is not all science - Intel provided a web center in the hotel with lots of nice computers equipped with Quake 3, so we could have big multiplayer deathmatches over the LAN. I also played cards more in the past week than I have in months, and generally just spent a lot of time hanging out with the other finalists.
Anyway, to get on to some of the comments the rest of you have made about Viviana's project. First, I will say that I'm not as familiar with her work as I am with some of the other projects.
She does attend a U.S. school - I think it's a public one but I'll have to look that up later. Personally, I attend a public magnet school (duPont Manual High School) and I know many of the other finalists do attend public schools.
It would probably be best if Viviana responded to your comments about DNA steganography, as I'm not an expert in the area. Still, the project did seem to be very well done and she did an excellent job of presenting it to the public.
As far as your comment about open source programmers... If an open source project involved a new algorithm or some other method that could be applied to science, then it would certainly stand a chance in the Intel competition. My wavelet code is open source, although at this point I haven't implemented enough features to make it very useful.
Also, you might be interested to know that the judging is not solely based on the research. The first stages are based on a research paper - out of about 1500 applicants, 300 semifinalists were chosen and then from those 300, forty were chosen as finalists.
The finalist judging is based on three 15-minute interviews in which judges ask questions related to science in general. Some questions are straightforward tests of scientific knowledge, others are more open-ended questions meant to see how well you can think. Some of the questions are things that no one knows...
These judging interviews took place on Thursday and Friday (the 9th and 10th). The next two days, March 11th and 12th, involved the public presentations, where we set up display boards at the National Academy of Science and talked about our research with judges, scientists, and anyone else who showed up. The judges talked to students on Saturday, and from what I understand had made all their decisions just before the dinner at Mr. K's (great Chinese restaurant) Saturday night. The winners were announced Monday evening.
So anyway, judging is based initially on the research, but the final awards are also based on general scientific knowledge and also ability to communicate that knowledge to others. The emphasis on communication is also evident in the Seaborg award, given to the student who best displays an excitement about science and a willingness to share that excitement - that award went to Eugene Simuni, who finished 5th. His work was all the more amazing because he's only lived in the U.S. for two years (he came here from Russia) and yet he's better at communicating science to the general public (in English, a language that he more or less taught himself) than most or maybe all of the rest of us who have been speaking English our whole lives.
Well, there is much more I could say, but I just wanted to give you a better idea of what this competition is all about. It's a great program, and I would recommend it to anyone. If you have any questions about the Intel STS, feel free to ask me.
Sorry for not making the "overkill" comment clear. No, the other stuff in Cygwin doesn't hurt. But for someone who simply wants to compile an open-source program for Windows, it isn't necessary. Someone who finds open-source C++ code for a useful program and wishes to run it under Windows, but who doesn't already have a C++ compiler for Windows, might find it to be too much trouble to use Cygwin.
I seem to disagree with most of you so here are my $0.02 on the issue.
Yes, Borland clearly wants people to buy their C++ Builder product. What's so evil about that? They're a business. They're trying to make money.
On the other hand, they have released a free compiler for Windows. Many of you imply this is useless. On the contrary, I think this is perhaps more useful than a Linux version of the compiler.
Why? Because Linux has gcc. gcc is a reasonably ANSI/ISO compliant C++ compiler. It includes a debugger, linker, etc. It's very useful. Borland C++, with no debugger, for Linux would not be particularly useful.
On the other hand, Windows has limited options for free C++ compilers. There is DJGPP, a DOS port of GNU tools. And then there's Cygwin. Cygwin is nice, but it's a large download and it's a complete set of tools. If all you want to do is compile C++ code, it's overkill. Especially since you have to distribute some large Cygwin DLLs.
Borland claims that this free compiler can support Open Source software. Most of you seem to disagree. I don't.
Imagine this scenario: someone writes open-source C++ software, using standard-compliant code and nothing operating-system specific. (It could even have a GUI, using something like GLUT or FLTK). They write this code on Linux. Compile it with gcc. Debug with gdb. Perhaps use one of the Linux IDEs.
Currently, this code would be difficult to use under Windows. You could compile with Cygwin, but this is overkill. Now, there's a new option. Compile the code with Borland's C++ compiler.
The lack of a debugger doesn't necessarily hurt this software. And this compiler does make it easier to write cross-platform, open-source software for Windows and Linux.
Go ahead and flame me. Say it's useless to write code that runs on Windows, or that it's supporting Micro$oft's evil empire. I disagree. Many people use Windows. Open-source software for these people isn't necessarily a bad thing. If they become interested enough in it, they can switch to Linux. And don't knock Borland for giving away free software without source. It's not the best thing we could have hoped for, true. But on the other hand, they are planning to develop their compiler software for Linux. Let's not change their minds by flaming them for trying to help out developers.
No, Visual C++ 6 is not yet ANSI compliant. Not if you're doing much of anything with templates. It doesn't support partial template specialization and member templates are still very screwed up. As are classes nested in templates. This code, for instance, won't compile:
// interface
template<typename T>
class A
{
public:
class B {
public:
int do_something();
};
};
// implementation
template<typename T>
int A<T>::B::do_something()
{
return 1;
};
On the other hand, inlining the code does work. But it's ugly and I don't like having to do it. It would be nice if Microsoft would produce a reasonably standard-compliant compiler for those of us who have to do Windows development...
My knowledge of this subject is limited (to what I've learned in AP psychology in high school) but I'll take a stab at this question.
The brain does need sleep. Furthermore, it needs dreams. Dreams happen in REM sleep and everyone dreams every night, if they sleep normally. Even if you don't remember it, you do.
(Another fact - sleepwalking does not happen in REM sleep and so sleepwalkers are not dreaming. Rather it happens in the deepest sleep stage. REM sleep happens in lighter sleep when the brain waves more closely resemble the alpha waves of awake but relaxed people. Sleepwalking involves deep sleep and delta waves.)
People who take sleeping pills will fail to have REM sleep, so this can be harmful. When taken off sleeping medication, people who fall asleep will spend much more time in REM sleep. Also, people who are sleep deprived will enter REM sleep very quickly. It is apparent that the brain needs REM sleep.
People tend to remember the preceding day's events better when they have more sleep, suggesting that REM sleep is involved in encoding short-term memories into long-term memory. Staying up all night to cram for a test might help but the information is not as likely to be remembered over a long period of time (what is forgotten will likely be forgotten very quickly - over time the amount of forgotten material levels off to a steady value - this is Ebbinghaus's forgetting curve).
The brain is probably not "exercising itself" through dreams. Rather, it is busily sorting through the day's events and storing the important material in long-term memory. Dreams appear to be the way that our mind superimposes a structure on this activity. This is the reason we see sudden changes in setting or events in a dream - our neurons are firing, and the brain tries to fit this somewhat random activity into a rational pattern through dreams.
Freud's ideas on latent and manifest content of dreams seem much less credible than the biological picture.
The events in dreams often echo the preceding day's events or things that have been on our minds. This lends support to the idea that the brain is encoding short-term memories into long-term storage.
Sleep deprivation may not hamper activity on some tasks - for instance, one research subject, a teenager who had been kept awake for several days, was able to beat the researcher in a game - pinball, if I remember correctly - even after several days without sleep.
From personal experience, it seems that sleep deprivation hurts mathematical ability but enhances creativity.
I don't have time to write more now, but hopefully this sheds a little light on the subject.