I'm sorry, I know I shouldn't respond, but I can't help it. It bothers me when people start spouting off about how everyone needs to believe in their religion or go to hell, etc.
I'm a very religious person, believe in God, and enjoy reading the Bible. So I take offense at this kind of attitude because it seems so unchristian to me. You talk about Jesus, but at the same time are being judgemental, and are being publicly "noisy" about how righteous you are, and so much more favored than the rest of us. I believe that the salvation Jesus Christ gives us isn't about the afterlife, but about here and now, obeying his teachings, following his example.
I wanted to post this because I didn't want others to think that rigid indoctrination is what religion is about. I don't believe the universe is only a few thousand years old, or that there needs to be any conflict between religion and science. I consider myself to be Christian mostly because I see great value in the Bible and Jesus' teachings. I think, for example, Matthew 5-7 encapsulates all of what Christianity is about. This is a fantastic moral discourse, and would bring world peace and prosperity if obeyed. But it's difficult and uncommon advice -- shockingly different from "worldly" (popular) philosophy.
Finally, I don't think it's a conflict to believe the Bible and also the Big Bang theory. I think the accounts of creation in the Bible are technically just ancient Hebrew folklore, but they belong in the Bible because they contain rich moral lessons, and describe the true nature God and the Universe. The Bible isn't a history book. It's not written that way. It's a religious work, and should be treated as such.
It feels wrong to reply to my own reply, but here are two or three more ideas:
You have to consider methods of heat transport when cooking. How quickly does the heat transfer from the inner surface of the lava to the bird? In between you mostly have charred leaves. This prevents radiative transfer (broiling) or convective transfer (baking). So you're left with conduction, which will be very slow through the charred leaves. This allows for a large difference in temperature between the rock and the bird.
If you want faster cooking, I suppose one could engineer a metal cage that keeps the lava a safe and uniform distance from the uncovered bird. A coarse wire mesh should do the trick since lava is so viscous and will quickly solidify. Then the bird would be cooked by radiation instead, very similar to broiling. You could get even heating from all sides, or at least from above and below. And the way the lava cools down works in your favor too, searing the bird and then cooking it. I doubt this would taste better than the wrapped-in-leaves method, but I'm sure it would be faster.
I don't know what the exact numbers are, but I know that lava is a fairly good insulator. So it's quite possible that when the inside cools down, the outside remains very hot for a long time. The main reason for this is that the lava contains disolved gasses that create bubbles as it cools. Lava rocks are very lightweight (well, surface flow lava rocks anyway) like solid foam.
I agree that the water in the leaves regulates the temperature inside, especially since pressure is not allowed to build up.
I think there's a simmilar reason that it takes about 25 minutes to barbecue corn in the husk, but only 3 minutes to boil it.
See the latest in innovation! Totally changes the way you work, and even think about working! Far superior to unsightly cabinets of dusty electronics, and for a price that is astounding! If you decide to keep it, we'll refund half the purchase price. This is so revolutionary that it makes the PC obsolete. Stop by, and you'll get a free onion!
Re:award winning, no less
on
Haiku vs Spam
·
· Score: 1
The English language has so much redundancy, you might be surprised!
I don't think that last post was off-topic at all! Mathematics is really a language that absolutely needs to be taught. There were a few quotes in that article that made my stomach churn, such as:
"I say that I would recommend this student repeat the course the following year, because I don't think that foundation is adequate for continuing in the math field."
Algebra isn't just for people working in the "math field" (whatever that is)! It's a basic building-block of a practical language and logic, and is necessary for people to function in today's society. I would concede that not everyone needs to be well-trained in calculus, but even then the concept of instantaneous rates and limits is vital.
Math is a language because we can communicate ideas with it. It's also a language that we can use as a thinking tool, just as spoken languages are used automatically as we think, to help crystalize abstract ideas. Algebra helps make a lot of basic abstract concepts more concrete, and helps people think more clearly. I don't think that students should be able to graduate from high school without a practical, working knowledge of algebra. Yes, I agree that integrating it with other subjects such as the sciences is also important to help students' knowledge become practical.
I didn't see anything on the web page explaining
what the author's goals for the language was. If
it's going to be popular, it will need to allow for
easy porting of existing software into BRiX.
I assume that it's more of a proof-of-principle
kind of thing, meant to push the envelope of OS
design but without any realistic hopes of being
popular. But it's gone farther than a simple
proof-of-principle would require. So, I'm
curious what the strategy and goals of BRiX are.
My other complaint is that Brand's new "for
dummies" text is not helpful, just arrogant.
That's what I was thinking too: How would most
gamers really use this feature? Most games suck
up all the CPU you have available, leaving nothing
left for encoding and writing to disk. A game would definitely
not be playable while recording. You might
be right about MPEG encoding cards handling the job,
but I'm sure that would require even more drivers
and possibly new hardware. Anyway, these things are
made for DV, which is only about 640x480
resulotion. Typical DV bandwidth is 29.5 Mbps (uncompressed). I'm
not sure what kind of bandwidth is allowed by
most video cards, but this may be nearly enough.
The article seemed to me to be a kind of glorified
whining, wishing that video card manufacturers would
take notice for some reason and satisfy the
author's personal desire. The case he makes for
the typical consumer wanting this feature is just
silly. Well, at least he made it on/.!
My own reasoning on this subject is you have two choices. You can realize that your mind is your own and that you can control the way you feel, or you can succomb to the idea that your mind is controlled by chemicals and you have no choice but to drug yourself to be satisfied.
I would rather live independant, happy, and free than to handcuff myself to a drug habit with its inevitable ups, downs, and the constant fear of doing permanent damage to my brain. This isn't a moral reason, it's just (un)common sense.
Boring? Of the people I know, the ones that use drugs a lot are the most boring. This is because they aren't looking for satisfaction in accomplishing things or doing anything fun and real. Instead, they only find satisfaction in being stupefyingly doped. Boring!
I need to admit at first that, although I'm curious about the subject, I don't really know enough about general relativity to answer your question.
There are a couple things I do know which might help: The curvature of spacetime is determined, in part, by the "energy density", not mass per se. Mass is one form of energy, and for ordinary objects mass is the dominant form of energy present. It makes a little more sense to me that something like energy density is what determines the shape of spacetime. But only a little -- I still don't have a satisfying understanding of this.
I can answer your question about mass in classical mechanics easily enough. One way to measure the mass of something is to fasten it to a spring. The other end of the spring is held in place (or, specifically, fastened to something much much more massive, for example, the Earth). Then you give the object a little push, and watch how fast it oscillates. A massive object will move with a slower frequency than a light one. This is (or was anyway) how astronauts "weighed" themselves when in space.
The next question is then: How do you describe mass? What is mass really, or where does it come from? Those are deep questions with possible answers coming from the next generation of particle physics experiments. In the next few years, we expect to have a direct observation of a Higgs particle if such a thing exists. This is intimately associated with the origins of mass, for reasons I can't do justice to here.
Thanks, that does help me understand quantum theories of gravity a bit more. I heard some argument at one point that gravity is a spin-2 field because there are no gravitational dipole moments. I'll have to think about that one a little bit, because it doesn't seem obvious to me. My background is experimental particle physics, so I'm weak on the theory here. The best reference I could find on gravitons was from the book by Peskin and Schroeder, on page 126. That gave me the impression that a tensor field was a candidate for gravity just because it was a singularly attractive potential.
I can still argue that your field-theoretical argument for gravity is a little odd since the tensor field is just an ad-hoc stand-in to reproduce GR in a certain (albeit reasonable) regime. If there were an independant reason to believe the field theory explanation of gravity, then it would be a different story.
I can also appreciate the line of reasoning that this antigrav research isn't completely unjustified. It's relatively low-cost I'm sure. But I have a hard time figuring out why they are reproducing Podkletnov's exact experiment instead of just putting together another random assemblage of cool devices. I would give his claims exactly zero credit until there is some reason to believe that he did his work carefully and honestly. On the other hand, there's no reason to build anything different, either.
One problem with his experiment is that it's complicated enough to make it hard to rule out the "usual suspects" (E&M effects) if any anti-grav-like effect is observed. Building the device and making the measurements sounds relatively easy, but interpreting the results could be nearly impossible. Well, at least it'll keep people busy.
One other thing: quantum-mechanically, it's not surprising that gravity is solely attractive: it's a tensor (spin-2) field, which IS solely attractive.
I think your terminology is correct here, but the reasoning is backward. There is no quantum field theory for gravity that has been tested in any way. People realized that a tensor boson would create an exclusively attractive force, so this is a candidate theory to explain the gravitational force. Hence the supposed "graviton". So to say that we know gravity is attractive based on quantum field theory is incorrect. We know that gravity is attractive based on experience. We have a candidate quantum field theory of gravity which has two major drawbacks: 1) it's untested (no exclusive predictions can be observed with our present technology). 2) it's inconsistent with GR, which has been tested to extremes.
I'm not an expert on general relativity, but AFAIK the equivalence principle, which is at the heart of GR, is in a sense the statement that gravitational mass and inertial mass are identical. In Newtonian theory, gravity is an external force that attract masses. In GR, Newton's gravitational force is a "fictitious force", not a force proper. A non-inertial reference frame is approximately the same as an inertial reference frame with an additional fictitious force. Mass (for some reason) creates curvature in spacetime, which is like a non-inertial reference frame in flat space-time.
I've never really understood the need for a quantum theory of gravity, since gravity is not a force to begin with. I hope that some string theorists can set me straight on this some time. (I just need the guts to walk down the hall and sask them point-blank. My fear is that I won't understand the answer.)
As for Podkletnov, I'm genuinely surprised that anybody is taking him seriously. (taking seriously = non-zero funding to investigate his claims.) The LA times article suggests that he is affraid of the credit being stolen if he publishes the details in a peer-reviewed journal. This is crazy since publishing the explicit experiment and its results is his only gaurantee that he will be recognized as the discoverer of the effect!
His other paper that he put on the preprint servers last year was a masterpiece of bogus science, and I can see why he has such a hard time holding a job or publishing anything. There were several logical flaws in that paper, and the experimental technique was horrible and imprecise. For example, there were no measurement errors quoted, which wouldn't even earn him a passing grade in a high school physics course.
My favorite line of reasoning in the paper was that the impulse imparted by his "anti-gravity beam" was proportional to the mass of the test subject. Thus, by extrapolation, if he were to put a hugely massive test subject in the beam, it would receive more kinetic energy than the amount of energy put into the beam. He then sites this as a violation of the equivalence principle! No, it's a violation of conservation of energy, and no one in their right mind would believe that he's observing violation of conservation of energy based on an absurd extrapolation, hundreds of times further than his actual data reaches. If you think about it, this "little goof" invalidates his whole anti-gravity explanation.
After reading that, I just shook my head in amazement. And now he's getting folks at NASA to take him seriously? NASA is desperately hurting for funding, and really shouldn't be dabbling in quackery right now.
This basically explains that the paper is secular, with the goal of unbiased reporting. It was started by Mary Baker Eddy, the founder of the Church of Christ, Scientist, in 1908. This was done more or less as a response to the "yellow journalism" of the day (much more about that in the above link). So far it has won six Pulitzer prizes for journalism.
I agree that most of the tangible results from this kind of research, and practically all basic science research in general, are the by-products of the process. Whenever you have a group of creative people trying to push on the frontiers of knowledge, trying to solve new kinds of problems, you get innovation. (Ouch, it's too bad Bill has brutalized that nice word in recent years!)
It happened with the Apollo missions jump-starting the integrated circuit, it happened with experimental particle physics when the unwieldy international collaborations came up with the web, and it happened in countless other instances.
But besides this, you should see the SETI search as a positive one regardless of the outcome. If they find something, well that's very interesting and opens up all kinds of possibilities. That's obviously worthwhile. But if they don't find anything, that's also valuable information since it helps to sets limits on some of the more slippery factors in Drake's equation. The search is probing a vast area of the sky and a generous frequency band on a scale that has never even been approached before. I think that the project is worthwhile on this basis alone.
Sorry to get political, but I feel it's important to plug basic science research like this as often as possible, since our society at large simply doesn't care or know about what's happening. The current US government is cutting down on basic research funding in an unprecidented way that will cripple progress for decades to come. I think other countries are beginning to follow the US's example and are doing the same. Recent graduates are finding better, more stable jobs in industry and the experience in these fields of science is being evaporated away. This impacts the education programs too, since fewer professors can be supported at universities, and thus the undergraduate science education suffers. This weakens the education of the ed majors too, many of whom will teach science at the high school level and earlier. This is already considered to be the primary weakness of the US education system, stemming from the fact that most students completing a 100-level physics course still lack even the most basic qualitative understanding of the material. Our society is built on technology, which is built on science. If you take away the foundation of a building, you can't expect it to remain standing.
I don't have much in the way of concrete, current numbers right now but obviously NASA is being cut to the bone, and particle physics is also suffering huge cutbacks. Virtually all new initiatives are being canned. I believe the situation is the same in most of the fields of basic science. This is because the public at large isn't putting any pressure on the government to continue supporting science funding. Research funding was not even an issue in the last Presidential election. I found only one statement in all of Bush's campaign materials that even mentioned basic science, which was a single sentence to the effect that it should all be privately funded. This statement is patently outrageous, but no one asked about it, knew about it, or even cared.
Oh, yes! Thanks, I couldn't remember the name of the language last night: Lolo.
I've heard about visual languages for several years, but the first one I've really seen and played with is LabVIEW (the visual programming language is just one aspect of this package). This isn't suited for early education since its focus is on measurement and data analysis. And it's expensive. But I'm sure it's not the only language of its kind. It is easy to learn and debug, and seems to provide just the right visualizations of programing concepts.
Hmm, I've heard a lot of good things about smalltalk, but I haven't learned it myself. So I can't really comment. But I won't let that stop me!:-)
Sounds like smalltalk is great for learning at the college level, or maybe possibly late in high school. But at ages 10-16, I doubt it would work as well.
I feel (though it can be argued either way) that children would learn better from a more procedural language. The first conceptual hurdle is how the computer follows your instructions to the letter. Tracing the flow of an OO program is more difficult than a simple procedural program. Even the concept of functions doesn't come at first. The concepts learned from procedural programming *do* carry over when advancing to OO techniques.
I guess a good young-learning language should be flexible, allowing students to learn as much technique and sophistication as they are ready to. But the entry level should be as simple as possible.
Are there any good programming languages for teaching children these days? Maybe I'm just out of the loop, but I haven't heard of any in the past ~15 years.
I basically agree with this article, but I also think that learning to use a computer at the right time, in the right way, and with ballance must be a good thing.
I learned to program in basic in 4th grade (on a TRS-80), and at the time it worked for me because computers were new and there weren't any more fun things to do on computers. There were practically no games and word processors were terrible. Since the "bar" was so low, any simple programs I could write seemed pretty cool, and kept me working at it. It was open-ended and fostered creativity. However, few other kids were interested in basic programming at that time.
When in high school, I tutored younger children who were learning to program in something like "turtle", where programs controled a drawing turtle. That was more attractive to students, since it was more artistic and gave a more concrete feedback.
But since then I haven't heard of anything for children, and the emphasis seems to be on learning to use software instead. I think that a little background in programming could be more helpful in the long term.
What programming languages are there that are appropriate for using at a young age? They should be relatively simple, allow for as much visualization as possible, and be powerful enough to make interesting things happen. This might just be a high-level library for an existing language, or something totally different.
A visual programming language, where you "draw" loops and such, might fit the bill. Are there any good ones?
It seems like the next step for the ISOS proposal is to model the basic system and see if it "works" economically and practically.
One could construct a computer model where there are N_x users of type X. The user model would include local usage patterns, down-time, CPU power, disk space, network connection, services requested, and budget. Users could include:
"home users" who have cable access and primarily evening usage. Many of these will turn off their computers periodically, experience crashes (if their using Windows), etc. Their service needs would be backups and downloads.
"business PCs" representing individual company/school-owned computers used mostly in the day and are always on, often with faster internet connections.
Business power-users who heavily rely on internet resources such as database retrieval and storage. These are the major consumers in the economic model.
Educational/research power-users like SETI, particle physics, and gene research. These have nearly unlimited needs in terms of both CPU and storage, but very little money. Perhaps users, acting as service-providers, could select which of these activities they would be willing to "sponsor".
Questions can then be asked of the model. What are the bandwidth requirements? Do the economics work out under any realistic configuration? By that I mean do companies save money over purchasing and maintaining server farms, and can home users afford their internet connection and services? If so, how "stable" is the "solution"?
If I weren't busy enough as it is, I would be tempted to work on such a model myself. Sounds like a lot of fun. It might be a good project for a CS major, except that it focuses on practical skills such as model building and analysis.
I totally agree with this post. The transcript sounded like it was making good arguments but none of them are really valid!
A few of my own objections:
Fallacy 2: Computers Allow People to Do things They Could Not Do Otherwise
Woah, I can't even begin to scratch the surface of the ways that computers have enabled myself and society at large. My perspective is that writing letters or adding up a couple numbers isn't what computers are for, and the fact that they can even help with this trivial stuff even argues their case further! I'm in experimental particle physics, which like many other branches of science has virtually exploded with the advent of computers. The experiment I'm collaborating on was basically designed in '94, and collected data in '96-'97. The latest computer technology (well, latest affordable tech anyway) in '94 was not capable of handling the rate of data aquisition necessary for the experiment to run, but it was designed with a little foresight and when it did run, it did spectacularly well. We collected about 30 TB of data. We performed the first round of processing (mostly reconstructing particle tracks) a year later, which required almost a full year to complete running on an average of 50 workstations around the clock.
Now, was this possible without computers? Not even remotely, in anyone's worst nightmares! The calculations alone, done with sliderulers, would probably take roughly 5 quadrillion man-hours. That's between 1 and 2 trillion man-years of stupefying labor!
Okay, most people don't do that much number-crunching on computers that they are aware of, but even just the advances of more mundane activities are worthwhile. Henning talks about letters on computer versus typed. Consider submitting a paper to a journal. In the bad old days one would carefully type up a draft, and hire a draftsman to draw the figures. This would take a week or so, mostly waiting for the drafting work. Then you would mail the paper to a journal, who would select reviewers and mail them copies. If modifications are required, well, you can see where I'm going with this! And you still run the risk of additional typos when the paper is typeset for the journal. My boss told me the story of years ago when he was submitting a paper with computer-printed plots for the first time. The journal refused to accept the figures, asking for the "originals". He tried several times to explain that anything more original than that was just a bunch of ones and zeroes, but they wouldn't accept that answer. So he eventually realized that the printer would print on draftsman's vellum. He printed the plots on that and sent them in. "I'm sorry, I was kidding before. These are the originals you wanted." "Thanks."
Okay this is too long already so I'll wrap up. Computers are much easier to use today compared to the 70's, programs do a lot more, GUIs vastly reduce the learning curve (when done well), computers are much faster and that's a Good Thing, programming techniques are rapidly developing (but at a realistic rate -- just be patient -- think of how much better Perl is than COBOL and it's been less than half a century), standards and openness are all good things, etc.
Obviously the real enemy of modern-day computing is bugs, not security exploits and viruses. I think we all suffer many times more (in terms of productivity loss) from program flaws and crashes than we do from system break-ins. Obviously this ratio varies a great deal from user to user, but overall I think bugs are a problem orders of magnitude greater than what we normally call "security".
I don't know anything about C#, but I do know that Java traps things like array overwrites at runtime. These are very common programming errors in C which can often take months to track down. So not only does Java's "security" in terms of pointer handling and bounds checking protect your system from exploits, it also protects you from programs gone haywire. And it helps programmers to debug their own code as well.
All the same logic about not running applications as root apples here as well. You do it more to protect the system from yourself than from others. I read the article as accusing C# of not providing the kid gloves that other modern languages do. Of course, bantering about "security" seems to attract a lot more attention than the more compelling arguments of "prevents you from screwing up your own system!" We have more fascination with security than we do with common sense.
The current model of software with a fixed purchase price handles one thing well, though. If you feel that an upgrade to a newer version is worth the price of that version, you will buy it. There are problems with this, of course. It encourages bugs in software that you can promise to fix in future versions, since that adds value to the upgrades. But anyway, that's the basic incentive mechanism for the software developer. The consumer has bargaining power by saying "no, that upgrade is too expensive and not worth it, so I won't pay."
The subscription model doesn't work here. If M$ got everyone on a subscription plan, they would have no incentive to improve their software, and every incentive to raise the subscription cost each year. If the consumer decides that they don't like the increased subscription cost, or that the upgrades are too pitiful to warrant the subscription, what can they do? If they refuse the subscription, then they loose the software they have and are sunk.
I like the article, and the idea of this distribution is very appealing to me, who foolishly tends to lean on the bleeding edge of software a bit too often (ouch)! But after finishing the article I was dissapointed that my immediate concerns about the distributions were never answered. Can anyone else fill in the blanks?
How much source code needs to be uploaded for a good, full-featured distribution? Can I do it over a dialup connection, or is a T3 line a prerequisite?
Much of the discussion here has revolved around how long it would take to compile everything. Well, the author just did that, but forgot to give the details about how long it took on his computer, and what computer he had.
But these are trivial issues of installation only. The more important questions deal with the maintanence and daily use of the OS. What about security? What about system stability? I don't think these words even appeared in the article!
I love the idea of a system that updates itself nightly, but the downside is that you could never be certain that the computer would still be working any given morning. This introduces a new kind of downtime: your computer is effectively down until patches come out that stabalize your system! If you ever use your computer for anything that's in any way important, this situation would not be tollerable. I'm playing the devil's advocate here in order to make the point that this is a major issue that needs to be addressed. It may turn out that sufficient testing is done with each software release and major screw-ups are extremely infrequent.
An automatically-updating system (or even manually-updated) is doing a lot of stuff as root, which has to be an enormous security risk. What mechanisms are in place to keep security tight? What are the risks? Since this is a niche distribution so far, I guess security is not a major concern yet. But if the author's predictions come true and this becomes a major distro, what then? How does this distribution concept scale?
As a physicist, I use Fortran a lot. I don't enjoy programming in it, but I do a lot of it because I'm working on a large experiment with tons of legacy code going back 20 years or more. The way I get my kicks, so to speak, is by doing as much as I can with Perl. I probably spend half my time programming in Perl, and the other half in Fortran. These are a great combo, since Fortran is great at the heavy-lifting (numerical stuff), and Perl is great with everything else. But I digress.
OOP techniques can provide amazing advantages for numerical computing, and I'm surprised that there isn't more written about it. When you deal with scientific computing, it's a combination of numerical computation and the usual comp-sci stuff figuring out how to implement simulations and the like. For "the usual stuff", we know very well how OOP can help. But for numerical programming, the benefits are still there.
In numerical work, you deal with a variety of mathematical objects such as complex numbers, vectors, matrices, tensors, algebras, and so on. In working with the math, we tend to use the more sophisticated objects because they are easier to deal with (laziness) and more meaningful. But when you go to code it up in, say, Fortran, all the beauty vanishes and the nitty-gritty details become a menace. Or, you could use OOP and make up things like vector classes, matrix classes, and so on. Then the code suddenly becomes simpler, you save a little time writing the code, and make many fewer mistakes. The code looks more like the math.
I hear that rogue wave makes some very fast C++ libraries for numerical work, though I haven't used them personally. There are other collections out there as well, I'm sure.
My main beef about other languages such as C++ and Java is that there is a blatant lack of concern for numerical programming. Here are some rants:
Complex numbers: In C++, there are a variety of ways of handling complex numbers, but there is no standard as far as I can tell. So by using complex numbers in C, you aren't writing portable code.
Powers: This is an old issue. Why on Earth can't modern compilers handle integer powers efficiently (Fortran's ** operator)?? You must realize that in numerical work, you use **n operations about half as often as addition and multiplication!!!
down-casting: 1.8 / 2 = 0? In Fortran, binary operators taking different types of operands generally do the "right thing" which is to promote integers to floats, floats to complex, single precision to double, etc. C and other languages do the opposite, which is almost always NOT what you want! This must be the #1 source of bugs in C numerical code.
I hate Fortran, I really do. But when doing pure numerical work Fortran's actually a lot better than most other languages I've seen. There's no good reason why other languages couldn't be reasonable in this regard.
Okay, I'll admit that the only two genuinely OO languages I know are Java and C++. And these both derive much of their behavior from C, which is not a very "modern" language any more. But it is still a lot younger than Fortran. (Yeah, Kernigan and Ritchie were going for speed and simplicity in the mid-70s, I understand.) Anyway, in spite of wanting to borrow syntax from C, why can't anyone simply decide to improve on the bad old features of C like those mentioned above?
Whew, okay. I'm done ranting now. I feel much better.
First, physics is inherently descriptive and approximate anyway, so if there is a "real" physics, it has yet to be discovered. Physicists are always working with approximations and numerical methods, but it's best to go as far as you can with "exact" solutions and generalities. Then, when you need to start doing numerical work, you have some place to start from. You can't do numerical integration unless you know what to integrate.
Second, the usual 100-level undergrad textbook in physics tells you a lot of things that you probably don't need to know when designing games (like E&M and some quantum mechanics), but also leaves out the more practical aspects of classical mechanics when dealing with less-than-ideal objects. Once you work with the motion of objects that are not spherically symmetric, you need mechanics at the next level, and you need to work with matrices and vectors. This stuff isn't difficult, but it's not in the typical undergrad textbook. And it does require a bit of mathematics, like most things that are worthwhile.
So it sounds to me like this book does have an important niche to fill, combining undergrad classical mechanics, a sampling of junior-level classical mechanics, and some numerical methods to boot.
Slashdot Mirror
I'm a very religious person, believe in God, and enjoy reading the Bible. So I take offense at this kind of attitude because it seems so unchristian to me. You talk about Jesus, but at the same time are being judgemental, and are being publicly "noisy" about how righteous you are, and so much more favored than the rest of us. I believe that the salvation Jesus Christ gives us isn't about the afterlife, but about here and now, obeying his teachings, following his example.
I wanted to post this because I didn't want others to think that rigid indoctrination is what religion is about. I don't believe the universe is only a few thousand years old, or that there needs to be any conflict between religion and science. I consider myself to be Christian mostly because I see great value in the Bible and Jesus' teachings. I think, for example, Matthew 5-7 encapsulates all of what Christianity is about. This is a fantastic moral discourse, and would bring world peace and prosperity if obeyed. But it's difficult and uncommon advice -- shockingly different from "worldly" (popular) philosophy.
Finally, I don't think it's a conflict to believe the Bible and also the Big Bang theory. I think the accounts of creation in the Bible are technically just ancient Hebrew folklore, but they belong in the Bible because they contain rich moral lessons, and describe the true nature God and the Universe. The Bible isn't a history book. It's not written that way. It's a religious work, and should be treated as such.
Sorry to rant. I'm done now.
It feels wrong to reply to my own reply, but here are two or three more ideas:
You have to consider methods of heat transport when cooking. How quickly does the heat transfer from the inner surface of the lava to the bird? In between you mostly have charred leaves. This prevents radiative transfer (broiling) or convective transfer (baking). So you're left with conduction, which will be very slow through the charred leaves. This allows for a large difference in temperature between the rock and the bird.
If you want faster cooking, I suppose one could engineer a metal cage that keeps the lava a safe and uniform distance from the uncovered bird. A coarse wire mesh should do the trick since lava is so viscous and will quickly solidify. Then the bird would be cooked by radiation instead, very similar to broiling. You could get even heating from all sides, or at least from above and below. And the way the lava cools down works in your favor too, searing the bird and then cooking it. I doubt this would taste better than the wrapped-in-leaves method, but I'm sure it would be faster.
- Topher
I don't know what the exact numbers are, but I know that lava is a fairly good insulator. So it's quite possible that when the inside cools down, the outside remains very hot for a long time. The main reason for this is that the lava contains disolved gasses that create bubbles as it cools. Lava rocks are very lightweight (well, surface flow lava rocks anyway) like solid foam.
I agree that the water in the leaves regulates the temperature inside, especially since pressure is not allowed to build up.
I think there's a simmilar reason that it takes about 25 minutes to barbecue corn in the husk, but only 3 minutes to boil it.
- Topher
See the latest in innovation! Totally changes the way you work, and even think about working! Far superior to unsightly cabinets of dusty electronics, and for a price that is astounding! If you decide to keep it, we'll refund half the purchase price. This is so revolutionary that it makes the PC obsolete. Stop by, and you'll get a free onion!
The English language
has so much redundancy,
you might be surprised!
Algebra isn't just for people working in the "math field" (whatever that is)! It's a basic building-block of a practical language and logic, and is necessary for people to function in today's society. I would concede that not everyone needs to be well-trained in calculus, but even then the concept of instantaneous rates and limits is vital.
Math is a language because we can communicate ideas with it. It's also a language that we can use as a thinking tool, just as spoken languages are used automatically as we think, to help crystalize abstract ideas. Algebra helps make a lot of basic abstract concepts more concrete, and helps people think more clearly. I don't think that students should be able to graduate from high school without a practical, working knowledge of algebra. Yes, I agree that integrating it with other subjects such as the sciences is also important to help students' knowledge become practical.
I assume that it's more of a proof-of-principle kind of thing, meant to push the envelope of OS design but without any realistic hopes of being popular. But it's gone farther than a simple proof-of-principle would require. So, I'm curious what the strategy and goals of BRiX are.
My other complaint is that Brand's new "for dummies" text is not helpful, just arrogant.
The article seemed to me to be a kind of glorified whining, wishing that video card manufacturers would take notice for some reason and satisfy the author's personal desire. The case he makes for the typical consumer wanting this feature is just silly. Well, at least he made it on /.!
Here's a late post, but what the heck.
My own reasoning on this subject is you have two choices. You can realize that your mind is your own and that you can control the way you feel, or you can succomb to the idea that your mind is controlled by chemicals and you have no choice but to drug yourself to be satisfied.
I would rather live independant, happy, and free than to handcuff myself to a drug habit with its inevitable ups, downs, and the constant fear of doing permanent damage to my brain. This isn't a moral reason, it's just (un)common sense.
Boring? Of the people I know, the ones that use drugs a lot are the most boring. This is because they aren't looking for satisfaction in accomplishing things or doing anything fun and real. Instead, they only find satisfaction in being stupefyingly doped. Boring!
I need to admit at first that, although I'm curious about the subject, I don't really know enough about general relativity to answer your question.
There are a couple things I do know which might help: The curvature of spacetime is determined, in part, by the "energy density", not mass per se. Mass is one form of energy, and for ordinary objects mass is the dominant form of energy present. It makes a little more sense to me that something like energy density is what determines the shape of spacetime. But only a little -- I still don't have a satisfying understanding of this.
I can answer your question about mass in classical mechanics easily enough. One way to measure the mass of something is to fasten it to a spring. The other end of the spring is held in place (or, specifically, fastened to something much much more massive, for example, the Earth). Then you give the object a little push, and watch how fast it oscillates. A massive object will move with a slower frequency than a light one. This is (or was anyway) how astronauts "weighed" themselves when in space.
The next question is then: How do you describe mass? What is mass really, or where does it come from? Those are deep questions with possible answers coming from the next generation of particle physics experiments. In the next few years, we expect to have a direct observation of a Higgs particle if such a thing exists. This is intimately associated with the origins of mass, for reasons I can't do justice to here.
Hope this helps -- a little!
- Topher
Thanks, that does help me understand quantum theories of gravity a bit more. I heard some argument at one point that gravity is a spin-2 field because there are no gravitational dipole moments. I'll have to think about that one a little bit, because it doesn't seem obvious to me. My background is experimental particle physics, so I'm weak on the theory here. The best reference I could find on gravitons was from the book by Peskin and Schroeder, on page 126. That gave me the impression that a tensor field was a candidate for gravity just because it was a singularly attractive potential.
I can still argue that your field-theoretical argument for gravity is a little odd since the tensor field is just an ad-hoc stand-in to reproduce GR in a certain (albeit reasonable) regime. If there were an independant reason to believe the field theory explanation of gravity, then it would be a different story.
I can also appreciate the line of reasoning that this antigrav research isn't completely unjustified. It's relatively low-cost I'm sure. But I have a hard time figuring out why they are reproducing Podkletnov's exact experiment instead of just putting together another random assemblage of cool devices. I would give his claims exactly zero credit until there is some reason to believe that he did his work carefully and honestly. On the other hand, there's no reason to build anything different, either.
One problem with his experiment is that it's complicated enough to make it hard to rule out the "usual suspects" (E&M effects) if any anti-grav-like effect is observed. Building the device and making the measurements sounds relatively easy, but interpreting the results could be nearly impossible. Well, at least it'll keep people busy.
- Topher
I think your terminology is correct here, but the reasoning is backward. There is no quantum field theory for gravity that has been tested in any way. People realized that a tensor boson would create an exclusively attractive force, so this is a candidate theory to explain the gravitational force. Hence the supposed "graviton". So to say that we know gravity is attractive based on quantum field theory is incorrect. We know that gravity is attractive based on experience. We have a candidate quantum field theory of gravity which has two major drawbacks: 1) it's untested (no exclusive predictions can be observed with our present technology). 2) it's inconsistent with GR, which has been tested to extremes.
I'm not an expert on general relativity, but AFAIK the equivalence principle, which is at the heart of GR, is in a sense the statement that gravitational mass and inertial mass are identical. In Newtonian theory, gravity is an external force that attract masses. In GR, Newton's gravitational force is a "fictitious force", not a force proper. A non-inertial reference frame is approximately the same as an inertial reference frame with an additional fictitious force. Mass (for some reason) creates curvature in spacetime, which is like a non-inertial reference frame in flat space-time.
I've never really understood the need for a quantum theory of gravity, since gravity is not a force to begin with. I hope that some string theorists can set me straight on this some time. (I just need the guts to walk down the hall and sask them point-blank. My fear is that I won't understand the answer.)
As for Podkletnov, I'm genuinely surprised that anybody is taking him seriously. (taking seriously = non-zero funding to investigate his claims.) The LA times article suggests that he is affraid of the credit being stolen if he publishes the details in a peer-reviewed journal. This is crazy since publishing the explicit experiment and its results is his only gaurantee that he will be recognized as the discoverer of the effect!
His other paper that he put on the preprint servers last year was a masterpiece of bogus science, and I can see why he has such a hard time holding a job or publishing anything. There were several logical flaws in that paper, and the experimental technique was horrible and imprecise. For example, there were no measurement errors quoted, which wouldn't even earn him a passing grade in a high school physics course.
My favorite line of reasoning in the paper was that the impulse imparted by his "anti-gravity beam" was proportional to the mass of the test subject. Thus, by extrapolation, if he were to put a hugely massive test subject in the beam, it would receive more kinetic energy than the amount of energy put into the beam. He then sites this as a violation of the equivalence principle! No, it's a violation of conservation of energy, and no one in their right mind would believe that he's observing violation of conservation of energy based on an absurd extrapolation, hundreds of times further than his actual data reaches. If you think about it, this "little goof" invalidates his whole anti-gravity explanation.
After reading that, I just shook my head in amazement. And now he's getting folks at NASA to take him seriously? NASA is desperately hurting for funding, and really shouldn't be dabbling in quackery right now.
- Topher
http://www.csmonitor.com/aboutus/about_the_monito
This basically explains that the paper is secular, with the goal of unbiased reporting. It was started by Mary Baker Eddy, the founder of the Church of Christ, Scientist, in 1908. This was done more or less as a response to the "yellow journalism" of the day (much more about that in the above link). So far it has won six Pulitzer prizes for journalism.
- Topher Cawlfield
I agree that most of the tangible results from this kind of research, and practically all basic science research in general, are the by-products of the process. Whenever you have a group of creative people trying to push on the frontiers of knowledge, trying to solve new kinds of problems, you get innovation. (Ouch, it's too bad Bill has brutalized that nice word in recent years!)
It happened with the Apollo missions jump-starting the integrated circuit, it happened with experimental particle physics when the unwieldy international collaborations came up with the web, and it happened in countless other instances.
But besides this, you should see the SETI search as a positive one regardless of the outcome. If they find something, well that's very interesting and opens up all kinds of possibilities. That's obviously worthwhile. But if they don't find anything, that's also valuable information since it helps to sets limits on some of the more slippery factors in Drake's equation. The search is probing a vast area of the sky and a generous frequency band on a scale that has never even been approached before. I think that the project is worthwhile on this basis alone.
Sorry to get political, but I feel it's important to plug basic science research like this as often as possible, since our society at large simply doesn't care or know about what's happening. The current US government is cutting down on basic research funding in an unprecidented way that will cripple progress for decades to come. I think other countries are beginning to follow the US's example and are doing the same. Recent graduates are finding better, more stable jobs in industry and the experience in these fields of science is being evaporated away. This impacts the education programs too, since fewer professors can be supported at universities, and thus the undergraduate science education suffers. This weakens the education of the ed majors too, many of whom will teach science at the high school level and earlier. This is already considered to be the primary weakness of the US education system, stemming from the fact that most students completing a 100-level physics course still lack even the most basic qualitative understanding of the material. Our society is built on technology, which is built on science. If you take away the foundation of a building, you can't expect it to remain standing.
I don't have much in the way of concrete, current numbers right now but obviously NASA is being cut to the bone, and particle physics is also suffering huge cutbacks. Virtually all new initiatives are being canned. I believe the situation is the same in most of the fields of basic science. This is because the public at large isn't putting any pressure on the government to continue supporting science funding. Research funding was not even an issue in the last Presidential election. I found only one statement in all of Bush's campaign materials that even mentioned basic science, which was a single sentence to the effect that it should all be privately funded. This statement is patently outrageous, but no one asked about it, knew about it, or even cared.
Okay, now my axe is a little bit lighter.
I forgot to preview that one!
Oh, yes! Thanks, I couldn't remember the name of the language last night: Lolo.
I've heard about visual languages for several years, but the first one I've really seen and played with is LabVIEW (the visual programming language is just one aspect of this package). This isn't suited for early education since its focus is on measurement and data analysis. And it's expensive. But I'm sure it's not the only language of its kind. It is easy to learn and debug, and seems to provide just the right visualizations of programing concepts.
Hmm, I've heard a lot of good things about smalltalk, but I haven't learned it myself. So I can't really comment. But I won't let that stop me! :-)
Sounds like smalltalk is great for learning at the college level, or maybe possibly late in high school. But at ages 10-16, I doubt it would work as well.
I feel (though it can be argued either way) that children would learn better from a more procedural language. The first conceptual hurdle is how the computer follows your instructions to the letter. Tracing the flow of an OO program is more difficult than a simple procedural program. Even the concept of functions doesn't come at first. The concepts learned from procedural programming *do* carry over when advancing to OO techniques.
I guess a good young-learning language should be flexible, allowing students to learn as much technique and sophistication as they are ready to. But the entry level should be as simple as possible.
Are there any good programming languages for teaching children these days? Maybe I'm just out of the loop, but I haven't heard of any in the past ~15 years.
I basically agree with this article, but I also think that learning to use a computer at the right time, in the right way, and with ballance must be a good thing.
I learned to program in basic in 4th grade (on a TRS-80), and at the time it worked for me because computers were new and there weren't any more fun things to do on computers. There were practically no games and word processors were terrible. Since the "bar" was so low, any simple programs I could write seemed pretty cool, and kept me working at it. It was open-ended and fostered creativity. However, few other kids were interested in basic programming at that time.
When in high school, I tutored younger children who were learning to program in something like "turtle", where programs controled a drawing turtle. That was more attractive to students, since it was more artistic and gave a more concrete feedback.
But since then I haven't heard of anything for children, and the emphasis seems to be on learning to use software instead. I think that a little background in programming could be more helpful in the long term.
What programming languages are there that are appropriate for using at a young age? They should be relatively simple, allow for as much visualization as possible, and be powerful enough to make interesting things happen. This might just be a high-level library for an existing language, or something totally different.
A visual programming language, where you "draw" loops and such, might fit the bill. Are there any good ones?
One could construct a computer model where there are N_x users of type X. The user model would include local usage patterns, down-time, CPU power, disk space, network connection, services requested, and budget. Users could include:
- "home users" who have cable access and primarily evening usage. Many of these will turn off their computers periodically, experience crashes (if their using Windows), etc. Their service needs would be backups and downloads.
- "business PCs" representing individual company/school-owned computers used mostly in the day and are always on, often with faster internet connections.
- Business power-users who heavily rely on internet resources such as database retrieval and storage. These are the major consumers in the economic model.
- Educational/research power-users like SETI, particle physics, and gene research. These have nearly unlimited needs in terms of both CPU and storage, but very little money. Perhaps users, acting as service-providers, could select which of these activities they would be willing to "sponsor".
Questions can then be asked of the model. What are the bandwidth requirements? Do the economics work out under any realistic configuration? By that I mean do companies save money over purchasing and maintaining server farms, and can home users afford their internet connection and services? If so, how "stable" is the "solution"?If I weren't busy enough as it is, I would be tempted to work on such a model myself. Sounds like a lot of fun. It might be a good project for a CS major, except that it focuses on practical skills such as model building and analysis.
A few of my own objections:
Fallacy 2: Computers Allow People to Do things They Could Not Do Otherwise
Woah, I can't even begin to scratch the surface of the ways that computers have enabled myself and society at large. My perspective is that writing letters or adding up a couple numbers isn't what computers are for, and the fact that they can even help with this trivial stuff even argues their case further! I'm in experimental particle physics, which like many other branches of science has virtually exploded with the advent of computers. The experiment I'm collaborating on was basically designed in '94, and collected data in '96-'97. The latest computer technology (well, latest affordable tech anyway) in '94 was not capable of handling the rate of data aquisition necessary for the experiment to run, but it was designed with a little foresight and when it did run, it did spectacularly well. We collected about 30 TB of data. We performed the first round of processing (mostly reconstructing particle tracks) a year later, which required almost a full year to complete running on an average of 50 workstations around the clock.
Now, was this possible without computers? Not even remotely, in anyone's worst nightmares! The calculations alone, done with sliderulers, would probably take roughly 5 quadrillion man-hours. That's between 1 and 2 trillion man-years of stupefying labor!
Okay, most people don't do that much number-crunching on computers that they are aware of, but even just the advances of more mundane activities are worthwhile. Henning talks about letters on computer versus typed. Consider submitting a paper to a journal. In the bad old days one would carefully type up a draft, and hire a draftsman to draw the figures. This would take a week or so, mostly waiting for the drafting work. Then you would mail the paper to a journal, who would select reviewers and mail them copies. If modifications are required, well, you can see where I'm going with this! And you still run the risk of additional typos when the paper is typeset for the journal. My boss told me the story of years ago when he was submitting a paper with computer-printed plots for the first time. The journal refused to accept the figures, asking for the "originals". He tried several times to explain that anything more original than that was just a bunch of ones and zeroes, but they wouldn't accept that answer. So he eventually realized that the printer would print on draftsman's vellum. He printed the plots on that and sent them in. "I'm sorry, I was kidding before. These are the originals you wanted." "Thanks."
Okay this is too long already so I'll wrap up. Computers are much easier to use today compared to the 70's, programs do a lot more, GUIs vastly reduce the learning curve (when done well), computers are much faster and that's a Good Thing, programming techniques are rapidly developing (but at a realistic rate -- just be patient -- think of how much better Perl is than COBOL and it's been less than half a century), standards and openness are all good things, etc.
Obviously the real enemy of modern-day computing is bugs, not security exploits and viruses. I think we all suffer many times more (in terms of productivity loss) from program flaws and crashes than we do from system break-ins. Obviously this ratio varies a great deal from user to user, but overall I think bugs are a problem orders of magnitude greater than what we normally call "security".
I don't know anything about C#, but I do know that Java traps things like array overwrites at runtime. These are very common programming errors in C which can often take months to track down. So not only does Java's "security" in terms of pointer handling and bounds checking protect your system from exploits, it also protects you from programs gone haywire. And it helps programmers to debug their own code as well.
All the same logic about not running applications as root apples here as well. You do it more to protect the system from yourself than from others. I read the article as accusing C# of not providing the kid gloves that other modern languages do. Of course, bantering about "security" seems to attract a lot more attention than the more compelling arguments of "prevents you from screwing up your own system!" We have more fascination with security than we do with common sense.
The current model of software with a fixed purchase price handles one thing well, though. If you feel that an upgrade to a newer version is worth the price of that version, you will buy it. There are problems with this, of course. It encourages bugs in software that you can promise to fix in future versions, since that adds value to the upgrades. But anyway, that's the basic incentive mechanism for the software developer. The consumer has bargaining power by saying "no, that upgrade is too expensive and not worth it, so I won't pay."
The subscription model doesn't work here. If M$ got everyone on a subscription plan, they would have no incentive to improve their software, and every incentive to raise the subscription cost each year. If the consumer decides that they don't like the increased subscription cost, or that the upgrades are too pitiful to warrant the subscription, what can they do? If they refuse the subscription, then they loose the software they have and are sunk.
How much source code needs to be uploaded for a good, full-featured distribution? Can I do it over a dialup connection, or is a T3 line a prerequisite?
Much of the discussion here has revolved around how long it would take to compile everything. Well, the author just did that, but forgot to give the details about how long it took on his computer, and what computer he had.
But these are trivial issues of installation only. The more important questions deal with the maintanence and daily use of the OS. What about security? What about system stability? I don't think these words even appeared in the article!
I love the idea of a system that updates itself nightly, but the downside is that you could never be certain that the computer would still be working any given morning. This introduces a new kind of downtime: your computer is effectively down until patches come out that stabalize your system! If you ever use your computer for anything that's in any way important, this situation would not be tollerable. I'm playing the devil's advocate here in order to make the point that this is a major issue that needs to be addressed. It may turn out that sufficient testing is done with each software release and major screw-ups are extremely infrequent.
An automatically-updating system (or even manually-updated) is doing a lot of stuff as root, which has to be an enormous security risk. What mechanisms are in place to keep security tight? What are the risks? Since this is a niche distribution so far, I guess security is not a major concern yet. But if the author's predictions come true and this becomes a major distro, what then? How does this distribution concept scale?
- Topher
OOP techniques can provide amazing advantages for numerical computing, and I'm surprised that there isn't more written about it. When you deal with scientific computing, it's a combination of numerical computation and the usual comp-sci stuff figuring out how to implement simulations and the like. For "the usual stuff", we know very well how OOP can help. But for numerical programming, the benefits are still there.
In numerical work, you deal with a variety of mathematical objects such as complex numbers, vectors, matrices, tensors, algebras, and so on. In working with the math, we tend to use the more sophisticated objects because they are easier to deal with (laziness) and more meaningful. But when you go to code it up in, say, Fortran, all the beauty vanishes and the nitty-gritty details become a menace. Or, you could use OOP and make up things like vector classes, matrix classes, and so on. Then the code suddenly becomes simpler, you save a little time writing the code, and make many fewer mistakes. The code looks more like the math.
I hear that rogue wave makes some very fast C++ libraries for numerical work, though I haven't used them personally. There are other collections out there as well, I'm sure. My main beef about other languages such as C++ and Java is that there is a blatant lack of concern for numerical programming. Here are some rants:
I hate Fortran, I really do. But when doing pure numerical work Fortran's actually a lot better than most other languages I've seen. There's no good reason why other languages couldn't be reasonable in this regard.
Okay, I'll admit that the only two genuinely OO languages I know are Java and C++. And these both derive much of their behavior from C, which is not a very "modern" language any more. But it is still a lot younger than Fortran. (Yeah, Kernigan and Ritchie were going for speed and simplicity in the mid-70s, I understand.) Anyway, in spite of wanting to borrow syntax from C, why can't anyone simply decide to improve on the bad old features of C like those mentioned above?
Whew, okay. I'm done ranting now. I feel much better.
First, physics is inherently descriptive and approximate anyway, so if there is a "real" physics, it has yet to be discovered. Physicists are always working with approximations and numerical methods, but it's best to go as far as you can with "exact" solutions and generalities. Then, when you need to start doing numerical work, you have some place to start from. You can't do numerical integration unless you know what to integrate.
Second, the usual 100-level undergrad textbook in physics tells you a lot of things that you probably don't need to know when designing games (like E&M and some quantum mechanics), but also leaves out the more practical aspects of classical mechanics when dealing with less-than-ideal objects. Once you work with the motion of objects that are not spherically symmetric, you need mechanics at the next level, and you need to work with matrices and vectors. This stuff isn't difficult, but it's not in the typical undergrad textbook. And it does require a bit of mathematics, like most things that are worthwhile.
So it sounds to me like this book does have an important niche to fill, combining undergrad classical mechanics, a sampling of junior-level classical mechanics, and some numerical methods to boot.