An idea I had a while ago was to mix rasterization and ray tracing to get the benefits of both. Rasterize the entire scene into simple color values, but also calculate intersection points and reflection, refraction, and lighting vertices during the rasterization process. That would hopefully benefit from data locality to generate all the initial ray intersections faster than the traditional approach of tracing one ray at a time. It looks like other people have actually published papers about doing that with current GPUs already and it seems to work.
It seems like large surface reflections and refractions could also benefit from an initial raster step before completing the ray tracing process. I think even in the case of global illumination most of the raster pipeline would still be useful for computing portions of the scene. I'd be interested if you have any more information on the topic since it looks like you've researched more than I have.
Sorry, I was referring to the number of primitives (lines, planes, spheres, triangles, etc.) in the scene.
Most rasterization methods require that every primitive be rendered, leading to an overall O(n) time complexity. Drawing only surfaces facing the viewer only reduces the workload by half, although visibility information can be computed from every possible viewpoint in the scene to any other portion of it (I think at least one of the Quakes did this) so that only a portion of the scene has to be rasterized. BSP trees can be used to exclude completely hidden surfaces. Even so, all the visible or partially visible primitives still have to be rendered.
Ray tracing begins with a fixed number of rays that's only dependent on the (constant) resolution and size of the display. The primitives in the scene are put into a space partitioning tree (oct-tree or BSP for example) and only the primitives that are in the same spaces the ray passes through are checked for intersection. By dividing the scene space into a tree, the number of intersections between ray and primitive that must be checked is O(log n), since each decent into the tree rules out at least half of the remaining primitives on average. Since ray tracing causes new rays to be spawned from intersection points for reflections, refractions, and lighting, the total number of rays is usually limited to a constant number for each initial ray, keeping the overall time in O(log n).
The constants in the time complexity obviously matter; It's much quicker to render a single pixel to the color and depth buffers than to trace an entire ray and its refractions through a scene. Even if a ray takes a thousand times longer to traverse a scene than to render a pixel, and each initial ray can spawn a thousand child rays, a scene with a billion elements could easily take less time to ray trace than to rasterize.
The guy spent $7500 on the case and forfeited the right to pursue further legal action against the city for being wrongfully arrested, but hey at least his criminal record is clean now? His record should have been clean the whole time, so there's no victory there either...
I bet the city ended up spending considerably more than $7500 to deal with the situation. That's a win in my book.
Common decency is one thing. Circuit City did not extend this courtesy to Michael, which is why he had no reason to extend the courtesy of showing his receipt to the "guard" at the door.
The time and place to examine merchandise leaving the store is at the cash register. All merchandise must be out in the open at that point, and if the cashier is too stupid or incompetent to make sure all the items are paid for, it's the fault of the cashier and the company that hired them. After the customer has exchanged payment for the items, the store has absolutely no claim on the items or on the person carrying them. Asking to see the receipt is not common decency, it's rude. The implication is that either the customer is a thief or the store is incompetent and undeserving of the customer's patronage. The guard at the door is to cover the sloppy asses of the underpaid cashiers at the inconvenience of the customer. Nothing more.
I would like to inspect all the cash registers and financial systems of the stores I shop at in order to ensure that they didn't overcharge me (how do I *really* know that they charged my credit card what it says on the terminal?). If I can't do that, then they can't ask to see my receipt after I've paid. It's as simple as that.
If you're concerned about common sense and decency, ask the stores to have some.
i know, that's some really wacky fascist thinking on my part, right?
Yes, the idea that normal people should listen to and obey anyone in authority without question is decidedly fascist. The degree of fascism is small, but who cares? Not seeing one measly receipt is no big deal for Circuit City either. They should just accept the fact that some people don't want to show their receipt and have the common sense and decency to accommodate them.
Other problems I see with raytraced games are the exponential increase of processing with higher resolutions or higher light source counts, the fact that poor raytracing actually looks worse on higher resolutions, the increased production and programming costs and the fact that graphics companies will not like seeing the investments made in their current GPU architectures melt away.
Ray tracing is actually more efficient. With the scene in a proper data structure, ray tracing can be O(log n) for a scene with n objects. Raster graphics is almost always O(n), even with advanced game engines that only draw portions of the game world. Every possibly visible object is rendered into a depth buffer to compose the entire scene, but with ray tracing only the x*y rays need to be traced through a tree of objects and checked for intersections. The tradeoff in ray tracing speed is how many reflections, refractions, and light sources to accumulate. Some complex scenes may require hundreds or thousands of rays for a single pixel on the screen, but that's an absolute maximum given the resolution. It matters much less if the scene has thousands, millions, or billions of objects in it. Look at the stuff OpenRT has done with small (8 to 16 core) clusters of generic CPUs. Some of the demos are faster than GPUs for large scenes, while others do things rasterization just can't do, like the accurate real time simulation of a reflective headlight.
Raytracing has no advantage over rasterizing for opaque surfaces. Rasterizers are faster there, since their performance is not tied directly to the screen resolution.
The advantages lie in refraction/reflection/shadows/translucency, which are painful to implement with rasterizers.
Actually, there's a big advantage. Raytracing is O(log n), but rasterization is O(n). OpenRT's demo of a 350 million triangle model of a Boeing rendered in real time on a single PC (without GPU support) is a good example. The entire model doesn't even fit in memory, so visible surfaces are cached. The result is still realtime (although only a few FPS) with incredible detail. Go slashdot the server and watch the movie. Modern raster based cards can only render that many triangles in a whole second with all their fancy hardware, if they're lucky.
Or is it? Simply means games will appear more eye-candy than they currently are. Gameplay will not change. EA will continue to use take last years sport game, through some new people into it, perhaps introduce some bug which makes it unusable and peddle it as The New Deluxe Edition. I wonder how many geometric objects it will be able to handle (and whether it handles transparancy with textures and patterns well) Having done a bit of raytracing I'm familiar with how quick things can bog down. It'll probably be a bit clunky at first, but get much better as horsepower and horsepower/dollar ratio improve.
With raytracing, there are lots of new possibilities. For one thing, reflection and refraction actually work like they do in real life. That means accurate mirrors, lenses, and water refraction. Lights can work accurately if you want them to, and radiosity can be precomputed for static scenes. That may just be eye candy to most people, but there are potentially game-play enhancements that make real life optics part of the game. Most of it (except good lenses) has been faked before with rasterization, but raytracing will actually let you set up a series of mirrors and telescopes to peek around corners in a FPS for instance. I can imagine a true hall of mirrors in an FPS would be at least a little more interesting than what we have now, too.
The other big technological benefit of raytracing is that it's asymptotically faster than rasterization. Raytracing is O(log n) versus O(n) for rasterization, which means that even though raytracing is currently slower (the constants involved in raytracing are higher), after the break even point is passed much less of the available computational power will be needed to render the scene and can instead be used for physics and AI.
Sure, until enterprising individuals build plants in the US to make the goods we were previously importing, but at a lower price. And those plants start hiring US workers.
You assume that's possible in the first place. Where's the profit to be made in building a factory here and paying an order of magnitude higher salary for fewer working hours with more breaks and benefits? How are we going to get $5 t-shirts from an American factory?
The fact is, Americans have had an asymmetric economic relationship with lots of foreign countries, essentially trading our higher standard of living for the low cost labor of people in other countries. In a closed economy it's much harder to do that. Not everyone can be middle class when the factories require cheap grunt work. The smaller middle class means fewer luxuries will be purchased, meaning the upper classes feel the impact just as bad as everyone else. Who's going to buy overpriced CDs, movies, computers, and iPods when they can only get a $5/hr factory job?
Returning to an industrial society is not a viable goal for the U.S.
My place can be small (one house) or large (an entire island). Just like property in SL is already. Sure, the transition will be a bit tricky (at what point exactly are people transfered to a different server, and how do they "see" the content inside/outside?), but that's a technical challenge that is, in principle, not that hard.
In fact, I'd be perfectly happy to have it work the Oblivion way (e.g. you click on the door, you are teleported inside. Windows both ways are faked with textures if at all.)
The traditional solution is to use portals. The client has to maintain connections to each realm or world that it can see, and it just renders objects from all possibly visible worlds that are visible through the portals, which act as interfaces between the worlds. You could do Escheresque things like portals that map onto
the inverse of themselves and other fun stuff. The load would be mostly client side, since the realms would not interact directly; only at the interface where objects transfer between them. Even that is a formality and can be handled entirely by the client if it can "own" the same object in both worlds; Each server knows that objects crossing the portal unspawn, and the client respawns the object in the new world. If the two worlds care about "continuity" at the portals the transfer can be mediated to appear seamless or to obey restrictions on transfer of objects, etc. If the servers sign (or HMAC) the objects being transferred, the client can be responsible for all the data transfer as well.
If the voter can't verify the hash of their vote receipt on the spot (which they can't), then the machine can easily issue them a bogus receipt that looks right as far as the human-verifiable portion of the receipt goes, and instead record a correctly hashed, correctly signed false vote. The voter later discovers their receipt doesn't match, and they have no way at all to prove that they voted the way they say they did. You have no way at all to prove that the authentic but intentionally malformed receipt you have is valid, and it becomes a matter of human vs machine's word.
Correct, the machine can always print out a bogus signature of a hash too.
Further, in order to expose problems in the system, the voter is required to expose the way they voted. If coercion is going on, this is exactly what you don't want to have as a predicate for detecting fraud. Also if there was some intention of throwing the election so that the entire results are considered invalid, because it's a hash chain, you would only have to invalidate or destroy the record of one very early vote, and the rest of the records would be un-verifiable.
Not necessarily. A photocopy of a tampered vote could be submitted anonymously, or even entered into a form on the Internet. If the tampering was a vote change after the polls closed, the claim could be automatically verified. Otherwise, the ballot could just be turned in to the local party who the voter voted for.
A hash chain is valid from any given point forward, and destroying receipts does not invalidate the chain. Destroying the entire chain itself would obviously destroy the record of votes. That's possible with paper ballots as well, and retaining the secret ballot makes it impossible to prevent against that scenario in a recoverable way. The only solution is a re-election.
Destroying a single entry in the chain would invalidate only the votes immediately after it before a valid receipt could be found. In general, destroying a ballot is election fraud and would require an investigation and possible re-vote anyway. It would also be much more detectable than losing a single paper ballot.
This doesn't gain you anything over a machine which simply records the vote in a dual human/computer readable format onto a card that the machine is physically given by the voter, and which is fed into a locked ballot box after verification by the voter. Part of the record on the card would be machine #, vote # by that machine, and time. You can't stuff the box because the stuffed cards would be out of sequence or else at a time when the polls have closed (or else a whole lot of votes 2 minutes before they closed). You could only forge this record if people physically entered the voting booth multiple times, which in an anonymous ballot is only defensible with human precautions which are already present.
A hash chain does provide vote verification; it's possible for anyone to sum all the votes and calculate the final totals while allowing every voter to ensure their vote was recorded. It's easier to "lose" paper ballots and their electronic equivalent. Having paper ballots does prevent electronic ballot stuffing, though.
If I haven't mentioned it before, I realize a hash chain is only a small part of a secure voting system. It is one way to provide verified vote counting, but there are better (complete) solutions out there. I think most of them rely on some form of chaining votes so that it becomes statistically impossible to alter votes.
I don't think any of them really address the man in the middle attack very well, as you pointed out. A machine can always pretend to record votes but actually perform the vote itself and issue a bogus receipt, just like a polling place can discard all its ballots and replace them with their own. If necessary, the polling place can print forged duplicates of its log books and ballots and use them in the election, then turn in the official ballots filled out as they desire. Replace th
Except that if the machine is programmed to alter a vote, and it knows the hash chain at the time it wants to alter the vote, it is able to record the altered vote, with a correct hash at the time the vote is recorded. This is what I meant when I said that if the machine has everything it needs to record the correct vote, it also has everything it needs to record an altered vote. The voter isn't able to manually verify the vote's hash, and even if every voter took a receipt and came back to verify it later, the vote on the receipt wouldn't agree with the machine, and you would have no way to reconcile it.
Didn't I specifically mention that receipts were generated with; 1) the latest hash value in the chain, and 2) the current vote? Each receipt is enough to verify one step of the hash chain. The voter doesn't need to manually verify the hash, because they can do it later at home with a simple web interface. The voting machine can't magically change the value on their receipt to match a forged hash chain. IF the hash printed on their receipt is in the hash chain, then it MUST be the result of hash(last_hash || vote). If the last hash in the chain before their vote (which the user can see), concatenated with the vote they picked (Which is also printed on their receipt) does not equal the hash printed on their receipt, then it's obvious the vote has been tampered with. Just as importantly, if a voter cannot find their signed hash in the hash chain, they know their vote was discarded and can prove it. The key is that the voter's actual vote is printed on the receipt; it's trivial to prove the error in the hash chain. The hash on each receipt can even be signed by the machine to prevent naughty voters from forging receipts. Statistics prevents vote tampering. For n voters, If a random sqrt(n) votes are changed and a random sqrt(n) number of people check their votes, the probability of the tampering being discovered is ~0.5. For larger numbers of tampered votes or checked receipts, the probability quickly tends toward 1. sqrt(n) is far too small a percentage to sway an election, but the number of computer geeks who can write their own correct implementations of hash functions almost certainly is less than sqrt(n) of most voting populations, not to mention "normal" people who could use a simple interface to verify their votes.
Hash chains verify that none of the data was changed after the chain was forged, but it doesn't guarantee anything about the integrity of the data as it is being chained. They assume that the author of the chain can be trusted, and the principle being protected against here is that we can't trust voting machines. That's why an inalterable record needs to be created which is verified at the time the vote is cast. The machine doesn't have access to the ballot cards, the voter gives one card to the machine per vote, so the machine is physically unable to alter, destroy, or insert votes. The cards once fed to the machine have the voting record printed on them, and the voter verifies the record is accurate. Once printed, the card gets inserted into a locked ballot box, or else returned to the voter if they indicate their vote was not as expected.
The receipts count as the unalterable record. Obviously a simple hash chain has voter coercion problems since every voter carries their own receipt which proves how they voted. A solution is to let voters print a receipt for *any* of the previous votes, thus letting them choose an existing vote to present as their own if they are being coerced. Another solution is to only print receipts for 10% of voters, or some other way of making it impossible for voters to be coerced. I agree that a physical voting token should also be required to prevent the machines from ballot stuffing. Preventing ballot stuffing is a harder problem than preventing vote changes, because it ultimately comes down to the problem of accurately and universally identifying every voter.
Richard Dawkins is about as rational as any of the hoplophobes at the VPC. Anyone who professes to be an atheist has voided any claim at rationality, period. It requires just as much, if not more, blind faith as any actual religion.
An atheist is just an agnostic who has concluded that the probability of encountering valid proof of the correct god (e.g. not the one that the other half of the world believes in) within his or her lifetime is so low that continuing to evaluate supposed evidence for a god is a waste of time. It's the same reason sane people are no longer searching for the philosopher's stone or the fountain of youth.
Don't take that as a knock against religion or belief in god either; evolution can make use of absolutely any variation in population to improve fitness. I am not convinced that religion is useless. It's quite possible that it was (or is) a necessary social mechanism required to move beyond simple hunter-gatherer tribal systems and develop technology. It is probably not irreplaceable.
This suggests that during a recount or investigation, every voter would show up with his or her vote receipt. Lose even one receipt and you can't verify the votes.
No, the voting machine stores the hash chain as well. The point is that if a vote is changed at any point, none of the hashes after that will match the receipt, forcing vote tampering to occur within the machine. If someone tried to tamper with a vote after all voters received their receipts, every vote after and including the tampered vote would not match the receipt of the voter, or would not generate a valid hash chain.
The idea is that you have to make the "authority" on which vote is which an immutable record. That is to say something which can't be changed after the vote has been cast. There's nothing in the computer world where this is the case. Not even cryptography would suffice since the voting machine does all the cryptography, and it could easily show you one cryptographically signed vote and record a different cryptographically signed vote. If it has everything it needs to do the original signing, it has everything it needs to forge the signing of different data.
This can be fixed by using a hash chain. The machine starts with an initial hash. After each vote, the value of the vote is salted and appended to the hash, which is then hashed to become the next hash in the chain. The final hash is reconstructible from the initial hash and the sequence of votes. The important thing is that every receipt has a copy of the current hash. There is no way to modify the sequence of votes without making at least one of the receipts invalid, and no way to certify the results without having a complete hash chain of all the votes. Obviously votes can be appended to the end (ballot stuffing), but that's a problem that cannot be fixed in general.
To check the results of voting, publish the entire hash chain without the vote or salt values. Everyone can verify that their vote is present in the hash chain by looking up the hash on their receipt in the chain. If someone can't find their hash, their vote was lost or changed. To prevent voter fraud, the machine can sign the hash on the receipt to ensure that only valid hashes can be looked up.
This scheme alone does not prevent vote-buying, since the receipt must have the salt and vote value on it to verify to the voter that the correct vote was recorded. However, it would be possible to issue only 1/p*sqrt(n) receipts to random voters so that by the birthday paradox no vote tampering could be accomplished that would affect more than sqrt(n) of the votes, where p is the percentage of people who would check their votes on average.
I might be able to prove it's in the final tally. You, too, might be able to prove it's in the final tally. 99% of the voting population, however, have not studied cryptography and would have to rely on an expert to check their vote (and, of course, such a system would have to be designed to make it impossible for the voter to prove to someone else which way they voted).
The 99% part doesn't matter. It's the birthday paradox; the probability that no tampered vote is verified is very low if the number of votes tampered with plus the number of votes verified is bigger than the square root of the total number of votes. The square root of the number of votes is far less than 1% of the voting population (about 11,000 people, given the turnout in 2004), meaning that effectively no vote tampering could occur. If even one vote is found to have been tampered with, it's enough to start a campaign of checking everyone's votes to see how many were really tampered with. I'm sure you could get more than 11,000 geeks to cryptographically verify their vote.
It's the same as buying a car. Yes, you are free to go buy a car from a dealer with no terms and conditions attached if you can negotiate that deal with the dealer. Or, if the dealer says "look, this is a brand new concept car that's not available for sale yet -- you can buy it, but you have to keep it in a closed garage and you can't take the bib and diaper off it until we tell you" then that's also a completely legal contract if you agree to buy that car. You don't get to tell the dealer "yes, I'll follow your terms" and then turn around and say "I own the car, I can do whatever I want." You entered into an agreement with the dealer and you are bound by it.
EULA's have no verbal or written agreement of the licensee. It's absurd to think that the mere action of opening or using something implies agreement to a contract. Obviously no judge has yet been handed evidence in a case in a sealed envelope with a EULA stating "by opening this envelope and reading the enclosed evidence you agree to rule in my favor." I think that would quickly turn the court's opinion against the validity of EULAs.
EULAs come down to property rights. If I have been sold a box or CD without agreeing to a contract, then it's my property. If I burn the shrink wrap license off the box, does that count as agreement to its terms? If I manually change the right areas in memory in *my* computer, does that count as agreement if the program starts working after the EULA is displayed? What if I change it so the EULA doesn't even display, or if I change the text and then agree to it? The courts have said that fair use covers usage of software. There's no legal rights that a EULA needs to grant in order to use a copy of software that you own, so long as you're just using it on your own computer and not accessing some service over the Internet.
Neural nets / genetic algorithms / learning systems These all belong to the family of hill-climbing optimizers. These approaches work on problems where continuous improvement via tweaking is helpful, but usually max out after a while. We still don't really understand how evolution makes favorable jumps. I once said to Koza's crowd that there's a Nobel Prize waiting for whomever figures that out. Nobody has won it yet.
Evolution gets stuck at local maxima for millions and billions of years. Cells as we see them today are basically the local maxima that evolution has been unable to surpass despite the obvious advantages of integrating any number of other elements to create stronger, faster, and more robust cellular structures. Medicine has made improvements upon evolution that were too far beyond the local maxima that blind evolution could achieve. The idea of species randomly cutting themselves open to try to fix things is absurd, except that it happened and we can now fix a host of problems with ourselves. Evolution had billions of years and trillions upon trillions of effective computational resources to create intelligence, and this is as far as it got. In roughly fifty years (a couple hundred if you count the concepts of Babbage and Leibniz) we have gone from simple computers to computers that can beat humans at chess, face recognition, information storage and retrieval, raw numerical ability, and are starting to match our ability in autonomous control of vehicles and many other areas.
AI is one of those fields, like fusion power, where the delivery date keeps getting further away. For this conference, the claim is "some time in the next century". Back in the 1980s, people in the field were saying 10-15 years.
Flight was like that. Human powered flight even more-so. People have dreamed of flight for thousands of years, but it took heat engines to achieve it. Human powered flight required advanced engineering and lightweight composites, and it still happened decades after powered flight. Just because a technology is always "10-15" years away doesn't mean anything about its possibility; it says more about the people doing the predictions than anything else.
Intelligence is not well defined. It is very hard to say how much of what we call "intelligence" is in fact the ability to make many connections between facts stored in a very sophisticated memory architecture. Simply building a machine able to process information very quickly achieves nothing because, without learning and a social context, it does not know what information to acquire and process. In human experience, academically brilliant people often fail because they work on the wrong problems, or without access to necessary knowledge.
You're being very anthropocentric in your definition of intelligence. Machines will probably have vastly different basic goals than humans, although they will probably have goals that are subgoals of what humans want. Creating a general intelligence that only wants to survive and reproduce is no different than trying to breed "better" humans; it leads to things like massive racism and genocide.
Nothing is actually achieved without creativity. We do not know what that is, or to what extent it is a social construct (i.e. it takes a developed society to have the necessary systems in place to translate an idea into a concrete reality.) And this leads onto the third point. It is no good having a highly intelligent, creative machine if its use of resources is such that it cannot replicate in large numbers. It may be that machine intelligence will ultimately replace human intelligence, but it may be that it will simply be too resource hungry. In effect, there may be a threshold of capability needed to solve some problems, and it may be that machine intelligence will run out of energy before it scales sufficiently to solve those problems. A machine society might, in effect, get stuck in the machine 19th century because coal or oil became a limiting resource. (In the same way, the energy and resources needed to be consumed to achieve a first independent space colony may exceed the total energy and resources available on Earth. It may be that a billion years or so of eukaryotic evolution has actually resulted in the optimum balance of intelligence, creativity and resource consumption, and that any attempt to exceed the present capability will tip us into declining resources faster than we can improve matters.
Creativity is just trying lots of different possible solutions to problems until one works. What humans usually call "creativity" is just normal intelligent humans trying a few wilder solutions than other people would. Most people do what their friends, neighbors, and ancestors have done. "Creative" people try whatever they can think of.
Nanotechnology will be able to achieve far more with fewer physical resources and less energy. Obviously there is a carrying capacity for intelligent species on the Earth; that's when space exploration begins in earnest. Then galactic exploration, and probably eventually a spread to as much of the universe as possible. In the end, the universe is finite. Humans and machines will have to realize that sooner or later and design societies that can remain stable (or shrink) happily. Ultimately the universe dies, and figuring out how to die gracefully will be the last big question for intelligent beings to answer. If it turns out that the universe is actually too low in energy to support space exploration, at least intelligent machines would have the ability to see that objectively without the emotional baggage humans would attach to that.
In many ways I hope this is wrong. But the argument that only one superior machine is necessary is, in fact, an inductive step too far. It is assuming that "intelligence" on its own can solve a class of problems which may involve a number of constraints which cannot be avoided - like the Laws of Thermodynamics, or the need for excessive amounts of energy.
I think the basis case is "Let humanity equal intelligent species number 0. If humanity can generate intelligent species number 1 then by induction intelligent species number x can generate intellig
FLASH used exactly that concept, and it lead to the discovery of the fact that there can be no quantum cloning. It was hypothesized that it was possible to tell whether an electron (or any other particle) was entangled or not, basically by seeing if a measurable property changed when another entangled property was measured. It relied on being able to make a perfect copy of the quantum state and measuring it repeatedly to get exact measurements in violation of the uncertainty principle. If that could be done, then measuring an entangled particle would transmit measurable information to the other entangled particle(s) immediately. It turns out that perfect (100% accurate) quantum cloning is not possible, so no matter which particle gets measured first there's no way to tell on the other end. It's interesting that the actual limit on the probability of success in quantum cloning (which has been measured physically and found to match the predicted value well) is exactly the limit necessary to prevent causal inconsistencies or faster than light information transfer.
Unfortunately, consumers brought DCMA on themselves.
See, "fair use" is part of a bargain between creators and consumers. Consumers are free to do whatever they like with creations as long as the remain within fair use. But starting around 1995, consumers began to violate "fair use" on a scale previously unimagined. Consumers violated the bargain, so DCMA was created to try to restore some balance. If consumers didn't want DCMA to come along then they shouldn't have been (and shouldn't be now) uploading and downloading every piece of IP they can get their hands on.
The DMCA has effectively stopped zero criminals. CSS is broken, AACS is broken, and any future DRM technology is broken by definition. Pre-release movies and songs flood the Internet. On the other hand, the DMCA is used to issue take-down notices by Scientology and other evil corporations, or to sue mod chip makers. Basically, the DMCA is used entirely for the manipulation of normal people and censorship of copyrighted works, and never to combat the majority of actual copyright infringement on the Internet.
Your assertion that customers brought the DMCA on themselves is true, but not for the reason you stated. The citizens of the U.S. brought the DMCA (and tons of other stupid laws and illegal wars to boot) on themselves by being lazy, pathetic excuses for informed citizens of a democratic nation.
From the perspective of an application developer, this is pure nonsense. I practically don't care wether my DB stores data in columns or rows or whatever.
You do care; you want it to be as fast as possible. This has less to do with how you interface with the database than it does with how fast reads are versus writes.
What I need is a good, consistant layer that can handle object-based tree structures - nothing more, nothing less. I want to dump my Java/Objective-C/C#/C++/PHP/Python objects in some storage layer, and I want to be able to get it back, search for it, etc.pp.
Look into operating systems with a single level store. They are not very common (one example is OS/400), but they are the perfect design for what you want. Objects are stored at permanent virtual addresses and everything that can't fit in memory is swapped out to secondary storage. No concept of loading or storing objects; they just exist until deleted.
Ugh. It was your sister's birthday, couldn't you drop the crusader bit for one lousy day?
I'm sorry, but this story just wreaks of petty selfishness. Don't let morality get in the way of being a good person.
Apparently these days people prefer pansies over patriots. A "good person" challenges authoritarian anti-freedom practices whenever they can. A much more important question is how could any officer in the U.S. arrest a man on his sister's birthday for exercising his rights? The officer is completely responsible for his own actions.
If you think that circuit city has a right to look through everyone's bags, and that police officers have a right to demand ID, just suppose that some terrorist watch list gets badly implemented everywhere instead of just in airports. How would you like having to show your ID every time you do anything or go anywhere, and having your vehicle and person searched at every opportunity simply because your name is similar to someone on the watch list and your license plates and credit cards get linked to the watch list as a result?
No one in their right mind would have thought that U.S. citizens would be subject to mandatory search and seizure of their liquids in order to travel. Not many today believe that in the near future everyone will be subject to random searches and mandatory identification, but it's obvious that it's coming. National IDs, overeager cops, terrorism fear mongering, the only thing missing is some sort of leeway (violation of the fourth amendment) to perform warrant-less searches. We already have warrant-less wiretapping, so it can't be very far down the road.
Slashdot Mirror
An idea I had a while ago was to mix rasterization and ray tracing to get the benefits of both. Rasterize the entire scene into simple color values, but also calculate intersection points and reflection, refraction, and lighting vertices during the rasterization process. That would hopefully benefit from data locality to generate all the initial ray intersections faster than the traditional approach of tracing one ray at a time. It looks like other people have actually published papers about doing that with current GPUs already and it seems to work.
It seems like large surface reflections and refractions could also benefit from an initial raster step before completing the ray tracing process. I think even in the case of global illumination most of the raster pipeline would still be useful for computing portions of the scene. I'd be interested if you have any more information on the topic since it looks like you've researched more than I have.
Sorry, I was referring to the number of primitives (lines, planes, spheres, triangles, etc.) in the scene.
Most rasterization methods require that every primitive be rendered, leading to an overall O(n) time complexity. Drawing only surfaces facing the viewer only reduces the workload by half, although visibility information can be computed from every possible viewpoint in the scene to any other portion of it (I think at least one of the Quakes did this) so that only a portion of the scene has to be rasterized. BSP trees can be used to exclude completely hidden surfaces. Even so, all the visible or partially visible primitives still have to be rendered.
Ray tracing begins with a fixed number of rays that's only dependent on the (constant) resolution and size of the display. The primitives in the scene are put into a space partitioning tree (oct-tree or BSP for example) and only the primitives that are in the same spaces the ray passes through are checked for intersection. By dividing the scene space into a tree, the number of intersections between ray and primitive that must be checked is O(log n), since each decent into the tree rules out at least half of the remaining primitives on average. Since ray tracing causes new rays to be spawned from intersection points for reflections, refractions, and lighting, the total number of rays is usually limited to a constant number for each initial ray, keeping the overall time in O(log n).
The constants in the time complexity obviously matter; It's much quicker to render a single pixel to the color and depth buffers than to trace an entire ray and its refractions through a scene. Even if a ray takes a thousand times longer to traverse a scene than to render a pixel, and each initial ray can spawn a thousand child rays, a scene with a billion elements could easily take less time to ray trace than to rasterize.
The guy spent $7500 on the case and forfeited the right to pursue further legal action against the city for being wrongfully arrested, but hey at least his criminal record is clean now? His record should have been clean the whole time, so there's no victory there either...
I bet the city ended up spending considerably more than $7500 to deal with the situation. That's a win in my book.
Common decency is one thing. Circuit City did not extend this courtesy to Michael, which is why he had no reason to extend the courtesy of showing his receipt to the "guard" at the door.
The time and place to examine merchandise leaving the store is at the cash register. All merchandise must be out in the open at that point, and if the cashier is too stupid or incompetent to make sure all the items are paid for, it's the fault of the cashier and the company that hired them. After the customer has exchanged payment for the items, the store has absolutely no claim on the items or on the person carrying them. Asking to see the receipt is not common decency, it's rude. The implication is that either the customer is a thief or the store is incompetent and undeserving of the customer's patronage. The guard at the door is to cover the sloppy asses of the underpaid cashiers at the inconvenience of the customer. Nothing more.
I would like to inspect all the cash registers and financial systems of the stores I shop at in order to ensure that they didn't overcharge me (how do I *really* know that they charged my credit card what it says on the terminal?). If I can't do that, then they can't ask to see my receipt after I've paid. It's as simple as that.
If you're concerned about common sense and decency, ask the stores to have some.
i know, that's some really wacky fascist thinking on my part, right?
Yes, the idea that normal people should listen to and obey anyone in authority without question is decidedly fascist. The degree of fascism is small, but who cares? Not seeing one measly receipt is no big deal for Circuit City either. They should just accept the fact that some people don't want to show their receipt and have the common sense and decency to accommodate them.
Other problems I see with raytraced games are the exponential increase of processing with higher resolutions or higher light source counts, the fact that poor raytracing actually looks worse on higher resolutions, the increased production and programming costs and the fact that graphics companies will not like seeing the investments made in their current GPU architectures melt away.
Ray tracing is actually more efficient. With the scene in a proper data structure, ray tracing can be O(log n) for a scene with n objects. Raster graphics is almost always O(n), even with advanced game engines that only draw portions of the game world. Every possibly visible object is rendered into a depth buffer to compose the entire scene, but with ray tracing only the x*y rays need to be traced through a tree of objects and checked for intersections. The tradeoff in ray tracing speed is how many reflections, refractions, and light sources to accumulate. Some complex scenes may require hundreds or thousands of rays for a single pixel on the screen, but that's an absolute maximum given the resolution. It matters much less if the scene has thousands, millions, or billions of objects in it. Look at the stuff OpenRT has done with small (8 to 16 core) clusters of generic CPUs. Some of the demos are faster than GPUs for large scenes, while others do things rasterization just can't do, like the accurate real time simulation of a reflective headlight.
Raytracing has no advantage over rasterizing for opaque surfaces. Rasterizers are faster there, since their performance is not tied directly to the screen resolution. The advantages lie in refraction/reflection/shadows/translucency, which are painful to implement with rasterizers.
Actually, there's a big advantage. Raytracing is O(log n), but rasterization is O(n). OpenRT's demo of a 350 million triangle model of a Boeing rendered in real time on a single PC (without GPU support) is a good example. The entire model doesn't even fit in memory, so visible surfaces are cached. The result is still realtime (although only a few FPS) with incredible detail. Go slashdot the server and watch the movie. Modern raster based cards can only render that many triangles in a whole second with all their fancy hardware, if they're lucky.
Or is it? Simply means games will appear more eye-candy than they currently are. Gameplay will not change. EA will continue to use take last years sport game, through some new people into it, perhaps introduce some bug which makes it unusable and peddle it as The New Deluxe Edition. I wonder how many geometric objects it will be able to handle (and whether it handles transparancy with textures and patterns well) Having done a bit of raytracing I'm familiar with how quick things can bog down. It'll probably be a bit clunky at first, but get much better as horsepower and horsepower/dollar ratio improve.
With raytracing, there are lots of new possibilities. For one thing, reflection and refraction actually work like they do in real life. That means accurate mirrors, lenses, and water refraction. Lights can work accurately if you want them to, and radiosity can be precomputed for static scenes. That may just be eye candy to most people, but there are potentially game-play enhancements that make real life optics part of the game. Most of it (except good lenses) has been faked before with rasterization, but raytracing will actually let you set up a series of mirrors and telescopes to peek around corners in a FPS for instance. I can imagine a true hall of mirrors in an FPS would be at least a little more interesting than what we have now, too.
The other big technological benefit of raytracing is that it's asymptotically faster than rasterization. Raytracing is O(log n) versus O(n) for rasterization, which means that even though raytracing is currently slower (the constants involved in raytracing are higher), after the break even point is passed much less of the available computational power will be needed to render the scene and can instead be used for physics and AI.
Sure, until enterprising individuals build plants in the US to make the goods we were previously importing, but at a lower price. And those plants start hiring US workers.
You assume that's possible in the first place. Where's the profit to be made in building a factory here and paying an order of magnitude higher salary for fewer working hours with more breaks and benefits? How are we going to get $5 t-shirts from an American factory?
The fact is, Americans have had an asymmetric economic relationship with lots of foreign countries, essentially trading our higher standard of living for the low cost labor of people in other countries. In a closed economy it's much harder to do that. Not everyone can be middle class when the factories require cheap grunt work. The smaller middle class means fewer luxuries will be purchased, meaning the upper classes feel the impact just as bad as everyone else. Who's going to buy overpriced CDs, movies, computers, and iPods when they can only get a $5/hr factory job?
Returning to an industrial society is not a viable goal for the U.S.
You can't make an omelet without breaking a few eggs.
You can't run a country without breaking a few legs, is that it?
Yours first, if you feel that way.
My place can be small (one house) or large (an entire island). Just like property in SL is already. Sure, the transition will be a bit tricky (at what point exactly are people transfered to a different server, and how do they "see" the content inside/outside?), but that's a technical challenge that is, in principle, not that hard. In fact, I'd be perfectly happy to have it work the Oblivion way (e.g. you click on the door, you are teleported inside. Windows both ways are faked with textures if at all.)
The traditional solution is to use portals. The client has to maintain connections to each realm or world that it can see, and it just renders objects from all possibly visible worlds that are visible through the portals, which act as interfaces between the worlds. You could do Escheresque things like portals that map onto the inverse of themselves and other fun stuff. The load would be mostly client side, since the realms would not interact directly; only at the interface where objects transfer between them. Even that is a formality and can be handled entirely by the client if it can "own" the same object in both worlds; Each server knows that objects crossing the portal unspawn, and the client respawns the object in the new world. If the two worlds care about "continuity" at the portals the transfer can be mediated to appear seamless or to obey restrictions on transfer of objects, etc. If the servers sign (or HMAC) the objects being transferred, the client can be responsible for all the data transfer as well.
Bah! Why didn't I preview!! Shows! Shows! She likes to watch shows not shoes! What is a TV shoe any way?
Bah, this is slashdot. We know you don't have a wife and that 'shoes' was a typo for porn!
If the voter can't verify the hash of their vote receipt on the spot (which they can't), then the machine can easily issue them a bogus receipt that looks right as far as the human-verifiable portion of the receipt goes, and instead record a correctly hashed, correctly signed false vote. The voter later discovers their receipt doesn't match, and they have no way at all to prove that they voted the way they say they did. You have no way at all to prove that the authentic but intentionally malformed receipt you have is valid, and it becomes a matter of human vs machine's word.
Correct, the machine can always print out a bogus signature of a hash too.
Further, in order to expose problems in the system, the voter is required to expose the way they voted. If coercion is going on, this is exactly what you don't want to have as a predicate for detecting fraud. Also if there was some intention of throwing the election so that the entire results are considered invalid, because it's a hash chain, you would only have to invalidate or destroy the record of one very early vote, and the rest of the records would be un-verifiable.
Not necessarily. A photocopy of a tampered vote could be submitted anonymously, or even entered into a form on the Internet. If the tampering was a vote change after the polls closed, the claim could be automatically verified. Otherwise, the ballot could just be turned in to the local party who the voter voted for.
A hash chain is valid from any given point forward, and destroying receipts does not invalidate the chain. Destroying the entire chain itself would obviously destroy the record of votes. That's possible with paper ballots as well, and retaining the secret ballot makes it impossible to prevent against that scenario in a recoverable way. The only solution is a re-election.
Destroying a single entry in the chain would invalidate only the votes immediately after it before a valid receipt could be found. In general, destroying a ballot is election fraud and would require an investigation and possible re-vote anyway. It would also be much more detectable than losing a single paper ballot.
This doesn't gain you anything over a machine which simply records the vote in a dual human/computer readable format onto a card that the machine is physically given by the voter, and which is fed into a locked ballot box after verification by the voter. Part of the record on the card would be machine #, vote # by that machine, and time. You can't stuff the box because the stuffed cards would be out of sequence or else at a time when the polls have closed (or else a whole lot of votes 2 minutes before they closed). You could only forge this record if people physically entered the voting booth multiple times, which in an anonymous ballot is only defensible with human precautions which are already present.
A hash chain does provide vote verification; it's possible for anyone to sum all the votes and calculate the final totals while allowing every voter to ensure their vote was recorded. It's easier to "lose" paper ballots and their electronic equivalent. Having paper ballots does prevent electronic ballot stuffing, though.
If I haven't mentioned it before, I realize a hash chain is only a small part of a secure voting system. It is one way to provide verified vote counting, but there are better (complete) solutions out there. I think most of them rely on some form of chaining votes so that it becomes statistically impossible to alter votes.
I don't think any of them really address the man in the middle attack very well, as you pointed out. A machine can always pretend to record votes but actually perform the vote itself and issue a bogus receipt, just like a polling place can discard all its ballots and replace them with their own. If necessary, the polling place can print forged duplicates of its log books and ballots and use them in the election, then turn in the official ballots filled out as they desire. Replace th
Except that if the machine is programmed to alter a vote, and it knows the hash chain at the time it wants to alter the vote, it is able to record the altered vote, with a correct hash at the time the vote is recorded. This is what I meant when I said that if the machine has everything it needs to record the correct vote, it also has everything it needs to record an altered vote. The voter isn't able to manually verify the vote's hash, and even if every voter took a receipt and came back to verify it later, the vote on the receipt wouldn't agree with the machine, and you would have no way to reconcile it.
Didn't I specifically mention that receipts were generated with; 1) the latest hash value in the chain, and 2) the current vote? Each receipt is enough to verify one step of the hash chain. The voter doesn't need to manually verify the hash, because they can do it later at home with a simple web interface. The voting machine can't magically change the value on their receipt to match a forged hash chain. IF the hash printed on their receipt is in the hash chain, then it MUST be the result of hash(last_hash || vote). If the last hash in the chain before their vote (which the user can see), concatenated with the vote they picked (Which is also printed on their receipt) does not equal the hash printed on their receipt, then it's obvious the vote has been tampered with. Just as importantly, if a voter cannot find their signed hash in the hash chain, they know their vote was discarded and can prove it. The key is that the voter's actual vote is printed on the receipt; it's trivial to prove the error in the hash chain. The hash on each receipt can even be signed by the machine to prevent naughty voters from forging receipts. Statistics prevents vote tampering. For n voters, If a random sqrt(n) votes are changed and a random sqrt(n) number of people check their votes, the probability of the tampering being discovered is ~0.5. For larger numbers of tampered votes or checked receipts, the probability quickly tends toward 1. sqrt(n) is far too small a percentage to sway an election, but the number of computer geeks who can write their own correct implementations of hash functions almost certainly is less than sqrt(n) of most voting populations, not to mention "normal" people who could use a simple interface to verify their votes.
Hash chains verify that none of the data was changed after the chain was forged, but it doesn't guarantee anything about the integrity of the data as it is being chained. They assume that the author of the chain can be trusted, and the principle being protected against here is that we can't trust voting machines. That's why an inalterable record needs to be created which is verified at the time the vote is cast. The machine doesn't have access to the ballot cards, the voter gives one card to the machine per vote, so the machine is physically unable to alter, destroy, or insert votes. The cards once fed to the machine have the voting record printed on them, and the voter verifies the record is accurate. Once printed, the card gets inserted into a locked ballot box, or else returned to the voter if they indicate their vote was not as expected.
The receipts count as the unalterable record. Obviously a simple hash chain has voter coercion problems since every voter carries their own receipt which proves how they voted. A solution is to let voters print a receipt for *any* of the previous votes, thus letting them choose an existing vote to present as their own if they are being coerced. Another solution is to only print receipts for 10% of voters, or some other way of making it impossible for voters to be coerced. I agree that a physical voting token should also be required to prevent the machines from ballot stuffing. Preventing ballot stuffing is a harder problem than preventing vote changes, because it ultimately comes down to the problem of accurately and universally identifying every voter.
Richard Dawkins is about as rational as any of the hoplophobes at the VPC. Anyone who professes to be an atheist has voided any claim at rationality, period. It requires just as much, if not more, blind faith as any actual religion.
An atheist is just an agnostic who has concluded that the probability of encountering valid proof of the correct god (e.g. not the one that the other half of the world believes in) within his or her lifetime is so low that continuing to evaluate supposed evidence for a god is a waste of time. It's the same reason sane people are no longer searching for the philosopher's stone or the fountain of youth.
Don't take that as a knock against religion or belief in god either; evolution can make use of absolutely any variation in population to improve fitness. I am not convinced that religion is useless. It's quite possible that it was (or is) a necessary social mechanism required to move beyond simple hunter-gatherer tribal systems and develop technology. It is probably not irreplaceable.
This suggests that during a recount or investigation, every voter would show up with his or her vote receipt. Lose even one receipt and you can't verify the votes.
No, the voting machine stores the hash chain as well. The point is that if a vote is changed at any point, none of the hashes after that will match the receipt, forcing vote tampering to occur within the machine. If someone tried to tamper with a vote after all voters received their receipts, every vote after and including the tampered vote would not match the receipt of the voter, or would not generate a valid hash chain.
The idea is that you have to make the "authority" on which vote is which an immutable record. That is to say something which can't be changed after the vote has been cast. There's nothing in the computer world where this is the case. Not even cryptography would suffice since the voting machine does all the cryptography, and it could easily show you one cryptographically signed vote and record a different cryptographically signed vote. If it has everything it needs to do the original signing, it has everything it needs to forge the signing of different data.
This can be fixed by using a hash chain. The machine starts with an initial hash. After each vote, the value of the vote is salted and appended to the hash, which is then hashed to become the next hash in the chain. The final hash is reconstructible from the initial hash and the sequence of votes. The important thing is that every receipt has a copy of the current hash. There is no way to modify the sequence of votes without making at least one of the receipts invalid, and no way to certify the results without having a complete hash chain of all the votes. Obviously votes can be appended to the end (ballot stuffing), but that's a problem that cannot be fixed in general.
To check the results of voting, publish the entire hash chain without the vote or salt values. Everyone can verify that their vote is present in the hash chain by looking up the hash on their receipt in the chain. If someone can't find their hash, their vote was lost or changed. To prevent voter fraud, the machine can sign the hash on the receipt to ensure that only valid hashes can be looked up.
This scheme alone does not prevent vote-buying, since the receipt must have the salt and vote value on it to verify to the voter that the correct vote was recorded. However, it would be possible to issue only 1/p*sqrt(n) receipts to random voters so that by the birthday paradox no vote tampering could be accomplished that would affect more than sqrt(n) of the votes, where p is the percentage of people who would check their votes on average.
I might be able to prove it's in the final tally. You, too, might be able to prove it's in the final tally. 99% of the voting population, however, have not studied cryptography and would have to rely on an expert to check their vote (and, of course, such a system would have to be designed to make it impossible for the voter to prove to someone else which way they voted).
The 99% part doesn't matter. It's the birthday paradox; the probability that no tampered vote is verified is very low if the number of votes tampered with plus the number of votes verified is bigger than the square root of the total number of votes. The square root of the number of votes is far less than 1% of the voting population (about 11,000 people, given the turnout in 2004), meaning that effectively no vote tampering could occur. If even one vote is found to have been tampered with, it's enough to start a campaign of checking everyone's votes to see how many were really tampered with. I'm sure you could get more than 11,000 geeks to cryptographically verify their vote.
It's the same as buying a car. Yes, you are free to go buy a car from a dealer with no terms and conditions attached if you can negotiate that deal with the dealer. Or, if the dealer says "look, this is a brand new concept car that's not available for sale yet -- you can buy it, but you have to keep it in a closed garage and you can't take the bib and diaper off it until we tell you" then that's also a completely legal contract if you agree to buy that car. You don't get to tell the dealer "yes, I'll follow your terms" and then turn around and say "I own the car, I can do whatever I want." You entered into an agreement with the dealer and you are bound by it.
EULA's have no verbal or written agreement of the licensee. It's absurd to think that the mere action of opening or using something implies agreement to a contract. Obviously no judge has yet been handed evidence in a case in a sealed envelope with a EULA stating "by opening this envelope and reading the enclosed evidence you agree to rule in my favor." I think that would quickly turn the court's opinion against the validity of EULAs.
EULAs come down to property rights. If I have been sold a box or CD without agreeing to a contract, then it's my property. If I burn the shrink wrap license off the box, does that count as agreement to its terms? If I manually change the right areas in memory in *my* computer, does that count as agreement if the program starts working after the EULA is displayed? What if I change it so the EULA doesn't even display, or if I change the text and then agree to it? The courts have said that fair use covers usage of software. There's no legal rights that a EULA needs to grant in order to use a copy of software that you own, so long as you're just using it on your own computer and not accessing some service over the Internet.
No one is going to carry a 5 terrabyte rainbow file around to try to crack a password that long.
At least not for a couple years until 5TB hard disks are available.
Neural nets / genetic algorithms / learning systems These all belong to the family of hill-climbing optimizers. These approaches work on problems where continuous improvement via tweaking is helpful, but usually max out after a while. We still don't really understand how evolution makes favorable jumps. I once said to Koza's crowd that there's a Nobel Prize waiting for whomever figures that out. Nobody has won it yet.
Evolution gets stuck at local maxima for millions and billions of years. Cells as we see them today are basically the local maxima that evolution has been unable to surpass despite the obvious advantages of integrating any number of other elements to create stronger, faster, and more robust cellular structures. Medicine has made improvements upon evolution that were too far beyond the local maxima that blind evolution could achieve. The idea of species randomly cutting themselves open to try to fix things is absurd, except that it happened and we can now fix a host of problems with ourselves. Evolution had billions of years and trillions upon trillions of effective computational resources to create intelligence, and this is as far as it got. In roughly fifty years (a couple hundred if you count the concepts of Babbage and Leibniz) we have gone from simple computers to computers that can beat humans at chess, face recognition, information storage and retrieval, raw numerical ability, and are starting to match our ability in autonomous control of vehicles and many other areas.
AI is one of those fields, like fusion power, where the delivery date keeps getting further away. For this conference, the claim is "some time in the next century". Back in the 1980s, people in the field were saying 10-15 years.
Flight was like that. Human powered flight even more-so. People have dreamed of flight for thousands of years, but it took heat engines to achieve it. Human powered flight required advanced engineering and lightweight composites, and it still happened decades after powered flight. Just because a technology is always "10-15" years away doesn't mean anything about its possibility; it says more about the people doing the predictions than anything else.
Intelligence is not well defined. It is very hard to say how much of what we call "intelligence" is in fact the ability to make many connections between facts stored in a very sophisticated memory architecture. Simply building a machine able to process information very quickly achieves nothing because, without learning and a social context, it does not know what information to acquire and process. In human experience, academically brilliant people often fail because they work on the wrong problems, or without access to necessary knowledge.
You're being very anthropocentric in your definition of intelligence. Machines will probably have vastly different basic goals than humans, although they will probably have goals that are subgoals of what humans want. Creating a general intelligence that only wants to survive and reproduce is no different than trying to breed "better" humans; it leads to things like massive racism and genocide.
Nothing is actually achieved without creativity. We do not know what that is, or to what extent it is a social construct (i.e. it takes a developed society to have the necessary systems in place to translate an idea into a concrete reality.) And this leads onto the third point. It is no good having a highly intelligent, creative machine if its use of resources is such that it cannot replicate in large numbers. It may be that machine intelligence will ultimately replace human intelligence, but it may be that it will simply be too resource hungry. In effect, there may be a threshold of capability needed to solve some problems, and it may be that machine intelligence will run out of energy before it scales sufficiently to solve those problems. A machine society might, in effect, get stuck in the machine 19th century because coal or oil became a limiting resource. (In the same way, the energy and resources needed to be consumed to achieve a first independent space colony may exceed the total energy and resources available on Earth. It may be that a billion years or so of eukaryotic evolution has actually resulted in the optimum balance of intelligence, creativity and resource consumption, and that any attempt to exceed the present capability will tip us into declining resources faster than we can improve matters.
Creativity is just trying lots of different possible solutions to problems until one works. What humans usually call "creativity" is just normal intelligent humans trying a few wilder solutions than other people would. Most people do what their friends, neighbors, and ancestors have done. "Creative" people try whatever they can think of.
Nanotechnology will be able to achieve far more with fewer physical resources and less energy. Obviously there is a carrying capacity for intelligent species on the Earth; that's when space exploration begins in earnest. Then galactic exploration, and probably eventually a spread to as much of the universe as possible. In the end, the universe is finite. Humans and machines will have to realize that sooner or later and design societies that can remain stable (or shrink) happily. Ultimately the universe dies, and figuring out how to die gracefully will be the last big question for intelligent beings to answer. If it turns out that the universe is actually too low in energy to support space exploration, at least intelligent machines would have the ability to see that objectively without the emotional baggage humans would attach to that.
In many ways I hope this is wrong. But the argument that only one superior machine is necessary is, in fact, an inductive step too far. It is assuming that "intelligence" on its own can solve a class of problems which may involve a number of constraints which cannot be avoided - like the Laws of Thermodynamics, or the need for excessive amounts of energy.
I think the basis case is "Let humanity equal intelligent species number 0. If humanity can generate intelligent species number 1 then by induction intelligent species number x can generate intellig
FLASH used exactly that concept, and it lead to the discovery of the fact that there can be no quantum cloning. It was hypothesized that it was possible to tell whether an electron (or any other particle) was entangled or not, basically by seeing if a measurable property changed when another entangled property was measured. It relied on being able to make a perfect copy of the quantum state and measuring it repeatedly to get exact measurements in violation of the uncertainty principle. If that could be done, then measuring an entangled particle would transmit measurable information to the other entangled particle(s) immediately. It turns out that perfect (100% accurate) quantum cloning is not possible, so no matter which particle gets measured first there's no way to tell on the other end. It's interesting that the actual limit on the probability of success in quantum cloning (which has been measured physically and found to match the predicted value well) is exactly the limit necessary to prevent causal inconsistencies or faster than light information transfer.
Unfortunately, consumers brought DCMA on themselves.
See, "fair use" is part of a bargain between creators and consumers. Consumers are free to do whatever they like with creations as long as the remain within fair use. But starting around 1995, consumers began to violate "fair use" on a scale previously unimagined. Consumers violated the bargain, so DCMA was created to try to restore some balance. If consumers didn't want DCMA to come along then they shouldn't have been (and shouldn't be now) uploading and downloading every piece of IP they can get their hands on.
The DMCA has effectively stopped zero criminals. CSS is broken, AACS is broken, and any future DRM technology is broken by definition. Pre-release movies and songs flood the Internet. On the other hand, the DMCA is used to issue take-down notices by Scientology and other evil corporations, or to sue mod chip makers. Basically, the DMCA is used entirely for the manipulation of normal people and censorship of copyrighted works, and never to combat the majority of actual copyright infringement on the Internet.
Your assertion that customers brought the DMCA on themselves is true, but not for the reason you stated. The citizens of the U.S. brought the DMCA (and tons of other stupid laws and illegal wars to boot) on themselves by being lazy, pathetic excuses for informed citizens of a democratic nation.
From the perspective of an application developer, this is pure nonsense. I practically don't care wether my DB stores data in columns or rows or whatever.
You do care; you want it to be as fast as possible. This has less to do with how you interface with the database than it does with how fast reads are versus writes.
What I need is a good, consistant layer that can handle object-based tree structures - nothing more, nothing less. I want to dump my Java/Objective-C/C#/C++/PHP/Python objects in some storage layer, and I want to be able to get it back, search for it, etc.pp.
Look into operating systems with a single level store. They are not very common (one example is OS/400), but they are the perfect design for what you want. Objects are stored at permanent virtual addresses and everything that can't fit in memory is swapped out to secondary storage. No concept of loading or storing objects; they just exist until deleted.
Ugh. It was your sister's birthday, couldn't you drop the crusader bit for one lousy day?
I'm sorry, but this story just wreaks of petty selfishness. Don't let morality get in the way of being a good person.
Apparently these days people prefer pansies over patriots. A "good person" challenges authoritarian anti-freedom practices whenever they can. A much more important question is how could any officer in the U.S. arrest a man on his sister's birthday for exercising his rights? The officer is completely responsible for his own actions.
If you think that circuit city has a right to look through everyone's bags, and that police officers have a right to demand ID, just suppose that some terrorist watch list gets badly implemented everywhere instead of just in airports. How would you like having to show your ID every time you do anything or go anywhere, and having your vehicle and person searched at every opportunity simply because your name is similar to someone on the watch list and your license plates and credit cards get linked to the watch list as a result?
No one in their right mind would have thought that U.S. citizens would be subject to mandatory search and seizure of their liquids in order to travel. Not many today believe that in the near future everyone will be subject to random searches and mandatory identification, but it's obvious that it's coming. National IDs, overeager cops, terrorism fear mongering, the only thing missing is some sort of leeway (violation of the fourth amendment) to perform warrant-less searches. We already have warrant-less wiretapping, so it can't be very far down the road.