That is not legal. Downloading it still making a copy, which is prohibited under copyright. Should it be covered under fair use? I'd say definitely, but I don't think that particular issue been tested by the courts yet.
Of course it's legal. Just look at all the shareware and commercial trials that can be downloaded over the Internet. It's also completely legal to accept the copy of copyrighted web pages that HTTP servers send you. The key is that the network transfer of the work is considered to be an actual transfer, just like getting handed a CD with the work on it. It has been upheld in several cases that caches, proxies, and store-and-forward routers do not infringe copyright by holding ephemeral copies of copyrighted works, so it makes sense that downloading something from the Internet is just as legal as accepting a CD from someone, e.g. if you know the CD is stolen, you can't accept it just like you can't download software from War3z-R-Us and claim it was a legitimate copy. In short, the burden of following copyright law falls almost entirely on the uploader.
So let them redo the internet into a new corporate-friendly version. Let them rape us six ways from sunday. After working in the industry as I have, I could just as easily walk away and leave it to other more patient and gullible folks to handle.
People like you are killing democracy. 'nuff said.
In reality, splitting the encrypted data or re-encrypting it using a one-time pad both accomplish the same goal:
They turn 1 piece of data into two, and neither piece alone has any usefulness to an attacker. Everything else is fallback in case the attacker recovers both pieces.
There's a proper way to do secret sharing. There's no point in splitting the ciphertext if a modern cipher has been used; it doesn't provide any more security than splitting the key with a secret sharing method and the latter is much more efficient.
I played with the picture recognition software and if you just make a dot and click add noise like 20 times it thinks it's a dog. I did that 4 or 5 times in a row with the same result.
And if you pour ink on a folded page and unfold it, people think they see all sorts of things in it. What was your point again?
data->strong encryption->split data into 2 files using a secret algorithm. The second step doesn't add a lot of security but it does make the adversary work harder if he wants to intercept the message in-transit.
Invariably, you would want to transfer your two files over the same network, which means the attacker can grab both files. Since it's merely a secret algorithm, the attacker just needs a copy of your secret program to put them back together. Relying on secret algorithms is silly, because all it takes is *one* successful intrusion to completely break them.
data->strong encryption->one-time pad encryption
This shows that you don't know much more about cryptography than the makers of the broken USB drive. One time pads are perfectly secure. There's no need to encrypt with another algorithm because it adds absolutely no security. Also, if a one time pad is reused or not generated from a purely random source, it's not a true one time pad. Managing one time pads for anything larger than short messages is essentially impossible. Even if you want to carry around a DVD full of random noise, is that really more secure in practice than a 20 byte passphrase you remember and never write down? No.
The reason for broken encryption utilities is that people think they can design secure encryption systems without having; 1) broken several cryptosystems for practice, 2) read as much available literature about known attacks on both algorithms and protocols. I would not try to design a secure cryptosystem. When I have tried, I have always found information later explaining why the ideas I had were broken. There are a lot of pitfalls in designing secure systems, and most of them have to do with the protocols (the data storage, user interface, threat model, etc.), not the basic secure cryptographic algorithms. A good rule of thumb is that if you cannot trust a single correct application of a secure cipher like AES, combined with a secure message integrity function like SHA1-HMAC, there is something fundamentally wrong with the design of the protocol or your understanding of the threat model.
What about OpenSSL? It's the base for SSH and dozens of web servers and clients. I would bet it's the second most used encryption product after the crypto that ships with Windows.
1. Keep the data in CPU registers and cache.
2. Split the keys up into smaller pieces, and spread them around when in memory.
1. Impossible, without incredibly slow decryption. The x86_64 chips *might* have enough registers to do AES with lots of extra computation, but all the fast (e.g. more than 1MB/sec) algorithms use a large key schedule that is directly derived from the key. Heck, with some SSE hacks it might even be possible for generic x86_32 processors to keep the key in registers. Additionally, it would be apparent that the key is being stored in the registers because it could not be disguised very easily.
2. Slightly more doable, by moving and "disguising" (probably XORing with constants) the key schedule it could be forced to never exist entirely as plaintext in memory, so the crackers would have to do a little bit of time analysis to figure out when to grab parts of the key schedule. It's still going to be incredibly difficult because every 16 byte block of ciphertext requires every byte of key schedule to decrypt it. Given the massive amount of data streaming from the disc, the player is going to sit in two main areas: AES decryption and MPEG decoding. MPEG decoding takes up the majority of the time, but AES will be a very noticeable percentage, and probably easy to identify because of the table lookups that are necessary. After the code is identified, it's obvious how to get the original key back, because it forms the first 16 bytes of the key schedule.
An interesting solution would be to find a mathematically equivalent algorithm for AES that uses different lookup table constants and a different key schedule. It would probably not be very hard to do, maybe requiring an extra prestage and afterstage to put things right. That could take a little while to figure out, but again because of the time spent decrypting it would probably be relatively obvious where the code was, and when it didn't look like AES the hackers would know to look for the fudging to figure out how to reverse it. It might just take a hacker who's good at linear algebra.
The real solution the media companies are aiming for is to skip the software players and have the video card act as a secure player, handling the AES and MPEG in hardware and using HDCP to talk directly to a "secure" monitor. That will basically close the digital hole, because the only place to get the plaintext signal would be at the interface cable going into the LCD panel. If they can get the drive to talk directly to the video card for key exchange, that cuts the main PC out of the loop entirely, meaning that special hardware would be required just to read the data off the disc.
The AACS scheme has the ability to revoke individual players - not individual models, but actual single units. They use a lot of fancy set theory to do it, but in essence each player is supposed to have a unique set of keys - possibly hundreds of keys out of a total of many thousands (hundreds of thousands perhaps). Each disc has the information on it to allow thousands of different keys to decrypt it. The way it works is that of all the keys on the disc, it is expected that each individual player will have at least one key that matches.
Doesn't matter, because every copy of a movie has the exact same key that encrypts the actual movie (pressing discs relies on having a static image for the majority of the data, obviously small areas can be written individually). Discovering that key allows all HD discs using that key (all discs of a given release of a movie, for instance) to be viewed without DRM, and posting it on the Internet does not allow the MAFIAA to identify which player was used to extract the volume key.
And the computers in a single segment of cable modem network can be considered a LAN, right? What about the computers on a single university campus?
It all depends on how the network administrators decided to assign IP address ranges. As far as I know, most places don't put much consideration into making numerically adjacent subnets physically adjacent. It's too much work. Instead, you'll see 192.168.4 in one city and 192.168.5 in another city across the country, and the subnet mask may be set to 192.168.0.0/16 for the entire organization anyway, making it difficult to determine where the physical network boundaries are.
The first-order approximation to this is to prefer peers who share a longer IPv4 address prefix, right?
I doubt it. Even within subnets it's possible for machines to be widely distributed around the globe, but probably not very common outside of LANs. Take a look at the map of IP addresses that neatly displays the adjacency of successive first octets and how widely distributed they are around the world. While there are some large contiguous blocks, there are numerous discontinuities where subnets next to each other are on other sides of the world. The class C addresses are probably the worst. Adjacency information really needs to come from routing tables, preferably with some cost attached, to be useful.
It wouldn't be a matter of coming down on behavior that I thought was bad because anybody who engaged in it was obviously doing something else wrong. It would simply be a matter of preserving the quality of service for all the customers, not just the high-bandwidth users. I'm not the personal T3 connection for someone who's paying me $50/mo.
Just a suggestion: Have off-peak hours during which the bandwidth cap doesn't apply. Make it just one or two hours in the wee hours of the morning even, but it will let customers who want to push or pull several gigabytes a day to do so when it won't hurt anyone else.
In terms of trying to maximize intra-ISP file sharing, you probably need either a huge customer base or lots of customers that like to download the same things. I don't know if it's possible to force most of the P2P clients to localize their searches to one subnet without actually modifying the software and running your own servers. For bittorrent you might be able to watch connections for bittorrent data and have your servers join the torrent in order to cache data locally and hopefully provide a node to help point other customers to the local network. If you really want to help bittorrents and you have gobs of disk space just run a tracker on your own server for the torrents that your customers have downloaded, and let them know where they can find a list of trackers that you're hosting. Once you have that, you can selectively pick peers from your own subnet for your customers to maximize the intra-ISP traffic.
In general, all P2P clients should have a way of valuing peers in their own subnet higher than other peers, at least for picking the initial list of peers to try. One could even use the Internet BGP tables to find the closest peers in the overall network, which would almost always provide higher speed for P2P applications. Making every client keep a copy of the BGP tables might be kind of a stretch though.
It's quite possible that a proper definition of what a human being "is" would disqualify fetuses and some babies. I don't think fear of that should necessarily stop us from defining it anyway. We can find other reasons to keep our kids around, like, say, because we love them.
Every living thing has the potential to become intelligent. It already happened at least once, after all. All it takes is time and ecological pressures, or of course human intervention. Intelligence amplification for animals would be a great way to understand more about the world from a different perspective.
The very idea of human rights is based on the premise that there is something intrinsically valuable in human beings, regardless of their mental capacities or physical abilities.
The ultimate question is where those intrinsic rights come from. Clearly in a naturalistic sense they have no source except as ideas in the heads of the creatures who assume they possess them. What's to say that chimps and other animals don't have their own ideas of self worth as a species? If so, shouldn't we respect their concept of value the same as we respect the human concept of value? There is ultimately nothing morally different between humans and animals except intelligence. Emotions are almost certainly as intense and rich in animals due to their source in the "lower" parts of the brain. Intelligence, as you said, is clearly not required for a human to have human rights. Why animals?
Eventually this will become more obvious when humans begin to modify themselves to be distinctly non-human either genetically or technologically. Clearly, they will continue to appreciate "human" rights, and it will become more apparent that animals are no different than us except in their level of intelligence. We can use animals as property now only because they are unorganized and relatively weak individually. As soon as some scientist puts more powerful brains in rats or the chimps, they will be able to organize animal rights campaigns themselves and humans will have to recognize them as intellectual and moral peers.
The Backwardistanis would need to prove that I've done anything inappropriate with the woman in question
In some fundamentalist countries, an unmarried woman traveling with a man who is not part of her immediate family is instantly a criminal. True, visiting such countries is probably a bad idea in the first place, but perhaps business or family needs might demand it, or even just dumb luck like plane trouble causing a diversion to the nearest airport.
Yes, the Internet DOES transport you if you specifically go after something physically located (the media or the server) in another location.
Riiight. That's why you pay taxes to the state or country the server lives that you buy things from online, and why you have to pay income tax for the items purchased from you on ebay to the state or country the buyer lives in. You are not touching the remote server, it's not even as physically connected as a phone line (most of which get packet switched for long distance calls anyway). You have no control over the route that your packets take, or the route return packets from the server take. Only specific laws about Internet traffic and computer access have jurisdiction, and only then in the country they are written. Extradition should be limited to crimes that can not be effectively prosecuted in the country where the defendant resides, not merely in the most politically convenient location. What good is your country if it won't protect you from the arbitrary laws of other countries?
If you buy from Lenovo, formerly IBM's PC division, they still recommend Windows XP professional over Vista, although Vista is selected by default (explain that). Perhaps the "Recommended by Lenovo" has been removed from the XP selection by now, or it only appears on computers without Vista drivers or something, but it's definitely still possible to get them shipped with XP.
I don't think that Microsoft was concealing anything. They were advertising a product with its niftiest features, but I think that about 15 minutes of research would have let someone know that they couldn't use the Aero interface. Microsoft used marketing and advertising to make their product look the best, that isn't the same as cheating someone.
To use your example, it would be like advertising cookie dough under the Vista trademark and clearly showing chocolate chips to be present in the cookies, but forgetting to mention that most "Vista Capable" ovens were actually incapable of baking cookies with chocolate chips in them. How is that not deceptive? Just because it's possible to discover (with 15 minutes of someone's personal time) that it's deceptive does not make it any better. How would a normal user even go about verifying whether all the features of an operating system would be available? Maybe a little sticker based on the technical specs of their computer? Hmmmmm.
Of course, I believe that any crime committed should warrant extradition, because that's just the right thing to do.
You have broken several country's laws today. Which one would you like to be extradited to first? Iran? China? North Korea?
I also believe that you should be judged on the laws of the land where the event took place, not on laws in your own homeland.
So when you touch down in Backwardistan on a connecting flight with your girlfriend, they can execute you for fornication even though you're an American citizen?
He acted "in" the USA, against US servers and security
I didn't know the Internet magically transported people to other countries. Does that mean the Chinese can arrest you for sending information about democracy to their country?
But I think it'd be -- or could be -- a visually spectacular sight to be able to perceive EM fields, because I imagine them being like auroras or the colors you see in plasma etch chambers. It'd make for a whole different form of art.
EM is directional as well as spatial. It would have to look like continuously flowing fluid so that you could see the magnitude and direction at every point in space. Color might give an indication of magnitude, but without the direction it would be confusing and somewhat useless, e.g. you couldn't sense polarity. Color might be useful to differentiate the electric and magnetic fields.
The classic example of that was the Kremer Prize for human-powered flight, won in 1977. Once that was done, interest in human-powered flight declined substantially. That effort didn't usher in an era of recreational pedal-powered flying.
In the article you linked to, you'll notice that the research to develop the Gossamer Condor lead to light, solar powered planes from NASA. Not every technology is immediately adopted by the masses, especially when human powered flight machines essentially require an athlete in peak condition to operate them.
The exact vendor selection criteria, often being secret, leaves vendors that had reasonable belief that they should have won completely baffled as to why they lost.
Maybe one of the criteria was "Cannot be unlocked with a hotel wet bar key."
That's the most interesting part of the article for me. Apart from 3-D rendering and folding@home, they are really pushed to find any real-world reason for having 4 cores.
Just wait for real time raytracing. From what I've seen of OpenRT, it only takes about 8 cores to get photorealistic real time raytracing for complex models entirely with software. Combining that with the advanced shaders on the GPU should make for some very visually stunning games with perfect reflections, refractions, shadows, and surface properties.
Normally I'd be on your side in the case like this but this time it seems you're just wasting the court's time fighting a battle you started. I agree that their message was overreaching, but does your want to prove that entitle you to duplicating and essentially broadcasting video they produced? Even if it did your methods amount to litigious entrapment.
It's a long, twisted road from fair use to litigious entrapment, don't you think? Or do you also think I shouldn't have been able to quote your post in my reply?
Re:What I'd like to see (and plan to implement soo
on
TrueCrypt 4.3 Released
·
· Score: 2, Informative
This will take/dev/md0, create an encrypted volume from it using the supplied 16 byte password in hex, and create/dev/mapper/encrypted_raid to mount as your root file system. Replace/dev/md0 and/dev/mapper paths with the appropriate locations your devices for your distro. A combination of pivot_root and chroot can be used to move the mounted encrypted raid device to the root of the filesystem. There are howtos out there, and apparently some easier (less manual) ways to get the encrypted root filesystem mounted by some distributions.
Sure, the idea of an abrupt Creation, or "Design," of the universe lets us joke about what God was doing before he got around to Creation, but the metaphor of water (or, let's suppose, some kind of cosmic stew) boiling into steam/universes leaves us with the same problem that we had in the first place: where in the [space larger than a universe] did the water/stew come from?
I think the universe is isomorphic to an infinite set. So far, we're on a fair track to proving that by finding very accurate mathematical models for the fundamental particles, and macroscopic behavior is very well defined. Even chaotic processes can be modeled. All this means is that the universe probably operates at some level like mathematical objects in set theory. The universe doesn't even have to be deterministic. A nondeterministic universe just splits into multiple parallel universes which is still relatively easy to express in set theory. Physics is just a set of equations that force the set to maintain the proper physical relationships between the matter and energy and topology of the universe.
What this means for existence and the beginning and end of the universe is that at any given instant in time, the universe can be represented as some subset of the entire universe, and there exist successor and predecessor functions that when applied to an instant in time will produce the "next" instant in time, or since the universe appears to be continuous at this point, any arbitrary time in the future or past. If the predecessor or successor function for an instant does not exist, those boundaries define the beginning or end of the universe respectively for all practical purposes. A nondeterministic universe is simply one where the successor and predecessor functions evaluate to a set of universe states, and each one of those states in turn has its own sets of states reachable through the successor and predecessor states.
Ultimately, only mathematics exists, which is to say that the relationships between hypothetical objects simply cause what is being represented by those objects to behave as if they existed. There is no difference between our universe, a simulation of our universe in a larger one, or a mathematical set that is isomorphic to our universe. A good example is the integers and the prime numbers. It's easy to prove that there are an infinite number of both integers and primes, and that means that even though our universe will never enumerate all the primes, their properties and relationships to other numbers still exist mathematically. By extension, in a model of a universe the atoms and fields have well defined positions at any given point in the future by using the mathematical rules of physics, and likewise the people in the model composed of those atoms and fields have well defined thoughts and actions over the entire domain of the universe. They exist, independent of being created or modeled or simulated, simply because it's possible to describe a mathematical system containing them. I can't find any counterproof to such an argument, but it is very difficult to prove from within our own universe. The best we can do is simulate small universes (and we do, usually of individual proteins or ideal gasses or even nuclear explosions) and extrapolate from that. Basically anything we can conceive of exists as its own universe, and all we have to do to see what happens in that universe is to simulate it. Within that universe, all the things that happen are already well defined by the initial conditions and the laws of physics.
Slashdot Mirror
That is not legal. Downloading it still making a copy, which is prohibited under copyright. Should it be covered under fair use? I'd say definitely, but I don't think that particular issue been tested by the courts yet.
Of course it's legal. Just look at all the shareware and commercial trials that can be downloaded over the Internet. It's also completely legal to accept the copy of copyrighted web pages that HTTP servers send you. The key is that the network transfer of the work is considered to be an actual transfer, just like getting handed a CD with the work on it. It has been upheld in several cases that caches, proxies, and store-and-forward routers do not infringe copyright by holding ephemeral copies of copyrighted works, so it makes sense that downloading something from the Internet is just as legal as accepting a CD from someone, e.g. if you know the CD is stolen, you can't accept it just like you can't download software from War3z-R-Us and claim it was a legitimate copy. In short, the burden of following copyright law falls almost entirely on the uploader.
So let them redo the internet into a new corporate-friendly version. Let them rape us six ways from sunday. After working in the industry as I have, I could just as easily walk away and leave it to other more patient and gullible folks to handle.
People like you are killing democracy. 'nuff said.
In reality, splitting the encrypted data or re-encrypting it using a one-time pad both accomplish the same goal: They turn 1 piece of data into two, and neither piece alone has any usefulness to an attacker. Everything else is fallback in case the attacker recovers both pieces.
There's a proper way to do secret sharing. There's no point in splitting the ciphertext if a modern cipher has been used; it doesn't provide any more security than splitting the key with a secret sharing method and the latter is much more efficient.
I played with the picture recognition software and if you just make a dot and click add noise like 20 times it thinks it's a dog. I did that 4 or 5 times in a row with the same result.
And if you pour ink on a folded page and unfold it, people think they see all sorts of things in it. What was your point again?
data->strong encryption->split data into 2 files using a secret algorithm. The second step doesn't add a lot of security but it does make the adversary work harder if he wants to intercept the message in-transit.
Invariably, you would want to transfer your two files over the same network, which means the attacker can grab both files. Since it's merely a secret algorithm, the attacker just needs a copy of your secret program to put them back together. Relying on secret algorithms is silly, because all it takes is *one* successful intrusion to completely break them.
data->strong encryption->one-time pad encryption
This shows that you don't know much more about cryptography than the makers of the broken USB drive. One time pads are perfectly secure. There's no need to encrypt with another algorithm because it adds absolutely no security. Also, if a one time pad is reused or not generated from a purely random source, it's not a true one time pad. Managing one time pads for anything larger than short messages is essentially impossible. Even if you want to carry around a DVD full of random noise, is that really more secure in practice than a 20 byte passphrase you remember and never write down? No.
The reason for broken encryption utilities is that people think they can design secure encryption systems without having; 1) broken several cryptosystems for practice, 2) read as much available literature about known attacks on both algorithms and protocols. I would not try to design a secure cryptosystem. When I have tried, I have always found information later explaining why the ideas I had were broken. There are a lot of pitfalls in designing secure systems, and most of them have to do with the protocols (the data storage, user interface, threat model, etc.), not the basic secure cryptographic algorithms. A good rule of thumb is that if you cannot trust a single correct application of a secure cipher like AES, combined with a secure message integrity function like SHA1-HMAC, there is something fundamentally wrong with the design of the protocol or your understanding of the threat model.
I trust exactly one encryption product: GnuPG.
What about OpenSSL? It's the base for SSH and dozens of web servers and clients. I would bet it's the second most used encryption product after the crypto that ships with Windows.
1. Keep the data in CPU registers and cache.
2. Split the keys up into smaller pieces, and spread them around when in memory.
1. Impossible, without incredibly slow decryption. The x86_64 chips *might* have enough registers to do AES with lots of extra computation, but all the fast (e.g. more than 1MB/sec) algorithms use a large key schedule that is directly derived from the key. Heck, with some SSE hacks it might even be possible for generic x86_32 processors to keep the key in registers. Additionally, it would be apparent that the key is being stored in the registers because it could not be disguised very easily.
2. Slightly more doable, by moving and "disguising" (probably XORing with constants) the key schedule it could be forced to never exist entirely as plaintext in memory, so the crackers would have to do a little bit of time analysis to figure out when to grab parts of the key schedule. It's still going to be incredibly difficult because every 16 byte block of ciphertext requires every byte of key schedule to decrypt it. Given the massive amount of data streaming from the disc, the player is going to sit in two main areas: AES decryption and MPEG decoding. MPEG decoding takes up the majority of the time, but AES will be a very noticeable percentage, and probably easy to identify because of the table lookups that are necessary. After the code is identified, it's obvious how to get the original key back, because it forms the first 16 bytes of the key schedule.
An interesting solution would be to find a mathematically equivalent algorithm for AES that uses different lookup table constants and a different key schedule. It would probably not be very hard to do, maybe requiring an extra prestage and afterstage to put things right. That could take a little while to figure out, but again because of the time spent decrypting it would probably be relatively obvious where the code was, and when it didn't look like AES the hackers would know to look for the fudging to figure out how to reverse it. It might just take a hacker who's good at linear algebra.
The real solution the media companies are aiming for is to skip the software players and have the video card act as a secure player, handling the AES and MPEG in hardware and using HDCP to talk directly to a "secure" monitor. That will basically close the digital hole, because the only place to get the plaintext signal would be at the interface cable going into the LCD panel. If they can get the drive to talk directly to the video card for key exchange, that cuts the main PC out of the loop entirely, meaning that special hardware would be required just to read the data off the disc.
The AACS scheme has the ability to revoke individual players - not individual models, but actual single units. They use a lot of fancy set theory to do it, but in essence each player is supposed to have a unique set of keys - possibly hundreds of keys out of a total of many thousands (hundreds of thousands perhaps). Each disc has the information on it to allow thousands of different keys to decrypt it. The way it works is that of all the keys on the disc, it is expected that each individual player will have at least one key that matches.
Doesn't matter, because every copy of a movie has the exact same key that encrypts the actual movie (pressing discs relies on having a static image for the majority of the data, obviously small areas can be written individually). Discovering that key allows all HD discs using that key (all discs of a given release of a movie, for instance) to be viewed without DRM, and posting it on the Internet does not allow the MAFIAA to identify which player was used to extract the volume key.
And the computers in a single segment of cable modem network can be considered a LAN, right? What about the computers on a single university campus?
It all depends on how the network administrators decided to assign IP address ranges. As far as I know, most places don't put much consideration into making numerically adjacent subnets physically adjacent. It's too much work. Instead, you'll see 192.168.4 in one city and 192.168.5 in another city across the country, and the subnet mask may be set to 192.168.0.0/16 for the entire organization anyway, making it difficult to determine where the physical network boundaries are.
The first-order approximation to this is to prefer peers who share a longer IPv4 address prefix, right?
I doubt it. Even within subnets it's possible for machines to be widely distributed around the globe, but probably not very common outside of LANs. Take a look at the map of IP addresses that neatly displays the adjacency of successive first octets and how widely distributed they are around the world. While there are some large contiguous blocks, there are numerous discontinuities where subnets next to each other are on other sides of the world. The class C addresses are probably the worst. Adjacency information really needs to come from routing tables, preferably with some cost attached, to be useful.
It wouldn't be a matter of coming down on behavior that I thought was bad because anybody who engaged in it was obviously doing something else wrong. It would simply be a matter of preserving the quality of service for all the customers, not just the high-bandwidth users. I'm not the personal T3 connection for someone who's paying me $50/mo.
Just a suggestion: Have off-peak hours during which the bandwidth cap doesn't apply. Make it just one or two hours in the wee hours of the morning even, but it will let customers who want to push or pull several gigabytes a day to do so when it won't hurt anyone else.
In terms of trying to maximize intra-ISP file sharing, you probably need either a huge customer base or lots of customers that like to download the same things. I don't know if it's possible to force most of the P2P clients to localize their searches to one subnet without actually modifying the software and running your own servers. For bittorrent you might be able to watch connections for bittorrent data and have your servers join the torrent in order to cache data locally and hopefully provide a node to help point other customers to the local network. If you really want to help bittorrents and you have gobs of disk space just run a tracker on your own server for the torrents that your customers have downloaded, and let them know where they can find a list of trackers that you're hosting. Once you have that, you can selectively pick peers from your own subnet for your customers to maximize the intra-ISP traffic.
In general, all P2P clients should have a way of valuing peers in their own subnet higher than other peers, at least for picking the initial list of peers to try. One could even use the Internet BGP tables to find the closest peers in the overall network, which would almost always provide higher speed for P2P applications. Making every client keep a copy of the BGP tables might be kind of a stretch though.
Just find a vendor that actually supports your needs. Buying PCs with XP when Vista is still available will make for some nice statistics, too.
It's quite possible that a proper definition of what a human being "is" would disqualify fetuses and some babies. I don't think fear of that should necessarily stop us from defining it anyway. We can find other reasons to keep our kids around, like, say, because we love them.
Every living thing has the potential to become intelligent. It already happened at least once, after all. All it takes is time and ecological pressures, or of course human intervention. Intelligence amplification for animals would be a great way to understand more about the world from a different perspective.
The very idea of human rights is based on the premise that there is something intrinsically valuable in human beings, regardless of their mental capacities or physical abilities.
The ultimate question is where those intrinsic rights come from. Clearly in a naturalistic sense they have no source except as ideas in the heads of the creatures who assume they possess them. What's to say that chimps and other animals don't have their own ideas of self worth as a species? If so, shouldn't we respect their concept of value the same as we respect the human concept of value? There is ultimately nothing morally different between humans and animals except intelligence. Emotions are almost certainly as intense and rich in animals due to their source in the "lower" parts of the brain. Intelligence, as you said, is clearly not required for a human to have human rights. Why animals?
Eventually this will become more obvious when humans begin to modify themselves to be distinctly non-human either genetically or technologically. Clearly, they will continue to appreciate "human" rights, and it will become more apparent that animals are no different than us except in their level of intelligence. We can use animals as property now only because they are unorganized and relatively weak individually. As soon as some scientist puts more powerful brains in rats or the chimps, they will be able to organize animal rights campaigns themselves and humans will have to recognize them as intellectual and moral peers.
The Backwardistanis would need to prove that I've done anything inappropriate with the woman in question
In some fundamentalist countries, an unmarried woman traveling with a man who is not part of her immediate family is instantly a criminal. True, visiting such countries is probably a bad idea in the first place, but perhaps business or family needs might demand it, or even just dumb luck like plane trouble causing a diversion to the nearest airport.
Yes, the Internet DOES transport you if you specifically go after something physically located (the media or the server) in another location.
Riiight. That's why you pay taxes to the state or country the server lives that you buy things from online, and why you have to pay income tax for the items purchased from you on ebay to the state or country the buyer lives in. You are not touching the remote server, it's not even as physically connected as a phone line (most of which get packet switched for long distance calls anyway). You have no control over the route that your packets take, or the route return packets from the server take. Only specific laws about Internet traffic and computer access have jurisdiction, and only then in the country they are written. Extradition should be limited to crimes that can not be effectively prosecuted in the country where the defendant resides, not merely in the most politically convenient location. What good is your country if it won't protect you from the arbitrary laws of other countries?
If you buy from Lenovo, formerly IBM's PC division, they still recommend Windows XP professional over Vista, although Vista is selected by default (explain that). Perhaps the "Recommended by Lenovo" has been removed from the XP selection by now, or it only appears on computers without Vista drivers or something, but it's definitely still possible to get them shipped with XP.
I don't think that Microsoft was concealing anything. They were advertising a product with its niftiest features, but I think that about 15 minutes of research would have let someone know that they couldn't use the Aero interface. Microsoft used marketing and advertising to make their product look the best, that isn't the same as cheating someone.
To use your example, it would be like advertising cookie dough under the Vista trademark and clearly showing chocolate chips to be present in the cookies, but forgetting to mention that most "Vista Capable" ovens were actually incapable of baking cookies with chocolate chips in them. How is that not deceptive? Just because it's possible to discover (with 15 minutes of someone's personal time) that it's deceptive does not make it any better. How would a normal user even go about verifying whether all the features of an operating system would be available? Maybe a little sticker based on the technical specs of their computer? Hmmmmm.
Of course, I believe that any crime committed should warrant extradition, because that's just the right thing to do.
You have broken several country's laws today. Which one would you like to be extradited to first? Iran? China? North Korea?
I also believe that you should be judged on the laws of the land where the event took place, not on laws in your own homeland.
So when you touch down in Backwardistan on a connecting flight with your girlfriend, they can execute you for fornication even though you're an American citizen?
He acted "in" the USA, against US servers and security
I didn't know the Internet magically transported people to other countries. Does that mean the Chinese can arrest you for sending information about democracy to their country?
But I think it'd be -- or could be -- a visually spectacular sight to be able to perceive EM fields, because I imagine them being like auroras or the colors you see in plasma etch chambers. It'd make for a whole different form of art.
EM is directional as well as spatial. It would have to look like continuously flowing fluid so that you could see the magnitude and direction at every point in space. Color might give an indication of magnitude, but without the direction it would be confusing and somewhat useless, e.g. you couldn't sense polarity. Color might be useful to differentiate the electric and magnetic fields.
The classic example of that was the Kremer Prize for human-powered flight, won in 1977. Once that was done, interest in human-powered flight declined substantially. That effort didn't usher in an era of recreational pedal-powered flying.
In the article you linked to, you'll notice that the research to develop the Gossamer Condor lead to light, solar powered planes from NASA. Not every technology is immediately adopted by the masses, especially when human powered flight machines essentially require an athlete in peak condition to operate them.
The exact vendor selection criteria, often being secret, leaves vendors that had reasonable belief that they should have won completely baffled as to why they lost.
Maybe one of the criteria was "Cannot be unlocked with a hotel wet bar key."
That's the most interesting part of the article for me. Apart from 3-D rendering and folding@home, they are really pushed to find any real-world reason for having 4 cores.
Just wait for real time raytracing. From what I've seen of OpenRT, it only takes about 8 cores to get photorealistic real time raytracing for complex models entirely with software. Combining that with the advanced shaders on the GPU should make for some very visually stunning games with perfect reflections, refractions, shadows, and surface properties.
Normally I'd be on your side in the case like this but this time it seems you're just wasting the court's time fighting a battle you started. I agree that their message was overreaching, but does your want to prove that entitle you to duplicating and essentially broadcasting video they produced? Even if it did your methods amount to litigious entrapment.
It's a long, twisted road from fair use to litigious entrapment, don't you think? Or do you also think I shouldn't have been able to quote your post in my reply?
Sure, the idea of an abrupt Creation, or "Design," of the universe lets us joke about what God was doing before he got around to Creation, but the metaphor of water (or, let's suppose, some kind of cosmic stew) boiling into steam/universes leaves us with the same problem that we had in the first place: where in the [space larger than a universe] did the water/stew come from?
I think the universe is isomorphic to an infinite set. So far, we're on a fair track to proving that by finding very accurate mathematical models for the fundamental particles, and macroscopic behavior is very well defined. Even chaotic processes can be modeled. All this means is that the universe probably operates at some level like mathematical objects in set theory. The universe doesn't even have to be deterministic. A nondeterministic universe just splits into multiple parallel universes which is still relatively easy to express in set theory. Physics is just a set of equations that force the set to maintain the proper physical relationships between the matter and energy and topology of the universe.
What this means for existence and the beginning and end of the universe is that at any given instant in time, the universe can be represented as some subset of the entire universe, and there exist successor and predecessor functions that when applied to an instant in time will produce the "next" instant in time, or since the universe appears to be continuous at this point, any arbitrary time in the future or past. If the predecessor or successor function for an instant does not exist, those boundaries define the beginning or end of the universe respectively for all practical purposes. A nondeterministic universe is simply one where the successor and predecessor functions evaluate to a set of universe states, and each one of those states in turn has its own sets of states reachable through the successor and predecessor states.
Ultimately, only mathematics exists, which is to say that the relationships between hypothetical objects simply cause what is being represented by those objects to behave as if they existed. There is no difference between our universe, a simulation of our universe in a larger one, or a mathematical set that is isomorphic to our universe. A good example is the integers and the prime numbers. It's easy to prove that there are an infinite number of both integers and primes, and that means that even though our universe will never enumerate all the primes, their properties and relationships to other numbers still exist mathematically. By extension, in a model of a universe the atoms and fields have well defined positions at any given point in the future by using the mathematical rules of physics, and likewise the people in the model composed of those atoms and fields have well defined thoughts and actions over the entire domain of the universe. They exist, independent of being created or modeled or simulated, simply because it's possible to describe a mathematical system containing them. I can't find any counterproof to such an argument, but it is very difficult to prove from within our own universe. The best we can do is simulate small universes (and we do, usually of individual proteins or ideal gasses or even nuclear explosions) and extrapolate from that. Basically anything we can conceive of exists as its own universe, and all we have to do to see what happens in that universe is to simulate it. Within that universe, all the things that happen are already well defined by the initial conditions and the laws of physics.