The automobile has revolutionized our society - changed family life, geography, etc. The car's impact has been huge. While not everything the car has brought us has been good, on the whole, I'd say it's been worth it.
Little early to say. The automobile might be America's downfall - you don't quite know yet. Cars encouraged American society to build expansive cities that essentially prohibit efficient public transportation (and basically discourage some of the most fuel-efficient forms of transportation: walking and bicycles). If you take Europe, for example, cities are smaller and cars are much less used, thanks to the high gas prices, but also due to the city design. Essentially, cars encouraged the design of cities that are just, from a very basic level, very inefficient.
The next few decades will kindof prove whether or not the automobile was really beneficial - whether or not the "design that cars forced" can adapt easily to a non-fossil fuel system. It may be that the cost of shifting an entire nation's automobile fleet is more than our economy can handle.
As a simple example, many pizza places are beginning to charge for delivery, and are losing delivery people because they can't make money anymore. As gas prices continue to rise, you could expect more and more businesses that relied on cheap gas-fueled vehicles to struggle. Whether or not those businesses can handle the shift to a new energy source is an open question. Or, better put, whether or not the US economy can handle the transition as easily as Europe's can is an open question.
It may be in fifty years, people point to Europe's high fuel taxes as the turning point in the world's economy, saying "this ensured that Europe did not become as dependent upon oil as the United States did, and thus was able to adapt much easier when the need to transition away from oil became critical." The main problem with viewing the automobile as a positive influence is that you're assuming the current situation is stable, when it definitely isn't.
But to actually curb energy "consumption" is outrageous. We need to find new, safe, and more plentiful ways to produce as much energy as possible.
I think you're misusing the term "energy". Consistent increase in energy production is also not exactly stable - Second Law of thermodynamics, and all that. You can't expect power plants to be generating the equivalent power of the Sun eventually, nor can you imagine households consuming gigawatts of power to fuel random household appliances. At some point, the goal to increase energy production has to shift towards maximizing efficiency, accepting that only a finite amount of energy is available. That point is long off, definitely, but I'm not sure it's as long off as people think it is.
Neither does the P2P uploader know if the recipient has a legit copy... but why should they be any less liable than Blockbuster? Hell, the P2P user doesn't even get paid for this...
You're missing the point. Someone renting from Blockbuster doesn't cause copyright laws to be violated (i.e. nothing new is copied). So Blockbuster can't be in trouble at all, because all they're doing is using the legitimate copy they purchased in whatever way they like. Right of first sale allows them to do that. Whatever illegal acts the user does, Blockbuster can't be held liable for, because they didn't do anything to violate copyright law.
For a P2P user, though, nothing in copyright law allows them to upload a file to another user in the act of downloading. They're making a copy. Only the copyright holder is authorized to make copies. It's not a question of "the downloader isn't responsible for someone else breaking copyright law" - the downloader is breaking copyright law.
When the tracker gives you an IP, it's implicitly giving you permission to upload to that IP.
Only if it says somewhere in the license agreement that it does. Which, I believe, it will. Otherwise, you have no right to make a copy and send it to the other IP, by the strict letter of copyright law.
They don't have to explain to anyone WHY they're not suing.
They do if they try to sue someone else for doing the same thing - they're implicitly accepting that it's not copyright violation to upload in a BitTorrent download. It won't get that person out of the one copyright violation, no, but it will prevent them from tacking on a hundred other violations. It's at least a good argument for a judge.
Doesn't matter. Unless it's in the license when you download, the downloader is not allowed to make a copy of the copyrighted work, and that's what they're doing. That's why I figure it will probably be in the license agreement. No way their lawyers are that stupid.
I would love if they would make the argument that if they're authorized to have it, then they are authorized to download, and upload as well: because that just justifies the previous argument.
If they say "yes, it's okay for a recipient of a legal copy of a copyrighted work to make a copy of it to give to another person who has purchased a license for said copyrighted work", that justifies a person downloading an MP3 of a CD that he owns.
Then, P2P networks of copyrighted material become legal - it just becomes illegal to download if you don't own it already (how many of those disclaimers have you seen?).
Yeah. But that's what happens in a normal light-going-through-a-medium case. I was under the impression that with a BEC, where there are no distinct atoms, it doesn't make sense to say that the photon was scattering off electrons.
There are no distinct atoms. That doesn't mean there are no distinct electrons.
It's not like the electrons converted into an odd non-charged particle: they're there, just in odd bound states with other electron-proton pairs. The photons are still scattering off electrons (well, electrons and protons): just electrons that are in a collapsed, much tighter state. Their fields are still there, and still interacting, and that's what causes the high index of refraction.
One way to think about it is to stop with the 'slowing down light thing', and instead conceive it as the BEC swallowing up photons for a while, storing the information, and then reconstructing a new photon which is exactly identical at the end.
But that's not what's really going on. Even in the case where you hit a hydrogen atom with a photon in its absorption band, it won't reemit an identical photon due to conservation of momentum (and spherical symmetry of the hydrogen atom). It was a pretty amazing feat earlier when a lab managed to "preserve" the state of a photon for a period of time, but that's totally different than a high index of refraction.
On a smaller scale, you could imagine the photon merely scattering constantly off of electrons, making it virtually impossible for it to continue forward progress.
On a more macroscopic scale, you could imagine the photon as a propagating electromagnetic wave, and the speed of the electromagnetic field being limited because of all of the electron electric fields. Essentially, the electromagnetic field is slowed because it needs to "drag" the electrons around as well (An electron can break the electromagnetic field - Cerenkov radiation - because it *isn't* just an EM field. A photon, however, *is* just an EM field, and so it can't move any faster than the field propagates).
Not just not without destroying them, but really not at all (if you do assume that all the properties really *are* the same)
No need to assume it. You can measure it.
The statistics are different between a Maxwell-Boltzmann case (distinguishable particles), Fermi-Dirac (indistinguishable, negative symmetry), and Bose-Einstein (indistinguishable, positive symmetry).
Just generate a gas of the particles, and measure the velocity dispersion and the temperature. If it follows Fermi-Dirac or Bose-Einstein characteristics, then the properties of the particles really are indistinguishable. The math wouldn't work otherwise.
It would be beneficial if the MPAA, RIAA or others embraced using p2p such as bittorrent because it helps to legitimize p2p. This doesn't mean protection for pirates
Actually, it could. Part of the problem with BitTorrent is the fact that downloading necessitates uploading - so even if you have a legal right to obtain a copy of something, in order to do it, you have to upload to other users (who may or may not have the right to have it). By embracing BitTorrent as a protocol, they're removing one of the possible objections - namely, they can't claim that a person downloading something from BitTorrent illegally is committing copyright infringement thousands of times - just once.
(Of course, this presumes that they won't recognize this problem and build it into some license agreement crap thing, which they probably will. It also doesn't avoid the objection that just because I have a legal license to one copy of a copyrighted work, and a fair right to make one copy of that copy, doesn't give me the right to make a copy of a different copy.)
Wait, if this is just refraction, then the light isn't slowed at all, right?
No, it is. Mentioning refraction is a little odd, as refraction is caused by the slowing of light, not the cause of the slowing of light.
Once you're out of free space, the speed that an electric field can move can be hugely affected by density, etc.
Think of it this way: in a high optical density material, light is so slow because it has to drag electrons around as it moves. Light's an electromagnetic field, after all, and electrons have an electric field.
Now, you could *also* consider on a very, very small scale (sub-sub-atomic) that the photons are in fact still traveling at the speed of light - it's just that they're interacting so often with the electrons present that their net speed is very, very, very low.
Then OSDL management violated the license by not making sure that their employees weren't sniffing packets off the LAN to develop a competing SCM. Especially after they were told that was happening, and by whom.
OSDL management has no legal or contractual requirement to do so. There's nothing in the commercial contract that says you can't develop a competing SCM - at least, nothing I could see. Heck, if Linus does have a commercial contract, he could've given Tridge all of the data.
I don't even think the free version would restrict them, either. It's Tridge who's working on a competing SCM (not really, but let's pretend), not the users, who agreed to the license and are bound by it.
These aren't word games. BitMover's trying to get away with a clause in a contract that isn't really tremendously enforceable. Go look up Lasercomb America vs. Reynolds, or similar cases. It's very similar, and very close to copyright misuse. BitMover is trying to get away with a copyright protected monopoly (on BitKeeper clients), and it's simply not legally or contractually possible.
Damn, we'll have to let them get away with it this time!
Uh, yes. That's the entire point of writing a clear contract. There are several loopholes in the GPL already - right of first sale creates one.
Maybe you would prefer to have licenses trump protected fair use. I, however, would prefer to keep right of first sale and reverse engineering, and that means that some licenses and business models are not going to work. Good.
So you're saying that Tridge has been secretly using someone else's BK server, and to this day, no one but Tridge knows whose BK server it is? I don't believe it.
No. For one thing, the BitKeeper servers don't contain a "if you allow unauthorized clients to connect, we terminate your license" clause (well, the commercial license doesn't have anything like that). I was just saying that you can't prevent unauthorized clients from connecting, fundamentally.
As it stands, there's nothing in the commercial contract that allows BitMover to terminate their contract just for letting Tridge do a tcpdump on BK traffic. I don't know about the free use contract, because I can't view it at their site.
Actually, Larry has refused to sell licenses to people involved with other SCM systems in the past; there is no reason to think that policy has changed. The options were probabably
Refusing to sell is one thing - but terminating Linus's contract is almost certainly misuse of copyright (see Lasercomb America vs. Reynolds). So Linus can keep using BK, but as others can't, it's all moot.
What if the owner of the BK server knowingly allows Tridge to connect to it?
In that case, they'd be in violation. But if Tridge did his job right, the owner of the BK server can't tell the difference between BitMover's client and Tridge's. That's the point. You can't tell the difference between a Windows SMB client and a Samba SMB client (or, at least, they could've made it so that you can't. Windows didn't contain a clause like this).
If BitMover claims that Tridge connected to the server, they have to be able to prove that the owner should have been able to tell, and prevent Tridge from connecting to it. That assumes that it is possible. If it's not, then that portion of the license is essentially pointless.
To be honest, I'm pretty sure that particular clause in BitMover's contract (the one that says you can't work on reverse engineering or duplicating this product) is completely unenforcable, as from Lasercomb America, Inc. vs. Reynolds - it's misuse of copyright, almost directly to the letter. Lasercomb had a clause that prevented employees from developing video game software. It's pretty much exactly the same as here (instead of "employees", you have "licensees", and instead of "video game software" you have "SCM software", but that's nothing important), and Lasercomb lost.
Violations of contract law are usually followed by lawsuits and a lot of money in damages.
Both ways. If BitMover terminates a contract illegally, it's just as bad as someone violating it. Can be worse, actually, as you can claim more damages.
This is a legal problem, not a technical one. If my door lock fails to prevent you from breaking into my house, does that mean that you're allowed to steal my TV?
There is a law that prevents breaking and entering private property.
Where is the law that says "No citizen of the United States may connect to a BitMover BitKeeper server unless it is with an authorized client?"
This isn't criminal law. It's contract law.
At best, BitMover would be able to terminate your license. But they can't really terminate licenses on a whim, because the user is complying with the license to the best of his ability. Numerous court cases have upheld that. If BitMover tried to terminate a user's license to run a server, that user could sue BitMover for breach of contract, and they would likely win.
Watch network traffic between the client and the server, and reimplement it based on what you see. It's easy enough. You don't have to be the one running the client - you just tap off of someone else.
That other person doesn't even have to know that you're doing it. Just plug the server into a hub, plug another computer into another port, and spy on all the traffic - all of which is completely legal.
Here, he's avoiding the bnetd problem, where they reimplemented it based on what they saw when they ran the game, which means they agreed to the Battle.net licenses. Tridge never saw, never knew of, and never agreed to the license.
It's even easier because Tridge knows fundamentally what the client is doing - accessing the Linux kernel and adding patches, for which the source code is available in both cases.
There were clients available already for the platform
Ah, yes, the "well, there are clients available for X platform, so why do you need to reverse engineer it?" argument.
Here's a hint: you don't need a reason to reverse engineer something. None. It's been explicitly protected many times in court proceedings. It's explicitly protected in copyright code when it comes to semiconductor chip design.
Look it up - for semiconductor designs, it's explicitly protected for one person to crack open a chip, write down the way the chip works, and for another person to reimplement the way the chip works based on the first person's documentation. They don't need a reason. They're explicitly allowed to do it.
If someone implements the chip without the first person's documentation, then it's even more explicitly okay - which is actually what Tridge did. No access to source code or workings whatsoever - just figuring it out from network traffic. Of course, maybe you'd like to say that someone can say "You may not use this product if your ethernet traffic may be captured" in which case, no one can use BitKeeper.
Sure, his solution would have been OSS, but that wasn't the drive behind it.
How do you know? How do you know his goal wasn't to make a BeOS client? But again, as I said above, you don't need a point to reverse engineer.
This client would sink their business.
I doubt it - and if that's true, then they're resting on their laurels far too much, because you're saying "they won't be able to compete with a reverse-engineered client." If that's true, that's competition for you. No one would've wanted to give IBM a free pass simply because it couldn't compete with Compaq's PCs.
Essentially, you're talking about granting them a monopoly. A monopoly on BitKeeper clients. If their business model relies on having a monopoly on BitKeeper clients, then they deserve to be sunk.
If the server does not provide tools for preventing non-authorized clients from connecting, I highly doubt the license is enforceable. Doing this is impossible, since how could it know whether or not a client is authorized?
The client can be engineered to look *exactly* like an authorized client (which... is the entire point, after all) and it will be entirely legal (there is no uber-DMCA, after all - it only applies to protecting copyright infringement).
BitMover: "You violated our license by letting Tridge access your server!" Admin: "But... how do you know he did? I can't tell. All the clients look the same." BitMover: "Well, we don't! But we suspect it! So we're pulling your license!"
By reverse enginerring Larry's code you're stealing his design.
Yah. Good call. Tridgell broke into Larry's house, stole a big box that said "Design" on it.
He didn't steal - nothing physical was taken. At best you're making a claim for copyright infringement, not theft - but you can't even do that, because Tridge was reverse engineering a design, and you can not copyright a design. You can only patent one.
And if you honestly think it's easy to clean-room reverse-engineer a proprietary design, you're nuts. At least with designing an SCM yourself, you have some knowledge of what you're doing.
While I may be obsessing over the whole "living in fear" deal,
Well, you're at least obsessing over fear more than you're understanding the scales involved.
A plane flies at what, 30,000 feet? This thing is 100,000 kilometers or so long. So a plane hits it. So what? You've just chopped off the bottom one one-hundredth of a percent of the elevator.
Plus, of course, the main point of the first elevator is to build more. The main cost of the elevator is lifting the first one. So what if one is knocked down? Lowering another one isn't that hard, and the first one is almost definitely recoverable anyway.
This is a space-based object we're talking about. Until terrorists get low-orbit missiles, don't worry about it.
The transistor speed that matters is the switching frequency (that is, how fast it can switch from a 1 to a 0, and vice versa). That's what the "GHz" means, right? OK.
The point here is that it would be pointless to have a CPU made of transistors that switched at 100 THz (optical) frequencies. It wouldn't be faster at all.
Why? Because that chip still has to be made of atoms, and atoms have a scale. That scale is on the order of nanometers. Let's just say 1 nanometer, to make things convenient.
At 100 THz, a signal can only travel about 3 microns in a clock cycle. If you've got a scale of 1 nanometer, that means every processing element has to be less than 3000 elements distance in order to keep things in sync - and there's no way you can build a CPU like that - it's just not enough components.
Scale is what limits the frequency of chips, not the method of signal propagation, and chips are very, very rapidly reaching the point where it's simply not feasible to scale anymore - not from a physical point of view, but from a signal propagation point of view.
Anyway, that's not the big bottleneck in processors anyway. Bandwidth is. A modern CPU spends a huge amount of time stalling and waiting on data. If there's anything that can speed up that data transfer, it'll have a much larger impact than CPU speed increases.
Also, I don't know of any physical device manufactuer who is even remotely claiming even a 1% efficiency for storage and manipulation of a bit. (That would be a huge marketing ploy if it ever were achieved.) Physical devices, even optical systems, are far less efficient than that. There is no way that they are > 100,000 times more energy efficient than conductor/semi-conductor systems if simply because it would have been done already if the savings were that substantial.
Optical systems are orders of magnitude (maybe not 5 orders, but at least 2 or 3) more efficient than electronic ones, simply because there's no resistance and because you can multiplex signals optically rather than needing to do it electronically.
A good talk, and she was willing to give away quite a few nanoseconds, much less picoseconds. It really gets the concept of distances in small times to a perspective that your mind can grasp real easily.
Speed of light is a foot per nanosecond (literally: google it - it's 0.98 ft/ns), or 10 mils per picosecond. Easy enough. Incredibly useful mnemonic if you need to generate a delayed signal. "Get me 20 feet of cable, I need a clock-cycle delay." (There's a factor of 2-ish in there for the signal speed, but it's not that big a deal, and easy enough to remember).
Still, even assuming that we can overcome some of the issues with FTL communication at some point in the future, Planck's constant is going to be lurking in the background ready to bite even if we are using individual quarks for gate switching.
You realize that you are considering the speed of light to be a "soft" issue (as opposed to the "hard" limits from energy concerns). What makes you think that the speed of light is any less "hard" than the other concerns? Or, put another way - if we can get around the speed of light issue, I guarantee we can get around the entropic considerations (for one thing, the two aren't independent constraints).
Something beyond a photomask on lithographed semi-conductors must be done to get another 1000x increase in CPU speeds. Manufacturers are already using X-rays to get the fine details that are needed for the device manufacturing. If the frequencies get much higher, it will move into the gamma-ray section of the EM band.
Nah, you use non-optical methods, like electron beam lithography (which is planned). But that's all scale issues. You can't get another thousand-fold increase in CPU speeds. It's not going to happen. Another tenfold increase is not going to happen. CPU speed increases were done and through the instant you saw "drive" stages appear in the chip design to compensate for speed of light delay. They have hit the wall. The main increase in computing power at this point will come via increases in computational ability, not computational frequency.
I don't think you need to clarify that it's an ellipse. I actually doubt that the correction from calling it an ellipse and not a parabola is much more significant than corrections for about a dozen other things, including air resistance.
But it was only Galileo in the 1600s AD, about 2000 years later, who observed this.
Now you're underestimating people's power of observation. Archers and artillery men knew perfectly well what path a projectile takes. Otherwise they couldn't aim. They just didn't know how to write it into words.
Modern basketball players are the same way. If they hit a basket, and then you put a barrier in the path that they just shot, I wouldn't doubt if many of them could alter their shot to avoid the barrier and still go in - which requires that you have to know what the shape of the path of the projectile is (and not just where it would land).
In exactly the same way, I think if you pointed to the moon during a lunar eclipse, and said "that is the Earth's shadow," people would realize that the Earth is round fairly quickly.
At least, I think they'd realize it much faster than if you tried to explain that the horizon was due to things being underneath the curvature of the Earth. In one example, you point up, and you see the shape of the Earth on the Moon. In another, you have to imagine the Earth's shape.
Anyway, it's kinda pointless - the Greeks knew what the shape of the Earth was. Eratosthenes calculated the Earth's radius in ~200 BC. Obviously they were more observant than you think. Several Greeks (notably Aristotle) just had an aversion to actually doing experiments.
There are fundimental information theory limits to how little energy can be consumed to flip that bit, and the formula of E=hf is a good place to start and try to figure out just how much energy must be used to change a one to a zero and back. The emmission and absorbtion of photons will increase entropy, and will eventually lead to a loss via emmision into the IR band. This generates heat.
Well, yes, but that energy is miniscule, even at optical frequencies. Even if you assume that it was changing a megabit of information every clock cycle for an entire second, that's still only 7 watts. And that's a processor that's well over a billion times more powerful than a modern processor. That's a perfectly efficient processor, yes, but you get the point.
And if the CPU clock frequency is increased to the degree claimed (100,000 times)
As I posted elsewhere, if the CPU clock frequency is that high, we're building computers out of individual atoms, which is, I believe, a harder task than dealing with dissipating 7 watts.
The physical constant which hurts you the most at these frequencies isn't Planck's constant. It's the speed of light.
Whoops, I'm an idiot. I carefully said "scale" originally (because I didn't want to confuse things), and then gave an example that said it scaled linearly. Oops.
Slashdot Mirror
The automobile has revolutionized our society - changed family life, geography, etc. The car's impact has been huge. While not everything the car has brought us has been good, on the whole, I'd say it's been worth it.
Little early to say. The automobile might be America's downfall - you don't quite know yet. Cars encouraged American society to build expansive cities that essentially prohibit efficient public transportation (and basically discourage some of the most fuel-efficient forms of transportation: walking and bicycles). If you take Europe, for example, cities are smaller and cars are much less used, thanks to the high gas prices, but also due to the city design. Essentially, cars encouraged the design of cities that are just, from a very basic level, very inefficient.
The next few decades will kindof prove whether or not the automobile was really beneficial - whether or not the "design that cars forced" can adapt easily to a non-fossil fuel system. It may be that the cost of shifting an entire nation's automobile fleet is more than our economy can handle.
As a simple example, many pizza places are beginning to charge for delivery, and are losing delivery people because they can't make money anymore. As gas prices continue to rise, you could expect more and more businesses that relied on cheap gas-fueled vehicles to struggle. Whether or not those businesses can handle the shift to a new energy source is an open question. Or, better put, whether or not the US economy can handle the transition as easily as Europe's can is an open question.
It may be in fifty years, people point to Europe's high fuel taxes as the turning point in the world's economy, saying "this ensured that Europe did not become as dependent upon oil as the United States did, and thus was able to adapt much easier when the need to transition away from oil became critical." The main problem with viewing the automobile as a positive influence is that you're assuming the current situation is stable, when it definitely isn't.
But to actually curb energy "consumption" is outrageous. We need to find new, safe, and more plentiful ways to produce as much energy as possible.
I think you're misusing the term "energy". Consistent increase in energy production is also not exactly stable - Second Law of thermodynamics, and all that. You can't expect power plants to be generating the equivalent power of the Sun eventually, nor can you imagine households consuming gigawatts of power to fuel random household appliances. At some point, the goal to increase energy production has to shift towards maximizing efficiency, accepting that only a finite amount of energy is available. That point is long off, definitely, but I'm not sure it's as long off as people think it is.
Neither does the P2P uploader know if the recipient has a legit copy... but why should they be any less liable than Blockbuster? Hell, the P2P user doesn't even get paid for this...
You're missing the point. Someone renting from Blockbuster doesn't cause copyright laws to be violated (i.e. nothing new is copied). So Blockbuster can't be in trouble at all, because all they're doing is using the legitimate copy they purchased in whatever way they like. Right of first sale allows them to do that. Whatever illegal acts the user does, Blockbuster can't be held liable for, because they didn't do anything to violate copyright law.
For a P2P user, though, nothing in copyright law allows them to upload a file to another user in the act of downloading. They're making a copy. Only the copyright holder is authorized to make copies. It's not a question of "the downloader isn't responsible for someone else breaking copyright law" - the downloader is breaking copyright law.
When the tracker gives you an IP, it's implicitly giving you permission to upload to that IP.
Only if it says somewhere in the license agreement that it does. Which, I believe, it will. Otherwise, you have no right to make a copy and send it to the other IP, by the strict letter of copyright law.
They don't have to explain to anyone WHY they're not suing.
They do if they try to sue someone else for doing the same thing - they're implicitly accepting that it's not copyright violation to upload in a BitTorrent download. It won't get that person out of the one copyright violation, no, but it will prevent them from tacking on a hundred other violations. It's at least a good argument for a judge.
Doesn't matter. Unless it's in the license when you download, the downloader is not allowed to make a copy of the copyrighted work, and that's what they're doing. That's why I figure it will probably be in the license agreement. No way their lawyers are that stupid.
I would love if they would make the argument that if they're authorized to have it, then they are authorized to download, and upload as well: because that just justifies the previous argument.
If they say "yes, it's okay for a recipient of a legal copy of a copyrighted work to make a copy of it to give to another person who has purchased a license for said copyrighted work", that justifies a person downloading an MP3 of a CD that he owns.
Then, P2P networks of copyrighted material become legal - it just becomes illegal to download if you don't own it already (how many of those disclaimers have you seen?).
Yeah. But that's what happens in a normal light-going-through-a-medium case. I was under the impression that with a BEC, where there are no distinct atoms, it doesn't make sense to say that the photon was scattering off electrons.
There are no distinct atoms. That doesn't mean there are no distinct electrons.
It's not like the electrons converted into an odd non-charged particle: they're there, just in odd bound states with other electron-proton pairs. The photons are still scattering off electrons (well, electrons and protons): just electrons that are in a collapsed, much tighter state. Their fields are still there, and still interacting, and that's what causes the high index of refraction.
One way to think about it is to stop with the 'slowing down light thing', and instead conceive it as the BEC swallowing up photons for a while, storing the information, and then reconstructing a new photon which is exactly identical at the end.
But that's not what's really going on. Even in the case where you hit a hydrogen atom with a photon in its absorption band, it won't reemit an identical photon due to conservation of momentum (and spherical symmetry of the hydrogen atom). It was a pretty amazing feat earlier when a lab managed to "preserve" the state of a photon for a period of time, but that's totally different than a high index of refraction.
On a smaller scale, you could imagine the photon merely scattering constantly off of electrons, making it virtually impossible for it to continue forward progress.
On a more macroscopic scale, you could imagine the photon as a propagating electromagnetic wave, and the speed of the electromagnetic field being limited because of all of the electron electric fields. Essentially, the electromagnetic field is slowed because it needs to "drag" the electrons around as well (An electron can break the electromagnetic field - Cerenkov radiation - because it *isn't* just an EM field. A photon, however, *is* just an EM field, and so it can't move any faster than the field propagates).
Not just not without destroying them, but really not at all (if you do assume that all the properties really *are* the same)
No need to assume it. You can measure it.
The statistics are different between a Maxwell-Boltzmann case (distinguishable particles), Fermi-Dirac (indistinguishable, negative symmetry), and Bose-Einstein (indistinguishable, positive symmetry).
Just generate a gas of the particles, and measure the velocity dispersion and the temperature. If it follows Fermi-Dirac or Bose-Einstein characteristics, then the properties of the particles really are indistinguishable. The math wouldn't work otherwise.
It would be beneficial if the MPAA, RIAA or others embraced using p2p such as bittorrent because it helps to legitimize p2p. This doesn't mean protection for pirates
Actually, it could. Part of the problem with BitTorrent is the fact that downloading necessitates uploading - so even if you have a legal right to obtain a copy of something, in order to do it, you have to upload to other users (who may or may not have the right to have it). By embracing BitTorrent as a protocol, they're removing one of the possible objections - namely, they can't claim that a person downloading something from BitTorrent illegally is committing copyright infringement thousands of times - just once.
(Of course, this presumes that they won't recognize this problem and build it into some license agreement crap thing, which they probably will. It also doesn't avoid the objection that just because I have a legal license to one copy of a copyrighted work, and a fair right to make one copy of that copy, doesn't give me the right to make a copy of a different copy.)
Wait, if this is just refraction, then the light isn't slowed at all, right?
No, it is. Mentioning refraction is a little odd, as refraction is caused by the slowing of light, not the cause of the slowing of light.
Once you're out of free space, the speed that an electric field can move can be hugely affected by density, etc.
Think of it this way: in a high optical density material, light is so slow because it has to drag electrons around as it moves. Light's an electromagnetic field, after all, and electrons have an electric field.
Now, you could *also* consider on a very, very small scale (sub-sub-atomic) that the photons are in fact still traveling at the speed of light - it's just that they're interacting so often with the electrons present that their net speed is very, very, very low.
Then OSDL management violated the license by not making sure that their employees weren't sniffing packets off the LAN to develop a competing SCM. Especially after they were told that was happening, and by whom.
OSDL management has no legal or contractual requirement to do so. There's nothing in the commercial contract that says you can't develop a competing SCM - at least, nothing I could see. Heck, if Linus does have a commercial contract, he could've given Tridge all of the data.
I don't even think the free version would restrict them, either. It's Tridge who's working on a competing SCM (not really, but let's pretend), not the users, who agreed to the license and are bound by it.
These aren't word games. BitMover's trying to get away with a clause in a contract that isn't really tremendously enforceable. Go look up Lasercomb America vs. Reynolds, or similar cases. It's very similar, and very close to copyright misuse. BitMover is trying to get away with a copyright protected monopoly (on BitKeeper clients), and it's simply not legally or contractually possible.
Damn, we'll have to let them get away with it this time!
Uh, yes. That's the entire point of writing a clear contract. There are several loopholes in the GPL already - right of first sale creates one.
Maybe you would prefer to have licenses trump protected fair use. I, however, would prefer to keep right of first sale and reverse engineering, and that means that some licenses and business models are not going to work. Good.
So you're saying that Tridge has been secretly using someone else's BK server, and to this day, no one but Tridge knows whose BK server it is? I don't believe it.
No. For one thing, the BitKeeper servers don't contain a "if you allow unauthorized clients to connect, we terminate your license" clause (well, the commercial license doesn't have anything like that). I was just saying that you can't prevent unauthorized clients from connecting, fundamentally.
As it stands, there's nothing in the commercial contract that allows BitMover to terminate their contract just for letting Tridge do a tcpdump on BK traffic. I don't know about the free use contract, because I can't view it at their site.
Actually, Larry has refused to sell licenses to people involved with other SCM systems in the past; there is no reason to think that policy has changed. The options were probabably
Refusing to sell is one thing - but terminating Linus's contract is almost certainly misuse of copyright (see Lasercomb America vs. Reynolds). So Linus can keep using BK, but as others can't, it's all moot.
What if the owner of the BK server knowingly allows Tridge to connect to it?
In that case, they'd be in violation. But if Tridge did his job right, the owner of the BK server can't tell the difference between BitMover's client and Tridge's. That's the point. You can't tell the difference between a Windows SMB client and a Samba SMB client (or, at least, they could've made it so that you can't. Windows didn't contain a clause like this).
If BitMover claims that Tridge connected to the server, they have to be able to prove that the owner should have been able to tell, and prevent Tridge from connecting to it. That assumes that it is possible. If it's not, then that portion of the license is essentially pointless.
To be honest, I'm pretty sure that particular clause in BitMover's contract (the one that says you can't work on reverse engineering or duplicating this product) is completely unenforcable, as from Lasercomb America, Inc. vs. Reynolds - it's misuse of copyright, almost directly to the letter. Lasercomb had a clause that prevented employees from developing video game software. It's pretty much exactly the same as here (instead of "employees", you have "licensees", and instead of "video game software" you have "SCM software", but that's nothing important), and Lasercomb lost.
Violations of contract law are usually followed by lawsuits and a lot of money in damages.
Both ways. If BitMover terminates a contract illegally, it's just as bad as someone violating it. Can be worse, actually, as you can claim more damages.
This is a legal problem, not a technical one. If my door lock fails to prevent you from breaking into my house, does that mean that you're allowed to steal my TV?
There is a law that prevents breaking and entering private property.
Where is the law that says "No citizen of the United States may connect to a BitMover BitKeeper server unless it is with an authorized client?"
This isn't criminal law. It's contract law.
At best, BitMover would be able to terminate your license. But they can't really terminate licenses on a whim, because the user is complying with the license to the best of his ability. Numerous court cases have upheld that. If BitMover tried to terminate a user's license to run a server, that user could sue BitMover for breach of contract, and they would likely win.
Watch network traffic between the client and the server, and reimplement it based on what you see. It's easy enough. You don't have to be the one running the client - you just tap off of someone else.
That other person doesn't even have to know that you're doing it. Just plug the server into a hub, plug another computer into another port, and spy on all the traffic - all of which is completely legal.
Here, he's avoiding the bnetd problem, where they reimplemented it based on what they saw when they ran the game, which means they agreed to the Battle.net licenses. Tridge never saw, never knew of, and never agreed to the license.
It's even easier because Tridge knows fundamentally what the client is doing - accessing the Linux kernel and adding patches, for which the source code is available in both cases.
There were clients available already for the platform
Ah, yes, the "well, there are clients available for X platform, so why do you need to reverse engineer it?" argument.
Here's a hint: you don't need a reason to reverse engineer something. None. It's been explicitly protected many times in court proceedings. It's explicitly protected in copyright code when it comes to semiconductor chip design.
Look it up - for semiconductor designs, it's explicitly protected for one person to crack open a chip, write down the way the chip works, and for another person to reimplement the way the chip works based on the first person's documentation. They don't need a reason. They're explicitly allowed to do it.
If someone implements the chip without the first person's documentation, then it's even more explicitly okay - which is actually what Tridge did. No access to source code or workings whatsoever - just figuring it out from network traffic. Of course, maybe you'd like to say that someone can say "You may not use this product if your ethernet traffic may be captured" in which case, no one can use BitKeeper.
Sure, his solution would have been OSS, but that wasn't the drive behind it.
How do you know? How do you know his goal wasn't to make a BeOS client? But again, as I said above, you don't need a point to reverse engineer.
This client would sink their business.
I doubt it - and if that's true, then they're resting on their laurels far too much, because you're saying "they won't be able to compete with a reverse-engineered client." If that's true, that's competition for you. No one would've wanted to give IBM a free pass simply because it couldn't compete with Compaq's PCs.
Essentially, you're talking about granting them a monopoly. A monopoly on BitKeeper clients. If their business model relies on having a monopoly on BitKeeper clients, then they deserve to be sunk.
If the server does not provide tools for preventing non-authorized clients from connecting, I highly doubt the license is enforceable. Doing this is impossible, since how could it know whether or not a client is authorized?
The client can be engineered to look *exactly* like an authorized client (which... is the entire point, after all) and it will be entirely legal (there is no uber-DMCA, after all - it only applies to protecting copyright infringement).
BitMover: "You violated our license by letting Tridge access your server!"
Admin: "But... how do you know he did? I can't tell. All the clients look the same."
BitMover: "Well, we don't! But we suspect it! So we're pulling your license!"
Yah, that might fly in court.
By reverse enginerring Larry's code you're stealing his design.
Yah. Good call. Tridgell broke into Larry's house, stole a big box that said "Design" on it.
He didn't steal - nothing physical was taken. At best you're making a claim for copyright infringement, not theft - but you can't even do that, because Tridge was reverse engineering a design, and you can not copyright a design. You can only patent one.
And if you honestly think it's easy to clean-room reverse-engineer a proprietary design, you're nuts. At least with designing an SCM yourself, you have some knowledge of what you're doing.
While I may be obsessing over the whole "living in fear" deal,
Well, you're at least obsessing over fear more than you're understanding the scales involved.
A plane flies at what, 30,000 feet? This thing is 100,000 kilometers or so long. So a plane hits it. So what? You've just chopped off the bottom one one-hundredth of a percent of the elevator.
Plus, of course, the main point of the first elevator is to build more. The main cost of the elevator is lifting the first one. So what if one is knocked down? Lowering another one isn't that hard, and the first one is almost definitely recoverable anyway.
This is a space-based object we're talking about. Until terrorists get low-orbit missiles, don't worry about it.
Or am I missing something here?
The transistor speed that matters is the switching frequency (that is, how fast it can switch from a 1 to a 0, and vice versa). That's what the "GHz" means, right? OK.
The point here is that it would be pointless to have a CPU made of transistors that switched at 100 THz (optical) frequencies. It wouldn't be faster at all.
Why? Because that chip still has to be made of atoms, and atoms have a scale. That scale is on the order of nanometers. Let's just say 1 nanometer, to make things convenient.
At 100 THz, a signal can only travel about 3 microns in a clock cycle. If you've got a scale of 1 nanometer, that means every processing element has to be less than 3000 elements distance in order to keep things in sync - and there's no way you can build a CPU like that - it's just not enough components.
Scale is what limits the frequency of chips, not the method of signal propagation, and chips are very, very rapidly reaching the point where it's simply not feasible to scale anymore - not from a physical point of view, but from a signal propagation point of view.
Anyway, that's not the big bottleneck in processors anyway. Bandwidth is. A modern CPU spends a huge amount of time stalling and waiting on data. If there's anything that can speed up that data transfer, it'll have a much larger impact than CPU speed increases.
Someone (perhaps with a perpetual motion machine) must think this is possible to accomplish.
And if that person has a perpetual motion machine and FTL travel, then he can get around both the speed of light issues and the heat issues!
Also, I don't know of any physical device manufactuer who is even remotely claiming even a 1% efficiency for storage and manipulation of a bit. (That would be a huge marketing ploy if it ever were achieved.) Physical devices, even optical systems, are far less efficient than that. There is no way that they are > 100,000 times more energy efficient than conductor/semi-conductor systems if simply because it would have been done already if the savings were that substantial.
Optical systems are orders of magnitude (maybe not 5 orders, but at least 2 or 3) more efficient than electronic ones, simply because there's no resistance and because you can multiplex signals optically rather than needing to do it electronically.
A good talk, and she was willing to give away quite a few nanoseconds, much less picoseconds. It really gets the concept of distances in small times to a perspective that your mind can grasp real easily.
Speed of light is a foot per nanosecond (literally: google it - it's 0.98 ft/ns), or 10 mils per picosecond. Easy enough. Incredibly useful mnemonic if you need to generate a delayed signal. "Get me 20 feet of cable, I need a clock-cycle delay." (There's a factor of 2-ish in there for the signal speed, but it's not that big a deal, and easy enough to remember).
Still, even assuming that we can overcome some of the issues with FTL communication at some point in the future, Planck's constant is going to be lurking in the background ready to bite even if we are using individual quarks for gate switching.
You realize that you are considering the speed of light to be a "soft" issue (as opposed to the "hard" limits from energy concerns). What makes you think that the speed of light is any less "hard" than the other concerns? Or, put another way - if we can get around the speed of light issue, I guarantee we can get around the entropic considerations (for one thing, the two aren't independent constraints).
Something beyond a photomask on lithographed semi-conductors must be done to get another 1000x increase in CPU speeds. Manufacturers are already using X-rays to get the fine details that are needed for the device manufacturing. If the frequencies get much higher, it will move into the gamma-ray section of the EM band.
Nah, you use non-optical methods, like electron beam lithography (which is planned). But that's all scale issues. You can't get another thousand-fold increase in CPU speeds. It's not going to happen. Another tenfold increase is not going to happen. CPU speed increases were done and through the instant you saw "drive" stages appear in the chip design to compensate for speed of light delay. They have hit the wall. The main increase in computing power at this point will come via increases in computational ability, not computational frequency.
I don't think you need to clarify that it's an ellipse. I actually doubt that the correction from calling it an ellipse and not a parabola is much more significant than corrections for about a dozen other things, including air resistance.
But it was only Galileo in the 1600s AD, about 2000 years later, who observed this.
Now you're underestimating people's power of observation. Archers and artillery men knew perfectly well what path a projectile takes. Otherwise they couldn't aim. They just didn't know how to write it into words.
Modern basketball players are the same way. If they hit a basket, and then you put a barrier in the path that they just shot, I wouldn't doubt if many of them could alter their shot to avoid the barrier and still go in - which requires that you have to know what the shape of the path of the projectile is (and not just where it would land).
In exactly the same way, I think if you pointed to the moon during a lunar eclipse, and said "that is the Earth's shadow," people would realize that the Earth is round fairly quickly.
At least, I think they'd realize it much faster than if you tried to explain that the horizon was due to things being underneath the curvature of the Earth. In one example, you point up, and you see the shape of the Earth on the Moon. In another, you have to imagine the Earth's shape.
Anyway, it's kinda pointless - the Greeks knew what the shape of the Earth was. Eratosthenes calculated the Earth's radius in ~200 BC. Obviously they were more observant than you think. Several Greeks (notably Aristotle) just had an aversion to actually doing experiments.
There are fundimental information theory limits to how little energy can be consumed to flip that bit, and the formula of E=hf is a good place to start and try to figure out just how much energy must be used to change a one to a zero and back. The emmission and absorbtion of photons will increase entropy, and will eventually lead to a loss via emmision into the IR band. This generates heat.
Well, yes, but that energy is miniscule, even at optical frequencies. Even if you assume that it was changing a megabit of information every clock cycle for an entire second, that's still only 7 watts. And that's a processor that's well over a billion times more powerful than a modern processor. That's a perfectly efficient processor, yes, but you get the point.
And if the CPU clock frequency is increased to the degree claimed (100,000 times)
As I posted elsewhere, if the CPU clock frequency is that high, we're building computers out of individual atoms, which is, I believe, a harder task than dealing with dissipating 7 watts.
The physical constant which hurts you the most at these frequencies isn't Planck's constant. It's the speed of light.
Whoops, I'm an idiot. I carefully said "scale" originally (because I didn't want to confuse things), and then gave an example that said it scaled linearly. Oops.