Maybe this was deliberate (though rather obvious) misinformation by a scientist who didn't want Hitler to get the bomb.
This diagram is exactly what I'd expect to find in a mainstream media article titled "How to build a nuke in your backyard." It's simple: you take around 8kg Pu from old reactor fuel, put them into a hollow sphere, put TNT around that and your nuke is ready. It just doesn't work without precise timing and an initiator.
I think, this is simply a fake. Deliberate misinformation, but created in 2005, not 1945.
I'm also in favor of abandoning the term "innovation" in discussions of OSS. The concept "innovation in computer programming" was coined by Microsoft when they realized they hadn't any invention to speak of. McVoy using the same NewSpeak is kind of fitting.
Well, sheet metal does in fact block microwaves and there are microwaves all around us that you actually might want to block. The emissions of your neighbor's wlan and bluetooth equipment as well as their wireless and mobile phones come to mind.
Do the microwaves cause health problems? Unlikely, I'd say, but the question is not settled yet. These people may be hypochondric, but they are not paranoid.
Anyway, the neighbor's wlan may not degrade your brain, but it does degrade your own wlan performance, and tin foil certainly helps there. It also makes eavesdropping harder, which may be a valid concern, too.
Lisp, Smalltalk, Haskell, Scheme, Ocaml, Erlang,..., combinations thereof. Oh, it has to be a "popular" or "major" programming language? Sorry, in this case, the project is doomed anyway.
Actually Fortran does have pointers. (Or something close to pointers, I don't know Fortran intimately enough.) Fortran has the same aliasing problems that C has. The difference is, if aliasing breaks an optimization in C or C++, the compiler is at fault. If the same happens in Fortran, the programmer is at fault, simply by definition. Amazing, but this small difference makes Fortran faster...
There is only one electron. The laser's energy is only sufficient to ionize one atom, but it is not clear if the first or second maximum actually freed the electron. The experiment shows that the atom is ionized by both maxima.
Also with the ordinary double slit experiment you try to measure position (which slit did it go through?) and you get a distribution over impulse (which direction is it going? hence the distributed interference pattern). Likewise when measuring time (when was the atom ionized?) you get a distribution over energy, which is what these guys plotted. Therefore there is no "collision point".
(To be honest, I don't quite see how this happens, however time and energy are connected in the same way as position and impulse.)
Well, actually nothing collapses and the measurement is nothing special. In a sense the photon took both ways around the lens. It always did so and all your measurement does is destroy the interference. Don't think of the photon as a particle and it all makes sense.
The voltage MUST be kept below the voltage of the battery or you will start doing funky things with the battery.
No, you need more voltage to charge a battery, otherwise no current will flow and nothing would happen. The voltage cannot be too high, for the high current would then cook the battery.
...and all of this has nothing to do with refuelling electric cars. Batteries are cells connected in series, which increases their voltage. And even without serial connection, there are switch mode regulators that would basically allow the car to be connected to some high voltage which is then reduced to something appropriate for the batteries without much loss.
I'm planing on building a high end HTPC setup using two 4:3 projectors
Nice idea, only SLI won't help you. Think about it: scanline interleave. You two projector setup doesn't interleave scanlines at all. What you are planning works with any two cards.
SLI can double the pixel throughput compared to a single card. It will not increase the polygon throughput, but the bus bandwidth needed will still double, along with the setup overhead.
Obviously Doom 3 taxes the pixel shaders heavily, Far Cry load the vertex shaders or something similar. Therefore, the former profits, the latter doesn't.
Anyway, 3dFX invented SLI 8 years or so ago with the Voodoo 2 cards. It was a stupid idea and 3dFX is gone now. 'nuff said.
I don't think the operating system could make much use of the APUs. The best that can be hoped for is an OS that somehow allocates apulets to the APUs, but since the APUs will work best if used as stream processors this allocation is... well... non-trivial.
However, given a way to allocate these units to userspace programs, there are lots of programs that could benefit. X and mplayer come to mind, provided someone implements the critical code for APUs, which may well mean coding in assembly.
What you need is a more general concept, probably at the programming language level, in which algorithmns can be expressed in such a way that the operating system can detect that they can be loaded into these subsidiary processors to be executed.
This will remain a dream for "general purpose languages" like C. However, I could imagine Parallel Haskell or something similar for the Cell. That would be way cool and could even work.
Anyway, the architecture without adequate software is quite useless. I'm still very much interested.
We use the heat generated by the decay of radioactive elements to fuel our generators. We do nothing like smashing atoms into smaller bits.
Damn well we do. We smash neutrons into uranium nuclei, splitting them in two, typically into krypton and barium, iirc. No, we can't split them into anything we like, but we do split nuclei.
Any typical C program is an unidiomatic and generally bad C++ program, and vice versa. Face it, the languages are different. C++ may almost be a superset of C, but that means that C++ has different idioms and evolves into a completely different direction.
My personal impression is that those self proclaimed "C/C++ programmers" write bad code in both these languages. They should stick to C/C++ and other hypothetical languages.
But note that the same considerations apply to C with the added complexity that you have to explicitly propagate your error code. Also note that error checking in C does cost performance, too, even if no error occured.
Instead, attach the full source code to each executable program. Install that program on all of the hardware you plan to lease. Then (and only then) zero out the parts of the program that contain the source code
Would you care to explain how you come to think that the program on this hardware is accompanied by its complete machine readable source code when it is placed in the hands of the user?
The reason there is a FAQ is that people don't understand the concept of a juridical entity. All copies belong to the company, so there is no transfer of rights from the company to their employees. All employees recieve and use the product as representatives of that legal entity.
Is this really true? The GPL talks about distribution of the program, not about some ephemeral right to use the program. So, the program is passed around within the company, and everyone who gets hold of it, has to get a license.
I also don't see how an individual can receive a program "as representative of a company". Employees are not representative of their employer, and the GPL also does not contain any wording that gives any weight to such a construction like "person as representative of something".
Section 2b of the GPL states: "You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License". So basically, if you distribute the program (to buyers of a device with the binary inside) you have to make the source available (to anyone who asks).
No you don't. The program being licensed under the GPL does not entitle anyone to anything. You only have a right to request the source for a GP-licensed program you already have the binary of. If you have the binary, only then do you get the right to distribute it, request the source, modify it.
That doesn't change the fact that any third party gets a license. But you can't run a license on your computer, you know?
Talk about exaggeration. This thing does not contain a single gram of plutonium, you cannot set it of even if you tried with both hands. The "nuclear material" in it is most likely just tritium and uranium. So what?
Tritium occurs in the atmosphere, it is continually produced by cosmic rays. With a half life of 12.5 years it is hardly worth worrying about. Just don't inhale large quantities of the stuff.
Uranium is rather plentiful in the oceans, about 4 gigatons (no, not TNT equivalent, real, massive 4000 million tons) estimated. Some hundred pound more rusting away slowly won't create glowfish or something.
And last but not least, the radioactivity in the area is 10 times the normal background level. You know what that means? Nothing at all. Some worms living in the sea bottom might get cancer, that's all.
Going off on another tangent, even terrorists could use this bomb only as a blueprint, if anything. The first phase is missing, it can't detonate without some pounds of plutonium. If detonated as a dirt bomb, the uranium will scare the living hell out of people, but hardly anyone would be harmed. A terrorist can wreak much more havoc using common household chemicals.
While I think using nuclear power to produce the hydrogen is in principle a good idea, you still end up investing the same amount of energy that is stored in the H2 to purify and pressurize it. Add to that the heavy pressure vessels needed to transport it, an H2 economy just sounds stupid. Better storage is needed, quite simply.
And alternative may be methanol. It's liquid, cheaply storable and can be produced from H2 and C02.
If so, something is wrong. Computers are good at repetition, humans can do better things. If your programming is repetitive, you are using the wrong tool. Heck, you should spend your time developing exactly this missing tool. Build a decent library or a preprocessor, then use it to automate away the repetition. It pays off soon. Compare http://c2.com/cgi/wiki?SharpenTheSaw and http://c2.com/cgi/wiki?ThreeStrikesAndYouRefactor.
Actually Linux is ready for the desktop. Those so called "regular users", the unwashed masses who could not think for themselves if their life depended on it, just aren't ready for Linux and never will be. Stop targeting them, it will only destroy Linux, if anything. They are happy with Windows, better don't disturb them.
It is a sad fact that a lot of web designers specifically target IE, whether out of the usefulness of some feature or out of sheer incompetence. Often the latter, but it simply doesn't matter: due to the market dominance of Windows there will always be things that are only possible on Windows. If Linux tries to catch up, that race is already lost.
Get over it. A large part of the desktop market is out of reach. A race for features with Microsoft cannot be won. The energy spent to mimic Windows is better invested in inventing new features unique to Linux or Unix. Let the Windows market alone and go on with different things. It's more rewarding.
Actually "$13 + 16%VAT" means 16% of $13, giving a total of $15 per computer, not $130. Given that private copying is legal in Germany exactly because copying machines and blank are taxed, I come to find that quite reasonable. Also nobody is called a thief, instead by buying a copying machine you also buy the right to actually make copies for private use. What's new is that the computer is considered a copying machine.
However, the situation is not as nice as that sounds. Even if we pay for the right to copy there is no way to enforce that right. More and more audio cds are sold with copy protection, despite the fact that we already pay for the copies we can no longer make.
Slashdot Mirror
Use vi and only the mouse. Go!
And think before you post, stupid!!
Maybe this was deliberate (though rather obvious) misinformation by a scientist who didn't want Hitler to get the bomb.
This diagram is exactly what I'd expect to find in a mainstream media article titled "How to build a nuke in your backyard." It's simple: you take around 8kg Pu from old reactor fuel, put them into a hollow sphere, put TNT around that and your nuke is ready. It just doesn't work without precise timing and an initiator.
I think, this is simply a fake. Deliberate misinformation, but created in 2005, not 1945.
Can anyone tell me if BitKeeper contains any innovations?
Not too much, I think arch (http://www.gnuarch.org/) and darcs (http://www.darcs.net/) were there first.
I'm also in favor of abandoning the term "innovation" in discussions of OSS. The concept "innovation in computer programming" was coined by Microsoft when they realized they hadn't any invention to speak of. McVoy using the same NewSpeak is kind of fitting.
Well, sheet metal does in fact block microwaves and there are microwaves all around us that you actually might want to block. The emissions of your neighbor's wlan and bluetooth equipment as well as their wireless and mobile phones come to mind.
Do the microwaves cause health problems? Unlikely, I'd say, but the question is not settled yet. These people may be hypochondric, but they are not paranoid.
Anyway, the neighbor's wlan may not degrade your brain, but it does degrade your own wlan performance, and tin foil certainly helps there. It also makes eavesdropping harder, which may be a valid concern, too.
Lisp, Smalltalk, Haskell, Scheme, Ocaml, Erlang, ..., combinations thereof. Oh, it has to be a "popular" or "major" programming language? Sorry, in this case, the project is doomed anyway.
Actually Fortran does have pointers. (Or something close to pointers, I don't know Fortran intimately enough.) Fortran has the same aliasing problems that C has. The difference is, if aliasing breaks an optimization in C or C++, the compiler is at fault. If the same happens in Fortran, the programmer is at fault, simply by definition. Amazing, but this small difference makes Fortran faster...
There is only one electron. The laser's energy is only sufficient to ionize one atom, but it is not clear if the first or second maximum actually freed the electron. The experiment shows that the atom is ionized by both maxima.
Also with the ordinary double slit experiment you try to measure position (which slit did it go through?) and you get a distribution over impulse (which direction is it going? hence the distributed interference pattern). Likewise when measuring time (when was the atom ionized?) you get a distribution over energy, which is what these guys plotted. Therefore there is no "collision point".
(To be honest, I don't quite see how this happens, however time and energy are connected in the same way as position and impulse.)
Well, actually nothing collapses and the measurement is nothing special. In a sense the photon took both ways around the lens. It always did so and all your measurement does is destroy the interference. Don't think of the photon as a particle and it all makes sense.
The voltage MUST be kept below the voltage of the battery or you will start doing funky things with the battery.
No, you need more voltage to charge a battery, otherwise no current will flow and nothing would happen. The voltage cannot be too high, for the high current would then cook the battery.
...and all of this has nothing to do with refuelling electric cars. Batteries are cells connected in series, which increases their voltage. And even without serial connection, there are switch mode regulators that would basically allow the car to be connected to some high voltage which is then reduced to something appropriate for the batteries without much loss.
BTW, what moron modded the parent "insightful"?
I'm planing on building a high end HTPC setup using two 4:3 projectors
Nice idea, only SLI won't help you. Think about it: scanline interleave. You two projector setup doesn't interleave scanlines at all. What you are planning works with any two cards.
SLI can double the pixel throughput compared to a single card. It will not increase the polygon throughput, but the bus bandwidth needed will still double, along with the setup overhead.
Obviously Doom 3 taxes the pixel shaders heavily, Far Cry load the vertex shaders or something similar. Therefore, the former profits, the latter doesn't.
Anyway, 3dFX invented SLI 8 years or so ago with the Voodoo 2 cards. It was a stupid idea and 3dFX is gone now. 'nuff said.
I don't think the operating system could make much use of the APUs. The best that can be hoped for is an OS that somehow allocates apulets to the APUs, but since the APUs will work best if used as stream processors this allocation is... well... non-trivial.
However, given a way to allocate these units to userspace programs, there are lots of programs that could benefit. X and mplayer come to mind, provided someone implements the critical code for APUs, which may well mean coding in assembly.
What you need is a more general concept, probably at the programming language level, in which algorithmns can be expressed in such a way that the operating system can detect that they can be loaded into these subsidiary processors to be executed.
This will remain a dream for "general purpose languages" like C. However, I could imagine Parallel Haskell or something similar for the Cell. That would be way cool and could even work.
Anyway, the architecture without adequate software is quite useless. I'm still very much interested.
For them to enjoy something they'd have to experience it and therefore have a consciousness.
As if that meant anything. How do you define consciousness and why do you think it has anything to do with the ability to experience joy?
I think, consciousness is the ability to model a world including a self-image. Or something like that. Whatever, I see no connection to joy.
Luckily, RS232 are dying ;-)
Yeah, but Ethernet repeated the same mistake and is sure to stay for a while.
We use the heat generated by the decay of radioactive elements to fuel our generators. We do nothing like smashing atoms into smaller bits.
Damn well we do. We smash neutrons into uranium nuclei, splitting them in two, typically into krypton and barium, iirc. No, we can't split them into anything we like, but we do split nuclei.
Any typical C program is an unidiomatic and generally bad C++ program, and vice versa. Face it, the languages are different. C++ may almost be a superset of C, but that means that C++ has different idioms and evolves into a completely different direction.
My personal impression is that those self proclaimed "C/C++ programmers" write bad code in both these languages. They should stick to C/C++ and other hypothetical languages.
But note that the same considerations apply to C with the added complexity that you have to explicitly propagate your error code. Also note that error checking in C does cost performance, too, even if no error occured.
Instead, attach the full source code to each executable program. Install that program on all of the hardware you plan to lease. Then (and only then) zero out the parts of the program that contain the source code
Would you care to explain how you come to think that the program on this hardware is accompanied by its complete machine readable source code when it is placed in the hands of the user?
The reason there is a FAQ is that people don't understand the concept of a juridical entity. All copies belong to the company, so there is no transfer of rights from the company to their employees. All employees recieve and use the product as representatives of that legal entity.
Is this really true? The GPL talks about distribution of the program, not about some ephemeral right to use the program. So, the program is passed around within the company, and everyone who gets hold of it, has to get a license.
I also don't see how an individual can receive a program "as representative of a company". Employees are not representative of their employer, and the GPL also does not contain any wording that gives any weight to such a construction like "person as representative of something".
What am I missing?
Section 2b of the GPL states: "You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License". So basically, if you distribute the program (to buyers of a device with the binary inside) you have to make the source available (to anyone who asks).
No you don't. The program being licensed under the GPL does not entitle anyone to anything. You only have a right to request the source for a GP-licensed program you already have the binary of. If you have the binary, only then do you get the right to distribute it, request the source, modify it.
That doesn't change the fact that any third party gets a license. But you can't run a license on your computer, you know?
I certainly wish people could learn to read.
He didn't. Linux users registered the trademark and gave it to Linus as a gift. Linux himself presumably wouldn't have bothered.
Talk about exaggeration. This thing does not contain a single gram of plutonium, you cannot set it of even if you tried with both hands. The "nuclear material" in it is most likely just tritium and uranium. So what?
Tritium occurs in the atmosphere, it is continually produced by cosmic rays. With a half life of 12.5 years it is hardly worth worrying about. Just don't inhale large quantities of the stuff.
Uranium is rather plentiful in the oceans, about 4 gigatons (no, not TNT equivalent, real, massive 4000 million tons) estimated. Some hundred pound more rusting away slowly won't create glowfish or something.
And last but not least, the radioactivity in the area is 10 times the normal background level. You know what that means? Nothing at all. Some worms living in the sea bottom might get cancer, that's all.
Going off on another tangent, even terrorists could use this bomb only as a blueprint, if anything. The first phase is missing, it can't detonate without some pounds of plutonium. If detonated as a dirt bomb, the uranium will scare the living hell out of people, but hardly anyone would be harmed. A terrorist can wreak much more havoc using common household chemicals.
While I think using nuclear power to produce the hydrogen is in principle a good idea, you still end up investing the same amount of energy that is stored in the H2 to purify and pressurize it. Add to that the heavy pressure vessels needed to transport it, an H2 economy just sounds stupid. Better storage is needed, quite simply.
And alternative may be methanol. It's liquid, cheaply storable and can be produced from H2 and C02.
The vast bulk of programming is just repetition.
.
If so, something is wrong. Computers are good at repetition, humans can do better things. If your programming is repetitive, you are using the wrong tool. Heck, you should spend your time developing exactly this missing tool. Build a decent library or a preprocessor, then use it to automate away the repetition. It pays off soon. Compare http://c2.com/cgi/wiki?SharpenTheSaw and http://c2.com/cgi/wiki?ThreeStrikesAndYouRefactor
Actually Linux is ready for the desktop. Those so called "regular users", the unwashed masses who could not think for themselves if their life depended on it, just aren't ready for Linux and never will be. Stop targeting them, it will only destroy Linux, if anything. They are happy with Windows, better don't disturb them.
It is a sad fact that a lot of web designers specifically target IE, whether out of the usefulness of some feature or out of sheer incompetence. Often the latter, but it simply doesn't matter: due to the market dominance of Windows there will always be things that are only possible on Windows. If Linux tries to catch up, that race is already lost.
Get over it. A large part of the desktop market is out of reach. A race for features with Microsoft cannot be won. The energy spent to mimic Windows is better invested in inventing new features unique to Linux or Unix. Let the Windows market alone and go on with different things. It's more rewarding.
Actually "$13 + 16%VAT" means 16% of $13, giving a total of $15 per computer, not $130. Given that private copying is legal in Germany exactly because copying machines and blank are taxed, I come to find that quite reasonable. Also nobody is called a thief, instead by buying a copying machine you also buy the right to actually make copies for private use. What's new is that the computer is considered a copying machine.
However, the situation is not as nice as that sounds. Even if we pay for the right to copy there is no way to enforce that right. More and more audio cds are sold with copy protection, despite the fact that we already pay for the copies we can no longer make.