Wasn't WordPerfect written in x86 assembly? I'm almost totally ignorant about Linux internals, but I suspect that
1) a program that assumes it is practically running on bare PC metal isn't going to work very well in a UNIX environment.
2) not too many people would be willing to go through x86 assembly code to figure out how to make it work with curses, e.g., and hook up to a real non-DOS filesystem.
WP's code might be wonderfully modular and easy to port, but any company that really was concerned about those metrics wouldn't have chosen assembly as their platform of choice, would they?
It's true that internet cafes are at risk from the authorities if they don't police the usage of chat rooms. I think the entrepreneurial impulse in urban China is going to find a way to overcome the political obstacles, perhaps through filtering technology, or more likely by greasing a few palms. Capitalists can officially join the Party now, you know. Still, my main point was to counter the view, which seems to be prevalent in this discussion, that internet users must have their own dial-up connection.
Internet cafes allow people to get on-line for pocket change; they don't need to foot the bill for a PC and phone line.
And in all of those countries, English is the primary second language. Widespread English competence is a huge marketing point for Singapore in attracting international business.
"From Taiwan-based company" != "manufactured/assembled in Taiwan"
The trend in Taiwan, as elsewhere, is to outsource more and more of the lower-tech work to the mainland. For those who are worried about the PRC's growing economic power, this is a major theme. For those who are worried about the ROC being able to maintain an independent existence from the PRC, this "hollowing out" of Taiwan industry is a serious concern.
For those who like to buy $400 PCs, it is a great thing.
But I would bet that most people who would use Hindi on the web are comfortable with English. In a country with hundreds of miscellaneous languages, with a common second language that happens to be the default world language, I don't think there is much incentive to use Hindi. Except to make a political point.
I don't think there is any practical way to expect that an open-source car design would be able to include sufficient information to allow one to manufacture the automobile profitably, if at all. There is too much process and industrial engineering expertise involved in bringing an automobile to production, not to mention the real expense of tooling, and so forth.
The design of modern automobiles is inseparable from the design of automotive assembly processes. Even GM doesn't make Corvettes all by themselves. A huge number of the components are procured from outside vendors on the basis of specifications, where each vendor has a large body of specialized experience in supplying parts to auto manufacturers. The whole flow of these products to the final assembly line is also a monstrous thing to organize.
Consider how difficult it can be to figure out why software doesn't compile when the usual "./configure; make; make install" fails. There is no analogous way to order that an automobile be produced; the problems when a car doesn't assemble itself are many orders of magnitude greater.
Sure, people can build their own cars in their garage; these cars are far different from the cars that GM makes, in terms of design sophistication, reliability, performance, availability of options, final cost to the consumer, and the ability for the average consumer to be able to drive it around and get it fixed when it stops working.
The point of a system of laws is to prohibit people from doing things which are technologically possible but have undesirable effects. When you say "no loss there" you miss the point. The loss is not that pre-existing information gets disseminated too widely, but rather that fewer people will put forth the effort to make useful information available at all if they cannot in someway be compensated for their effort.
Do you really want every viable information medium to have the signal-to-noise ratio of Slashdot?
Now I am sure you are completely wrong. A correct analysis of the problem shows that the travelling twin ages less than the stay-at-home twin. The Lorentz transformation does not depend on the sign of v; travelling toward Earth or away from Earth makes the clock move equally slowly on the ship, as observed from Earth.
The reason the situation is not symmetric is that the twin on the spaceship *turns around*, i.e., undergoes acceleration, and thereby changes inertial reference frames between the outbound and inbound journey. Accelerations are detectable; anything that isn't strapped down in the spaceship gets thrown around by the acceleration. Meanwhile, on Earth, nothing gets thrown around. To the extent that the Earth is not accelerating (i.e. to high accuracy), it remains in the same inertial frame throughout. This is different from the situation on the ship. The symmetry between the twins is an illusion, and is the source of confusion that leads to people calling it a paradox.
To repeat: there is NO TIME CONTRACTION ON THE RETURN JOURNEY. Please learn relativity correctly before posting again. Thank you.
I just have to say, for the benefit of anyone reading, that you are either completely wrong, or confused and irrelevant. The departure and arrival events in hypothetical FTL travel are separated by a spacelike interval, and cannot be causally connected, because causes have to come before effects. This is *including* any correction due to the delay of light travelling from the point of the event to the point of observation.
The speed of light in vacuum is c, regardless of the speed of the emitter or receiver. Your sound comment is a red herring. There is a distinguished frame of reference for sound, in which the medium of sound propagation is fixed. The speed of sound is not the same for observers moving with respect to the medium.
Also, in the twin "paradox," for physically realizable velocities less than c, time is dilated (as observed by the stay-at-home twin) in both directions for the travelling twin. Moving clocks run slow. They never run fast. *That* is the basis of the twin "paradox" which is, apart from an unfortunate name, is not a paraodox at all. I don't understand what you mean by "time-contraction." Yes, signals from the travelling twin to the stay-at-home twin are Doppler shifted, as are the signals in the opposite direction. However, neither observes the clock running faster in the other frame, after correcting for the signal transit time. That correction is assumed in any use of the word "observe" in relativity. Travelling twins age less than the stay-at-home twin. Re-read your relativity notes.
The real reason to believe that photons have zero rest mass is because of the inverse square law in electromagnetics.
The general idea is that interactions mediated by fields like electromagnetic fields (or weak nuclear forces) have a range which is determined by the mass of the force-carrying boson. The bosons mediating the inter-nuclear forces have mass, and are short-range (they have an exponential tail, which falls to zero quickly). But massless force-carrying particles should have inverse square behavior. (I.e. gravity and electromagnetism, where gravity is transmitted, in this kind of physical theory, by hypothetical "gravitons" analogous to the "photons" that transmit electromagnetic forces.)
The experiments to verify inverse-square behavior take the form of verifying Coulomb's law, which in turn takes the form of verifying that the free charge in a conductor resides on the surface. Read Jackson's _Classical Electrodynamics_ for some more description. (Chapter 0 or 1, I forget.)
Now, you might not be convinced: "but what if the connection between mass and force-coupling is not correct. The logic chain breaks down." Sure, but if that logic chain breaks down, then we don't even know what the word "photon" is supposed to refer to, much less what it would mean for such a "non-gauge-field photon" to have mass or not.
OK, thanks for the explanation, although the snide remark is unnecessary; I've gotten through plenty of classes without condescending to those others who have missed some point.
Still, the side point stands, which is that if asc(chr(x)) is not an identity, then the API has not provided the right tools. I.e., anybody actually using chr() or asc() is setting themselves up for a bug in Chinese Windows. Why have them around at all, except as a trap to the unwary? He claims this was a bug in his code; actually, it's probably a lurking bug in thousands of pieces of code. Given that APIs are, in principle, designed to enable third-party software, and not the other way around, it ought to be classified as a bug in Windows. Instead, thousands get it wrong, and a few persistent people get to lose sleep in order to figure out the mystical incantation that does the right thing.
If you expected asc(chr(x)) to be an identity operation, why didn't you just use x? If some conversion was necessary to satisfy the platform's API, why doesn't the platform provide a conversion that is guaranteed to work on all the frigging versions that are out there?
What is the point of a widespread platform if you potentially have to special case for every known variation?
If Windows hasn't gotten beyond this kind of inane kludginess, it's no wonder things are such a mess.
The reason this is forbidden is because of relativity of simultaneity. I don't know why this part of relativity is less well-remembered than the relativistic length-contraction and time-dilation, but it is really the key to almost all relativistic "paradoxes."
In any FTL travel, there are two events, A: leave the origin; B: arrive at the destination. FTL travel is believed to be impossible because observers in different inertial frames of reference would disagree about whether A or B happened first! Since it is paradoxical to arrive before you have left, the events cannot be causally connected.
Yeah, given the proper axioms, one can prove that 1+1=2, etc. But, in some sense, 1+1=2 is just re-stating the axioms that you felt it was convenient to assume for what you wished to call "sets" for instance.
There isn't any real way to be certain that those axioms do or do not apply to anything realistic. Sure, in concrete terms, I'm pretty sure I can count piles of apples and have the results make sense, but that is really just a gut feeling, and not provable.
Who knows if the universe obeys any logical rules at all? No one can. There's no way to know that God won't show up on Tuesday to change everything, while laughing in your face.
Of course, I suspect that physics in general is on to something when one can calculate the electron's gyromagnetic ratio to umpteen places, and agree with a variety of different experiments to measure it. I also know that new particle physics, no matter how perfect, is extremely unlikely to change how chemists, for instance, work. The SSC never got built, for instance, but things don't seem to have ground to a halt in other fields. Most scientists never missed it. Only particle physicists and their groupies really worry about not having found the Higgs boson yet.
My point is only that that "side effect" is actually only a symptom of a similar "side effect" on your electric bill. So it makes you feel warm, cozy, and AMD-loving, but leaves a hole in your wallet, whether you realize it or not.
If you had a CPU that weren't so power-hungry, you could put on a sweater to keep warm, and keep your electric bill low.
Corning spun off their Pyrex cookware and glass dishes, etc., to form WorldKitchen, Inc.. So they don't know as much about glass cookware as they know about optical fiber.
Because you want to heat the air (or, actually your body) to keep you warm, not heat the floor under the CPU. Sure it comes out as heat, but if it gets conducted away through the floor, you don't feel it. An electric heater radiates a substantial portion in the infrared, so that you can feel it, and has a wide open grille to circulate room air through it. Its design is not constrained by EMI requirements to keep apertures minimized.
Conduction through air is pretty lousy. That CPU can get pretty hot, and heat up the air nearby without heating up the air near you very much at all (the temperature gradient is steep). You need good *convection* to stir that hot air around to warm up the room as a whole, or *radiation* to pass through the air and warm the objects in the room.
I don't understand this position. Do you pay your electric bill? Your heating bill? Electric heat is generally the most expensive way to heat your living space, and using a CPU instead of an electric heater has to be less efficient still.
I understand that it is probably tongue-in-cheek, but still, you are paying the electric bill for that heat, and that cost is significant if the heat is enough to warm your room.
Yes, I'm aware that electronic media are cheap and easy to reproduce. So are most pharmaceutical products. However, what isn't cheap and easy to reproduce is the R&D that go into these products.
If the price is forced to zero to match the marginal cost of the media, there isn't any way to recoup these costs. Most people don't like putting their capital where it is guaranteed to get zero return, so guess what, few people do serious development work on either of these things unless they can use intellectual property rights to monopolize the end result, and get to charge something more like the average cost.
For instance, exactly how do you envision people developing (not just manufacturing) AIDS drugs if they aren't allowed exclusive patent rights? Or is it better to not have AIDS drugs because they don't reflect "current economic reality"?
That open-source developers (or their regular employers) often give away their scarce development talent gratis does not prove that their development talent isn't scarce, that there is anything "artificial" about this scarcity, or that there is something more "natural" about giving stuff away rather than building a profitable business.
"Artificial scarcity" is rarely an accurate description. It usually means one is mis-identifying the actual scarce resource. For instance, ignoring important and risky capital investments or research expenditures.
I should not have been so glib with this comment. Given the large number of chips produced for these lower cost systems, the increased engineering cost might, when amortized over the product lifetime, be relatively cheap.
However, I think there is still an important point here, which is that future generations of microprocessors are unlikely to have dramatically higher volumes than today's processors, and therefore, increased engineering costs are going to be difficult to offset by increasing the amount of product shipped.
Slashdot Mirror
Wasn't WordPerfect written in x86 assembly? I'm almost totally ignorant about Linux internals, but I suspect that
1) a program that assumes it is practically running on bare PC metal isn't going to work very well in a UNIX environment.
2) not too many people would be willing to go through x86 assembly code to figure out how to make it work with curses, e.g., and hook up to a real non-DOS filesystem.
WP's code might be wonderfully modular and easy to port, but any company that really was concerned about those metrics wouldn't have chosen assembly as their platform of choice, would they?
It's true that internet cafes are at risk from the authorities if they don't police the usage of chat rooms. I think the entrepreneurial impulse in urban China is going to find a way to overcome the political obstacles, perhaps through filtering technology, or more likely by greasing a few palms. Capitalists can officially join the Party now, you know. Still, my main point was to counter the view, which seems to be prevalent in this discussion, that internet users must have their own dial-up connection.
Internet cafes allow people to get on-line for pocket change; they don't need to foot the bill for a PC and phone line.
And in all of those countries, English is the primary second language. Widespread English competence is a huge marketing point for Singapore in attracting international business.
"From Taiwan-based company" != "manufactured/assembled in Taiwan"
The trend in Taiwan, as elsewhere, is to outsource more and more of the lower-tech work to the mainland. For those who are worried about the PRC's growing economic power, this is a major theme. For those who are worried about the ROC being able to maintain an independent existence from the PRC, this "hollowing out" of Taiwan industry is a serious concern.
For those who like to buy $400 PCs, it is a great thing.
Two words: internet cafes.
But I would bet that most people who would use Hindi on the web are comfortable with English. In a country with hundreds of miscellaneous languages, with a common second language that happens to be the default world language, I don't think there is much incentive to use Hindi. Except to make a political point.
I don't think there is any practical way to expect that an open-source car design would be able to include sufficient information to allow one to manufacture the automobile profitably, if at all. There is too much process and industrial engineering expertise involved in bringing an automobile to production, not to mention the real expense of tooling, and so forth.
The design of modern automobiles is inseparable from the design of automotive assembly processes. Even GM doesn't make Corvettes all by themselves. A huge number of the components are procured from outside vendors on the basis of specifications, where each vendor has a large body of specialized experience in supplying parts to auto manufacturers. The whole flow of these products to the final assembly line is also a monstrous thing to organize.
Consider how difficult it can be to figure out why software doesn't compile when the usual "./configure; make; make install" fails. There is no analogous way to order that an automobile be produced; the problems when a car doesn't assemble itself are many orders of magnitude greater.
Sure, people can build their own cars in their garage; these cars are far different from the cars that GM makes, in terms of design sophistication, reliability, performance, availability of options, final cost to the consumer, and the ability for the average consumer to be able to drive it around and get it fixed when it stops working.
The point of a system of laws is to prohibit people from doing things which are technologically possible but have undesirable effects. When you say "no loss there" you miss the point. The loss is not that pre-existing information gets disseminated too widely, but rather that fewer people will put forth the effort to make useful information available at all if they cannot in someway be compensated for their effort.
Do you really want every viable information medium to have the signal-to-noise ratio of Slashdot?
Now I am sure you are completely wrong. A correct analysis of the problem shows that the travelling twin ages less than the stay-at-home twin. The Lorentz transformation does not depend on the sign of v; travelling toward Earth or away from Earth makes the clock move equally slowly on the ship, as observed from Earth.
The reason the situation is not symmetric is that the twin on the spaceship *turns around*, i.e., undergoes acceleration, and thereby changes inertial reference frames between the outbound and inbound journey. Accelerations are detectable; anything that isn't strapped down in the spaceship gets thrown around by the acceleration. Meanwhile, on Earth, nothing gets thrown around. To the extent that the Earth is not accelerating (i.e. to high accuracy), it remains in the same inertial frame throughout. This is different from the situation on the ship. The symmetry between the twins is an illusion, and is the source of confusion that leads to people calling it a paradox.
To repeat: there is NO TIME CONTRACTION ON THE RETURN JOURNEY. Please learn relativity correctly before posting again. Thank you.
I just have to say, for the benefit of anyone reading, that you are either completely wrong, or confused and irrelevant. The departure and arrival events in hypothetical FTL travel are separated by a spacelike interval, and cannot be causally connected, because causes have to come before effects. This is *including* any correction due to the delay of light travelling from the point of the event to the point of observation.
The speed of light in vacuum is c, regardless of the speed of the emitter or receiver. Your sound comment is a red herring. There is a distinguished frame of reference for sound, in which the medium of sound propagation is fixed. The speed of sound is not the same for observers moving with respect to the medium.
Also, in the twin "paradox," for physically realizable velocities less than c, time is dilated (as observed by the stay-at-home twin) in both directions for the travelling twin. Moving clocks run slow. They never run fast. *That* is the basis of the twin "paradox" which is, apart from an unfortunate name, is not a paraodox at all. I don't understand what you mean by "time-contraction." Yes, signals from the travelling twin to the stay-at-home twin are Doppler shifted, as are the signals in the opposite direction. However, neither observes the clock running faster in the other frame, after correcting for the signal transit time. That correction is assumed in any use of the word "observe" in relativity. Travelling twins age less than the stay-at-home twin. Re-read your relativity notes.
The real reason to believe that photons have zero rest mass is because of the inverse square law in electromagnetics.
The general idea is that interactions mediated by fields like electromagnetic fields (or weak nuclear forces) have a range which is determined by the mass of the force-carrying boson. The bosons mediating the inter-nuclear forces have mass, and are short-range (they have an exponential tail, which falls to zero quickly). But massless force-carrying particles should have inverse square behavior. (I.e. gravity and electromagnetism, where gravity is transmitted, in this kind of physical theory, by hypothetical "gravitons" analogous to the "photons" that transmit electromagnetic forces.)
The experiments to verify inverse-square behavior take the form of verifying Coulomb's law, which in turn takes the form of verifying that the free charge in a conductor resides on the surface. Read Jackson's _Classical Electrodynamics_ for some more description. (Chapter 0 or 1, I forget.)
Now, you might not be convinced: "but what if the connection between mass and force-coupling is not correct. The logic chain breaks down." Sure, but if that logic chain breaks down, then we don't even know what the word "photon" is supposed to refer to, much less what it would mean for such a "non-gauge-field photon" to have mass or not.
OK, thanks for the explanation, although the snide remark is unnecessary; I've gotten through plenty of classes without condescending to those others who have missed some point.
Still, the side point stands, which is that if asc(chr(x)) is not an identity, then the API has not provided the right tools. I.e., anybody actually using chr() or asc() is setting themselves up for a bug in Chinese Windows. Why have them around at all, except as a trap to the unwary? He claims this was a bug in his code; actually, it's probably a lurking bug in thousands of pieces of code. Given that APIs are, in principle, designed to enable third-party software, and not the other way around, it ought to be classified as a bug in Windows. Instead, thousands get it wrong, and a few persistent people get to lose sleep in order to figure out the mystical incantation that does the right thing.
I don't get your example.
If you expected asc(chr(x)) to be an identity operation, why didn't you just use x? If some conversion was necessary to satisfy the platform's API, why doesn't the platform provide a conversion that is guaranteed to work on all the frigging versions that are out there?
What is the point of a widespread platform if you potentially have to special case for every known variation?
If Windows hasn't gotten beyond this kind of inane kludginess, it's no wonder things are such a mess.
Nice troll. Anyhow, those "truths" are just your axioms, which you can pick or choose.
Quick: is the Axiom of Choice "true" or "false"?
Have a nice day.
By the way, it sounds like you've been studying too hard. Have a beer.
The signature of your metric is wrong. The time and space components should have opposite signs.
The reason this is forbidden is because of relativity of simultaneity. I don't know why this part of relativity is less well-remembered than the relativistic length-contraction and time-dilation, but it is really the key to almost all relativistic "paradoxes."
In any FTL travel, there are two events, A: leave the origin; B: arrive at the destination. FTL travel is believed to be impossible because observers in different inertial frames of reference would disagree about whether A or B happened first! Since it is paradoxical to arrive before you have left, the events cannot be causally connected.
Yeah, given the proper axioms, one can prove that 1+1=2, etc. But, in some sense, 1+1=2 is just re-stating the axioms that you felt it was convenient to assume for what you wished to call "sets" for instance.
There isn't any real way to be certain that those axioms do or do not apply to anything realistic. Sure, in concrete terms, I'm pretty sure I can count piles of apples and have the results make sense, but that is really just a gut feeling, and not provable.
Who knows if the universe obeys any logical rules at all? No one can. There's no way to know that God won't show up on Tuesday to change everything, while laughing in your face.
Of course, I suspect that physics in general is on to something when one can calculate the electron's gyromagnetic ratio to umpteen places, and agree with a variety of different experiments to measure it. I also know that new particle physics, no matter how perfect, is extremely unlikely to change how chemists, for instance, work. The SSC never got built, for instance, but things don't seem to have ground to a halt in other fields. Most scientists never missed it. Only particle physicists and their groupies really worry about not having found the Higgs boson yet.
My point is only that that "side effect" is actually only a symptom of a similar "side effect" on your electric bill. So it makes you feel warm, cozy, and AMD-loving, but leaves a hole in your wallet, whether you realize it or not.
If you had a CPU that weren't so power-hungry, you could put on a sweater to keep warm, and keep your electric bill low.
Corning spun off their Pyrex cookware and glass dishes, etc., to form WorldKitchen, Inc.. So they don't know as much about glass cookware as they know about optical fiber.
Because you want to heat the air (or, actually your body) to keep you warm, not heat the floor under the CPU. Sure it comes out as heat, but if it gets conducted away through the floor, you don't feel it. An electric heater radiates a substantial portion in the infrared, so that you can feel it, and has a wide open grille to circulate room air through it. Its design is not constrained by EMI requirements to keep apertures minimized.
Conduction through air is pretty lousy. That CPU can get pretty hot, and heat up the air nearby without heating up the air near you very much at all (the temperature gradient is steep). You need good *convection* to stir that hot air around to warm up the room as a whole, or *radiation* to pass through the air and warm the objects in the room.
I don't understand this position. Do you pay your electric bill? Your heating bill? Electric heat is generally the most expensive way to heat your living space, and using a CPU instead of an electric heater has to be less efficient still.
I understand that it is probably tongue-in-cheek, but still, you are paying the electric bill for that heat, and that cost is significant if the heat is enough to warm your room.
Yes, I'm aware that electronic media are cheap and easy to reproduce. So are most pharmaceutical products. However, what isn't cheap and easy to reproduce is the R&D that go into these products.
If the price is forced to zero to match the marginal cost of the media, there isn't any way to recoup these costs. Most people don't like putting their capital where it is guaranteed to get zero return, so guess what, few people do serious development work on either of these things unless they can use intellectual property rights to monopolize the end result, and get to charge something more like the average cost.
For instance, exactly how do you envision people developing (not just manufacturing) AIDS drugs if they aren't allowed exclusive patent rights? Or is it better to not have AIDS drugs because they don't reflect "current economic reality"?
That open-source developers (or their regular employers) often give away their scarce development talent gratis does not prove that their development talent isn't scarce, that there is anything "artificial" about this scarcity, or that there is something more "natural" about giving stuff away rather than building a profitable business.
I'm not sure this is exactly what you are talking about but you can buy
six-foot long flexible drill bits or
six-foot long flexible extensions to use with drill bits you already have.
"Artificial scarcity" is rarely an accurate description. It usually means one is mis-identifying the actual scarce resource. For instance, ignoring important and risky capital investments or research expenditures.
I should not have been so glib with this comment. Given the large number of chips produced for these lower cost systems, the increased engineering cost might, when amortized over the product lifetime, be relatively cheap.
However, I think there is still an important point here, which is that future generations of microprocessors are unlikely to have dramatically higher volumes than today's processors, and therefore, increased engineering costs are going to be difficult to offset by increasing the amount of product shipped.