Yes, but if you add, say, 1.0e0 and 1.0e-18, the result may get rounded down to 1.0e0 (no change) as the real result, 1.000000000000000001, is not distinctly representable in the number of bits available. To avoid losing precision all the bits after the leading '1' (which is assumed) have to fit within the mantissa, which is of finite size. This applies regardless of the exponent, except for the special cases of zero, NaN, and infinity.
For a problem like this one tends to be better off using fixed-point notation, with a word size large enough to represent both the smallest increment and the largest magnitude you expect to work with. That way you don't run into the case where a small increment may or may not get rounded off depending on the magnitude of the other operand.
In this case, however, any directions from the flight crew already have the force of law. If the airline simply made it a policy not to allow cell phone use it would be just as legally binding as an act of Congress, while retaining far greater flexibility -- for example, the airline could separate the cell phone users into their own section so as not to bother the rest of the passengers, as suggested in the summary.
This is similar to the concept of preferring municipal or state laws over federal ones for local issues. A ruling body closer to the problem will tend to come up with a better solution, and if the decision turns out to be a poor one in retrospect it only affects that one area (city, state, or, in this case, airline) instead of everyone.
... Internet connectivity isn't a luxury, it's a necessity for many people...
1) Even if these people's lives were to depend on immediate, unlimited 3G wireless Internet access, I would still argue that they cannot force the providers to accept their terms. They still need the provider's permission to use the network. The provider owns the infrastructure, and thus has every right to set the conditions for its use, regardless of what anyone else wants or "needs".
2) Internet access -- by any means, much less 3G wireless data plans specifically -- never has been, is not, and never will be anything but a luxury.
Introducing, the $15 laptop:... Note the 5-line backlit display...
I said "laptop-style display". A 5x12 text-only LCD doesn't count. This is (one reason) why the site you linked to called it a "personal organizer" rather than a laptop.
Assembler or C on an 8-bit CPU is no more difficult than any other chip. What's confusing you is the fact that there's no IDEs or libraries to make it quick to put a high-level program together.
That's not "confusing" me, it's exactly what I was referring to. My job involves low-level system programming in C and PowerPC assembly, and embedded CPU design (on FPGAs) is something of a hobby of mine; I'm well aware of the requirements. Programming an embedded system with limited resources and no IDEs, standard libraries, or interactive debugging is a far more demanding task than normal PC application programming, and consequently draws on a much smaller (and thus more expensive) supply of skilled labor.
A pointing device is completely unnecessary. It's used with desktops only because better input methods never caught on.
Pointing devices are an essential part of any modern, general-purpose computer. For symbolic input a keyboard is sufficient, even preferable, but some tasks demand an analogue input device. Could you imaging anyone choosing to enter screen coordinates via a keyboard to operate a graphics package, for example? Or even CAD? The mouse was invented for a reason, and not just because users don't properly "appreciate" the power of the keyboard. However, I am willing to grant that analogue input is likely to be less useful for a UMPC than for a normal desktop system, and may be omitted in this context.
As for network support, I was referring to the standard 802.11 protocols (wired and wireless) for compatibility with existing systems, but if you don't mind breaking compatibility please at least use one of the numerous existing point-to-point systems (such as this 2.4GHz Nordic module) rather than inflicting yet another incompatible wireless standard on an unsuspecting world...
What if you were unfortunate enough to live in an area where AT&T were the ONLY operators?
Irrelevant. If you want to use their network you need their permission, which means agreeing to their terms.
What if you have an iPhone?
Then you agreed to the terms when you purchased the device. Didn't you read your service contract before agreeing to it? I'm sure you had every opportunity to do so.
What if you've already signed up to their UNLIMITED package and just started a 12-month contract only to find it's not quite so Unlimited?
Again, did you bother to read the contract? The only case you might be able to make here is for misleading advertising -- of the "bait & switch" variety -- but even then it should have been obvious what you were getting into at some point before you signed up. If a contract is "too complicated" then don't sign it; that's the only way to limit both your liability and the general proliferation of long and complex contracts. By signing it you're claiming to understand what the contract entails.
... you could get a $10 laptop, if you really were willing to restrict it to all 8-bit software and the inherent limitations.
So you've got a $1 CPU. Where are you going to get a laptop-sized (readable) flat screen, plus the control module(s) and even a minimal GPU, for less than the remaining $9? Not to mention a keyboard, pointing device, wired or (preferably) wireless NIC, case, and assorted essential glue hardware?
$100, sure. Maybe even $50, given bulk pricing. But $10? As you yourself pointed out, an electronic thermostat costs more than that, and it doesn't even have a laptop-style display or a QWERTY keyboard.
Software would be an issue as well. It requires a much more specialized skill set to code anything worthwhile for a 50MHz 8-bit CPU than for even a low-end PowerPC or XScale processor. You'd probably be much better off spending a few extra dollars per unit, if only for the sake of encouraging more third-party developers to write software for your platform. Even a free laptop is little more than a paperweight without decent applications.
My guess would be that Firefox gives you the option to trust the signed certificate but not the signing certificate.
If you're referring to the dialog (or pseudo-page) that appears when the certificate is first encountered, that is correct. One can always go into the preferences and add the signing certificate as a trusted root certificate authority, however. The process isn't automated due to the obvious security implications, but it is an option nonetheless.
OK, although that isn't technically a problem with KDE itself, but rather with the FreeType rendering library. Still, it's nice to know that one can have both clear fonts and (mostly) accurate spacing.
Would you happen to know why KDE and GNOME behave differently in that regard, given that they use the same font rendering engine? Or is it as I said, that the reason GNOME's fonts look better to the GP is that GNOME doesn't enable sub-pixel positioning by default?
It's not "a lack of professionalism and attention to detail", it's a consequence of aligning edges with the nearest pixel. You can either have blurry fonts with precise spacing or clear fonts with the sort of sub-pixel spacing variations you describe. You prefer the blurry fonts; that's fine. You can configure KDE to render fonts that way as well. Similarly, GNOME can be configured to perform the subpixel alignment demonstrated in your screenshot.
KDE and GNOME use the same libraries for font rendering; the differences in appearance are purely a matter of preference, not capability.
But the point is to communicate information, which is abstract. The physical changes, as I said, are secondary, having no value beyond that of the abstract concepts they represent. Mental processes often involve physical changes -- it would be hard to imagine one that didn't involve at least neurons firing in the brain, and the creation of patterns of graphite on paper is also very common -- but these changes are not the purpose of the process.
In the same way the purpose of the algorithm covered by a software patent is not a particular pattern of lights on a screen, dots of toner on paper, etc., but rather computation and communication of abstract concepts. One can patent an abacus, but not the mathematical formula x+y, nor the process of evaluating it. It makes no difference whether you evaluate the formula in your head or on a device. A software patent covers an abstract algorithm, in essence a mathematical formula, and it should make no difference whether you evaluate that formula in your head or on a computer.
And there are lots of worldviews where killing innocents is not considered wrong, but tell me is it their's or my point of view thats [sic] skewed.
It's still subjective. That view is incompatible with yours or mine, of course, but to call it "skewed" implies the existence of some objective standard against which it can be measured, when no such standard exists.
What is objective, however, is that one who holds such a view cannot argue that it is wrong for others to kill him or her in response to the killing of an innocent. An action can either be right or wrong; if one claims that an action was right then one cannot subsequently claim that it is wrong for others to do the same; if one claims that the action was wrong one is admitting that punishment is justified. Ultimately it doesn't matter what the murderer's worldview is, as the maximum possible response is justified either way.
For lesser offenses the response is necessarily limited by the nature of the offense; for copyright infringement, for example, a reciprocal response might be to invalidate any copyright claims of the infringer.
Given a general-purpose robot, everything can be reduced to "a pure mental process -- automated by means of a general-purpose computer".
That's not true. I even pointed out the different between mental processes and chemical processes, but to elaborate: mental processes deal with abstract concepts; any physical changes, e.g. electricity running through circuits, are purely secondary effects valued only to the extent that they represent the abstract concepts. The numbers and/or symbols they represent are the real inputs and outputs of the process. Production processes, whether chemical, manufacturing, or your "general-purpose robot", may follow an algorithm but are designed to effect some specific change in the physical world.
The USPTO already has a policy of rejecting applications involving only abstract concepts; this distinction already exists, it just isn't always enforced. Why should the same process, still involving only abstract concepts but "with a computer!!!", be any different?
The real issue, which most people avoid addressing, is that there is no practical distinction between software patents, chemical process patents, or machinery patents as a necessary consequence of basic theory.... As every computer geek should know, there is no theoretical distinction between the machine, the program, and the data.
There is a practical distinction, though. Sure, in the context of a running program, machine, program and data are all different aspects of the same unit. In the context of a patent application, however, they are separated by a number of factors, not least of which is the fact that the application itself does not include any claims relating to the machine or data, just the program. It cannot include such claims, because the machine was invented by others, and a patent limited to specific input data would be extremely limited in scope.
What we're left with is an algorithm -- not a method of production, as in a chemical process patent, but a pure mental process -- automated by means of a general-purpose computer. How does an existing device, used as designed, plus an unpatentable mental process possibly qualify for a patent?
The fact that you built a new widget using tools designed for the purpose of building widgets does not make the widget unpatentable.
Sure, but in that case what you're building is itself a novel device independent from the tools used to build it. In the case of software the "device" you're "building" is essentially a mental process (an algorithm) automated by a machine. The machine is already patented, and mental processes are not considered patentable. How, then, can the combination of a machine (used as intended) and a mental process be separately patentable?
Chemical process patents may indeed be algorithms, as you say, but they aren't mental processes -- the point of the algorithm is to produce something, not to perform abstract calculations.
Here's an idea: limit the number of patents granted to a small number per year. At the end of the year only the 50-100 most novel and useful applications are granted; everyone else is welcome to try again in a year. Why reward mediocrity, especially at such a high cost to society in general? I imagine this would considerably reduce the proportion of software and business method patents as well.
What the GP did is definitely copyright infringement and illegal in the U.S. and a number of other countries. That is an objective fact. Whether it is wrong or not is subjective, because the basic principles of morality are themselves subjective. There exist numerous worldviews by which copyright infringement is not considered wrong.
The thing about winmodems is that reverse engineering isn't enough. It's not a hardware interface problem.
A winmodem is basically just a basic sound card tied to a phone line, with the driver generating the tones in software. The protocol itself is well-known, but certain parts essential to at least the higher baud rates are patented, and thus can't be included in an open-source driver. With a hardware modem the driver doesn't need to know anything about the details of the protocol, just the standard AT command set. For bonus points, a hardware modem doesn't drop the connection whenever the computer has to swap to disk or runs out of spare CPU cycles.
Some manufacturers supply their own drivers, but the ones I've seen are all binary-only, don't work in SMP configurations (Core 2 Duo? Not compatible...), and tend to be limited to 14.4 kbps or lower. That's when you can get them to work at all.
My personal favorite silly DRM law is the one which sets out massive penalties for circumventing a DRM mechanism - making anyone who holds the shift key while loading a CD into Windoze box a felon.
Technically it also makes the act of telling people that holding down the Shift key will bypass DRM a felony. Of course, the Powers that Be may find it a bit difficult to locate a court willing to convict given the obvious 1st Amendment implications; but then again, maybe not. The courts are a branch of the government, after all, and thus ultimately on the same side most of the time.
In which case, there is no benefit to a gold-backed standard from a control perspective.
No, because the scope of that control is also relevant. Short of an obvious act of aggression, such as confiscating privately-owned gold or banning it outright, the maximum amount any entity can interfere in a commodity-gold-based market is limited, on the one hand, by the amount of gold it already possesses, and on the other by the amount it can acquire in trade for services or existing non-gold assets. That's a lot, granted, but actually turning to any of those sources is very costly; the more gold employed to manipulate the market, the less there is for future market operations. Increasing taxes to make up the difference burns significant political capital. In a system based on fiat currency, on the other hand, there are no limits; the government has an effectively infinite supply of paper money which it can use at minimal cost, financial or political. If they feel, allowing for the inevitable political repercussions, that it's in their best interests to make Federal Reserve Notes worthless overnight there's nothing to prevent them from doing just that.
Fiat currency a bit like democracy. It's the worst system out there, except for all the others.
That was a mistake. You see, I'm an libertarian; some would say an "anarcho-capitalist". To me, individual sovereignty and personal responsibility beat tyranny of the majority any day. The thing about that quote--and it's a good one, I like it a lot--is that it implies that the only alternative better than democracy is no system at all. No government; no governing; no rulers. Individual sovereignty does not equate to lawlessness or chaos; it simply means that each individual is responsible for his or her own actions, without any exceptions due to birth, wealth, status, strength, popularity, etc.
Democracy removes an obvious conflict present in other systems by (presumably) making the rulers representatives of the majority being governed. That makes it more stable, since any resistance with sufficient support to take over essentially already has taken over via popular elections. It's no more just than any other system, however, and tends to become far more intrusive into everyday life over time than most societies would tolerate in a non-democratic government.
As a commodity, the supply of gold via mining is self-regulating. If the price of gold is higher than the cost of mining it (per unit mass) then mining will continue until the price decreases--or the cost increases--such that mining is no longer profitable. In the long run, mining gold is no more or less profitable than any other enterprise. The "commodity producers" thus don't bother me; under normal circumstances they have no incentive, or even ability, to flood the market with newly-mined gold.
As for the commodities market, anyone can play at that. If you feel you're overextended on cash -- which is what gold would be under a gold commodity currency -- then trade for some other commodity to distribute your risk. In any event, under a gold currency there wouldn't be a significant gold commodity market -- you'd just be trading gold for gold, which is kind of pointless. The only fluctuations would be in the value of refining and/or minting, not the gold itself.
The only groups that have traditionally been involved in manipulating the supply of gold are governments and their close allies, national central banks. Their actions with respect to commodity currencies are at least limited to the amount of such currencies they actually possess; these are, after all, the same groups that have complete control over the supply of fiat currency. Your options for controlling intervention in the money supply are exactly the same under a commodity currency as under a fiat one.
If you bothered to read the article you linked to, the crash was attributed to excessive, risky loans to railroad developers (a government-backed project, BTW) coupled with insufficient reserves in the banks to cover their depositor's accounts. It had nothing to do with the choice of currency.
The only difference today is that, instead of a few depositors losing most of their accounts, the FDIC has the Treasury print up however much new currency is required to cover the bank's lack of reserves, devaluing everyone's money by the same total amount and distributing the loss across the entire economy. The risk of investing in a low-reserve bank is thus paid by everyone, not just those who chose to accept the risk and/or failed to perform due diligence before investing.
If you can get 20 cheeseburgers for a pound of gold in Detroit, then you should also be able to get 20 cheeseburgers in Guatemala (...).
That's completely wrong. It assumes the price of a good (cheeseburgers here) is necessarily the same in different countries, when the price can differ depending on the local supply, cost of shipping (higher for perishables!), cost of labor to put it together, etc.
As for the value of money itself... any constant amount of money is sufficient to satisfy the demand, provided it's suitably divisible (as both gold and fiat currency are). The breaking point comes in an inflationary system when people come to expect ever-increasing rates of inflation. At that point hyperinflation sets in and the currency becomes worthless.
random hits to the valuation of a currency due to influx of more resources
Sure, if by "random" you mean "not controlled by the Federal Reserve". The U.S. dollar has lost over 95% of its value since the Fed was created and a deliberate policy of continual inflation instituted. The rate of inflation has itself varied quite a bit from year to year as well. Compared to the post-1913 dollar gold commodity currencies are an absolute paragon of both short- and long-term stability, with "gold standard" systems somewhere in between. The limited periods of disruption in that stability can all be directly traced to either natural disasters or man-made ones, typically in the form of government intervention and/or massive fraud by banks.
The main reason precious metals don't make sense as money is the fact that they don't account for the growth of the economy.... What happens when there is no more gold left to pile up in Fort Knox? Does the economy stop growing at that instant?
This has got to be the most oft-repeated argument against a commodity gold currency, and yet it's plainly inconsistent with even that most basic of economic principles, supply and demand. The supply of gold currency is the entire amount available for use in trade. The demand for currency (including gold) is the total value of everything offered in exchange for currency, i.e. everything currency can buy. Ergo, the price of currency varies dynamically such that the value of the total supply of currency tends toward the total value of everything currency can buy.
If the economy grows, and the amount of gold stays fixed, then the value of a given quantity of gold -- what you can buy with it -- increases. There's no risk of ever "running out of gold". A limited supply of gold does not limit economic growth in any way. The same applies to paper currency: printing more paper money does not create wealth, or allow or encourage wealth to be created.
If you're going to argue against using gold as money, at least use an argument that makes sense. I'm not personally a fanatic about using gold in everyday transactions, but at the same time I have significant misgivings about using a currency subject to someone else's manipulation. Currency fluctuations carry meaning, particularly in the financial markets (e.g. interest rates), and an actively managed currency sends all the wrong signals.
The argument that inflation keeps people from "hoarding" is rather instructive. Why would anyone willingly choose a currency that can and will be trivially manipulated by others, for the express purpose of driving them to choose riskier behaviors they would rather avoid? A commodity currency at least raises a barrier against that kind of manipulation.
A manipulated currency, moreover, is worse than useless for financial planning, accounting, economic forecasting, etc. A free currency has understandable variations due to known economic conditions; a manipulated currency is necessarily chaotic and unpredictable. If people could predict the changes the manipulation would have no effect.
The reason people abandoned the gold standard was because of two things:
No, the reason people abandoned the gold standard was primarily that the government confiscated all the privately-owned gold and made it illegal to own gold except as a collector's item. Having no other viable choice, people switched to using unbacked paper currency. It is no longer illegal -- technically -- to own gold or use it in trade, but (a) changing the primary currency of an entire society is a monumental task, which only happened before due to a monumental threat of force; (b) having been disrupted for so long, gold currency lacks marketability in small quantities (very few sellers accept it directly); and (c) the government keeps doing things like this from time to time, making the perfectly legal ownership and use of gold currency a nonetheless risky proposition.
It took centuries, millenia, for gold to become established as a near-universal currency. It's far too premature to say that it's been "abandoned" when it's only been quasi-legal here for a few decades.
Slashdot Mirror
Yes, but if you add, say, 1.0e0 and 1.0e-18, the result may get rounded down to 1.0e0 (no change) as the real result, 1.000000000000000001, is not distinctly representable in the number of bits available. To avoid losing precision all the bits after the leading '1' (which is assumed) have to fit within the mantissa, which is of finite size. This applies regardless of the exponent, except for the special cases of zero, NaN, and infinity.
For a problem like this one tends to be better off using fixed-point notation, with a word size large enough to represent both the smallest increment and the largest magnitude you expect to work with. That way you don't run into the case where a small increment may or may not get rounded off depending on the magnitude of the other operand.
In this case, however, any directions from the flight crew already have the force of law. If the airline simply made it a policy not to allow cell phone use it would be just as legally binding as an act of Congress, while retaining far greater flexibility -- for example, the airline could separate the cell phone users into their own section so as not to bother the rest of the passengers, as suggested in the summary.
This is similar to the concept of preferring municipal or state laws over federal ones for local issues. A ruling body closer to the problem will tend to come up with a better solution, and if the decision turns out to be a poor one in retrospect it only affects that one area (city, state, or, in this case, airline) instead of everyone.
1) Even if these people's lives were to depend on immediate, unlimited 3G wireless Internet access, I would still argue that they cannot force the providers to accept their terms. They still need the provider's permission to use the network. The provider owns the infrastructure, and thus has every right to set the conditions for its use, regardless of what anyone else wants or "needs".
2) Internet access -- by any means, much less 3G wireless data plans specifically -- never has been, is not, and never will be anything but a luxury.
I said "laptop-style display". A 5x12 text-only LCD doesn't count. This is (one reason) why the site you linked to called it a "personal organizer" rather than a laptop.
That's not "confusing" me, it's exactly what I was referring to. My job involves low-level system programming in C and PowerPC assembly, and embedded CPU design (on FPGAs) is something of a hobby of mine; I'm well aware of the requirements. Programming an embedded system with limited resources and no IDEs, standard libraries, or interactive debugging is a far more demanding task than normal PC application programming, and consequently draws on a much smaller (and thus more expensive) supply of skilled labor.
Pointing devices are an essential part of any modern, general-purpose computer. For symbolic input a keyboard is sufficient, even preferable, but some tasks demand an analogue input device. Could you imaging anyone choosing to enter screen coordinates via a keyboard to operate a graphics package, for example? Or even CAD? The mouse was invented for a reason, and not just because users don't properly "appreciate" the power of the keyboard. However, I am willing to grant that analogue input is likely to be less useful for a UMPC than for a normal desktop system, and may be omitted in this context.
As for network support, I was referring to the standard 802.11 protocols (wired and wireless) for compatibility with existing systems, but if you don't mind breaking compatibility please at least use one of the numerous existing point-to-point systems (such as this 2.4GHz Nordic module) rather than inflicting yet another incompatible wireless standard on an unsuspecting world...
You'd need a universal legal system, not universal currency, to accomplish that goal.
Irrelevant. If you want to use their network you need their permission, which means agreeing to their terms.
Then you agreed to the terms when you purchased the device. Didn't you read your service contract before agreeing to it? I'm sure you had every opportunity to do so.
Again, did you bother to read the contract? The only case you might be able to make here is for misleading advertising -- of the "bait & switch" variety -- but even then it should have been obvious what you were getting into at some point before you signed up. If a contract is "too complicated" then don't sign it; that's the only way to limit both your liability and the general proliferation of long and complex contracts. By signing it you're claiming to understand what the contract entails.
So you've got a $1 CPU. Where are you going to get a laptop-sized (readable) flat screen, plus the control module(s) and even a minimal GPU, for less than the remaining $9? Not to mention a keyboard, pointing device, wired or (preferably) wireless NIC, case, and assorted essential glue hardware?
$100, sure. Maybe even $50, given bulk pricing. But $10? As you yourself pointed out, an electronic thermostat costs more than that, and it doesn't even have a laptop-style display or a QWERTY keyboard.
Software would be an issue as well. It requires a much more specialized skill set to code anything worthwhile for a 50MHz 8-bit CPU than for even a low-end PowerPC or XScale processor. You'd probably be much better off spending a few extra dollars per unit, if only for the sake of encouraging more third-party developers to write software for your platform. Even a free laptop is little more than a paperweight without decent applications.
If you're referring to the dialog (or pseudo-page) that appears when the certificate is first encountered, that is correct. One can always go into the preferences and add the signing certificate as a trusted root certificate authority, however. The process isn't automated due to the obvious security implications, but it is an option nonetheless.
OK, although that isn't technically a problem with KDE itself, but rather with the FreeType rendering library. Still, it's nice to know that one can have both clear fonts and (mostly) accurate spacing.
Would you happen to know why KDE and GNOME behave differently in that regard, given that they use the same font rendering engine? Or is it as I said, that the reason GNOME's fonts look better to the GP is that GNOME doesn't enable sub-pixel positioning by default?
It's not "a lack of professionalism and attention to detail", it's a consequence of aligning edges with the nearest pixel. You can either have blurry fonts with precise spacing or clear fonts with the sort of sub-pixel spacing variations you describe. You prefer the blurry fonts; that's fine. You can configure KDE to render fonts that way as well. Similarly, GNOME can be configured to perform the subpixel alignment demonstrated in your screenshot.
KDE and GNOME use the same libraries for font rendering; the differences in appearance are purely a matter of preference, not capability.
But the point is to communicate information, which is abstract. The physical changes, as I said, are secondary, having no value beyond that of the abstract concepts they represent. Mental processes often involve physical changes -- it would be hard to imagine one that didn't involve at least neurons firing in the brain, and the creation of patterns of graphite on paper is also very common -- but these changes are not the purpose of the process.
In the same way the purpose of the algorithm covered by a software patent is not a particular pattern of lights on a screen, dots of toner on paper, etc., but rather computation and communication of abstract concepts. One can patent an abacus, but not the mathematical formula x+y, nor the process of evaluating it. It makes no difference whether you evaluate the formula in your head or on a device. A software patent covers an abstract algorithm, in essence a mathematical formula, and it should make no difference whether you evaluate that formula in your head or on a computer.
It's still subjective. That view is incompatible with yours or mine, of course, but to call it "skewed" implies the existence of some objective standard against which it can be measured, when no such standard exists.
What is objective, however, is that one who holds such a view cannot argue that it is wrong for others to kill him or her in response to the killing of an innocent. An action can either be right or wrong; if one claims that an action was right then one cannot subsequently claim that it is wrong for others to do the same; if one claims that the action was wrong one is admitting that punishment is justified. Ultimately it doesn't matter what the murderer's worldview is, as the maximum possible response is justified either way.
For lesser offenses the response is necessarily limited by the nature of the offense; for copyright infringement, for example, a reciprocal response might be to invalidate any copyright claims of the infringer.
That's not true. I even pointed out the different between mental processes and chemical processes, but to elaborate: mental processes deal with abstract concepts; any physical changes, e.g. electricity running through circuits, are purely secondary effects valued only to the extent that they represent the abstract concepts. The numbers and/or symbols they represent are the real inputs and outputs of the process. Production processes, whether chemical, manufacturing, or your "general-purpose robot", may follow an algorithm but are designed to effect some specific change in the physical world.
The USPTO already has a policy of rejecting applications involving only abstract concepts; this distinction already exists, it just isn't always enforced. Why should the same process, still involving only abstract concepts but "with a computer!!!", be any different?
There is a practical distinction, though. Sure, in the context of a running program, machine, program and data are all different aspects of the same unit. In the context of a patent application, however, they are separated by a number of factors, not least of which is the fact that the application itself does not include any claims relating to the machine or data, just the program. It cannot include such claims, because the machine was invented by others, and a patent limited to specific input data would be extremely limited in scope.
What we're left with is an algorithm -- not a method of production, as in a chemical process patent, but a pure mental process -- automated by means of a general-purpose computer. How does an existing device, used as designed, plus an unpatentable mental process possibly qualify for a patent?
Sure, but in that case what you're building is itself a novel device independent from the tools used to build it. In the case of software the "device" you're "building" is essentially a mental process (an algorithm) automated by a machine. The machine is already patented, and mental processes are not considered patentable. How, then, can the combination of a machine (used as intended) and a mental process be separately patentable?
Chemical process patents may indeed be algorithms, as you say, but they aren't mental processes -- the point of the algorithm is to produce something, not to perform abstract calculations.
Here's an idea: limit the number of patents granted to a small number per year. At the end of the year only the 50-100 most novel and useful applications are granted; everyone else is welcome to try again in a year. Why reward mediocrity, especially at such a high cost to society in general? I imagine this would considerably reduce the proportion of software and business method patents as well.
Life hint: not everyone shares your worldview.
What the GP did is definitely copyright infringement and illegal in the U.S. and a number of other countries. That is an objective fact. Whether it is wrong or not is subjective, because the basic principles of morality are themselves subjective. There exist numerous worldviews by which copyright infringement is not considered wrong.
The thing about winmodems is that reverse engineering isn't enough. It's not a hardware interface problem.
A winmodem is basically just a basic sound card tied to a phone line, with the driver generating the tones in software. The protocol itself is well-known, but certain parts essential to at least the higher baud rates are patented, and thus can't be included in an open-source driver. With a hardware modem the driver doesn't need to know anything about the details of the protocol, just the standard AT command set. For bonus points, a hardware modem doesn't drop the connection whenever the computer has to swap to disk or runs out of spare CPU cycles.
Some manufacturers supply their own drivers, but the ones I've seen are all binary-only, don't work in SMP configurations (Core 2 Duo? Not compatible...), and tend to be limited to 14.4 kbps or lower. That's when you can get them to work at all.
Technically it also makes the act of telling people that holding down the Shift key will bypass DRM a felony. Of course, the Powers that Be may find it a bit difficult to locate a court willing to convict given the obvious 1st Amendment implications; but then again, maybe not. The courts are a branch of the government, after all, and thus ultimately on the same side most of the time.
No, because the scope of that control is also relevant. Short of an obvious act of aggression, such as confiscating privately-owned gold or banning it outright, the maximum amount any entity can interfere in a commodity-gold-based market is limited, on the one hand, by the amount of gold it already possesses, and on the other by the amount it can acquire in trade for services or existing non-gold assets. That's a lot, granted, but actually turning to any of those sources is very costly; the more gold employed to manipulate the market, the less there is for future market operations. Increasing taxes to make up the difference burns significant political capital. In a system based on fiat currency, on the other hand, there are no limits; the government has an effectively infinite supply of paper money which it can use at minimal cost, financial or political. If they feel, allowing for the inevitable political repercussions, that it's in their best interests to make Federal Reserve Notes worthless overnight there's nothing to prevent them from doing just that.
That was a mistake. You see, I'm an libertarian; some would say an "anarcho-capitalist". To me, individual sovereignty and personal responsibility beat tyranny of the majority any day. The thing about that quote--and it's a good one, I like it a lot--is that it implies that the only alternative better than democracy is no system at all. No government; no governing; no rulers. Individual sovereignty does not equate to lawlessness or chaos; it simply means that each individual is responsible for his or her own actions, without any exceptions due to birth, wealth, status, strength, popularity, etc.
Democracy removes an obvious conflict present in other systems by (presumably) making the rulers representatives of the majority being governed. That makes it more stable, since any resistance with sufficient support to take over essentially already has taken over via popular elections. It's no more just than any other system, however, and tends to become far more intrusive into everyday life over time than most societies would tolerate in a non-democratic government.
As a commodity, the supply of gold via mining is self-regulating. If the price of gold is higher than the cost of mining it (per unit mass) then mining will continue until the price decreases--or the cost increases--such that mining is no longer profitable. In the long run, mining gold is no more or less profitable than any other enterprise. The "commodity producers" thus don't bother me; under normal circumstances they have no incentive, or even ability, to flood the market with newly-mined gold.
As for the commodities market, anyone can play at that. If you feel you're overextended on cash -- which is what gold would be under a gold commodity currency -- then trade for some other commodity to distribute your risk. In any event, under a gold currency there wouldn't be a significant gold commodity market -- you'd just be trading gold for gold, which is kind of pointless. The only fluctuations would be in the value of refining and/or minting, not the gold itself.
The only groups that have traditionally been involved in manipulating the supply of gold are governments and their close allies, national central banks. Their actions with respect to commodity currencies are at least limited to the amount of such currencies they actually possess; these are, after all, the same groups that have complete control over the supply of fiat currency. Your options for controlling intervention in the money supply are exactly the same under a commodity currency as under a fiat one.
If you bothered to read the article you linked to, the crash was attributed to excessive, risky loans to railroad developers (a government-backed project, BTW) coupled with insufficient reserves in the banks to cover their depositor's accounts. It had nothing to do with the choice of currency.
The only difference today is that, instead of a few depositors losing most of their accounts, the FDIC has the Treasury print up however much new currency is required to cover the bank's lack of reserves, devaluing everyone's money by the same total amount and distributing the loss across the entire economy. The risk of investing in a low-reserve bank is thus paid by everyone, not just those who chose to accept the risk and/or failed to perform due diligence before investing.
That's completely wrong. It assumes the price of a good (cheeseburgers here) is necessarily the same in different countries, when the price can differ depending on the local supply, cost of shipping (higher for perishables!), cost of labor to put it together, etc.
As for the value of money itself... any constant amount of money is sufficient to satisfy the demand, provided it's suitably divisible (as both gold and fiat currency are). The breaking point comes in an inflationary system when people come to expect ever-increasing rates of inflation. At that point hyperinflation sets in and the currency becomes worthless.
Sure, if by "random" you mean "not controlled by the Federal Reserve". The U.S. dollar has lost over 95% of its value since the Fed was created and a deliberate policy of continual inflation instituted. The rate of inflation has itself varied quite a bit from year to year as well. Compared to the post-1913 dollar gold commodity currencies are an absolute paragon of both short- and long-term stability, with "gold standard" systems somewhere in between. The limited periods of disruption in that stability can all be directly traced to either natural disasters or man-made ones, typically in the form of government intervention and/or massive fraud by banks.
Not a problem.
This has got to be the most oft-repeated argument against a commodity gold currency, and yet it's plainly inconsistent with even that most basic of economic principles, supply and demand. The supply of gold currency is the entire amount available for use in trade. The demand for currency (including gold) is the total value of everything offered in exchange for currency, i.e. everything currency can buy. Ergo, the price of currency varies dynamically such that the value of the total supply of currency tends toward the total value of everything currency can buy.
If the economy grows, and the amount of gold stays fixed, then the value of a given quantity of gold -- what you can buy with it -- increases. There's no risk of ever "running out of gold". A limited supply of gold does not limit economic growth in any way. The same applies to paper currency: printing more paper money does not create wealth, or allow or encourage wealth to be created.
If you're going to argue against using gold as money, at least use an argument that makes sense. I'm not personally a fanatic about using gold in everyday transactions, but at the same time I have significant misgivings about using a currency subject to someone else's manipulation. Currency fluctuations carry meaning, particularly in the financial markets (e.g. interest rates), and an actively managed currency sends all the wrong signals.
The argument that inflation keeps people from "hoarding" is rather instructive. Why would anyone willingly choose a currency that can and will be trivially manipulated by others, for the express purpose of driving them to choose riskier behaviors they would rather avoid? A commodity currency at least raises a barrier against that kind of manipulation.
A manipulated currency, moreover, is worse than useless for financial planning, accounting, economic forecasting, etc. A free currency has understandable variations due to known economic conditions; a manipulated currency is necessarily chaotic and unpredictable. If people could predict the changes the manipulation would have no effect.
No, the reason people abandoned the gold standard was primarily that the government confiscated all the privately-owned gold and made it illegal to own gold except as a collector's item. Having no other viable choice, people switched to using unbacked paper currency. It is no longer illegal -- technically -- to own gold or use it in trade, but (a) changing the primary currency of an entire society is a monumental task, which only happened before due to a monumental threat of force; (b) having been disrupted for so long, gold currency lacks marketability in small quantities (very few sellers accept it directly); and (c) the government keeps doing things like this from time to time, making the perfectly legal ownership and use of gold currency a nonetheless risky proposition.
It took centuries, millenia, for gold to become established as a near-universal currency. It's far too premature to say that it's been "abandoned" when it's only been quasi-legal here for a few decades.