You'd have to do this at the hardware level. Are there any JTAG pads on the board? If not, clipping onto the firmware flash chip with the appropriate tool may be necessary. That, or some means to prevent the existing firmware from loading while loading a substitute into RAM, which will then reload the firmware flash.
First to figure this out might get a little military contract:-)
So... you are saying... software is hard to do, so let's go to the least reliable source for it... ?
Both commercial software (off the shelf... COTS) and open source (also off the shell... FOTS) are full of bugs. At least open source is subject to peer review (in a wider peer space) and gets bugs fixed sooner (there's rarely a coverup of bugs in open source, unlike commercial).
One big problem is that the internal review process, that still has to be done inside the government, will be weaker at this job because the people who would know how best to do that won't be working in the kinds of jobs that would be in a track to the analyst positions that can do these reviews. At least one reason to have an in-house programming team in the government is so that some of them can move up to being top level analysts without being biased in favor of certain commercial interests.
... located at each individual airport. The airport already knows exactly where it is. It can receive the GPS signal and see how far off it is... specifically for that airport. Then it would transmit that correction data in real time over a local UHF frequency that can serve approaching planes out to some distance (perhaps 100km). Nearby airports use different frequencies which get selected when the target airport is selected and GPS indicates they are within range.
They could also spend more money and put up a triangulation based TPS that would allow accurate terrestrial positioning independent of GPS. That would be in addition to final approach guidance systems. That is, of course, if you feel warm and cozy about having extra redundant systems serving the airplane you are on.
Without patents at all, every inventor has equal rights to what he invents. He just doesn't have any protection from those that steal rights from those who don't invent. What a patent really does is, in the name of protecting an inventor's rights from those who steal, it actually steals the rights from subsequent inventors (just because they didn't get a patent first). There is no property right in a patent as that right is always held. Instead, there is the expectation that the government will block the thieves who would steal... and also those who happened to have invented the same thing. That expectation is being treated as a property.
The original purpose of a patent is to benefit the nation by encouraging invention that would otherwise not happen. Invention that would have happened, anyway, would benefit the nation, anyway. Patenting inventions that would have happened, anyway, actually hurts the nation, because it steals from the alternative inventors, destroys competition (for the term of the patent), and even discourages invention efforts (because of so many inventors doing so many things now, there is little assurance you could be first to the patent office).
Any real property right is not in patents held, but rather, in the fact that others don't hold a patent that takes your property away. So when no one else has the patent, you get to keep the invention you made.
There is another benefit of having a separate yellow. Something that is yellow (because the signal says to turn on both red and green at that spot) can come out as a monochromatic yellow, which will be sharper in resolution than using separate red and green... which will be even more separated in the spectrum by efforts to increase the gamut. The more separated the colors are in the spectrum, the less sharp the images will be (due to chromatic aberration in the eye lens as well as corrective lenses that might be in use). Even if the color is yellowish-green or orange, you still form that color from primaries that are closer in wavelength than from those that are further apart. Basically, this allows sharpness and saturation to be less of a compromise against each other.
Or having a boatload of student loans costing thousands of a month, and having 5 kids and an unemployed spouse to feed at home, which is mortgaged under water.
The level that fees would need to be to stop the large corporation patent application mills from churning out junk would be so drastically high that they would completely lock out the small inventor entirely.
OTOH, raise them to the point where the excess taken in from patent applications could entirely and continuously fund free universal health care for everyone in the country, and I could be encouraged to look away.
Being in software is irrelevant. They are granting patents for junk inventions of all kinds, which just encourages more applications to pile up. Of course, software does cloud the issue and makes it harder to determine if something be applied for is truly innovative. Statistically, it's likely not innovative (since most patents aren't). While software will have an even higher probability of being not innovative, that does not rule out true innovation that can be, or should be, implemented in software.
The USPTO has (decades ago) lost all sense of what patents are even for. They are NOT for recognizing who invented something (first). Instead, they are for granting exclusive rights to an invention, for a period of time, in exchange for having the invention made available in the first place. Patents actually take away the rights of inventors who happen to not be first. This is considered an acceptable tradeoff when the public gets to have (most) inventions that would otherwise have not been available.
The truly innovative invention is statistically unlikely for someone else to invent it anytime soon. Eventually, at some point in the future, lots of people could invent it. But then, it's not innovative, anymore. When it's not innovative, there's no public benefit of patents since the invention would have happened, anyway. Once dozens of inventors could invent something, someone will just do it without patents (and a patent at this point only destroys a competitive business environment).
The USPTO needs to get back to the original idea of patents, which existed even before there was such a thing as a corporation. They need to be rejecting applications for everything that is not innovative, and deny the patent.
For each application, ask the question "Is this something that is likely to be invented by someone else within the next X years who would not bother to apply for a patent on it?"... where X is half the term of the patent that would be issued. Also ask "Is this invention possibly obvious to at least a few inventors educated or experienced in this field?" And ask "Does issuing a patent for this invention encourage other inventions from OTHER inventors that would otherwise never be invented by anyone were this patent not issued?" If either of the first two is answered YES, or the last is answered NO, then the patent should NOT be granted.
Remember that NOT granting a patent does NOT take away the rights of the inventor. What is different is that not granting a patent also does not take away any rights of any other inventor, either.
The value of a patent SHOULD be in protecting an EXCLUSIVE inventor from copycats would would just take the idea they did not invent, and unfairly compete. And as such, it should only be used where it is clear there likely is just ONE inventor. Things that are obvious, would have more than one inventor (given a short period of time, once the need for it emerges) and no patent is needed for the public to gain from it.
USPTO... just stop granting patents for junk inventions... which are the vast majority of your applications... and the vast majority of your past grants. The applications just keep piling in because you are just summarily granting anything that isn't an obvious duplicate of something previous. Do continue checking for duplicates. But also check for obviousness, and public benefit of the grant. You shouldn't be issuing more than about 2% to 5% of what has been issued in the past. Even better, set a finite quota of 1000 patents per month and issue them only for the best applications (figured according to the questions above).
Slashdotters... quit jumping on software patents. It's the patent system itself that is broken, regardless of whether implementations can be done in hardware or software. Truly innovative patents (a tiny fraction of what gets granted these days) should be granted, regardless of how they can be invented... in hardware or software.
Alice makes some interesting content well over 12 minutes in length, compresses it in H.264, and sells copies of it to Bill, Cincy, and Dave. Bill sells 88,000 copies for $55 each, giving nothing to Alice. Cindy sells 96,000 copies for $48 each, giving nothing to Alice. Dave sells 93,400 copies for $53.95 each and offers $186,800 to Alice, who refuses to take it. Who is MPEG LA going to go after?
If the H.264 code binary can be run in user space, non-root, in a chroot jail, then my issues with it are just philosophical and not enough to prevent me from running it. I prefer open source. But I'm not opposed to running binary code. I'm also not opposed to paying for it.
What I am opposed to is borging my computer by running un-inspectable code as a kernel module, root process, or even an unjailed user process. I do not trust corporations to do things right. I'm not going to give permissions to untrusted code. And if I can't read the source, it's untrusted... by definition.
Choosing between a primarily closed source insecure system and a closed source system that is likely insecure and has more evil corporate hooks? I just hate having to choose between the lesser of evils while they keep getting more and more evil. So what if Dirac and Theora are not the most efficient. At least they work and I can patch the security holes when they are found.
Maybe... hopefully... Oracle will spin-off or sell-off the low end Sparc hardware business. That would give them some cash for a business sector that apparently are not interested in. Then they could focus on competing for the high end database server market. IBM is diverse enough and experienced enough to carry out a wide range of business and still compete against Oracle for high-end database machines (they've been doing this for decades, with hardware... mainframes). Oracle isn't experienced in that front. Sun certainly brings some in, but more work still needs to be done (including paying attention to what they bought).
Most, almost all, other computer manufacturers do not do this. Sun itself did not do this until it was borged by Larry. In the sense of Oracle's approach and business model of shaking everyone down for every penny in their pocket, it makes sense. Except for the very top end giant servers that would be running Oracle software even if Oracle had not bought Sun, this is going to decimate the Sun market that is, for the most part, not accustomed to this much aggressive gouging. IBM now has an opportunity to push PPC based machines as the alternative to x86 architectures. I can only hope they do that.
CPU operations are limited to a certain number of bits for their operations. Programming languages like C/C++ perform their basic arithmetic operations at the machine level, so they inherit the same limitations. These bounds are not a limit either through library/template facilities at the C/C++ level, or with basic operations in high level (particular object oriented) languages such as Pike and Python.
I can tell you libgmp is not stuck with bcd. But the bcd aspect will exist because some kinds of uses for extended precision are financial/money based, and conversion to and from an external decimal format is sufficiently frequent that it's easier/faster to just do the arithmetic directly on decimal, even if tightly squeezed into 4 bits per digit. This has been going on since early computers. FYI, the ancient IBM model 1620 computer could do this in hardware. As you can see from the code in the links I posted earlier, a choice of language can hide the fact that the underlying architecture has fixed width arithmetic.
BTW, for fun, compare the speeds of those two programs, which are implementations of the same algorithm.
64 bit ints are for wimps. Real number geeks play with gigabit ints (equivalent to over 330 million decimal digits). For example, calculate square roots using extreme integers in Pike and Python.
That's an exceeding narrow role for restore points. Just because it comes up does not mean it is usable, or stays usable indefinitely when it is usable for a time. There should be a clear dialog for policy, with the option to specify how far back to keep restore points indefinitely (expressible by number, or space required, or maybe even flagging specific restore points). I would definitely keep at least some restore points across reboots.
Having been on the receiving end of bonuses, I'd say that is not true. My bonuses were IT projects completion bonuses. If it was done on time, and worked, I got my bonus. If it were to have been late, I would have gotten a pro-rated bonus. If it didn't work, I would have gotten squat. If left in the middle of the project, I would have gotten squat. Instead, I delivered the projects, working, and on time. I got my bonus and took a vacation.
BTW, in two of the projects, the terms, including deadline and bonus amount, were negotiated ahead of time.
A bonus is just a part of a more complex compensation structure. And in many cases, it makes the executive management feel better about it more so than the employees who receive it. Maybe most cases. Maybe all cases.
So when the economy goes down and every business in a particular industry loses money, shouldn't you consider the business that lost the least as a "success" and compensate its management accordingly? I see no problem with bonuses in a down economy. They kept the company from tanking in a bad situation. What bonuses did Lehman Brothers give out after they went belly up? Sure, business does have the objective to make profit. But when that objective cannot be met at certain times, the clear objective is to SURVIVE. If you contribute to the company's survival, you should be compensated.
Bonuses are also a nice way to be sure someone doesn't change jobs in the middle of a crucial project. Been there, done that.
Slashdot Mirror
You'd have to do this at the hardware level. Are there any JTAG pads on the board? If not, clipping onto the firmware flash chip with the appropriate tool may be necessary. That, or some means to prevent the existing firmware from loading while loading a substitute into RAM, which will then reload the firmware flash.
First to figure this out might get a little military contract :-)
So ... you are saying ... software is hard to do, so let's go to the least reliable source for it ... ?
Both commercial software (off the shelf ... COTS) and open source (also off the shell ... FOTS) are full of bugs. At least open source is subject to peer review (in a wider peer space) and gets bugs fixed sooner (there's rarely a coverup of bugs in open source, unlike commercial).
One big problem is that the internal review process, that still has to be done inside the government, will be weaker at this job because the people who would know how best to do that won't be working in the kinds of jobs that would be in a track to the analyst positions that can do these reviews. At least one reason to have an in-house programming team in the government is so that some of them can move up to being top level analysts without being biased in favor of certain commercial interests.
... located at each individual airport. The airport already knows exactly where it is. It can receive the GPS signal and see how far off it is ... specifically for that airport. Then it would transmit that correction data in real time over a local UHF frequency that can serve approaching planes out to some distance (perhaps 100km). Nearby airports use different frequencies which get selected when the target airport is selected and GPS indicates they are within range.
They could also spend more money and put up a triangulation based TPS that would allow accurate terrestrial positioning independent of GPS. That would be in addition to final approach guidance systems. That is, of course, if you feel warm and cozy about having extra redundant systems serving the airplane you are on.
Where's BGPSEC when you need it?
Without patents at all, every inventor has equal rights to what he invents. He just doesn't have any protection from those that steal rights from those who don't invent. What a patent really does is, in the name of protecting an inventor's rights from those who steal, it actually steals the rights from subsequent inventors (just because they didn't get a patent first). There is no property right in a patent as that right is always held. Instead, there is the expectation that the government will block the thieves who would steal ... and also those who happened to have invented the same thing. That expectation is being treated as a property.
The original purpose of a patent is to benefit the nation by encouraging invention that would otherwise not happen. Invention that would have happened, anyway, would benefit the nation, anyway. Patenting inventions that would have happened, anyway, actually hurts the nation, because it steals from the alternative inventors, destroys competition (for the term of the patent), and even discourages invention efforts (because of so many inventors doing so many things now, there is little assurance you could be first to the patent office).
Any real property right is not in patents held, but rather, in the fact that others don't hold a patent that takes your property away. So when no one else has the patent, you get to keep the invention you made.
There is another benefit of having a separate yellow. Something that is yellow (because the signal says to turn on both red and green at that spot) can come out as a monochromatic yellow, which will be sharper in resolution than using separate red and green ... which will be even more separated in the spectrum by efforts to increase the gamut. The more separated the colors are in the spectrum, the less sharp the images will be (due to chromatic aberration in the eye lens as well as corrective lenses that might be in use). Even if the color is yellowish-green or orange, you still form that color from primaries that are closer in wavelength than from those that are further apart. Basically, this allows sharpness and saturation to be less of a compromise against each other.
Or having a boatload of student loans costing thousands of a month, and having 5 kids and an unemployed spouse to feed at home, which is mortgaged under water.
The level that fees would need to be to stop the large corporation patent application mills from churning out junk would be so drastically high that they would completely lock out the small inventor entirely.
OTOH, raise them to the point where the excess taken in from patent applications could entirely and continuously fund free universal health care for everyone in the country, and I could be encouraged to look away.
Being in software is irrelevant. They are granting patents for junk inventions of all kinds, which just encourages more applications to pile up. Of course, software does cloud the issue and makes it harder to determine if something be applied for is truly innovative. Statistically, it's likely not innovative (since most patents aren't). While software will have an even higher probability of being not innovative, that does not rule out true innovation that can be, or should be, implemented in software.
The USPTO has (decades ago) lost all sense of what patents are even for. They are NOT for recognizing who invented something (first). Instead, they are for granting exclusive rights to an invention, for a period of time, in exchange for having the invention made available in the first place. Patents actually take away the rights of inventors who happen to not be first. This is considered an acceptable tradeoff when the public gets to have (most) inventions that would otherwise have not been available.
The truly innovative invention is statistically unlikely for someone else to invent it anytime soon. Eventually, at some point in the future, lots of people could invent it. But then, it's not innovative, anymore. When it's not innovative, there's no public benefit of patents since the invention would have happened, anyway. Once dozens of inventors could invent something, someone will just do it without patents (and a patent at this point only destroys a competitive business environment).
The USPTO needs to get back to the original idea of patents, which existed even before there was such a thing as a corporation. They need to be rejecting applications for everything that is not innovative, and deny the patent.
For each application, ask the question "Is this something that is likely to be invented by someone else within the next X years who would not bother to apply for a patent on it?" ... where X is half the term of the patent that would be issued. Also ask "Is this invention possibly obvious to at least a few inventors educated or experienced in this field?" And ask "Does issuing a patent for this invention encourage other inventions from OTHER inventors that would otherwise never be invented by anyone were this patent not issued?" If either of the first two is answered YES, or the last is answered NO, then the patent should NOT be granted.
Remember that NOT granting a patent does NOT take away the rights of the inventor. What is different is that not granting a patent also does not take away any rights of any other inventor, either.
The value of a patent SHOULD be in protecting an EXCLUSIVE inventor from copycats would would just take the idea they did not invent, and unfairly compete. And as such, it should only be used where it is clear there likely is just ONE inventor. Things that are obvious, would have more than one inventor (given a short period of time, once the need for it emerges) and no patent is needed for the public to gain from it.
USPTO ... just stop granting patents for junk inventions ... which are the vast majority of your applications ... and the vast majority of your past grants. The applications just keep piling in because you are just summarily granting anything that isn't an obvious duplicate of something previous. Do continue checking for duplicates. But also check for obviousness, and public benefit of the grant. You shouldn't be issuing more than about 2% to 5% of what has been issued in the past. Even better, set a finite quota of 1000 patents per month and issue them only for the best applications (figured according to the questions above).
Slashdotters ... quit jumping on software patents. It's the patent system itself that is broken, regardless of whether implementations can be done in hardware or software. Truly innovative patents (a tiny fraction of what gets granted these days) should be granted, regardless of how they can be invented ... in hardware or software.
Alice makes some interesting content well over 12 minutes in length, compresses it in H.264, and sells copies of it to Bill, Cincy, and Dave. Bill sells 88,000 copies for $55 each, giving nothing to Alice. Cindy sells 96,000 copies for $48 each, giving nothing to Alice. Dave sells 93,400 copies for $53.95 each and offers $186,800 to Alice, who refuses to take it. Who is MPEG LA going to go after?
If you want to make a system fully legal, you need the license for H.264. The problem is, they won't license any open source implementations.
They can, but they may not be the least cost way.
If the H.264 code binary can be run in user space, non-root, in a chroot jail, then my issues with it are just philosophical and not enough to prevent me from running it. I prefer open source. But I'm not opposed to running binary code. I'm also not opposed to paying for it.
What I am opposed to is borging my computer by running un-inspectable code as a kernel module, root process, or even an unjailed user process. I do not trust corporations to do things right. I'm not going to give permissions to untrusted code. And if I can't read the source, it's untrusted ... by definition.
Choosing between a primarily closed source insecure system and a closed source system that is likely insecure and has more evil corporate hooks? I just hate having to choose between the lesser of evils while they keep getting more and more evil. So what if Dirac and Theora are not the most efficient. At least they work and I can patch the security holes when they are found.
Send both where they belong ... nethernetherland.
Maybe ... hopefully ... Oracle will spin-off or sell-off the low end Sparc hardware business. That would give them some cash for a business sector that apparently are not interested in. Then they could focus on competing for the high end database server market. IBM is diverse enough and experienced enough to carry out a wide range of business and still compete against Oracle for high-end database machines (they've been doing this for decades, with hardware ... mainframes). Oracle isn't experienced in that front. Sun certainly brings some in, but more work still needs to be done (including paying attention to what they bought).
Most, almost all, other computer manufacturers do not do this. Sun itself did not do this until it was borged by Larry. In the sense of Oracle's approach and business model of shaking everyone down for every penny in their pocket, it makes sense. Except for the very top end giant servers that would be running Oracle software even if Oracle had not bought Sun, this is going to decimate the Sun market that is, for the most part, not accustomed to this much aggressive gouging. IBM now has an opportunity to push PPC based machines as the alternative to x86 architectures. I can only hope they do that.
CPU operations are limited to a certain number of bits for their operations. Programming languages like C/C++ perform their basic arithmetic operations at the machine level, so they inherit the same limitations. These bounds are not a limit either through library/template facilities at the C/C++ level, or with basic operations in high level (particular object oriented) languages such as Pike and Python.
I can tell you libgmp is not stuck with bcd. But the bcd aspect will exist because some kinds of uses for extended precision are financial/money based, and conversion to and from an external decimal format is sufficiently frequent that it's easier/faster to just do the arithmetic directly on decimal, even if tightly squeezed into 4 bits per digit. This has been going on since early computers. FYI, the ancient IBM model 1620 computer could do this in hardware. As you can see from the code in the links I posted earlier, a choice of language can hide the fact that the underlying architecture has fixed width arithmetic.
BTW, for fun, compare the speeds of those two programs, which are implementations of the same algorithm.
64 bit ints are for wimps. Real number geeks play with gigabit ints (equivalent to over 330 million decimal digits). For example, calculate square roots using extreme integers in Pike and Python.
That's an exceeding narrow role for restore points. Just because it comes up does not mean it is usable, or stays usable indefinitely when it is usable for a time. There should be a clear dialog for policy, with the option to specify how far back to keep restore points indefinitely (expressible by number, or space required, or maybe even flagging specific restore points). I would definitely keep at least some restore points across reboots.
Having been on the receiving end of bonuses, I'd say that is not true. My bonuses were IT projects completion bonuses. If it was done on time, and worked, I got my bonus. If it were to have been late, I would have gotten a pro-rated bonus. If it didn't work, I would have gotten squat. If left in the middle of the project, I would have gotten squat. Instead, I delivered the projects, working, and on time. I got my bonus and took a vacation.
BTW, in two of the projects, the terms, including deadline and bonus amount, were negotiated ahead of time.
I would have taken the money.
A bonus is just a part of a more complex compensation structure. And in many cases, it makes the executive management feel better about it more so than the employees who receive it. Maybe most cases. Maybe all cases.
So when the economy goes down and every business in a particular industry loses money, shouldn't you consider the business that lost the least as a "success" and compensate its management accordingly? I see no problem with bonuses in a down economy. They kept the company from tanking in a bad situation. What bonuses did Lehman Brothers give out after they went belly up? Sure, business does have the objective to make profit. But when that objective cannot be met at certain times, the clear objective is to SURVIVE. If you contribute to the company's survival, you should be compensated.
Bonuses are also a nice way to be sure someone doesn't change jobs in the middle of a crucial project. Been there, done that.
I thought it was the other way around.
Who cares about three-eyed monsters. Snuff 'em.