That backwards tricycle sounds like the Buckminster Fuler's Dymaxian Car. That beast was designed for minimum air resistance. Also having the two wheels in front provides better stability when cornering during hard braking. Still, tricycles do have some roll-over stability problems because the CG is closer to the sides of the wheelbase.
Chaotic systems are actually quite controlloable in a very interesting way. The key property that makes a chaotic system so unpredictable is divergence -- if two copies of the system differ by delta, that delta will grow exponentially in time (doubling according to a coefficient call the Lyapunov coefficient). Yet, the divergence is never arbitrary. Instead, the divergence in chaotic systems happen within a space called the strange attractor - the diverging trajectories stay within in the attractor zone even as the split from each other.
If you map the strange attractor and nudge the system are the right point of the cycle, you can push the system into what ever mode of behaviro you want. Although you cannot predict the longterm behavior of the chaotic system, you can perturb it periodicaly to stabiize it or rapidlly shift its behavior. Scientists are looking at how to use this chaotic control theory to control unstable systems such as ultrahigh power lasers, manuerable jet aircraft, and heart tissue.
The key controlling a chaotic system is to understand how the chaotic system diverges (the shape of the strange attractor) and use that knowledge to deftly inject perturbations at just the right moment.
Ease-of-use is not unidimensional. The terseness of *nix commands provide ease-of-use for hardcore CLI users. This is an ease-of-power dimension of ease-of-use. The graphical menus of GUIs provide ease-of-use for more maintstream users. This is an ease-of-learning dimension of ease-of-use.
The challenge for OSS is that it is, by and large, nerds writing software for nerds. As such, they create software that emphasizes ease-of-power. They ignore the learning curve because they assume users will want to take the time to learn the software and that they will use the software intensively enough that they do not forget the commands. These assumption are false for mainstream users.
If OSS is to become mainstream, it will need to put ease-of-learning above ease-of-power (perhaps providing a dual interface such as OS X's GUI + CLI design). Adoption will depend on the extent that OSS front-end and near-front-end architecture is driven by usability experts/advocates instead of hardcore programmers.
I think that the mineral oils used with microscopy have an index that matches or exceeds quartz (around 1.5). If one wants to get fancier (and more dangerous) one could use carbon disulfide (n = 1.63) or methylene iodide (n = 1.74). Whatever one uses, it must be a nonsolid since it must provide an airless interface between the wafer and the lens. (BTW, for the wafer fab application, dispersion is not a problem because the wafers are exposed with monochromatic light)
High-end microscopes go a step further than these chip makers and use high-index low-dispersion oils instead of water as explained at formulas and intro tutorial. Replacing the air between the lens and wafer with a denser high-index fluid increases magnifaction and increases the efective aperture of the optical systems. A larger aperture increase the theoretical resolution of the system.
Sounds like these timecards should be written to WORM media or handled by a secure bitemporal database that can track ALL changes for audit purposes. I'm surprised the SEC, IRS, OSHA, etc. even allow critical financial records to be stored on standard PCs and applications. This is how we get fiascos like Enron, Parmalat, and Barings Bank -- its too easy to cook the books because the books are kept on insecure systems.
Proudly ignorant or TCO-conscious?
on
Gates on Winsecurity
·
· Score: 5, Insightful
Apple computers are created for, and solely used by people who know, and want to know nothing about computers, the "proudly ignorants"
Every extra hour that I am forced to spend learning how make make a computer do what it should have done in the first place adds $50 to the TCO of that machine. So if I have spend even one hour per week figuring out how to keep my machine safe from exploits, I've added $2500 to for the cost of that machine for that year.
I am not proudly ignorant, I only realize that my time is limited and that spending it patching gaping holes in a badly designed product is not top of my list of either fun or productive things to do. At best, you could call me resentfully ignorant because I resent that ignorance should be a problem.
I'm not even sure how you can blame Apple for much of the Internet's current dismal state of affairs. What percentage of viruses, trojans, spam, etc. are distributed via Apple machines?
But, as long as we are playing the blame game, I might as well burn a few karma points. Lets add some more culprits to the list:
1. All the IT vendors that touted software and internet services.
2. All the businesses and organizations that listened to IT vendor's hype and gave PCs to all their employees.
3. The original internet standards designer who gave us naive, overly-trusting standards that make it too easy for anonymous blackhats and spammers to send out untraceable virus packets and spam
4. CPU makers (and Gordon Moore) for giving us such a rapid pace of performance growth that no platform ever matures before it is replaced by another exploit-ridden next generation OS
I suspect that many will not like self-driving cars because they will not drive agressively enough. For example, many dislike the automatic speed-matching systems that maintain a "safe" distance to the next car because they leave too much distance to the next car. Tailgaters honk at the automated cars because they wont close the gap and the others cut into the large gap created by these systems. What the system (and safety experts and the car maker's insurance companies) consider "safe" is too tame for most drivers.
While many drivers are comfortable in taking risks, the corporate creators of these systems will be risk averse. That excessive risk averseness will hinder public acceptance.
If the employees also got to rate each other on trustworthiness or teamwork, then the backstabbing would drop. It sound liek the current system rewards backstabbing. If you change the ranking mechanism so that screwing someone gets you a low rank, then you won't do it.
Ranking systems are not neccessarily bad, they just need to be designed to provide incentives for desirable behavior. If a company wants teamwork, then make that part of the ranking .
If everyone was willing to settle for older or slower hardware, demand for it, and thus prices, would be higher. Did you ever stop to wonder why older or lower-end stuff is so cheap? The people buying the new stuff at much higher prices are essentially subsidizing it.
Very good point. Its very true that early adopters pay high prices that subsidize the R&D and plant&equipment investments needed for the low-end stuff.
I'm not sure I agree with the demand necessarily equals price -- the law of supply and demand assumes no economies of scale. If demand picks up, prices do rise temporarily, but then someone realizes they produce the higher volumes at much lower costs per unit and then price drops. For example, although demand for cars with "luxuries" like power windows, power seats, etc. has increased, prices have fallen as car makers have become very efficient at creating ever more complex cars.
The question is will consumers buy the "free" stuff that Bill Gates mentions or will they always be willing to spend $1000-$2000 for a main computing/workstation/media/internet device and perhaps $200-$500 for smaller consumer electronics devices (cellphone/PDA/digital camera, etc.). Those devices will grow in performance to fit these standard budgets, but the low-end may never be popular.
I suspect that your point is that the hardware can never be free, but for different reasons. Either there is a widely adopted high-end and a not-so widely adopted near-free low-end. Or there is a universally adopted middle point with modest features and modestly high prices (with everyone paying to support the investment required to develop and manufacture these devices). My only point is that many consumers eschew the low-end, free or otherwise.
Hardware will never be free as long as software continues to bloat with an ever-expanding list of core features and peopel continue to believe the MHz myth.
At one level, hardware can already be free. I saw a small PDA with about the same specs as the original Palm Pilot selling for $19.95. Yet such devices are NOT popular because everyone wants the latest wiz-bang features on their PDA.
Its the same reason why laptops get such aweful battery life. I'm sure that someone could create a very functional laptop with a 50 MHz processor that does a competent job running a basic office suite and have superb battery life. As a real-life example, my Psion 5Mx gets 30 hours on 2 AAs and does a great job of basic office work on a 37 MHz ARM processor. You don't need battery-sucking GHz to do the job.
Yet nobody wants to buy "under-powered" devices because they have been trained for 2 decades by Wintel that they must have the fastest machine to get decent performance.
If the WTO can force the U.S to admit offshore online casinos, perhaps the WTO can force China to admit offshore information services. The Chinese consumers should be able to access any commercial internet site (including a paid weblog service like Typepad) as a free trade issue.
Having not read this book, I don't know if the author addresses the issue, but one key potential weakness in many crypto systems is the math at the core of them. Although the difficulating in cracking many crypto systems scales exponentially with the number of bits in the key, no one can gaurantee that a given size key is intractable. The formula for the time required is a*b^N, but nobody can gaurantee that a and b aren't small numbers.
If some mathematician creates an easy way to factor large numbers (and they have been finding better and better ways to do this), then systems like RSA become vulnerable even if they use umpteen bits. Any math-based crypto system faces this challenge. Ironically, the strength of the system is, in part, based on the weakness of out understand of the math. Unless someone can prove that a lower bounds exists, the system is of unknown uncrackability.
The key to lowering the cost of launches is mass production and that means emphasizing manufacturing design, rather than rocket design. Yes, you must build something that will fly. But if you don't do a good job building the systems (the factory) that build the systems (the rockets), you will be stuck forever in a high-cost hell of precision, one-off, hand-assembled, hand-tweaked machines. This means using standaridzed parts, designing custom parts that can be mass-produced at low cost, and design easy-to-assemble, easy-to-lauch rockets.
It also means having enough volume that you can afford to invest in factory. This is the real chicken-and-egg problem. Without a high volume of launches, you can't justify the invetsment in a multi-billion dollar rocket factory and streamlined launch process. And without the rocket factory, you can't get the launch price low enough to create the launch volume. I do hope that some of the remaining wealthy internet entrepeneurs invest their collective billions in this endevour.
Cost of Choice: Social Pressure, Societal Scale
on
The Paradox of Choice
·
· Score: 4, Insightful
I do agree with concept that we have too much choice in our society, or rather, we are deep in information overload. Too much choice is not a problem if you can quickly whittle down what you want and what you don't want. The problem is when the choices become confusing and ambiguous - and I think that has happened for the average individual.
Very good points. I see the issue in terms of 4 factors:
Rising Cost of Decision Making: Excessive options and excessive information on each option drive up the cost of choice. The cost of decision making can easily exceed the marginal benefit of making the decison.
Psychological Risk of Decision Making: Some people are more comfortable without choice because it absolves them of responsibility. If you have only one choice, you get to bitch about it. If you have multiple choices and you chose incorrectly, you have only yourself to blame.
Cost of Competition: We seem to live in a competative, judgemental socitey in which people are judged by the choices they make. This increases the importance of every minor decision. Faced with a number of reasonably good options, people often spend too much time deciding. They feel compelled to do this because of the perceived social penalty of making the wrong choice. Nobody wants to pick the second-best option even if it is nearly as good as the #1 option.
Scale of Society: The bigger problem is the increasing scale of society. Many might think that have umpteen types of mustard, text editors, or cars is too much. But there is no unanimous agreement on which alternatives to remove.
This problem will only get worse. I would wager that in most industries, the number of economically viable choices scales with the log of the market-accessible population. With global trade and rising standards of living, we will only see more choices.
At about $2 a subscriber per month, its one of the most expensive channel groups. And it will only get more expensive as sports leagues have been upping license fees. The cost of sports programming on cable rose 59% between 1999 and 2002.
Now if I could only opt-out of those sales taxes and tourist taxes that are squandered on sports stadiums, I'd be a happy camper.
I was thinking more of a system that would catch this sort of error in advance, rather than after the damage had already been done by a data-entry error.
Absolutely! A competent network simulation would help predict the impact of a change and catch errors before they are committed. A monitoring system that correlated changes in system behavior against changes in configuration provides added safety in the likely event that the simulation is imperfect or the technician enters values different from that simulated.
Its a tricky problem because at some micolevel, the technician's entries were valid because the box he was configuring would have no way of knowing that it should not route calls to the bank (the bank and 911 system are just 2 indistingushable high-volume trunk lines). Only at the application layer would the system notice that 911 calls weren't getting answered.
I do agree that simulation of the ffects of command/config and inline warnings of the impact of a change woudl be good.
As I read the article, it is obvious that NYC's system is fraught with deep flaws in its design and management. These include:
1. False redundancy: Although the NYC system has a backup central offices and call centers, it apparently routes all calls from the affected area through a single Verizon subsystem. Their system is fully redundant except where its not.
2. Organizational silos in a coupled system: The City claimed that its 911 system was fine because "an error like the one made by Verizon could not necessarily have been prevented because it was not a flaw in the 911 system itself." Yet the Verizon circuits, systems, and procedures are an integral part of the 911 system. The City (and Verizon) maintain a fiction that they are independent entities when, in fact, they are tightly coupled. This division of responsibility is fine for playing the CYA Blame Game, but does not create a robust system.
3. User Interface Flaws I don't know what kind of user interface that technician was using, but it obviously has some terrible flaws if it did not warn him of the implications of the data entries. I also suspect that he was manually retyping some numbers off a computer print-out when he should have had some mechanism to download a set of proofread, verified, double-checked entries.
I don't fault NYC or Verizon in particular, they are probably no worse that anyone else. I only get angry that these types of structural insecurities are probably more widespread than anyone realizes.
This type of error is a classic problem of the computer assuming that the human was right. We create machines to give us power, but use that power to cavalierly.
The idea of having a second person "double-check" is nice in theory, but I will wager that the second person will let errors through too. If the first person is careful, the second person is faced with a long list of matching, correct entries to check. The second person soon becomes fatigued and keeps hitting the "OK" button even when there is a discrepancy. Unless the second person is offered an outsized reward (and the first person is penalized by an even greater amount), its to easy to become apathetic or non-vigilant. (Also, the double-checking process assumes that the original set of command directives was correct).
The real solution is a meta system that logs any changes to the system (like a config change), monitors dependencies of that change, and cross-checks them during exceptions. When an exception occurs, such as a bunch of 911 busy signals, the system would trace through the code and config files and correlate the fact that the onset of busy-911 calls corresponded with the insertion of the erroneous numbers. The system would then either roll-back the changes that caused a fault or alert someone of the list of likely culprits.
Seriously, who is going to buy downloaded music from Wal*Mart? Their store customer base for the most part is not sophisticated enough to get the download model and if they can't attract them, who are they going to attract? I won't buy from Wal*Mart based upon a savings of 10 cents per song when its in WMA format and I cannot be sure the track hasn't been edited.
You may not be price sensitive, but Wal-Mart's success (now at over a quarter trillion dollars a year in sales) suggests that many people do like WalMart's "Always Low Prices."
WalMart might be the only online music retailer that's making a profit - even at 88 cents/song. WalMart is the #1 retailer of CDs. They handle 14% of all music sales world wide. Thus WalMart has a huge negotiating advantage when it talks to the record lables about online sales rights -- you want your CDs on WalMart's shelves? Then license them for online sale at WalMart.com for a competative price. In contrast, Apple, Microsoft, Napster, etc. can only offer a rather meager carrot to the record labels. Thus, I'd bet that Wal-Mart pays less for its licence than do the other online music stores.
Its a case of too many players (online music sites) and too few players (downloaders). I suspect the industry will converge down to 4 or 5 major online music sites. Initial survivors of the first round of consolidation will include: Apple (they've got the iPod, nice interface, and early lead), Microsoft (they've got the desktop monopoly), and Wal-Mart (they've got the low cost structure). Perhps a couple of others might surive by having a nice sales model (e.g., subscription) or novel technology (i.e., a better way to find new interesting music).
Remember, once you have a large enough amount of capital, any advantage over 50% is garaunteed to make you money.
This is true, but you need to have amounts of money approaching or exceeding the capitalization of the casino (the ratio of the sizes is important). IIRC, big casinos are usually capitalized at over $10 billion to avoid the problem of losing streaks. With a only a slight advantage and a modest starting stake, too many random walks of bets end in gambler's ruin. And if you pick a tiny casino, then the most you can win is modest. (And if you pick any casino, they will throw you out if you win too much.)
I believe that approximations are the best we can do. ....
So, I don't even believe in 1 as a physically measurable number!
Cool, I can tell that you and I are on a similar wavelength. Whether that wavelenth is representable as a real number or as a 24-bit color is another matter.;)
I say again, every measurement is an approximation. Ergo, choose N large enough that no one can practically tell the difference. Then the approximation becomes reality.
And I agree 100% that a large enough N creates a indistinguishably fine approximation with one important exception. If the physical system violates the axioms of the mathematical system used in measurement, then there will be physical states or dynamical behaviors that have no corresponding mathematical state or admissible mathematical transform. Thus, for example, there are physical and perceptual colors for which there are no 24-bit approximations (the gamut problem). Moreover, the inverse problem occurs too. A mathematical system can have states with no corresponding physical state (see the problem of illegal colors)
The extent that the physical and mathematical systems lack a bijective (1-to-1) mapping of both states and admissible transforms is the extent that mathematical reasoning has short-comings. Math is great. As an engineer who has studied math extensively, I can vouch for the power of math to construct axiomatic systems that represent novel physical systems. I can also vouch for the weakness of math in misconstructing those axiomatic analogs of physical systems.
Math is like a chainsaw -- very powerful at cutting into problems, but also very dangerous if one is not careful.
But, maybe the real world fits into complex binary categories. For example, suppose I ask you to pick a real number between 0 and 1
Excellent point. But again, I'm not sure that the real world actually obeys the laws of real numbers either. Again, wave-particle duality makes a mess of mathematically notions of pure discrete and pure continuous. Some theories of physics suggest the existence of a quantum mechanical foam at dimensions of about 10^-33 meters. Perhaps the physical world is neither continuous (in the infinite-digit real number sense) nor discrete (in the exactly N-bits binary sense) Perhaps continuous real numbers are a good approximation, but whether real numbers are real (or just a very convenient mathematical construct) is debatable
Similarly, a question asking the color of something (which has finitely many answers) could be reformulated as a sequence of yes/no questions. For example, if the color is in 24-bit format, start with: Is the first bit a 1? and so on.
An interesting example. Yet real-world colors aren't 24-bit, although they can be approximated with a 24-bit color measuring systems. Its a crude approximation, unfortunately. I don't even know of a 24-bit system that has the color gamut of human vision, let alone one that properly measures the hyperspectral reflectance, transflectance, absorption, & flourescence properties of real-world materials. Yes, if you assume a 24-bit approximation, then binary yes/no questions suffice. My point is that one is forced to make a big (sometime right, sometime wrong) assumption in reducing the physical world to any N-bit approximation.
After all, everything you do on a computer, from playing video games to chatting via Instant messaging, ultimately gets reduced to binary form.
So very true.
To me, the deeper issue is whether the real world obeys the mathematicaly axioms of an algebra, Boolean or otherwise. The real world is nonlinear and that throws a wrench in the axioms right there. I also wonder about the axiom of closure -- that interactions of physical quantities in physical systems have consequences outside the algebraic variables of the system.
Again, I'm sure that algebras and real numbers or N-bit numbers are excellent approximations as long as we don't forget that they are only approxmations.
Slashdot Mirror
That backwards tricycle sounds like the Buckminster Fuler's Dymaxian Car. That beast was designed for minimum air resistance. Also having the two wheels in front provides better stability when cornering during hard braking. Still, tricycles do have some roll-over stability problems because the CG is closer to the sides of the wheelbase.
Chaotic systems are actually quite controlloable in a very interesting way. The key property that makes a chaotic system so unpredictable is divergence -- if two copies of the system differ by delta, that delta will grow exponentially in time (doubling according to a coefficient call the Lyapunov coefficient). Yet, the divergence is never arbitrary. Instead, the divergence in chaotic systems happen within a space called the strange attractor - the diverging trajectories stay within in the attractor zone even as the split from each other.
If you map the strange attractor and nudge the system are the right point of the cycle, you can push the system into what ever mode of behaviro you want. Although you cannot predict the longterm behavior of the chaotic system, you can perturb it periodicaly to stabiize it or rapidlly shift its behavior. Scientists are looking at how to use this chaotic control theory to control unstable systems such as ultrahigh power lasers, manuerable jet aircraft, and heart tissue.
The key controlling a chaotic system is to understand how the chaotic system diverges (the shape of the strange attractor) and use that knowledge to deftly inject perturbations at just the right moment.
Ease-of-use is not unidimensional. The terseness of *nix commands provide ease-of-use for hardcore CLI users. This is an ease-of-power dimension of ease-of-use. The graphical menus of GUIs provide ease-of-use for more maintstream users. This is an ease-of-learning dimension of ease-of-use.
The challenge for OSS is that it is, by and large, nerds writing software for nerds. As such, they create software that emphasizes ease-of-power. They ignore the learning curve because they assume users will want to take the time to learn the software and that they will use the software intensively enough that they do not forget the commands. These assumption are false for mainstream users.
If OSS is to become mainstream, it will need to put ease-of-learning above ease-of-power (perhaps providing a dual interface such as OS X's GUI + CLI design). Adoption will depend on the extent that OSS front-end and near-front-end architecture is driven by usability experts/advocates instead of hardcore programmers.
I think that the mineral oils used with microscopy have an index that matches or exceeds quartz (around 1.5). If one wants to get fancier (and more dangerous) one could use carbon disulfide (n = 1.63) or methylene iodide (n = 1.74). Whatever one uses, it must be a nonsolid since it must provide an airless interface between the wafer and the lens. (BTW, for the wafer fab application, dispersion is not a problem because the wafers are exposed with monochromatic light)
High-end microscopes go a step further than these chip makers and use high-index low-dispersion oils instead of water as explained at formulas and intro tutorial. Replacing the air between the lens and wafer with a denser high-index fluid increases magnifaction and increases the efective aperture of the optical systems. A larger aperture increase the theoretical resolution of the system.
Sounds like these timecards should be written to WORM media or handled by a secure bitemporal database that can track ALL changes for audit purposes. I'm surprised the SEC, IRS, OSHA, etc. even allow critical financial records to be stored on standard PCs and applications. This is how we get fiascos like Enron, Parmalat, and Barings Bank -- its too easy to cook the books because the books are kept on insecure systems.
Apple computers are created for, and solely used by people who know, and want to know nothing about computers, the "proudly ignorants"
Every extra hour that I am forced to spend learning how make make a computer do what it should have done in the first place adds $50 to the TCO of that machine. So if I have spend even one hour per week figuring out how to keep my machine safe from exploits, I've added $2500 to for the cost of that machine for that year.
I am not proudly ignorant, I only realize that my time is limited and that spending it patching gaping holes in a badly designed product is not top of my list of either fun or productive things to do. At best, you could call me resentfully ignorant because I resent that ignorance should be a problem.
I'm not even sure how you can blame Apple for much of the Internet's current dismal state of affairs. What percentage of viruses, trojans, spam, etc. are distributed via Apple machines?
But, as long as we are playing the blame game, I might as well burn a few karma points. Lets add some more culprits to the list:
1. All the IT vendors that touted software and internet services.
2. All the businesses and organizations that listened to IT vendor's hype and gave PCs to all their employees.
3. The original internet standards designer who gave us naive, overly-trusting standards that make it too easy for anonymous blackhats and spammers to send out untraceable virus packets and spam
4. CPU makers (and Gordon Moore) for giving us such a rapid pace of performance growth that no platform ever matures before it is replaced by another exploit-ridden next generation OS
I'm sure there are others.
I suspect that many will not like self-driving cars because they will not drive agressively enough. For example, many dislike the automatic speed-matching systems that maintain a "safe" distance to the next car because they leave too much distance to the next car. Tailgaters honk at the automated cars because they wont close the gap and the others cut into the large gap created by these systems. What the system (and safety experts and the car maker's insurance companies) consider "safe" is too tame for most drivers.
While many drivers are comfortable in taking risks, the corporate creators of these systems will be risk averse. That excessive risk averseness will hinder public acceptance.
If the employees also got to rate each other on trustworthiness or teamwork, then the backstabbing would drop. It sound liek the current system rewards backstabbing. If you change the ranking mechanism so that screwing someone gets you a low rank, then you won't do it.
Ranking systems are not neccessarily bad, they just need to be designed to provide incentives for desirable behavior. If a company wants teamwork, then make that part of the ranking .
If everyone was willing to settle for older or slower hardware, demand for it, and thus prices, would be higher. Did you ever stop to wonder why older or lower-end stuff is so cheap? The people buying the new stuff at much higher prices are essentially subsidizing it.
Very good point. Its very true that early adopters pay high prices that subsidize the R&D and plant&equipment investments needed for the low-end stuff.
I'm not sure I agree with the demand necessarily equals price -- the law of supply and demand assumes no economies of scale. If demand picks up, prices do rise temporarily, but then someone realizes they produce the higher volumes at much lower costs per unit and then price drops. For example, although demand for cars with "luxuries" like power windows, power seats, etc. has increased, prices have fallen as car makers have become very efficient at creating ever more complex cars.
The question is will consumers buy the "free" stuff that Bill Gates mentions or will they always be willing to spend $1000-$2000 for a main computing/workstation/media/internet device and perhaps $200-$500 for smaller consumer electronics devices (cellphone/PDA/digital camera, etc.). Those devices will grow in performance to fit these standard budgets, but the low-end may never be popular.
I suspect that your point is that the hardware can never be free, but for different reasons. Either there is a widely adopted high-end and a not-so widely adopted near-free low-end. Or there is a universally adopted middle point with modest features and modestly high prices (with everyone paying to support the investment required to develop and manufacture these devices). My only point is that many consumers eschew the low-end, free or otherwise.
Hardware will never be free as long as software continues to bloat with an ever-expanding list of core features and peopel continue to believe the MHz myth.
At one level, hardware can already be free. I saw a small PDA with about the same specs as the original Palm Pilot selling for $19.95. Yet such devices are NOT popular because everyone wants the latest wiz-bang features on their PDA.
Its the same reason why laptops get such aweful battery life. I'm sure that someone could create a very functional laptop with a 50 MHz processor that does a competent job running a basic office suite and have superb battery life. As a real-life example, my Psion 5Mx gets 30 hours on 2 AAs and does a great job of basic office work on a 37 MHz ARM processor. You don't need battery-sucking GHz to do the job.
Yet nobody wants to buy "under-powered" devices because they have been trained for 2 decades by Wintel that they must have the fastest machine to get decent performance.
If the WTO can force the U.S to admit offshore online casinos, perhaps the WTO can force China to admit offshore information services. The Chinese consumers should be able to access any commercial internet site (including a paid weblog service like Typepad) as a free trade issue.
Having not read this book, I don't know if the author addresses the issue, but one key potential weakness in many crypto systems is the math at the core of them. Although the difficulating in cracking many crypto systems scales exponentially with the number of bits in the key, no one can gaurantee that a given size key is intractable. The formula for the time required is a*b^N, but nobody can gaurantee that a and b aren't small numbers.
If some mathematician creates an easy way to factor large numbers (and they have been finding better and better ways to do this), then systems like RSA become vulnerable even if they use umpteen bits. Any math-based crypto system faces this challenge. Ironically, the strength of the system is, in part, based on the weakness of out understand of the math. Unless someone can prove that a lower bounds exists, the system is of unknown uncrackability.
The key to lowering the cost of launches is mass production and that means emphasizing manufacturing design, rather than rocket design. Yes, you must build something that will fly. But if you don't do a good job building the systems (the factory) that build the systems (the rockets), you will be stuck forever in a high-cost hell of precision, one-off, hand-assembled, hand-tweaked machines. This means using standaridzed parts, designing custom parts that can be mass-produced at low cost, and design easy-to-assemble, easy-to-lauch rockets.
It also means having enough volume that you can afford to invest in factory. This is the real chicken-and-egg problem. Without a high volume of launches, you can't justify the invetsment in a multi-billion dollar rocket factory and streamlined launch process. And without the rocket factory, you can't get the launch price low enough to create the launch volume. I do hope that some of the remaining wealthy internet entrepeneurs invest their collective billions in this endevour.
I do agree with concept that we have too much choice in our society, or rather, we are deep in information overload. Too much choice is not a problem if you can quickly whittle down what you want and what you don't want. The problem is when the choices become confusing and ambiguous - and I think that has happened for the average individual.
Very good points. I see the issue in terms of 4 factors:
Rising Cost of Decision Making: Excessive options and excessive information on each option drive up the cost of choice. The cost of decision making can easily exceed the marginal benefit of making the decison.
Psychological Risk of Decision Making: Some people are more comfortable without choice because it absolves them of responsibility. If you have only one choice, you get to bitch about it. If you have multiple choices and you chose incorrectly, you have only yourself to blame.
Cost of Competition: We seem to live in a competative, judgemental socitey in which people are judged by the choices they make. This increases the importance of every minor decision. Faced with a number of reasonably good options, people often spend too much time deciding. They feel compelled to do this because of the perceived social penalty of making the wrong choice. Nobody wants to pick the second-best option even if it is nearly as good as the #1 option.
Scale of Society: The bigger problem is the increasing scale of society. Many might think that have umpteen types of mustard, text editors, or cars is too much. But there is no unanimous agreement on which alternatives to remove.
This problem will only get worse. I would wager that in most industries, the number of economically viable choices scales with the log of the market-accessible population. With global trade and rising standards of living, we will only see more choices.
At about $2 a subscriber per month, its one of the most expensive channel groups. And it will only get more expensive as sports leagues have been upping license fees. The cost of sports programming on cable rose 59% between 1999 and 2002.
Now if I could only opt-out of those sales taxes and tourist taxes that are squandered on sports stadiums, I'd be a happy camper.
I was thinking more of a system that would catch this sort of error in advance, rather than after the damage had already been done by a data-entry error.
Absolutely! A competent network simulation would help predict the impact of a change and catch errors before they are committed. A monitoring system that correlated changes in system behavior against changes in configuration provides added safety in the likely event that the simulation is imperfect or the technician enters values different from that simulated.
Its a tricky problem because at some micolevel, the technician's entries were valid because the box he was configuring would have no way of knowing that it should not route calls to the bank (the bank and 911 system are just 2 indistingushable high-volume trunk lines). Only at the application layer would the system notice that 911 calls weren't getting answered.
I do agree that simulation of the ffects of command/config and inline warnings of the impact of a change woudl be good.
As I read the article, it is obvious that NYC's system is fraught with deep flaws in its design and management. These include:
1. False redundancy: Although the NYC system has a backup central offices and call centers, it apparently routes all calls from the affected area through a single Verizon subsystem. Their system is fully redundant except where its not.
2. Organizational silos in a coupled system: The City claimed that its 911 system was fine because "an error like the one made by Verizon could not necessarily have been prevented because it was not a flaw in the 911 system itself." Yet the Verizon circuits, systems, and procedures are an integral part of the 911 system. The City (and Verizon) maintain a fiction that they are independent entities when, in fact, they are tightly coupled. This division of responsibility is fine for playing the CYA Blame Game, but does not create a robust system.
3. User Interface Flaws I don't know what kind of user interface that technician was using, but it obviously has some terrible flaws if it did not warn him of the implications of the data entries. I also suspect that he was manually retyping some numbers off a computer print-out when he should have had some mechanism to download a set of proofread, verified, double-checked entries.
I don't fault NYC or Verizon in particular, they are probably no worse that anyone else. I only get angry that these types of structural insecurities are probably more widespread than anyone realizes.
This type of error is a classic problem of the computer assuming that the human was right. We create machines to give us power, but use that power to cavalierly.
The idea of having a second person "double-check" is nice in theory, but I will wager that the second person will let errors through too. If the first person is careful, the second person is faced with a long list of matching, correct entries to check. The second person soon becomes fatigued and keeps hitting the "OK" button even when there is a discrepancy. Unless the second person is offered an outsized reward (and the first person is penalized by an even greater amount), its to easy to become apathetic or non-vigilant. (Also, the double-checking process assumes that the original set of command directives was correct).
The real solution is a meta system that logs any changes to the system (like a config change), monitors dependencies of that change, and cross-checks them during exceptions. When an exception occurs, such as a bunch of 911 busy signals, the system would trace through the code and config files and correlate the fact that the onset of busy-911 calls corresponded with the insertion of the erroneous numbers. The system would then either roll-back the changes that caused a fault or alert someone of the list of likely culprits.
Although I am in favor of broader adoption of broadband, I do see a couple of downsides:
1. More telecom taxes to support universal service (including taxes on VoIP)
2. more zombie boxen and virus datastorms from clueless broadband users
We shall see if universal service improves the economies of scale enough to cover the increased costs of taxes and AV/firewall.
Seriously, who is going to buy downloaded music from Wal*Mart? Their store customer base for the most part is not sophisticated enough to get the download model and if they can't attract them, who are they going to attract? I won't buy from Wal*Mart based upon a savings of 10 cents per song when its in WMA format and I cannot be sure the track hasn't been edited.
You may not be price sensitive, but Wal-Mart's success (now at over a quarter trillion dollars a year in sales) suggests that many people do like WalMart's "Always Low Prices."
WalMart might be the only online music retailer that's making a profit - even at 88 cents/song. WalMart is the #1 retailer of CDs. They handle 14% of all music sales world wide. Thus WalMart has a huge negotiating advantage when it talks to the record lables about online sales rights -- you want your CDs on WalMart's shelves? Then license them for online sale at WalMart.com for a competative price. In contrast, Apple, Microsoft, Napster, etc. can only offer a rather meager carrot to the record labels. Thus, I'd bet that Wal-Mart pays less for its licence than do the other online music stores.
Its a case of too many players (online music sites) and too few players (downloaders). I suspect the industry will converge down to 4 or 5 major online music sites. Initial survivors of the first round of consolidation will include: Apple (they've got the iPod, nice interface, and early lead), Microsoft (they've got the desktop monopoly), and Wal-Mart (they've got the low cost structure). Perhps a couple of others might surive by having a nice sales model (e.g., subscription) or novel technology (i.e., a better way to find new interesting music).
Remember, once you have a large enough amount of capital, any advantage over 50% is garaunteed to make you money.
This is true, but you need to have amounts of money approaching or exceeding the capitalization of the casino (the ratio of the sizes is important). IIRC, big casinos are usually capitalized at over $10 billion to avoid the problem of losing streaks. With a only a slight advantage and a modest starting stake, too many random walks of bets end in gambler's ruin. And if you pick a tiny casino, then the most you can win is modest. (And if you pick any casino, they will throw you out if you win too much.)
I believe that approximations are the best we can do.
....
;)
So, I don't even believe in 1 as a physically measurable number!
Cool, I can tell that you and I are on a similar wavelength. Whether that wavelenth is representable as a real number or as a 24-bit color is another matter.
I say again, every measurement is an approximation. Ergo, choose N large enough that no one can practically tell the difference. Then the approximation becomes reality.
And I agree 100% that a large enough N creates a indistinguishably fine approximation with one important exception. If the physical system violates the axioms of the mathematical system used in measurement, then there will be physical states or dynamical behaviors that have no corresponding mathematical state or admissible mathematical transform. Thus, for example, there are physical and perceptual colors for which there are no 24-bit approximations (the gamut problem). Moreover, the inverse problem occurs too. A mathematical system can have states with no corresponding physical state (see the problem of illegal colors)
The extent that the physical and mathematical systems lack a bijective (1-to-1) mapping of both states and admissible transforms is the extent that mathematical reasoning has short-comings. Math is great. As an engineer who has studied math extensively, I can vouch for the power of math to construct axiomatic systems that represent novel physical systems. I can also vouch for the weakness of math in misconstructing those axiomatic analogs of physical systems.
Math is like a chainsaw -- very powerful at cutting into problems, but also very dangerous if one is not careful.
But, maybe the real world fits into complex binary categories. For example, suppose I ask you to pick a real number between 0 and 1
Excellent point. But again, I'm not sure that the real world actually obeys the laws of real numbers either. Again, wave-particle duality makes a mess of mathematically notions of pure discrete and pure continuous. Some theories of physics suggest the existence of a quantum mechanical foam at dimensions of about 10^-33 meters. Perhaps the physical world is neither continuous (in the infinite-digit real number sense) nor discrete (in the exactly N-bits binary sense) Perhaps continuous real numbers are a good approximation, but whether real numbers are real (or just a very convenient mathematical construct) is debatable
Similarly, a question asking the color of something (which has finitely many answers) could be reformulated as a sequence of yes/no questions. For example, if the color is in 24-bit format, start with: Is the first bit a 1? and so on.
An interesting example. Yet real-world colors aren't 24-bit, although they can be approximated with a 24-bit color measuring systems. Its a crude approximation, unfortunately. I don't even know of a 24-bit system that has the color gamut of human vision, let alone one that properly measures the hyperspectral reflectance, transflectance, absorption, & flourescence properties of real-world materials. Yes, if you assume a 24-bit approximation, then binary yes/no questions suffice. My point is that one is forced to make a big (sometime right, sometime wrong) assumption in reducing the physical world to any N-bit approximation.
After all, everything you do on a computer, from playing video games to chatting via Instant messaging, ultimately gets reduced to binary form.
So very true.
To me, the deeper issue is whether the real world obeys the mathematicaly axioms of an algebra, Boolean or otherwise. The real world is nonlinear and that throws a wrench in the axioms right there. I also wonder about the axiom of closure -- that interactions of physical quantities in physical systems have consequences outside the algebraic variables of the system.
Again, I'm sure that algebras and real numbers or N-bit numbers are excellent approximations as long as we don't forget that they are only approxmations.