Or that urban myth about that Finnish guy that built a Unix-like kernel from scratch. There's a considerable difference between rolling your own software and rolling your own hardware. It's called "tooling costs." If you design your own mobo to fit your own case, it will cost over $5,000 to get _one_ bare board built. (This assumes you managed to borrow the software needed to do such a design -- it costs $5,000 to $50,000 -- and ignores the fact that it takes several years of designing boards to get good enough to have any chance of a design like this working on the first try.) Adding the parts to that bare board by hand will take you several days, and probably cost a few thousand, since small quantity prices are much higher than buying parts by the reel. And some parts are only sold in reels of 1,000 pieces or more. Machine-shop charges for a custom case, $500 to 1,000, assuming you didn't make it too complicated. Etc.
You could indeed use PC104 or another one of the standard bus families. You pay more for less performance, but you get good reliability since these cards are industrial grade. One issue is that your box will probably come out more cubical than lap-top shaped. Any of the rack-mounted systems are automatically at least 5 inches thick. PC104 cards are designed to stack one on top of the other, but if the clock speed is low it is possible to connect them side by side with ribbon cables. I'm not sure how low it has to be -- maybe 20MHz? But if you can live with that sort of performance, it would certainly help the battery life.
There are so many instances where a "higher up" takes credit for something someone else has done. Hell, some people have made a career out of it. Like when Gore "created" the internet?
IRA..... aaah...... they haven`t destroy any big building and their acts are localized on the North Ireland. The IRA is not restricted to North Ireland, they set off a bomb in England every few weeks. Mostly they are small bombs that don't kill or maim very many people, but there have been so many of them... Their goal seems to be to kill just enough Brits each time to remain a continual irritation, and it seems to be working; the English have removed most of the public waste receptacles because these have been a favorite place to drop off a small bomb. Maybe the 9/11 attacks killed more Brits, but the IRA has kept them from feeling safe in their own country for at least a century...
As for the rest of Europe, take Germany as an example: The Bader-Meinhoff gang was a small but quite violent terrorist organization, which took over a decade to round up. That leaves the Germans with two more problems: foreign (mostly mid-Eastern) terrorists, and German neo-nazi gangs. Apparently foreign terrorists have blown up German nightclubs to get American servicemen, planted a bomb in an airplane to explode over Scotland, and frequently pass through Germany on their way to commit mayhem elsewhere. The neo-nazis beat up and sometimes kill foreign workers -- maybe they are upset about being cut out of the toilet-cleaning jobs?
tourist farms... Being a Nashvillian, I've often wondered where those brightly-colored perennial oddities are grown. I grew up in a resort area in Michigan. I think they're manufactured in Detroit, probably in the old Edsel factory. 8-)
I'm Irish, spent years living in various German cities until 1996 and I can assure that I encounter far more racism, institutionalised and general, here in the USA than I ever did in Germany. In the 3 weeks I've spent in Germany, I didn't see any racism -- but I didn't see any people of color either. Racism is easier to detect when there are more people to be racist towards... But I have heard of apartment buildings burned down in Germany, with Turkish families inside. There is a history in the US of black homes and churches burned, but the last racial-arson deaths I recall were in the 1960's. We've still got nutcases who might want to do things like this, but they know they won't get away with it.
No I don't have statistics, and I don't trust government statistics unless I'm quite sure that the agencies collecting them aren't trying to conceal problems. If you want to know how many blacks were lynched in the US in the 1890's, don't look at crime reports -- the cops didn't consider it a crime. The best data you'll find, I think, are estimates compiled by historians several decades later.
If you already have a mind-set of a racist or hater, this person will NOT search for balanced information and discussion, it will search for the most effective propaganda there is to support his views. Just how will letting some government official regulate which propaganda is allowed correct this problem?
Large-scale outbreaks of racist or religious murders in the last 50 years have always been preceded by gov't sponsored propaganda campaigns to stir up the hate. (the middle East, Burundi, Bosnia, Croatia, & Serbia.) The US has (greatly to our shame) had a few hundred hate-crime murders in the last 50 years, and a few thousand in the last 100, but we just had 7,000 murdered by foreigners raised on a program of hate in Saudi government schools. With government control of the media 60 years ago, Germany murdered at least 12 million just because of their ethnic origin; it's a lot harder to find accurate information about the Soviets, but probably Stalin killed more of his countrymen than the Germans did, sometimes for ethnic reasons. Present Western European gov'ts do't do this sort of thing, but will you bet your life that your countrymen will never elect another Hitler or Milosevich?
In the US, we drag the hate out into the open and discuss it, and only a few real wackos still express their hatred violently. How do countries that suppress this discussion do? I suggest you try to find how many Turkish guest-workers have been murdered in Germany lately. German law-enforcement is generally very efficient, and not bound up in so many constitutional restrictions as in the USA, but they seem to be either unable or unwilling to stop racist attacks on people of color living within their borders.
... because they're PRISONERS. As far as I'm concerned, they have no rights whatsoever. They are the ones that violated OUR rights in one way or another. They are detainees, NOT convicted of anything, and in general NOT EVEN CHARGED with anything. One man was held for three weeks because he called the FBI and told them that he had rented a room to one of the hijackers for a while...
I guess landlords are automatically criminals, eh?
if you're spending $100,000,000 for a launch it doesn't make sense to save $50,000 by using a cheap antenna which is more likely to break.
Is $100M what it costs to launch the Shuttle? (It seems like the right order of magnitude.) But the USNA satellite didn't ride the shuttle, it shared a small single-use rocket with 3 other satellites. Launch cost was much less, maybe around $1M, and since it was sharing a ride with other, larger items, it's share of the launch cost may have been as low as $100K. (Probably the Navy didn't have to pay it.) Note that the most common cost quote is $5000/lb; the satellite was "the size of a TV set", so maybe it was 20 lb weight = $100K cost. And it probably would have made more sense to spend more to get longer lasting electronics, but since the thing was designed by college students maybe expensive parts would not have been that well utilized.
AFAIK, the Shuttle can carry up to 60,000 pounds payload, so if you can load it fully and it costs $100M to launch, this works out to $1667 per pound. However, 60,000 pound loads are pretty rare (maybe major pieces of the ISS), the rest of the time it carries as many smaller satellites and on-board experiments as can be deployed or ran in one flight, the total payload is considerably less, and the cost per pound considerably more. In addition, a significant part of the inflated launch cost comes from keeping the 500 or more people needed to launch it on salary for months between each launch; if you had the shuttles and the payloads to launch every few days, the per launch cost would go down quite a lot. In other words, the Shuttle is too damned big for efficient operations, unless you are building something BIG up there (cough, **starwars, cough).
By the way, a metal tape-measure sounds like a quite good solution when you want a simple quarter-wave-dipole antenna in space. It's cheap and stores compactly. By removing the case and mounting it properly, I think you could get the spring action of the tape to provide the motive power for deployment (uncoiling). It's unlikely anything would bump it in space, but if it did a tape measure would bend and then spring back, where a telescoping antenna would break.
When a major bug was found, not only did they pull the installer *immediately*, but they fixed the bug and had a new one up in its place Excuse me, had a new bug up in it's place???
On the other hand, if it was Microsoft, the patch probably would be a new bug. 8-(
If we place total faith in the security of an ID card, then all it takes is one good attack to steal quite a few identities and wreak havoc. When I was in the Air Force, there was a guy caught with a half-dozen drivers licenses, under a half-dozen names, and apparently all actually issued by various states with his picture. He was wanted for armed robbery under some of those identities. The only question was which prison he'd be living in for the next 20 years...
Put through the national ID, and guys like this will just have a half-dozen separate identities in the national database, and each will appear perfectly legitimate as long as they don't show more than one card at a time. All the electronics will do is make the cards _look_ more trustworthy.
Many insurance companies take in less money than they pay out... [T]hey can take in large amounts of money and hold it for a while, paying out... some time later. This allows them a period of time in which to invest the money that comes through.
True, but the real point is that insurance customers would do considerably better on the average if they put the premiums into safe investments in their own names and paid the cssualty losses themselves out of the accumulated deposits + interest & gains. However, since when, and in some cases whether, these losses occur is unpredictable, there would be no assurance that you personally would have enough saved up. This is most clear with something like fire insurance -- most people don't have one house burn down in their lifetime, so the lifetime accumulation of one family's fire insurance premiums + interest is less than the value of their house. And it just might burn down the day after you signed the mortgage...
So you pay for risk reduction. That shouldn't be too strange of a concept -- it's also why the interest on bonds and other loans increases with the risk of the borrower going bankrupt, and why the average gain on stocks is much higher than on bonds.
Auto insurance is mandatory, not for the sake of the idiot that causes an accident, but to ensure that there is money to compensate his victims. If you've got enough cash, you should be able to "self-ensure" by putting it in an escrow account, and you are almost sure to come out ahead in the long run. (You aren't paying for insurance company personnel, offices, computers, and profits, just for the accidents.) But most people don't want to wait until they've saved $50,000 or more _plus_ a car downpayment to start driving...
Term life insurance is another example. The risk you are insuring against is dying too soon, leaving spouse and children without adequate income. Most people would wind up with more money if they invested instead of buying life insurance, and be able to cash in the investments while they were alive to spend it -- but dying young would leave your family broke.
However, there are "insurance" policies that mix other things in with the risk reduction. Whole life policies combine coverage against dying too young and a (rather poor) long-term investment plan; buying term life only while your kids are young and investing the difference will probably build a larger retirement account... Medical insurance normally covers both catastrophic illnesses that many people never have, and routine (and not overly unpredictable) procedures like physicals, vaccinations, and treatment for the sniffles. About half of medical premiums go to pay the bureacracy to track and authorize all of those minor expenses. If you actually pay the whole cost of your medical insurance, you can do a whole lot better by buying just major medical coverage (deductible of $5K, say), and budgeting for everything else. But most people don't directly pay their entire premiums, and even those that do get tax deductions for them, so the real costs are hidden. (Because of employer-paid premiums, you get paid less and pay more for your purchases; because of tax deductions on one thing, you pay more tax otherwise.)
I could legally in US buy drugs so long as the transaction is carried out in Holland. I don't think things are quite that wide open in Holland (actually The Netherlands) yet that you can legally run a web-site selling drugs, but if they are, I'm applauding... If you travel to the Netherlands, buy some pot, and smoke it there, that should be the concern only of Netherlands law. If you order pot from a web-site in the Netherlands to be delivered to the USA, then the moment it crosses the border it's a violation of US laws. But the USA should be prosecuting the US resident who ordered it and whoever tried to ship it to a US address (if the Dutch will extradite), not shut down the foreign web site. (The Clinton and both Bush administrations have tended to overreach their legal powers, I love it when one of our courts develops a spine and overrules them, and I would love to also see them slapped down in foreign courts.)
Similarly, the French police can arrest a French resident who bypasses the French Yahoo site to reach the US Yahoo server and orders Nazi memorabilia. They might also go after the shipper (which is not Yahoo, but some Yahoo customer), although US courts might not consider selling a swastika to be an extraditable offense...
No, the speech was just an opportunity to arrest him in this country. Arresting someone just for making a speech is such an obvious 1st amendment violation that even Ashcroft has to work up to it by stages... The basis for the charges seem to be that the program is sold by a Russian company on the web, therefore Americans can buy it.
So what would U.S. District Court Judge Jeremy Fogel think of the Dmitry Skylarov case? (If you've been asleep for six months, Skylarov is a Russian citizen who is charged with violating the DMCA by writing a program _in Russia_.)
It's impossible to conceive of Linux _needing_ that big a hard drive. But think of how fast Microsoft code bloats. Every few years M$ has to invent a new file system to properly handle the larger drives needed to hold Windows & Office. And so who knows how big common disk drives will be in 10 years? But Linux is ready NOW...;-)
It's not a "vapor refrigerator" but a heat pipe. It's non-powered, except for the fans that blow air through the radiator part of it. You could eliminate the fans by making the radiator larger.
Cost/performance ratio doesn't sound so good though. ($85 for cooling 10 degrees C better than an ordinary heatsink/fan, and not quite as good as the best performing (and enormous, I suppose) HSF's. Heat pipes are NASA technology, and there are few cost-effective applications on Earth. But if you really want a _quiet_ Pentium 3 or 4 system, your choices right now seem to be either immersing the motherboard in liquid, or a heatpipe to carry the heat out to a big fan-less radiator. Or else run non-bloated software on a CPU that doesn't need extreme heatsinking...
Hydrogen is an energy storage strategy and not an energy supply strategy. It's also an energy _shipment_ strategy. You can put H2 in a pipeline and ship it far further than is practical for electricity. This is important because rooftop solar panels in the southwest could probably supply the USA's non-mobile energy needs, but more than half of the energy consumption is over 1,000 miles away.
hydrogen fuel cells are better than batteries because of the rate they can deliver energy.
Batteries can deliver enough current to instantly melt heavy wires. Their momentary output is limited mainly by the current-carrying capacity of the terminals and internal connections. Of course, they run down pretty fast when you use them this way, but batteries are definitely not the limiting factor in an acceleration test. The reasons electric cars might be slow:
1) The batteries are heavy, so the electric car is considerably heavier than the same model with gasoline power.
2) To get more range out of the batteries, the designers may have limited the motor power.
3) If you do run a battery-powered car at high speed and accelerations, you won't get far.
Fuel cells could help on all three points, not because they put out more current, but because they allow you to store more energy in less mass.
I am no expert, but I would expect short-term fuel cell output to be limited by the rate at which the fuel and air can diffuse to the electrodes and react. This is going to be a lower rate than you get from batteries where the lead and acid are already at the electrodes. Sustained output of both fuel cells and batteries may be limited by heat -- when you draw too much current, you increase the losses in the cells, and lost energy becomes heat, which must be removed or eventually the cells will self-destruct.
MS opens all protocols so that competitors products can communicate effectively. Except that if it's security related, they don't have to. Leave that in there, and in two years _everything_ will be allegedly security related.
If Bin Laden wants to blow up Comdex (which seems unlikely), they won't bother smuggling in a little bomb disguised as a laptop. With his budget, they can rent a booth and bring in a few hundred pounds of explosives tucked inside computer cases...
I can estimate my debugging time pretty closely. If all I have to find is my own mistakes, the time works out to 40% design, 20% coding, 40% test and debug. (If you are willing to let your customers find many of the bugs, you could cut the test time down...) Where things get less predictable (aside from the effects of fuzzy specs and requirement bloat) is that often I'm interfacing to someone else's work, and either their description of the interface is not entirely accurate and so I have to find out how to make it work by trial and error, or they didn't get all their bugs out and I have to work around that. So I have to pad the estimate out to account for that.
The other issue is that you send through an estimate that says 2 months for design, 1 for coding, and 3 for debugging, and the PHB's get all upset...
Ultimately the project will be released too soon because of the lack of communication and the bugs will cause another rapid release, and the cycle continues.
Don't forget that leaving lots of bugs in rev 1.0 gives marketing something to brag about in 1.1 ("50% of it works like it ought to...").
Slashdot Mirror
Or that urban myth about that Finnish guy that built a Unix-like kernel from scratch. There's a considerable difference between rolling your own software and rolling your own hardware. It's called "tooling costs." If you design your own mobo to fit your own case, it will cost over $5,000 to get _one_ bare board built. (This assumes you managed to borrow the software needed to do such a design -- it costs $5,000 to $50,000 -- and ignores the fact that it takes several years of designing boards to get good enough to have any chance of a design like this working on the first try.) Adding the parts to that bare board by hand will take you several days, and probably cost a few thousand, since small quantity prices are much higher than buying parts by the reel. And some parts are only sold in reels of 1,000 pieces or more. Machine-shop charges for a custom case, $500 to 1,000, assuming you didn't make it too complicated. Etc.
You could indeed use PC104 or another one of the standard bus families. You pay more for less performance, but you get good reliability since these cards are industrial grade. One issue is that your box will probably come out more cubical than lap-top shaped. Any of the rack-mounted systems are automatically at least 5 inches thick. PC104 cards are designed to stack one on top of the other, but if the clock speed is low it is possible to connect them side by side with ribbon cables. I'm not sure how low it has to be -- maybe 20MHz? But if you can live with that sort of performance, it would certainly help the battery life.
There are so many instances where a "higher up" takes credit for something someone else has done. Hell, some people have made a career out of it. Like when Gore "created" the internet?
IRA ..... aaah ...... they haven`t destroy any big building and their acts are localized on the North Ireland. The IRA is not restricted to North Ireland, they set off a bomb in England every few weeks. Mostly they are small bombs that don't kill or maim very many people, but there have been so many of them... Their goal seems to be to kill just enough Brits each time to remain a continual irritation, and it seems to be working; the English have removed most of the public waste receptacles because these have been a favorite place to drop off a small bomb. Maybe the 9/11 attacks killed more Brits, but the IRA has kept them from feeling safe in their own country for at least a century...
As for the rest of Europe, take Germany as an example: The Bader-Meinhoff gang was a small but quite violent terrorist organization, which took over a decade to round up. That leaves the Germans with two more problems: foreign (mostly mid-Eastern) terrorists, and German neo-nazi gangs. Apparently foreign terrorists have blown up German nightclubs to get American servicemen, planted a bomb in an airplane to explode over Scotland, and frequently pass through Germany on their way to commit mayhem elsewhere. The neo-nazis beat up and sometimes kill foreign workers -- maybe they are upset about being cut out of the toilet-cleaning jobs?
tourist farms... Being a Nashvillian, I've often wondered where those brightly-colored perennial oddities are grown. I grew up in a resort area in Michigan. I think they're manufactured in Detroit, probably in the old Edsel factory. 8-)
I'm Irish, spent years living in various German cities until 1996 and I can assure that I encounter far more racism, institutionalised and general, here in the USA than I ever did in Germany. In the 3 weeks I've spent in Germany, I didn't see any racism -- but I didn't see any people of color either. Racism is easier to detect when there are more people to be racist towards... But I have heard of apartment buildings burned down in Germany, with Turkish families inside. There is a history in the US of black homes and churches burned, but the last racial-arson deaths I recall were in the 1960's. We've still got nutcases who might want to do things like this, but they know they won't get away with it.
No I don't have statistics, and I don't trust government statistics unless I'm quite sure that the agencies collecting them aren't trying to conceal problems. If you want to know how many blacks were lynched in the US in the 1890's, don't look at crime reports -- the cops didn't consider it a crime. The best data you'll find, I think, are estimates compiled by historians several decades later.
If you already have a mind-set of a racist or hater, this person will NOT search for balanced information and discussion, it will search for the most effective propaganda there is to support his views. Just how will letting some government official regulate which propaganda is allowed correct this problem?
Large-scale outbreaks of racist or religious murders in the last 50 years have always been preceded by gov't sponsored propaganda campaigns to stir up the hate. (the middle East, Burundi, Bosnia, Croatia, & Serbia.) The US has (greatly to our shame) had a few hundred hate-crime murders in the last 50 years, and a few thousand in the last 100, but we just had 7,000 murdered by foreigners raised on a program of hate in Saudi government schools. With government control of the media 60 years ago, Germany murdered at least 12 million just because of their ethnic origin; it's a lot harder to find accurate information about the Soviets, but probably Stalin killed more of his countrymen than the Germans did, sometimes for ethnic reasons. Present Western European gov'ts do't do this sort of thing, but will you bet your life that your countrymen will never elect another Hitler or Milosevich?
In the US, we drag the hate out into the open and discuss it, and only a few real wackos still express their hatred violently. How do countries that suppress this discussion do? I suggest you try to find how many Turkish guest-workers have been murdered in Germany lately. German law-enforcement is generally very efficient, and not bound up in so many constitutional restrictions as in the USA, but they seem to be either unable or unwilling to stop racist attacks on people of color living within their borders.
... because they're PRISONERS. As far as I'm concerned, they have no rights whatsoever. They are the ones that violated OUR rights in one way or another. They are detainees, NOT convicted of anything, and in general NOT EVEN CHARGED with anything. One man was held for three weeks because he called the FBI and told them that he had rented a room to one of the hijackers for a while...
I guess landlords are automatically criminals, eh?
if you're spending $100,000,000 for a launch it doesn't make sense to save $50,000 by using a cheap antenna which is more likely to break.
Is $100M what it costs to launch the Shuttle? (It seems like the right order of magnitude.) But the USNA satellite didn't ride the shuttle, it shared a small single-use rocket with 3 other satellites. Launch cost was much less, maybe around $1M, and since it was sharing a ride with other, larger items, it's share of the launch cost may have been as low as $100K. (Probably the Navy didn't have to pay it.) Note that the most common cost quote is $5000/lb; the satellite was "the size of a TV set", so maybe it was 20 lb weight = $100K cost. And it probably would have made more sense to spend more to get longer lasting electronics, but since the thing was designed by college students maybe expensive parts would not have been that well utilized.
AFAIK, the Shuttle can carry up to 60,000 pounds payload, so if you can load it fully and it costs $100M to launch, this works out to $1667 per pound. However, 60,000 pound loads are pretty rare (maybe major pieces of the ISS), the rest of the time it carries as many smaller satellites and on-board experiments as can be deployed or ran in one flight, the total payload is considerably less, and the cost per pound considerably more. In addition, a significant part of the inflated launch cost comes from keeping the 500 or more people needed to launch it on salary for months between each launch; if you had the shuttles and the payloads to launch every few days, the per launch cost would go down quite a lot. In other words, the Shuttle is too damned big for efficient operations, unless you are building something BIG up there (cough, **starwars, cough).
By the way, a metal tape-measure sounds like a quite good solution when you want a simple quarter-wave-dipole antenna in space. It's cheap and stores compactly. By removing the case and mounting it properly, I think you could get the spring action of the tape to provide the motive power for deployment (uncoiling). It's unlikely anything would bump it in space, but if it did a tape measure would bend and then spring back, where a telescoping antenna would break.
When a major bug was found, not only did they pull the installer *immediately*, but they fixed the bug and had a new one up in its place Excuse me, had a new bug up in it's place???
On the other hand, if it was Microsoft, the patch probably would be a new bug. 8-(
If we place total faith in the security of an ID card, then all it takes is one good attack to steal quite a few identities and wreak havoc. When I was in the Air Force, there was a guy caught with a half-dozen drivers licenses, under a half-dozen names, and apparently all actually issued by various states with his picture. He was wanted for armed robbery under some of those identities. The only question was which prison he'd be living in for the next 20 years...
Put through the national ID, and guys like this will just have a half-dozen separate identities in the national database, and each will appear perfectly legitimate as long as they don't show more than one card at a time. All the electronics will do is make the cards _look_ more trustworthy.
Many insurance companies take in less money than they pay out... [T]hey can take in large amounts of money and hold it for a while, paying out ... some time later. This allows them a period of time in which to invest the money that comes through.
True, but the real point is that insurance customers would do considerably better on the average if they put the premiums into safe investments in their own names and paid the cssualty losses themselves out of the accumulated deposits + interest & gains. However, since when, and in some cases whether, these losses occur is unpredictable, there would be no assurance that you personally would have enough saved up. This is most clear with something like fire insurance -- most people don't have one house burn down in their lifetime, so the lifetime accumulation of one family's fire insurance premiums + interest is less than the value of their house. And it just might burn down the day after you signed the mortgage...
So you pay for risk reduction. That shouldn't be too strange of a concept -- it's also why the interest on bonds and other loans increases with the risk of the borrower going bankrupt, and why the average gain on stocks is much higher than on bonds.
Auto insurance is mandatory, not for the sake of the idiot that causes an accident, but to ensure that there is money to compensate his victims. If you've got enough cash, you should be able to "self-ensure" by putting it in an escrow account, and you are almost sure to come out ahead in the long run. (You aren't paying for insurance company personnel, offices, computers, and profits, just for the accidents.) But most people don't want to wait until they've saved $50,000 or more _plus_ a car downpayment to start driving...
Term life insurance is another example. The risk you are insuring against is dying too soon, leaving spouse and children without adequate income. Most people would wind up with more money if they invested instead of buying life insurance, and be able to cash in the investments while they were alive to spend it -- but dying young would leave your family broke.
However, there are "insurance" policies that mix other things in with the risk reduction. Whole life policies combine coverage against dying too young and a (rather poor) long-term investment plan; buying term life only while your kids are young and investing the difference will probably build a larger retirement account... Medical insurance normally covers both catastrophic illnesses that many people never have, and routine (and not overly unpredictable) procedures like physicals, vaccinations, and treatment for the sniffles. About half of medical premiums go to pay the bureacracy to track and authorize all of those minor expenses. If you actually pay the whole cost of your medical insurance, you can do a whole lot better by buying just major medical coverage (deductible of $5K, say), and budgeting for everything else. But most people don't directly pay their entire premiums, and even those that do get tax deductions for them, so the real costs are hidden. (Because of employer-paid premiums, you get paid less and pay more for your purchases; because of tax deductions on one thing, you pay more tax otherwise.)
See this User Friendly cartoon.
I could legally in US buy drugs so long as the transaction is carried out in Holland. I don't think things are quite that wide open in Holland (actually The Netherlands) yet that you can legally run a web-site selling drugs, but if they are, I'm applauding... If you travel to the Netherlands, buy some pot, and smoke it there, that should be the concern only of Netherlands law. If you order pot from a web-site in the Netherlands to be delivered to the USA, then the moment it crosses the border it's a violation of US laws. But the USA should be prosecuting the US resident who ordered it and whoever tried to ship it to a US address (if the Dutch will extradite), not shut down the foreign web site. (The Clinton and both Bush administrations have tended to overreach their legal powers, I love it when one of our courts develops a spine and overrules them, and I would love to also see them slapped down in foreign courts.)
Similarly, the French police can arrest a French resident who bypasses the French Yahoo site to reach the US Yahoo server and orders Nazi memorabilia. They might also go after the shipper (which is not Yahoo, but some Yahoo customer), although US courts might not consider selling a swastika to be an extraditable offense...
Would you want French web sites to restrict sexual content according to what is acceptable in Saudi Arabia?
No, the speech was just an opportunity to arrest him in this country. Arresting someone just for making a speech is such an obvious 1st amendment violation that even Ashcroft has to work up to it by stages... The basis for the charges seem to be that the program is sold by a Russian company on the web, therefore Americans can buy it.
From the article: The most vital, "mission-critical" systems that keep corporations running still rely on traditional Unix systems.
So what would U.S. District Court Judge Jeremy Fogel think of the Dmitry Skylarov case? (If you've been asleep for six months, Skylarov is a Russian citizen who is charged with violating the DMCA by writing a program _in Russia_.)
It's impossible to conceive of Linux _needing_ that big a hard drive. But think of how fast Microsoft code bloats. Every few years M$ has to invent a new file system to properly handle the larger drives needed to hold Windows & Office. And so who knows how big common disk drives will be in 10 years? But Linux is ready NOW... ;-)
It's not a "vapor refrigerator" but a heat pipe. It's non-powered, except for the fans that blow air through the radiator part of it. You could eliminate the fans by making the radiator larger.
Cost/performance ratio doesn't sound so good though. ($85 for cooling 10 degrees C better than an ordinary heatsink/fan, and not quite as good as the best performing (and enormous, I suppose) HSF's. Heat pipes are NASA technology, and there are few cost-effective applications on Earth. But if you really want a _quiet_ Pentium 3 or 4 system, your choices right now seem to be either immersing the motherboard in liquid, or a heatpipe to carry the heat out to a big fan-less radiator. Or else run non-bloated software on a CPU that doesn't need extreme heatsinking...
Hydrogen is an energy storage strategy and not an energy supply strategy. It's also an energy _shipment_ strategy. You can put H2 in a pipeline and ship it far further than is practical for electricity. This is important because rooftop solar panels in the southwest could probably supply the USA's non-mobile energy needs, but more than half of the energy consumption is over 1,000 miles away.
hydrogen fuel cells are better than batteries because of the rate they can deliver energy.
Batteries can deliver enough current to instantly melt heavy wires. Their momentary output is limited mainly by the current-carrying capacity of the terminals and internal connections. Of course, they run down pretty fast when you use them this way, but batteries are definitely not the limiting factor in an acceleration test. The reasons electric cars might be slow:
1) The batteries are heavy, so the electric car is considerably heavier than the same model with gasoline power.
2) To get more range out of the batteries, the designers may have limited the motor power.
3) If you do run a battery-powered car at high speed and accelerations, you won't get far.
Fuel cells could help on all three points, not because they put out more current, but because they allow you to store more energy in less mass.
I am no expert, but I would expect short-term fuel cell output to be limited by the rate at which the fuel and air can diffuse to the electrodes and react. This is going to be a lower rate than you get from batteries where the lead and acid are already at the electrodes. Sustained output of both fuel cells and batteries may be limited by heat -- when you draw too much current, you increase the losses in the cells, and lost energy becomes heat, which must be removed or eventually the cells will self-destruct.
MS opens all protocols so that competitors products can communicate effectively. Except that if it's security related, they don't have to. Leave that in there, and in two years _everything_ will be allegedly security related.
If Bin Laden wants to blow up Comdex (which seems unlikely), they won't bother smuggling in a little bomb disguised as a laptop. With his budget, they can rent a booth and bring in a few hundred pounds of explosives tucked inside computer cases...
I can estimate my debugging time pretty closely. If all I have to find is my own mistakes, the time works out to 40% design, 20% coding, 40% test and debug. (If you are willing to let your customers find many of the bugs, you could cut the test time down...) Where things get less predictable (aside from the effects of fuzzy specs and requirement bloat) is that often I'm interfacing to someone else's work, and either their description of the interface is not entirely accurate and so I have to find out how to make it work by trial and error, or they didn't get all their bugs out and I have to work around that. So I have to pad the estimate out to account for that.
The other issue is that you send through an estimate that says 2 months for design, 1 for coding, and 3 for debugging, and the PHB's get all upset...
Ultimately the project will be released too soon because of the lack of communication and the bugs will cause another rapid release, and the cycle continues.
Don't forget that leaving lots of bugs in rev 1.0 gives marketing something to brag about in 1.1 ("50% of it works like it ought to...").