You seem to be implying that he should try it first. My response is "why?"
You never posted on/. until you did. You never drove a car until you tried. You never worked as a programmer until you got hired. Those, and millions more, are examples which make your question unnecessary.
A comparison to a hair stylist is useless because there is no transfer of skill from a programming job, for example. However an artist could do very well in hair styling, it pays well at the top.
Basically, a good coder knows more than he can implement with his own pair of hands. Then the coder gets promoted to a designer, team lead, a manager - where he is given an extension of his hands to do the job that he holds in his head and implements with help of younger, less experienced engineers. Any other use of his time would be a waste.
What makes these morons above him think he has any skill at management, or even any interest?
I would not be so quick to label people I don't know as "morons". It might be unfair, on occasion:-) Besides, if he was chosen out of hundreds of possible candidates then probably there is a reason to try him in this role? Managers are usually good at understanding people.
So why don't they hire managerial people for management roles?
Because most people, and even some managers, understands that Dilbert's PHB is not always optimal;-)
But, control of the IP remains with the owner who may move production at anytime.
Yes, that's the lullaby that outsourcing US corporations like to sing to themselves. However they are one USB thumb drive away from losing even this last sliver of ownership. The rest is already overseas. Besides, the factories that now bake CPUs for us can continue doing so for a long, long time - and if our Asian overlords want they can completely deny Intel and Motorola access to their own factories (because, as they say, possession is 90% of the law.) Where would you move the production to? There are no spare semiconductor fabs, and there are no suitable low-cost 3rd world countries left to move into.
What then those suddenly fabless semiconductor companies will do? Build a new factory in the USA? You'd need equipment that is made in Asia, and if it is not sold then it's just too bad. The proper answer to "what to do" is just this: to rebuild the indigenous high-tech industry from the bottom up. How long do you think this will take, considering that US has only brains, and very little in terms of hands? Decades, and you'd have to start with schools. By the time US has that done where do you think China and India will be, after having such a head start? They already have more engineers and scientists just because their population is so much larger, and in this business only the absolute number of scientists matters (as opposed to doctors, for example.) China just doesn't want to rock the boat yet, but they'd be fools if they don't consider hundreds of possible scenarios, short- and long-term.
I don't believe for a moment that China has no access to the technology that it should not have. That's what spies are for, and it's not that hard to find a spy if every 5th man on the planet is Chinese.
IMO the ITAR restrictions are not always productive. They are like that restriction on export of strong encryption products with keys longer than $n. Can anyone honestly say that those regulations stopped any country in the world from using the latest encryption? There are probably more PhDs in China than americans on the whole planet:-) Surely one of those PhDs could just read a math paper and write 100 lines of simple code that AES, or DES, or any other strong cipher requires. The tables and the proof and the reference implementation are published, and in reality if Chinese government wanted PGP 2.x (which was not allowed for export) it could always ask one of its diplomats to download the code within the USA (and to lie when the server asks if you are a foreigner - the server has no way to check!) and bring it in on a floppy. Big deal, spies do worse things every day. And now a 10-minute download wrecked the whole sector of US software industry, and as a "bonus" you can be sure that Chinese use the clean PGP, without any backdoors or "forgotten" NSA keys.
Ineffective laws like that only prevent US companies from selling what the USA excels in, while not really impeding the progress of the other side. It could be even making things worse. For example, Iran has a lot of american airplanes (F-15 or something.) Pentagon put the stop on selling parts, and the Iran's airforce is suddenly in tatters. But imagine that Iran could not buy those F-15s back then. Iranians would then buy from China, from Russia, from France - or if all else fails they'd make their own, not as good but 10 times as cheap (and plentiful) and now all of a sudden you have no leverage, no spare parts to pull from the market, and no control whatsoever. How would that be better?
Inquiring minds do not have the need to know. Besides, any US startup can raise funds and buy almost any machine, even one that requires ten times its weight in paperwork to ship to Canada. See ITAR for details.
The problem here is that we don't really make much to sell to them.
And to make things worse, ITAR stops many US companies from selling some high-tech stuff that China would love to have, but which has dual use or military use. And much of US export is military and high-tech stuff. The USA can't compete with China on rice, for example, or on metals; not even on cars - China flooded Russia with cheap cars, and if anyone wants something better then Honda and Toyota are just a ferry ride away, and always glad to sell.
This has raised concerns among American government officials about the risks to national security in transferring high technology to China
I think the horse has not only left the barn, it's off the planet by now. What were those "government officials" thinking for last decades? And this process is not [easily] reversible - China has all the factories now, and rephrasing Mao, "Power comes out of the gates of the factory." This much we see already.
The airplane will not fall unless the pilot allows it to fall, due to lack of understanding of the situation, or because of inability to control the flight. But it surely can fall, one way or another, if you don't know which way you are going. In this thread we have even a firsthand account of flying a presumably GA airplane with no power. The trick is to remain in control, and larger commercial airplanes (which climb to 40K feet) are often all fly by wire. Smaller airplanes are closer to early designs of 1900's which required no electricity at all, except ignition for the engine. But even if you have barometric altimeter and a pneumatic airspeed meter, you probably still need a flashlight to see them (do they still paint the needles with phosphorescent paint these days? I used to have some that were painted such, they were bright enough to see in darkness.)
You are thinking that falling will cause low gravity and that can be sensed. However if you are also moving in a spiral the centrifugal force will compensate, and that's how you can fly into the ground without knowing it. Basically, all you can sense is the direction of the acceleration vector, but if it points downward it doesn't at all mean you are safe. If you follow the link above you will see that the spiral dive is specifically characterized by the acceleration (gravity + spin) vector being normal to the floor of the cabin, and the only hint you have (outside of the artificial horizon in front of your nose being tilted out of wazoo) is that this "gravity" may be somewhat stronger or weaker - which could also happen as you ascend or descend.
How much then the external lights take? What provides the power to fuel pumps? How the pilot is to know what RPMs of remaining engines are (kinda important if you have one or more off,) and whether an engine is already on fire or not? The airplane will not fall like a rock without any and all power only if the pilot has the gyroscopes running, the light to see them, and hydraulic pressure to operate the surfaces. Hydraulics is powered by electric pumps, not by power of will. Different airplanes have more or less of electrical control of surfaces. Deicing can be done in many ways, and air bleed is only used in one; application of alcohol based fluids is common often on the ground, but some older airplanes are designed for this method of deicing in flight as well.
All in all, completely losing power is unacceptable, but in case you lose all your primary generators the airplane gets dark fast. I do not recall for how long the batteries ought to suffice, but your figure (30 min.) is close enough to what I said. Most of the battery's power will be spent on mechanically controlling the airplane.
Well, this thing is no different from an internet cafe, and we know that those are popular enough. It's probably because most users of an internet cafe do not care about their email passwords. Business users are typically issued company notebooks which connect to the company's mail server through a VPN. But if a company permits direct access to its internal webmail then it assumes the risk of passwords being stolen one way or another, since you type them into the browser. Any keylogger, or a custom build of Firefox, would do the stealing easily, since the cafe owner is the sysadmin on all the computers. But all things considered, I would trust a large airline more than a shady Internet cafe owner.
It's quite possible that there are no executables on the system other than the browser and the OpenOffice. If the USB drive is mounted with '-o noexec' then you can't run your own apps either, and you can't execute anything from your $HOME either (can be also mounted noexec.)
I use embedded Linux at work, and the hardware that we run it on has a jumper which allows to electrically prevent any and all writes into the Flash. If you want to upgrade the software, press a button inside. No software can compensate for a WP# pin on Flash being tied to the ground; you'd need to do the iPhone-style hack with a soldering iron, and I don't see this as likely during a flight:-)
Usually each engine has a primary generator that is powering the airplane's circuits. This ensures that an engine failure (or shutdown) does not turn the entire airplane off, but only reduces the available power. An airplane needs an awful amount of electricity to fly; even your common strobe lights, mandatory in flight, consume kilowatts, and the landing lights are usually more powerful and must be on during the landing. But there is plenty of more important hardware on board, such as engine and flight control, navigation, radio, deicing, gear, fire suppression, etc. Imagine losing all power at 40,000 ft at night - you could be falling like a stone and not knowing it... because of that every passenger airplane made in last 50 years carries some batteries, and if they are used then they only feed the flight support group (essential instruments) and only for so many minutes (10-20) because of the current needed.
Also many airplanes have an auxiliary power plant, as you say placed at the tail, it is usually needed to provide fast moving air to spin the main engines, but can produce electricity as well. It is started by an electrical motor, which is powered from the truck on the ground. Batteries may be used, but only as an emergency measure.
Also some airplanes have a small external generator which can be used in an emergency. If you lose lots of power from engines at 40,000 ft you often have more altitude and more speed than you need (depends on where you plan to land,) and if so you can drop both and at the same time get some electrical power.
But in any case, an airplane is well provided with power, except in emergencies. A 50W here and there do not count, and besides the main cabin's lights and entertainment are the first to go if a power source fails.
Linux is the most suitable OS for that, since all you need to do is to reset the box, and everything is back to factory defaults. Even ThinkNIC (which I happen to have) used to do that, though it had a small (32 MB) persistent storage for bookmarks and such. You could also use a truly embedded kiosk system, but it wouldn't have many apps already written for it.
Of course it's possible, and that's how Linux-based embedded systems work. Your/home/$USERNAME can be created in RAM and deleted (recreated from a skeleton) after you log out (or the system restarted.) There is nothing else writeable on the whole box. This is necessary in embedded systems to prevent Flash wearing out, and to ensure reliability. Same needs here.
I understand your point, and it is valid. However the concern can be easily alleviated by just warning the passenger: "Sir, I do not believe this is the fastest route, and I know very well where the airport is, so are you absolutely sure you want me to turn left here? If we do that, we'll have to merge on a busy highway through a one-lane ramp and that may take a while, considering it's 5:30pm now..."
If the passenger insists to take the turn, it's his choice, and in fact his right to tell you where he wants you to go. Your only recourse, as a cab driver, is to refuse to go there, and I do not know what legal rights the passenger then gets against you (since you dumped him in the middle of nowhere.)
But on a larger scale of things, your comparison to a programmer is not exactly fitting. A programmer creates a product - a permanent entity that will be around for a while, will be used by thousands of users, will be modified and improved by tens of programmers in the future. You want to put some work into a permanent construction. In addition to that, a programmer is specifically selected and hired to use his personal skills and knowledge to create a better product. If a programmer doesn't do that but instead goes along with a poorer solution he is not doing his job.
On the other hand, a trip from point A to point B is a one-time service, offered to one person, and after the service is complete there will be no direct trace of it left in history. You, as a taxi driver, are hired by the cab company based on your decent knowledge of the city and on your reasonable ability to drive. However your duty to the company (and to yourself, if you rent the cab) is to earn money. Your duty to the passenger is to offer your best judgment on how to get from here to there under current conditions. This duty may conflict with the financial one, and you can find taxi drivers who pick one over another, it's a personal choice. However if the passenger refuses your offer of route planning, even despite you advising him to not do that, then you are in the clear, morally. You did your job, you offered a better way, and you were refused. Very well, then let the fool tell you where to go - there is plenty of gas in the tank and you are familiar with the area for hundreds of miles around.
not make me empowered to make demands that we "turn left here."
If I were driving a cab and you were a passenger, I'd welcome your guidance as long as you pay for the whole route. I know that I can't possibly earn less than a fair fare (such as the fastest/cheapest route), and I also know that most passengers are not local and will not come up with an optimal route. As matter of fact, many would just become lost, on a one-way street, or on a highway, moving very fast *somewhere*, next exit ten miles:-) There are plenty of places where you can follow the map, approach an intersection and see the desired road 30 feet above you, with no ramp to get there:-) However you put it, passenger's guidance can only make me richer.
Yes, GPS is a passive system. However the receiver can easily record the position as often as every second, thus providing the complete route (and the speed) to whoever reviews the data later. The data can be either stored on a dirt cheap SD/MMC flash card, or just wirelessly transmitted when the taxicab returns to the garage.
how come we never saw magnetic drives with builtin RAM caches in the GB scale
In terms of lazy write, the controller probably wants to write all the cache onto the platters in case of power failure. This takes time, and you can write only so much until the power becomes too low to write.
In terms of read caching, it would require a DDR2 design on the controller board, and those controllers aren't high-tech enough for that. DDR2 is very difficult to connect, requires picky controllers, and consumes a lot of power. Why would anyone do that if the computer already has tons of RAM, and you can add more if you want?
Besides, caching in main RAM is better because the CPU knows what is where and can control caching. A RAM cache is a virtual page away from being physically accessible to your needy process; nanoseconds, in other words. But if you do your caching in the HDD controller then you are doing it after squeezing through the [S]ATA bottleneck, and the controller only has bits and pieces of the information. For example, the main CPU can recognize that you are reading a file, and can pre-read subsequent sectors anticipating your future request. The HDD controller doesn't know what a file is, and so it can only react to events.
"The only thing is that the moonshiners won't give up their lucrative trade willingly. They will market their alcohol as "better" or "stronger" than the stuff you get at a liquor store. It will still sell for a lot. And the alcoholics will still burgle and rob to get the money to buy it."
On this planet, however, "good enough" is good enough for any alcohol drinker (or a drug user.) Getting an affordable drug when one needs it surely beats robbing a store and potentially getting killed. Drug users may be reckless but still not suicidal. Some addicts would be glad to stop, but their bodies changed to require the drug, and if forced to abstain they feel extreme pain. Under the threat of such pain an addict will rob and kill; however given an option I believe many would accept the government-sponsored drug, the pain will be gone just as well as when using a street drug.
Slashdot Mirror
You never posted on /. until you did. You never drove a car until you tried. You never worked as a programmer until you got hired. Those, and millions more, are examples which make your question unnecessary.
A comparison to a hair stylist is useless because there is no transfer of skill from a programming job, for example. However an artist could do very well in hair styling, it pays well at the top.
Basically, a good coder knows more than he can implement with his own pair of hands. Then the coder gets promoted to a designer, team lead, a manager - where he is given an extension of his hands to do the job that he holds in his head and implements with help of younger, less experienced engineers. Any other use of his time would be a waste.
What makes these morons above him think he has any skill at management, or even any interest?
I would not be so quick to label people I don't know as "morons". It might be unfair, on occasion :-) Besides, if he was chosen out of hundreds of possible candidates then probably there is a reason to try him in this role? Managers are usually good at understanding people.
So why don't they hire managerial people for management roles?
Because most people, and even some managers, understands that Dilbert's PHB is not always optimal ;-)
Yes, that's the lullaby that outsourcing US corporations like to sing to themselves. However they are one USB thumb drive away from losing even this last sliver of ownership. The rest is already overseas. Besides, the factories that now bake CPUs for us can continue doing so for a long, long time - and if our Asian overlords want they can completely deny Intel and Motorola access to their own factories (because, as they say, possession is 90% of the law.) Where would you move the production to? There are no spare semiconductor fabs, and there are no suitable low-cost 3rd world countries left to move into.
What then those suddenly fabless semiconductor companies will do? Build a new factory in the USA? You'd need equipment that is made in Asia, and if it is not sold then it's just too bad. The proper answer to "what to do" is just this: to rebuild the indigenous high-tech industry from the bottom up. How long do you think this will take, considering that US has only brains, and very little in terms of hands? Decades, and you'd have to start with schools. By the time US has that done where do you think China and India will be, after having such a head start? They already have more engineers and scientists just because their population is so much larger, and in this business only the absolute number of scientists matters (as opposed to doctors, for example.) China just doesn't want to rock the boat yet, but they'd be fools if they don't consider hundreds of possible scenarios, short- and long-term.
IMO the ITAR restrictions are not always productive. They are like that restriction on export of strong encryption products with keys longer than $n. Can anyone honestly say that those regulations stopped any country in the world from using the latest encryption? There are probably more PhDs in China than americans on the whole planet :-) Surely one of those PhDs could just read a math paper and write 100 lines of simple code that AES, or DES, or any other strong cipher requires. The tables and the proof and the reference implementation are published, and in reality if Chinese government wanted PGP 2.x (which was not allowed for export) it could always ask one of its diplomats to download the code within the USA (and to lie when the server asks if you are a foreigner - the server has no way to check!) and bring it in on a floppy. Big deal, spies do worse things every day. And now a 10-minute download wrecked the whole sector of US software industry, and as a "bonus" you can be sure that Chinese use the clean PGP, without any backdoors or "forgotten" NSA keys.
Ineffective laws like that only prevent US companies from selling what the USA excels in, while not really impeding the progress of the other side. It could be even making things worse. For example, Iran has a lot of american airplanes (F-15 or something.) Pentagon put the stop on selling parts, and the Iran's airforce is suddenly in tatters. But imagine that Iran could not buy those F-15s back then. Iranians would then buy from China, from Russia, from France - or if all else fails they'd make their own, not as good but 10 times as cheap (and plentiful) and now all of a sudden you have no leverage, no spare parts to pull from the market, and no control whatsoever. How would that be better?
Inquiring minds do not have the need to know. Besides, any US startup can raise funds and buy almost any machine, even one that requires ten times its weight in paperwork to ship to Canada. See ITAR for details.
And to make things worse, ITAR stops many US companies from selling some high-tech stuff that China would love to have, but which has dual use or military use. And much of US export is military and high-tech stuff. The USA can't compete with China on rice, for example, or on metals; not even on cars - China flooded Russia with cheap cars, and if anyone wants something better then Honda and Toyota are just a ferry ride away, and always glad to sell.
I think the horse has not only left the barn, it's off the planet by now. What were those "government officials" thinking for last decades? And this process is not [easily] reversible - China has all the factories now, and rephrasing Mao, "Power comes out of the gates of the factory." This much we see already.
The airplane will not fall unless the pilot allows it to fall, due to lack of understanding of the situation, or because of inability to control the flight. But it surely can fall, one way or another, if you don't know which way you are going. In this thread we have even a firsthand account of flying a presumably GA airplane with no power. The trick is to remain in control, and larger commercial airplanes (which climb to 40K feet) are often all fly by wire. Smaller airplanes are closer to early designs of 1900's which required no electricity at all, except ignition for the engine. But even if you have barometric altimeter and a pneumatic airspeed meter, you probably still need a flashlight to see them (do they still paint the needles with phosphorescent paint these days? I used to have some that were painted such, they were bright enough to see in darkness.)
You are thinking that falling will cause low gravity and that can be sensed. However if you are also moving in a spiral the centrifugal force will compensate, and that's how you can fly into the ground without knowing it. Basically, all you can sense is the direction of the acceleration vector, but if it points downward it doesn't at all mean you are safe. If you follow the link above you will see that the spiral dive is specifically characterized by the acceleration (gravity + spin) vector being normal to the floor of the cabin, and the only hint you have (outside of the artificial horizon in front of your nose being tilted out of wazoo) is that this "gravity" may be somewhat stronger or weaker - which could also happen as you ascend or descend.
The app itself must exist. Take QNX, for example - it has a GUI, but no apps.
All in all, completely losing power is unacceptable, but in case you lose all your primary generators the airplane gets dark fast. I do not recall for how long the batteries ought to suffice, but your figure (30 min.) is close enough to what I said. Most of the battery's power will be spent on mechanically controlling the airplane.
Well, this thing is no different from an internet cafe, and we know that those are popular enough. It's probably because most users of an internet cafe do not care about their email passwords. Business users are typically issued company notebooks which connect to the company's mail server through a VPN. But if a company permits direct access to its internal webmail then it assumes the risk of passwords being stolen one way or another, since you type them into the browser. Any keylogger, or a custom build of Firefox, would do the stealing easily, since the cafe owner is the sysadmin on all the computers. But all things considered, I would trust a large airline more than a shady Internet cafe owner.
So you don't have the answers.
It's quite possible that there are no executables on the system other than the browser and the OpenOffice. If the USB drive is mounted with '-o noexec' then you can't run your own apps either, and you can't execute anything from your $HOME either (can be also mounted noexec.)
I use embedded Linux at work, and the hardware that we run it on has a jumper which allows to electrically prevent any and all writes into the Flash. If you want to upgrade the software, press a button inside. No software can compensate for a WP# pin on Flash being tied to the ground; you'd need to do the iPhone-style hack with a soldering iron, and I don't see this as likely during a flight :-)
Also many airplanes have an auxiliary power plant, as you say placed at the tail, it is usually needed to provide fast moving air to spin the main engines, but can produce electricity as well. It is started by an electrical motor, which is powered from the truck on the ground. Batteries may be used, but only as an emergency measure.
Also some airplanes have a small external generator which can be used in an emergency. If you lose lots of power from engines at 40,000 ft you often have more altitude and more speed than you need (depends on where you plan to land,) and if so you can drop both and at the same time get some electrical power.
But in any case, an airplane is well provided with power, except in emergencies. A 50W here and there do not count, and besides the main cabin's lights and entertainment are the first to go if a power source fails.
Linux is the most suitable OS for that, since all you need to do is to reset the box, and everything is back to factory defaults. Even ThinkNIC (which I happen to have) used to do that, though it had a small (32 MB) persistent storage for bookmarks and such. You could also use a truly embedded kiosk system, but it wouldn't have many apps already written for it.
Of course it's possible, and that's how Linux-based embedded systems work. Your /home/$USERNAME can be created in RAM and deleted (recreated from a skeleton) after you log out (or the system restarted.) There is nothing else writeable on the whole box. This is necessary in embedded systems to prevent Flash wearing out, and to ensure reliability. Same needs here.
If the passenger insists to take the turn, it's his choice, and in fact his right to tell you where he wants you to go. Your only recourse, as a cab driver, is to refuse to go there, and I do not know what legal rights the passenger then gets against you (since you dumped him in the middle of nowhere.)
But on a larger scale of things, your comparison to a programmer is not exactly fitting. A programmer creates a product - a permanent entity that will be around for a while, will be used by thousands of users, will be modified and improved by tens of programmers in the future. You want to put some work into a permanent construction. In addition to that, a programmer is specifically selected and hired to use his personal skills and knowledge to create a better product. If a programmer doesn't do that but instead goes along with a poorer solution he is not doing his job.
On the other hand, a trip from point A to point B is a one-time service, offered to one person, and after the service is complete there will be no direct trace of it left in history. You, as a taxi driver, are hired by the cab company based on your decent knowledge of the city and on your reasonable ability to drive. However your duty to the company (and to yourself, if you rent the cab) is to earn money. Your duty to the passenger is to offer your best judgment on how to get from here to there under current conditions. This duty may conflict with the financial one, and you can find taxi drivers who pick one over another, it's a personal choice. However if the passenger refuses your offer of route planning, even despite you advising him to not do that, then you are in the clear, morally. You did your job, you offered a better way, and you were refused. Very well, then let the fool tell you where to go - there is plenty of gas in the tank and you are familiar with the area for hundreds of miles around.
If I were driving a cab and you were a passenger, I'd welcome your guidance as long as you pay for the whole route. I know that I can't possibly earn less than a fair fare (such as the fastest/cheapest route), and I also know that most passengers are not local and will not come up with an optimal route. As matter of fact, many would just become lost, on a one-way street, or on a highway, moving very fast *somewhere*, next exit ten miles :-) There are plenty of places where you can follow the map, approach an intersection and see the desired road 30 feet above you, with no ramp to get there :-) However you put it, passenger's guidance can only make me richer.
And what do you do when the passenger explains that s/he is with law enforcement or IRS?
Or what do you do if the passenger agrees to off the meter fixed fare, and upon arrival chooses to pay what the meter shows (zero) ?
Yes, GPS is a passive system. However the receiver can easily record the position as often as every second, thus providing the complete route (and the speed) to whoever reviews the data later. The data can be either stored on a dirt cheap SD/MMC flash card, or just wirelessly transmitted when the taxicab returns to the garage.
In other words, you refuse to be a photographer :-)
In terms of lazy write, the controller probably wants to write all the cache onto the platters in case of power failure. This takes time, and you can write only so much until the power becomes too low to write.
In terms of read caching, it would require a DDR2 design on the controller board, and those controllers aren't high-tech enough for that. DDR2 is very difficult to connect, requires picky controllers, and consumes a lot of power. Why would anyone do that if the computer already has tons of RAM, and you can add more if you want?
Besides, caching in main RAM is better because the CPU knows what is where and can control caching. A RAM cache is a virtual page away from being physically accessible to your needy process; nanoseconds, in other words. But if you do your caching in the HDD controller then you are doing it after squeezing through the [S]ATA bottleneck, and the controller only has bits and pieces of the information. For example, the main CPU can recognize that you are reading a file, and can pre-read subsequent sectors anticipating your future request. The HDD controller doesn't know what a file is, and so it can only react to events.
This quote simply proves that Martian microbes probably are not "terrestrial kinds of cells." Wasn't it obvious to begin with?
On this planet, however, "good enough" is good enough for any alcohol drinker (or a drug user.) Getting an affordable drug when one needs it surely beats robbing a store and potentially getting killed. Drug users may be reckless but still not suicidal. Some addicts would be glad to stop, but their bodies changed to require the drug, and if forced to abstain they feel extreme pain. Under the threat of such pain an addict will rob and kill; however given an option I believe many would accept the government-sponsored drug, the pain will be gone just as well as when using a street drug.