Now that Microsoft has admitted that Linux is one of their primary threats, we can attempt to deduce their opinion of the SCO-IBM lawsuit.
Fact 0: Microsoft could buy SCO for a single day's worth of revenue. Fact 1: SCO claims that without their permission, nobody can use Linux. Fact 2: Microsoft knows that Linux is one of their biggest threat to profits. Fact 3: Microsoft has not bought SCO.
The natural conclusion of these facts is that Microsoft feels SCO's claim has no merit, and will be struck down in court. Rather than buying SCO and expediting the court-case so that Linux can be quashed immediately, they've chosen to sit back and allow the unsettled allegation to stir up uncertainty and dissuade potential Linux adopters.
Note: this doesn't mean that Microsoft considers it impossible for SCO to win the case- only that they don't think there's a high probabilty of victory. They benefit from allowing the FUD to continue for as long as possible before the dice are rolled in court. In fact, there's another way they benefit from holding off the verdict: if some companies deploy Linux and then have their operations interrupted by C&D orders in the wake of an SCO victory, it will discourage future corporate adoption of all kinds of Open Source software.
What will happen with Palladium is that the hardware will refuse to load untrusted software. If you were somehow able to recompile Windows, it would be an untrusted binary (unless you made zero changes from Microsoft's source code).
So prehaps you could recompile it, but there'd be no hardware to run it on. If motherboard DRM catches on, we can expect (and fear) 100% of inexpensive PCs to use it. Only $2000+ "engineering" systems will be able to load arbitrary OSes.
Of course, recompiling Windows is impossible for most people anyway... however, motherboard DRM could also be used with Linux! This is a real threat to the Open Source philosophy- a loophole in the licenses, if you will. If vendors include (modified) Linux with their hardware, they have to give customers the source. But if the hardware only loads signed binaries, the customers have little use for the source (unless they want to build their own hardware)
They also do it because heavy users of Microsoft Windows(r) software except to click through a license as they install. If they don't see 10 pages of lawyer-speak to ignore, they get confused and think they've missed a step.
Seriously, the kits used to make installation packages for Windows software (like InstallShield, or Winzip installer) create the Accept/Decline license box automatically, and just ask the developer to paste in his license.
I think a keen trick would be to put the GPL in a click-to-agree box, but with multiple buttons: "Yes", "No", "Maybe", "I'll read it later", "I'm under 18, or am drunk, or am otherwise not allowed to enter a contract". Of course, any one of those buttons would proceed with the installation as normal...
But a library is different, because the FSF considers any form of linkage to be derivation.
What is the basis for your claim?
The GPL is viral because its common interpretation, including the interpretation of the FSF, goes beyond copyright law and effects regulations on non-derivative works.
The GPL says that the definition of a derivative work is exactly what copyright law does. Look at GPL section 0:... a "work based on the Program" means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it...
Now, this does bring up an interesting point. In US law, short-term violations of copyright for purposes of normal use of a work are legal (this enables you, for example, to copy a program from your hard drive into RAM for execution without needing a separate license from the publisher. One reason you're free to ignore shrinkwrap EULAs). That means that dynamically linking a library is not a copyright violation, although it does temporarily create a derived work.
In Britain, however, "incidental copying as necessary to normally access a work" is not protected by fair use. There, to dynamically link code together would be a copyright violation, and you'd need permission from each author to accomplish it.
Ironically, if US law changes to make EULAs binding, then the GPL (version 3) would be able to firmly prevent people from dynamically linking to libraries. Without such a change, it'll have no way to "reach beyond copyright law" (as you put it), since it's power is based only on copyright.
My application is not a recasting, transformation or adaptation of the GPL library. It contains no editorial revisions, annotations, elaborations or other modifications to the library.
Oh no? No "modifications"?
The law you quote is wordy with many examples of common types of derivative works, but it all comes down to one phrase: "or other modifications".
If you copy someone else's work (as in by linking a library), and your product is non-identical to the original (you've changed it in any way, such as adding a "main()" function which calls their library routines), then it is a derivative work!
If you made no modifications, then it's just a 100% copy. In either case, you can't publish it without permission from the original copyright holder.
There is no way you can claim that copying substantial persons of someone's work into your own (which is what static linking does) does not give them some control of your product.
The GPL is a tool to make YOUR software free, not someone else's software. It's not designed to "destroy copyright from within". If you think so, you are a retarded assmonkey who needs to shut up and read what Richard Stallman has to say about the goals of the FSF and the GPL.
It is intended to make all software free. That is the goal of the FSF and the GPL. Eric Raymond of the Open Source Initiative takes a more moderate view, and explains that some categories of software should not be free... but that's not the GPL's goal.
By use of the GPL, RMS hoped to make all software Free by providing quality GPL code as an incentive for new programs to be GPLed too. Otherwise, they'd be missing out on many cheaply available features. The idea was that GPL use would snowball- at some point, when the preponderance of useful libraries are GPLed, then creating a non-GPL program that can't use them would be an exercise in futile money-wasting.
RMS doesn't like the LGPL for this reason- he does not want people to be able to link to Free libraries without Freeing up their code. That's why he renamed it from "Library GPL" to "Lesser GPL"- to emphasize disapproval.
The word "viral" to describe the GPL is of course incorrect- unless one also agrees that the copyright system is viral itself (according to the legal definition of a derived work, which are "infected" with the copyright of the previous author).
The LGPL says 'the work' (i.e. the library licensed under the LGPL), not 'your work'.
Yes, it says "the work". And it means your new program, not "the library licensed under the GPL". In the license text, that library is called "the Library".
Look at it in context: 6. As an exception to the Sections above, you may also combine or link a "work that uses the Library" with the Library to produce a work containing portions of the Library, and distribute that work under terms of your choice, provided that the terms permit modification of the work for the customer's own use and reverse engineering for debugging such modifications.
I have highlighted where "the work" occurs. We can clearly see that "work" refers to the new program using the "Library".
Re:CDBurners not the end for high-capacity Zip dri
on
DVD Burner Round-up
·
· Score: 1
CDBurners were very much the end for ZIP drives.
After CDBurners became common, Zip drives were still often used by people who needed to carry around large files (from MS Office or Adobe programs) and edit them on multiple computers. Offices that had poor ethernet connectivity were the last bastion of the Zip disk. But today, USB keychains are conquering that territory.
It does seem that CD-RW drives could've invaded that space, if using software that convincingly emulated a normal read-write filesystem. But apparently such software never reached the level of penetration/reliability needed to succeed.
Adobe was able to sue Macromedia because Macromedia copied the implementation verbatim.
They did not.
"Verbatim" copying of a portion of a computer program would be to look at the source code and make an exact duplicate of every relevant word (aka "token"). Obviously they didn't do this, it would've been copyright violation. There's no need for a patent to protect a really specific implementation- you could always sue for a slam-dunk copyright charge instead.
That's one of the reasons software patents shouldn't exist.
There will be a time in the not-too-distant future when portable devices will contain GPS by default, and automatically switch locations and users on the fly. Apple is doing the right thing here, formalizing their design via the patent system.
Apple is doing the wrong thing. It makes sense businesswise, prehaps, but they are perpetuating the abuse of our patent system. The idea of a portable computer automatically reconfiguring itself depending on location is an obvious one. It merely requires the maturation of two separate technologies (small GPS, and small powerful computers) to converge on the inevitable result. There is no innovative idea in there worthy of patent protection.
with the expectation that you can optimise the production pipeline, so that you can make a profit on the boxes sold later.
The pipeline optimizations usually don't switch the consoles to a profitable item- those savings just keep pace with reducing the price consumers pay. The buyers won't keep paying the original price for 6+ month old hardware.
Ask your favorite compiler vendor to implement whatever additional feature you like.
That request should really be directed to a library vendor. It's already possible (even easy) to write bound-checked array classes in C++.
A guiding principle of C++ development was not to force features into the language itself if they could be implemented in terms of the language. (This is why C++ has std::string, unlike Java's native String which behaves unlike any class a library could provide)
If a compiler author accedes to this request, and you use it in your programs, then what you write can no longer be called "C++ code". That's a bad portent for future portability. Write to an optional C++ library, and you can at least expect to move the program to another compiler someday.
For those who don't know, Murphy's Law was first observed during the assembly of the spinning chairs to which pilots can be strapped to prepare them for the experience of pulling 9 gravities in jet-powered dogfights.
The RPM indicator which the operator watches as he controls the device could be plugged in either correctly, or upside down. When the latter happened, the operator could inadvertendly accelerate when attempting to deactivate the machine.
This may not have been comfortable for the pilots.
It is slightly tricky to package, but NVidia has accomplished it well. You need to split the driver into it's binary core, but also ship the functions which call that core as source code, so your installer routine can recompile them to work with whatever Linux it finds on the target system.
One standard complaint about Linux is that it makes no effort to support compatibility with binary kernel drivers from different (older/newer) versions. Thus NVidia had to go through this trickiness.
a flock of people go running to improve the drivers.
"Improving the drivers", in this case, is low-hanging fruit. The current drivers are only available for a small number of Linux kernel revisions. Anyone with the source code could trivially recompile them for newer/older Linux versions.
The drivers provided by NVidia corporation aren't "Linux drivers" either. They are "x86/IA64/AMD64 Linux drivers". If they were genuine Linux drivers, you could use them on any platform supporting both Linux and AGP cards- such as a Macintosh/powerpc.
(However, the unaccelerated open-source drivers for NVidia cards can work on non-x86 Linux)
"Apocalypse Now" was derived from multiple sources. The Korean and Vietnam wars gave Coppola an opportunity to transplant an old tale into a modern milieu.
which was originally published in 1898.
1901.
and is in fact mostly identical to
The book had Kurtz working for the UK, not US, and in Africa, not Southeast Asia.
In fact, the movie Kurtz was an amalgamation of not only the Conrad character, but also of real-life US guerrilla leaders: Rheault, Rexer, and Poshepny. Poshepny was similar to the movie character in that he had most firmly established himself as a king-like figure. Rheault was the one whose actions actually brought him Army prosecution- his case was well known to the screenwriter.
The natures of those mens' deeds were widespread rumor in the 60s, although names were not accurately attached to the tales that filtered to the public.
Unarmed civilians *are* the most helpless victims possible. Including but not limited to full grown adults.
It's not about the targeting. Adults are slightly easier to hit than children, although slightly more likely to survive a small caliber round. Children are more vulnerable because they can't run away as fast, they're less mentally focused when deciding to run away (choosing wrong direction), and they're unlikely to crouch quietly if they find a good hiding spot. They're short, so the front of a crowd won't shield the rest of it (not like bullets won't penetrate some anyhow)
Plus, there's the matter of their physical in-offensiveness. If a murderer wades into a crowd of victims, or turns a corner and smacks facefirst into a disoriented civilian, he faces a risk of his gun being pulled away. That's particularly important when reloading. The more adults present, the more likely one of them can overpower the attacker. Children pose no such risk.
There's no sport in it.
If a potential mass-murderer wants "sport", he should setup with a sniper rifle outside Fort Knox. That's sporting!
Police officers are woefully unprepared for an attack like this, and as such are more likely to get themselves killed than anything else.
There's little reason airport guards would need that much firepower, except to respond to a major attack that is well organized and well armed. (Which requires a different caliber of personnel than the common terrorist groups can recruit just now... but no matter). An enemy like that will attack with surprise, so the first guards they meet will be killed regardless of their armament. The airport should leave their normal guards with light armament (pistol + nightstick), and station any personnel with heavier weapons in concealed rooms. They should be able to monitor the terminal floor by camera, and reach it quickly if an attack occurs, but their position should be hidden from civilian users of the airport (some of whom may be terrorist scouts)
Striding back and forth with an SMG all day might make tourists feel like they've got powerful protectors, but it just ensures the attackers will be careful to kill that guard in the first volley.
Since then, we've been working on ways Al Queda could increase the body count. I figure hitting a packed football stadium (no roof, of course)would easily result in >3,000 deaths, when you think about how big a 747 is, and how far burning jet fuel will spread upon impact against what is essentially a flat surface....
Actually that's not quite the best way. A stadium isn't really a flat surface- the seats are slanted. You'll just get one quadrant of spectators (15-25% of the total), and the rest of the fuel will just pool down onto the field.
Stadia are excellent as a terrorism target, however. The best way to attack one is less specatular than a plane crash, but more lethal. I'd suggest a two-pronged conventional and chembio attack. First send 3-5 agents to release a device spraying a deadly substance outside each major exit. (My fondness is for sarin*, but anthrax may be better in some regards. It's safer to handle, and if weaponized properly you need less mass of contaminant. But both are easy to get). As soon as they release (and move away from the scene), fire off a mortar into the stands. The mortar can be 3 km away in a vacant lot- the operator just needs to put 4-5 shells into various points of the stadium, and then he can flee as well.
What this will accomplish is to start a relentless, semi-panicked evacuation from the stadium. The people fleeing the explosion will be distracted, confused, and pressed forward by the crowd behind them. So even if they detect the toxic cloud (it will become obvious in time, at least by the pile of fresh corpses), they won't be able to turn back from passing through it.
With this approach, not only could 6 terrorists kill 20,000 American victims, but they'd have a moderate chance of escaping with their lives.
So- if you're attending the Superbowl and small bombs start falling from the sky, don't rush outside immediately. You might be funneled to killing zones. Climb a post to get free of the mob, then find a sturdy indoor bathroom to hole up in, and wait for further developments to evaluate your next move.
*Statisitcally, 1 gram of sarin on the palm of your hand is more lethal than a single bullet to the heart. If he were transported to an ICU inside 10 seconds, the doctors could possibly save the gunshot victim (the wound is evident and treament is rehearsed), but the poison victim wouldn't even be diagnosed until long dead.
Multiple-murders are a tiny, tiny proportion of all "violence". Poverty is positively correlated with violence in general, but wealth contributes to serial or mass attacks.
Are you referring to the popular "Apocalypse Now" movie? Or is that a reference to the real Anthony Poshepny, who just passed away? (I'd be suprised if many US citizens have heard of him- even his obituary didn't circulate much in the American press)
the easiest way to take out the most people by myself with a limited number of weapons.
You do all that work, and then don't publish your results? (For academic interest only, of course)
My view: For maximum kill-count, target selection is somewhat more important than weapons. You want the most helpless victims possible. Kindergarden is when children are first congregated in large numbers, so try for that.
And then as to the "virtues" of an AK-47... it's not the best choice. The rounds are too large and too noisy. You'll be firing at small, soft targets at ranges much less than 100m (often less than 5), so the power of a larger cartridge is wasted. A bullet which passes through a target is actually less deadly than one which stops inside (preferably after rotating in passage). It's also so loud that potential targets will be scattering away before you can get to most of them.
Optimally, you'd want a silenced submachine gun. An Uzi, or something similar but more reliable. Look at what the police carry in US airports today. Advantages of silenced SMG over an assault rifle: quieter, ammo is smaller so you can carry more shots, shorter barrel is easier to aim at close quarters (when there are wailing mothers hanging off of you, for instance), shorter barrel is more concealable as you make your approach.
If you think the shooting just targeted bears and bison, then you played the later, expurgiated version of the game.
The original Oregon Trail featured monochrome stick-figures of Indian Braves for you to defeat.
It was as aggressively violent as anything of the era, and more realistically than most. (The setting was real world, unlike other games except the originalCastle Wolfenstein... and the Nazi opponents in that game were more definitively evil than the hapless Native American resistance)
Slashdot Mirror
Now that Microsoft has admitted that Linux is one of their primary threats, we can attempt to deduce their opinion of the SCO-IBM lawsuit.
Fact 0: Microsoft could buy SCO for a single day's worth of revenue.
Fact 1: SCO claims that without their permission, nobody can use Linux.
Fact 2: Microsoft knows that Linux is one of their biggest threat to profits.
Fact 3: Microsoft has not bought SCO.
The natural conclusion of these facts is that Microsoft feels SCO's claim has no merit, and will be struck down in court. Rather than buying SCO and expediting the court-case so that Linux can be quashed immediately, they've chosen to sit back and allow the unsettled allegation to stir up uncertainty and dissuade potential Linux adopters.
Note: this doesn't mean that Microsoft considers it impossible for SCO to win the case- only that they don't think there's a high probabilty of victory. They benefit from allowing the FUD to continue for as long as possible before the dice are rolled in court. In fact, there's another way they benefit from holding off the verdict: if some companies deploy Linux and then have their operations interrupted by C&D orders in the wake of an SCO victory, it will discourage future corporate adoption of all kinds of Open Source software.
You're both kind of right.
What will happen with Palladium is that the hardware will refuse to load untrusted software. If you were somehow able to recompile Windows, it would be an untrusted binary (unless you made zero changes from Microsoft's source code).
So prehaps you could recompile it, but there'd be no hardware to run it on. If motherboard DRM catches on, we can expect (and fear) 100% of inexpensive PCs to use it. Only $2000+ "engineering" systems will be able to load arbitrary OSes.
Of course, recompiling Windows is impossible for most people anyway... however, motherboard DRM could also be used with Linux! This is a real threat to the Open Source philosophy- a loophole in the licenses, if you will. If vendors include (modified) Linux with their hardware, they have to give customers the source. But if the hardware only loads signed binaries, the customers have little use for the source (unless they want to build their own hardware)
They also do it because heavy users of Microsoft Windows(r) software except to click through a license as they install. If they don't see 10 pages of lawyer-speak to ignore, they get confused and think they've missed a step.
Seriously, the kits used to make installation packages for Windows software (like InstallShield, or Winzip installer) create the Accept/Decline license box automatically, and just ask the developer to paste in his license.
I think a keen trick would be to put the GPL in a click-to-agree box, but with multiple buttons: "Yes", "No", "Maybe", "I'll read it later", "I'm under 18, or am drunk, or am otherwise not allowed to enter a contract". Of course, any one of those buttons would proceed with the installation as normal...
But a library is different, because the FSF considers any form of linkage to be derivation.
... a "work based on the Program" means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it ...
What is the basis for your claim?
The GPL is viral because its common interpretation, including the interpretation of the FSF, goes beyond copyright law and effects regulations on non-derivative works.
The GPL says that the definition of a derivative work is exactly what copyright law does. Look at GPL section 0:
Now, this does bring up an interesting point.
In US law, short-term violations of copyright for purposes of normal use of a work are legal (this enables you, for example, to copy a program from your hard drive into RAM for execution without needing a separate license from the publisher. One reason you're free to ignore shrinkwrap EULAs). That means that dynamically linking a library is not a copyright violation, although it does temporarily create a derived work.
In Britain, however, "incidental copying as necessary to normally access a work" is not protected by fair use. There, to dynamically link code together would be a copyright violation, and you'd need permission from each author to accomplish it.
Ironically, if US law changes to make EULAs binding, then the GPL (version 3) would be able to firmly prevent people from dynamically linking to libraries. Without such a change, it'll have no way to "reach beyond copyright law" (as you put it), since it's power is based only on copyright.
My application is not a recasting, transformation or adaptation of the GPL library. It contains no editorial revisions, annotations, elaborations or other modifications to the library.
Oh no? No "modifications"?
The law you quote is wordy with many examples of common types of derivative works, but it all comes down to one phrase: "or other modifications".
If you copy someone else's work (as in by linking a library), and your product is non-identical to the original (you've changed it in any way, such as adding a "main()" function which calls their library routines), then it is a derivative work!
If you made no modifications, then it's just a 100% copy. In either case, you can't publish it without permission from the original copyright holder.
There is no way you can claim that copying substantial persons of someone's work into your own (which is what static linking does) does not give them some control of your product.
There's no ambiguity. "The work" means the new work using the library. The library itself is called "the Library".
The poster you are responding to is completely wrong.
The GPL is a tool to make YOUR software free, not someone else's software. It's not designed to "destroy copyright from within". If you think so, you are a retarded assmonkey who needs to shut up and read what Richard Stallman has to say about the goals of the FSF and the GPL.
It is intended to make all software free. That is the goal of the FSF and the GPL. Eric Raymond of the Open Source Initiative takes a more moderate view, and explains that some categories of software should not be free... but that's not the GPL's goal.
By use of the GPL, RMS hoped to make all software Free by providing quality GPL code as an incentive for new programs to be GPLed too. Otherwise, they'd be missing out on many cheaply available features. The idea was that GPL use would snowball- at some point, when the preponderance of useful libraries are GPLed, then creating a non-GPL program that can't use them would be an exercise in futile money-wasting.
RMS doesn't like the LGPL for this reason- he does not want people to be able to link to Free libraries without Freeing up their code. That's why he renamed it from "Library GPL" to "Lesser GPL"- to emphasize disapproval.
The word "viral" to describe the GPL is of course incorrect- unless one also agrees that the copyright system is viral itself (according to the legal definition of a derived work, which are "infected" with the copyright of the previous author).
The LGPL says 'the work' (i.e. the library licensed under the LGPL), not 'your work'.
Yes, it says "the work". And it means your new program, not "the library licensed under the GPL". In the license text, that library is called "the Library".
Look at it in context:
6. As an exception to the Sections above, you may also combine or link a "work that uses the Library" with the Library to produce a work containing portions of the Library, and distribute that work under terms of your choice, provided that the terms permit modification of the work for the customer's own use and reverse engineering for debugging such modifications.
I have highlighted where "the work" occurs. We can clearly see that "work" refers to the new program using the "Library".
CDBurners were very much the end for ZIP drives.
After CDBurners became common, Zip drives were still often used by people who needed to carry around large files (from MS Office or Adobe programs) and edit them on multiple computers. Offices that had poor ethernet connectivity were the last bastion of the Zip disk. But today, USB keychains are conquering that territory.
It does seem that CD-RW drives could've invaded that space, if using software that convincingly emulated a normal read-write filesystem. But apparently such software never reached the level of penetration/reliability needed to succeed.
Adobe was able to sue Macromedia because Macromedia copied the implementation verbatim.
They did not.
"Verbatim" copying of a portion of a computer program would be to look at the source code and make an exact duplicate of every relevant word (aka "token"). Obviously they didn't do this, it would've been copyright violation. There's no need for a patent to protect a really specific implementation- you could always sue for a slam-dunk copyright charge instead.
That's one of the reasons software patents shouldn't exist.
There will be a time in the not-too-distant future when portable devices will contain GPS by default, and automatically switch locations and users on the fly. Apple is doing the right thing here, formalizing their design via the patent system.
Apple is doing the wrong thing. It makes sense businesswise, prehaps, but they are perpetuating the abuse of our patent system. The idea of a portable computer automatically reconfiguring itself depending on location is an obvious one. It merely requires the maturation of two separate technologies (small GPS, and small powerful computers) to converge on the inevitable result. There is no innovative idea in there worthy of patent protection.
with the expectation that you can optimise the production pipeline, so that you can make a profit on the boxes sold later.
The pipeline optimizations usually don't switch the consoles to a profitable item- those savings just keep pace with reducing the price consumers pay. The buyers won't keep paying the original price for 6+ month old hardware.
Ask your favorite compiler vendor to implement whatever additional feature you like.
That request should really be directed to a library vendor. It's already possible (even easy) to write bound-checked array classes in C++.
A guiding principle of C++ development was not to force features into the language itself if they could be implemented in terms of the language. (This is why C++ has std::string, unlike Java's native String which behaves unlike any class a library could provide)
If a compiler author accedes to this request, and you use it in your programs, then what you write can no longer be called "C++ code". That's a bad portent for future portability. Write to an optional C++ library, and you can at least expect to move the program to another compiler someday.
I think C++ needs stuff *removed* more than it needs anything added.
Funny, the Java creators are *adding* stuff they learned from C++.
Java 1.4.1 just got templates, for instance.
For those who don't know, Murphy's Law was first observed during the assembly of the spinning chairs to which pilots can be strapped to prepare them for the experience of pulling 9 gravities in jet-powered dogfights.
The RPM indicator which the operator watches as he controls the device could be plugged in either correctly, or upside down. When the latter happened, the operator could inadvertendly accelerate when attempting to deactivate the machine.
This may not have been comfortable for the pilots.
It shouldn't be that hard to repackage.
It is slightly tricky to package, but NVidia has accomplished it well. You need to split the driver into it's binary core, but also ship the functions which call that core as source code, so your installer routine can recompile them to work with whatever Linux it finds on the target system.
One standard complaint about Linux is that it makes no effort to support compatibility with binary kernel drivers from different (older/newer) versions. Thus NVidia had to go through this trickiness.
a flock of people go running to improve the drivers.
"Improving the drivers", in this case, is low-hanging fruit. The current drivers are only available for a small number of Linux kernel revisions. Anyone with the source code could trivially recompile them for newer/older Linux versions.
The drivers provided by NVidia corporation aren't "Linux drivers" either. They are "x86/IA64/AMD64 Linux drivers". If they were genuine Linux drivers, you could use them on any platform supporting both Linux and AGP cards- such as a Macintosh/powerpc.
(However, the unaccelerated open-source drivers for NVidia cards can work on non-x86 Linux)
"Apocalypse Now" was derived from multiple sources. The Korean and Vietnam wars gave Coppola an opportunity to transplant an old tale into a modern milieu.
which was originally published in 1898.
1901.
and is in fact mostly identical to
The book had Kurtz working for the UK, not US, and in Africa, not Southeast Asia.
In fact, the movie Kurtz was an amalgamation of not only the Conrad character, but also of real-life US guerrilla leaders: Rheault, Rexer, and Poshepny. Poshepny was similar to the movie character in that he had most firmly established himself as a king-like figure. Rheault was the one whose actions actually brought him Army prosecution- his case was well known to the screenwriter.
The natures of those mens' deeds were widespread rumor in the 60s, although names were not accurately attached to the tales that filtered to the public.
Unarmed civilians *are* the most helpless victims possible. Including but not limited to full grown adults.
It's not about the targeting. Adults are slightly easier to hit than children, although slightly more likely to survive a small caliber round. Children are more vulnerable because they can't run away as fast, they're less mentally focused when deciding to run away (choosing wrong direction), and they're unlikely to crouch quietly if they find a good hiding spot. They're short, so the front of a crowd won't shield the rest of it (not like bullets won't penetrate some anyhow)
Plus, there's the matter of their physical in-offensiveness. If a murderer wades into a crowd of victims, or turns a corner and smacks facefirst into a disoriented civilian, he faces a risk of his gun being pulled away. That's particularly important when reloading. The more adults present, the more likely one of them can overpower the attacker. Children pose no such risk.
There's no sport in it.
If a potential mass-murderer wants "sport", he should setup with a sniper rifle outside Fort Knox. That's sporting!
Police officers are woefully unprepared for an attack like this, and as such are more likely to get themselves killed than anything else.
There's little reason airport guards would need that much firepower, except to respond to a major attack that is well organized and well armed. (Which requires a different caliber of personnel than the common terrorist groups can recruit just now... but no matter). An enemy like that will attack with surprise, so the first guards they meet will be killed regardless of their armament. The airport should leave their normal guards with light armament (pistol + nightstick), and station any personnel with heavier weapons in concealed rooms. They should be able to monitor the terminal floor by camera, and reach it quickly if an attack occurs, but their position should be hidden from civilian users of the airport (some of whom may be terrorist scouts)
Striding back and forth with an SMG all day might make tourists feel like they've got powerful protectors, but it just ensures the attackers will be careful to kill that guard in the first volley.
Since then, we've been working on ways Al Queda could increase the body count. I figure hitting a packed football stadium (no roof, of course)would easily result in >3,000 deaths, when you think about how big a 747 is, and how far burning jet fuel will spread upon impact against what is essentially a flat surface. ...
Actually that's not quite the best way. A stadium isn't really a flat surface- the seats are slanted. You'll just get one quadrant of spectators (15-25% of the total), and the rest of the fuel will just pool down onto the field.
Stadia are excellent as a terrorism target, however. The best way to attack one is less specatular than a plane crash, but more lethal. I'd suggest a two-pronged conventional and chembio attack. First send 3-5 agents to release a device spraying a deadly substance outside each major exit. (My fondness is for sarin*, but anthrax may be better in some regards. It's safer to handle, and if weaponized properly you need less mass of contaminant. But both are easy to get). As soon as they release (and move away from the scene), fire off a mortar into the stands. The mortar can be 3 km away in a vacant lot- the operator just needs to put 4-5 shells into various points of the stadium, and then he can flee as well.
What this will accomplish is to start a relentless, semi-panicked evacuation from the stadium. The people fleeing the explosion will be distracted, confused, and pressed forward by the crowd behind them. So even if they detect the toxic cloud (it will become obvious in time, at least by the pile of fresh corpses), they won't be able to turn back from passing through it.
With this approach, not only could 6 terrorists kill 20,000 American victims, but they'd have a moderate chance of escaping with their lives.
So- if you're attending the Superbowl and small bombs start falling from the sky, don't rush outside immediately. You might be funneled to killing zones. Climb a post to get free of the mob, then find a sturdy indoor bathroom to hole up in, and wait for further developments to evaluate your next move.
*Statisitcally, 1 gram of sarin on the palm of your hand is more lethal than a single bullet to the heart. If he were transported to an ICU inside 10 seconds, the doctors could possibly save the gunshot victim (the wound is evident and treament is rehearsed), but the poison victim wouldn't even be diagnosed until long dead.
Multiple-murders are a tiny, tiny proportion of all "violence". Poverty is positively correlated with violence in general, but wealth contributes to serial or mass attacks.
They dont cluster bomb the civilians in the first place. Clever, huh?
Over 50 years, the US has advanced from atomic bombs killing 20,000 to cluster bombs killing 400. Give them a little credit!
Are you referring to the popular "Apocalypse Now" movie? Or is that a reference to the real Anthony Poshepny, who just passed away? (I'd be suprised if many US citizens have heard of him- even his obituary didn't circulate much in the American press)
the easiest way to take out the most people by myself with a limited number of weapons.
You do all that work, and then don't publish your results? (For academic interest only, of course)
My view: For maximum kill-count, target selection is somewhat more important than weapons. You want the most helpless victims possible. Kindergarden is when children are first congregated in large numbers, so try for that.
And then as to the "virtues" of an AK-47... it's not the best choice. The rounds are too large and too noisy. You'll be firing at small, soft targets at ranges much less than 100m (often less than 5), so the power of a larger cartridge is wasted. A bullet which passes through a target is actually less deadly than one which stops inside (preferably after rotating in passage). It's also so loud that potential targets will be scattering away before you can get to most of them.
Optimally, you'd want a silenced submachine gun. An Uzi, or something similar but more reliable. Look at what the police carry in US airports today. Advantages of silenced SMG over an assault rifle: quieter, ammo is smaller so you can carry more shots, shorter barrel is easier to aim at close quarters (when there are wailing mothers hanging off of you, for instance), shorter barrel is more concealable as you make your approach.
If you think the shooting just targeted bears and bison, then you played the later, expurgiated version of the game.
The original Oregon Trail featured monochrome stick-figures of Indian Braves for you to defeat.
It was as aggressively violent as anything of the era, and more realistically than most. (The setting was real world, unlike other games except the originalCastle Wolfenstein... and the Nazi opponents in that game were more definitively evil than the hapless Native American resistance)