1) First, this is a drop in the bucket. CERN is one of only a few entities using this kind of power. Take a look at the 55,000 home figure. Say these are your Asian homes, and you end up with 550,000 homes that could be powered by the energy used at CERN. Given the masses of humanity, that adds up to NOTHING. Certainly not enough to justify the potentially disastrous consequences to humanity that could result if we stop this kind of research. The average utility bill at Fermilab (US counterpart to CERN) is 1.5 million dollars. Over the course of a year, thats $18 million. In comparison, the USAID annual budget is $6 billion. The Federal Highway program's budget on the other hand, is $26 billion. Yes, people would rather fix potholes than develop countries. Such is the world we live in. But its not like research money is coming out of the mouths of the poor. The real problem isn't that the world spends too much on research (if only). The real problem is that people have no clue about anything outside the bounds of their tiny meaningless existance. For example, most people, when survayed, said that they thought the US spends too much on foreign aid. When asked how much they thought was appropriate, they said 5% of the budget. The real figure is one-tenth of that number. There are dozens of things you can do to make up the cost of this research, including improving distribution methods, bringing down cultural barries that make access to healthcare inefficient, reforming patent conventions that jack up the cost of medications, etc. In the end, science is not the thing to sacrifice for humanity. BTW, I grew up learning stuff about international development from my dad (that's his line of work), and I was born in Thailand and spent part of my childhood in Bangladesh. I *do* know what I'm talking about.
2) We're not talking about the "many marvelous inventions in the last, say, 20 years." We're talking about how physics has redefined the universe was we know it over the last 200 years. A large percentage of the modern economy owes its existance to quantum physics. The work at CERN is simply an extension of the very ancient search for knowledge about the structure of matter. Anytime you get a cat-scan or an X-ray, take medicine, use a computer, drive a car, watch TV, etc, you're directly benifeting from that research. Even those in inpovrished countries should thank this research for allowing scientists to use advanced imaging tehnologies to create things like TB vaccines that sells for dollars per dose. Going into the future, the only sure way to relieve poverty is to find more resources. It is not possible for the human population to keep growing, expanding, evolving, reaching towards a higher state of being, without more raw matter. So yes, that warp drive space ship WILL help the guy living in poverty, if you stop being so short-sighted. Giving a man a fish is not the only way to help him.
Yes. Because they advance humanity. It is impossible, with our current resources, to pull people out of poverty. It is work like this that will one day give us the chance to pull humanity out of the gutter. Back when some crazy's were fooling around with dead human bodies, nobody cared. Now, we call them doctors and thank them for delivering us from the horrors of disease. Besides, don't you think there are better uses for the power used by your computer than posting on/.?
I always love morons who criticize things when they aren't part of their intended audience. Unless you're a gamer, you have no reason to be at those sites. If you are, those sites are invaluable. Quick, what's faster, the Radeon 8500 or the GeForce4 MX? I don't really care, my gaming is limited to some RPGs and the occasional CounterStrike match. However when I wanted to buy my brother a new graphics card, you bet he was happy that I did my research at those sites first. I used to frequent those sites a lot, back when I did a lot of 3D programming and gaming. I honestly needed the performance, and those sights provided a lot of good guidence, as well as useful ways to get cheap hardware to do cool things (overclocking Celeron 300A's and running them in SMP for example, or joining the bridges on Athlon XPs to get them running in SMP). As for the bit about advertisers, that's bull. The internet keeps these sites honest. A GeForce 4 TI 4600 really is a lot faster than a Radeon 8500. If a site says otherwise because of advertising, a second opinion (or third or fourth) is just a click away. In the future, it would be wise to provide specific proof before criticizing people of lacking journalistic integrity.
Um, disk isn't a bottleneck on my system. I do lots of C++ programming, and everything gets cached in RAM after the first build. Thus, my bottleneck becomes gcc...err... CPU and memory bandwidth:) Same thing when I'm doing 3D rendering (which had better fit in RAM or else) or playing 3D games and whatnot. I thought moving to a 4200 RPM laptop hard drive was going to be bad, after my nice 7200 RPM IDE RAID. In truth, thanks to the Linux VM, I don't notice the difference after the first half hour of using the system. However, I did notice the big boost that came with moving from a PC100 memory system to a PC266, even just palying around with GUI wigets (resizing and whatnot).
If American engineers can't compete and they lose their jobs, well too bad, such are the costs of capitalism. And don't tell me these are real engineers with real families, because so are they, and they need the jobs more than we do.
Actually, if you had been keeping up, some guy had broken XFS up into nice tiny patches, while also converting it to use the generic I/O routines. In the end, there were four lines of changes to non-XFS code. There is an OS news article about this, with a link to the relavent lkml thread.
Example: IPsec. Not included in the standard kernel. In order to get it working, I'll have to patch, configure, and recompile kernels for half a dozen different machines. For handhelds running Linux, this will be even more of a chore. >>>>>> If you need IPsec, then you can certainly recompile your kernel. The thing is that external stuff like IPsec is entirely analagous to external stuff in Windows. For example, Windows prior to Win98 SE didn't come with any NAT capability. Installing an add-on like raspppoe took a good bit of work. IPsec isn't a part of the standard Linux installation. Thus, it is to be expected that installing it takes some extra work.
Yes, and that is one of the problems: rather than fixing the kernel, kernel developers just stick more and more drivers into the kernel source tree. >>>>>>>>>> There is no problem to fix. If the kernel developers wanted a standard ABI, they could have done it long ago. But they didn't for a reason. PC hardware changes too quickly for that to be feasible. Look at Windows and how long it takes them to support advances to PC hardware. Why? Because they have to be very careful that any changes they make don't break old binary drivers. Both Linux and Windows were designed at a time when drivers needed to know nothing about power management or ACPI or hotplugging. Now, drivers need to know these things. To support this, Windows has all sorts of strange interfaces (writing Windows drivers is a lot harder than writing UNIX drivers. I/O request packets and whatnot are a bitch). Linux has had to break interfaces, but the end result is much cleaner and more managable. As for adding more drivers, that's a good thing. The more drivers that are in the standard tree, the more support there is for hardware out of box.
I agree 100%. And the software they need that isn't working is the drivers and other kernel modules they need to get their hardware working and communicate with the rest of the world. >>>>>>>>>> Again, what hardware drivers are you talking about? If you're using a modern distribution, everything should be supported out of box. If it isn't, then consider it a piece of hardware that Linux doesn't support by default. Just as Windows doesn't support certain hardware out of box. Both take some work to get running.
If, after 10 years of hacking, it's not possible to provide a basic set of APIs for drivers, file systems, and other common kernel components, then the design is at fault. If not anything else, the Linux kernel could have two sets of APIs: stable and experimental. Other kernel architectures, involving message passing, RPC, or objects, also force people to think about this rather than keep changing things around haphazardly. >>>>>>>>>>>>&g t; Before you go faulting the design of the kernel, I'd ask you, what do you know that everyone hacking on the kernel doesn't. Face it, hardware changes, the goals of the OS change. Even now, if the driver API were frozen, it might be (for example) unsuitable for the NUMA machines that Linux is trying to target. Keeping the driver API fluid allows developers to fix stuff as needed, instead of being a slave to backwards compatibility. One thing an open source kernel gains over a closed-source one is more freedom with interfaces. GCC can keep breaking binary compatibility (and hence keep improving the ABI) because people can just recompile their software. Closed source OSs can't do that, and there is no reason for Linux to try to emulate that undesireable behavior. Umm, the same thing happens with most OSs.
I don't know what "most" means, but there are certainly many ways of avoiding that problem. For example, if you write the kernel in something other than assembly or C/C++, >>>>>>>> You've immediatly lost all credibility right there. In the real world, people don't use sissy languages like Scheme to do OS programming. Its ASM and C, live with it.
it gets much harder to crash the kernel accidentally. >>>> And you lose all semblence of performance.
If you build a message passing kernel, you can transparently move drivers in and out of kernel address space, trading off performance and safety as needed. >>>>> If you're drivers are bothering you that much, you've got a problem. I've used some pretty flaky drivers in the past (NVIDIA's early kernel drivers) and I have yet to crash the kernel due to a driver problem. This is a dead horse. People long ago figured out that existing architectures were just not designed for microkernel systems, and that the saftey of a message passing interface did not justify the overhead required to give drivers access to the hardware. Even OS-X realized this, and put the whole kernel back in kernel-space, and replaced messaging with procedure calls.
It's only monolithic kernels written in an unsafe language that are this sensitive. >>>>>>>>> Which is basically all of them. And they are that way for a reason. Besides, the fact that you use the word monolithic identifies you as a throwback to the 1990's. There are no monolithic kernels anymore, they're all modular. They don't have the safety of microkernels, but have all the advantage of seperating out components, and that's always been the real win.
Don't get me wrong: Linux has been a reliable workhorse for many years, and the functionality in it is wonderful. But I think these issues are really becoming the biggest obstacle to its more widespread adoption and use on the desktop, and it's only going to get worse. >>>>>>>>> People don't go, "I don't use Linux because its not a microkernel written in Scheme," people say "I don't use Linux because it doesn't have the software I need." I mean what problem exactly are you trying to solve? Instability? Come on, not even MS claims that Linux is unstable. How about ease of use? Nope. As long as you're using an easy distro like Redhat or Mandrake, driver updates are a simple urpmi kernel-2.4.XX away. These days, there is very little mainstream hardware not supported in the stock kernel. On my Inspiron 8200 laptop, for example, every single gadget I have, from my USB mouse to my Pocket PC to my QuickCam is supported in the stock kernel. There is simply no reason to install outside drivers. And because there is no reason to do that, it makes no sense to limit the kernel developers just so the 3 people that distribute seperate drivers can have a stable ABI.
You didn't know what to expect. Now if they could have just gotten someone besides Kirsten Dunst... >>>>>>>>>>>>&g t; Say it ain't so! Kirsten Dunst was great. Not entirely true to the character, but she's Kirsten Dunst!
Face the music: there are not enough users on Linux to justify having any developers work on a port of, say, Photoshop. It would take millions of dollars to port, and nobody will buy it. Given that Linux has maybe 0.5% of the desktop, and that maybe 1% of that will ever buy software that costs more than $30, I doubt the expense is justified. >>>>>>>>>> Actually, given the number of development houses that are switching their workstation's to Linux, stuff like Photoshop is what has a market on Linux. What really won't sell on Linux is home user stuff, like Mavis Beacon's Typing Tutor.
Actually, Darwin is about half as fast as Linux on the same hardware for basic kernel operations (mmap, open, etc). And its subsystems (especially the VM) aren't anything to brag about in comparison to their counterparts in FreeBSD and NetBSD.
Because, empirically, kernel modules seem to end up being very dependent on kernel versions; if they weren't, distributions like Debian wouldn't ship with different collections of most kernel modules for each kernel, they would ship with one kernel module per package for each function/driver, without much of a notion of a "kernel version". >>>>>> That's more attributable to the fact that Linus doesn't want to freeze the driver API rather than any fault of the design. BeOS (and Windows!) both use dynamically loaded drivers into kernel space, and they work just fine (Windows' other design weirdness aside).
Another reason is that one bug in one kernel modules brings down the whole thing. That's unnecessary and makes driver development a huge pain. >>>>>>>> Umm, the same thing happens with most OSs. If it works for a 64-proc Solaris box, it sure as hell is good enough for my laptop! Given that OS X runs everything in kernel space, it isn't immune to this either.
1) Innovation only increases the cost of product for a limited period of time after the innovation. Once that period is over, market forces will bring selling prices back in line with production costs.
2) Monopolies are illegal, exactly because monopolies lead to prices that don't reflect production costs. Competition leads to lowering of production costs, which leads to lower prices.
3) Cartel's are also illegal.
4) Advertising can be looked as an increasing in production costs. It allows for some level of increase in price, but doesn't allow it to stay at levels vastly in excess of production costs.
Now, if you take a look at the actual market place, you'll see that this is generally true. Most products do not sell that far in excess of production costs. Cars, industrial machines, and even computers, at the large scale, have very low profit margins. Computer manufacturers, for example, make only a few percent profit per machine. Now, software on the other hand, has huge profit margins. Why? Because people don't understand it, and because its new. Once the novelty of software wears off, and it becomes commodity (and people figure out that software monopolies like Microsoft are just as bad as any steel or oil monopoly ever was) the cost of software will decrease to the point where it is somewhat in line with production costs.
The idea that a few months of subsidized work could undermine a huge sector of the economy is a pretty new one. >>>>>>>>> Methinks that if a few months of subsidized work can undermine a huge sector of the economy, that sector of the economy should not have become as large as it is, and eventually would have crashed anyway. Don't forget, at some point in economics, there is reality involved. If software just isn't worth the price people are charging for it (and my guess is that the $40 billion in cash in Microsoft's bank indicates the cost of labor is far outstripped by the final selling costs) then eventually something will happen to stabilize the situation again.
Err, I have no such illusions. I expect the Hammer to be about 20-30% faster at a given clock than existing chips, which is somewhat optimistic, but entirely within the expectations for the chip. I want a 64-bit machine because there are some things in OS development that are more fun when you have 64-bits of address space. Things like single-address space operating systems and persistant virtual memory stores become feasible when with 64-bits of address space while they aren't so nice to implement with only 32 bits.
Because the sooner it comes out, the sooner I get to play with a 64-bit OS development on a machine that gets top performance and doesn't cost $20,000. That alone is reason enough for AMD to ship it sooner.
Slashdot Mirror
GDP already stands for Gross Domestic Product. Can we call it GnuDP instead?
1) First, this is a drop in the bucket. CERN is one of only a few entities using this kind of power. Take a look at the 55,000 home figure. Say these are your Asian homes, and you end up with 550,000 homes that could be powered by the energy used at CERN. Given the masses of humanity, that adds up to NOTHING. Certainly not enough to justify the potentially disastrous consequences to humanity that could result if we stop this kind of research. The average utility bill at Fermilab (US counterpart to CERN) is 1.5 million dollars. Over the course of a year, thats $18 million. In comparison, the USAID annual budget is $6 billion. The Federal Highway program's budget on the other hand, is $26 billion. Yes, people would rather fix potholes than develop countries. Such is the world we live in. But its not like research money is coming out of the mouths of the poor. The real problem isn't that the world spends too much on research (if only). The real problem is that people have no clue about anything outside the bounds of their tiny meaningless existance. For example, most people, when survayed, said that they thought the US spends too much on foreign aid. When asked how much they thought was appropriate, they said 5% of the budget. The real figure is one-tenth of that number. There are dozens of things you can do to make up the cost of this research, including improving distribution methods, bringing down cultural barries that make access to healthcare inefficient, reforming patent conventions that jack up the cost of medications, etc. In the end, science is not the thing to sacrifice for humanity. BTW, I grew up learning stuff about international development from my dad (that's his line of work), and I was born in Thailand and spent part of my childhood in Bangladesh. I *do* know what I'm talking about.
2) We're not talking about the "many marvelous inventions in the last, say, 20 years." We're talking about how physics has redefined the universe was we know it over the last 200 years. A large percentage of the modern economy owes its existance to quantum physics. The work at CERN is simply an extension of the very ancient search for knowledge about the structure of matter. Anytime you get a cat-scan or an X-ray, take medicine, use a computer, drive a car, watch TV, etc, you're directly benifeting from that research. Even those in inpovrished countries should thank this research for allowing scientists to use advanced imaging tehnologies to create things like TB vaccines that sells for dollars per dose. Going into the future, the only sure way to relieve poverty is to find more resources. It is not possible for the human population to keep growing, expanding, evolving, reaching towards a higher state of being, without more raw matter. So yes, that warp drive space ship WILL help the guy living in poverty, if you stop being so short-sighted. Giving a man a fish is not the only way to help him.
What's TCP-C? Can somebody convert it into the Pop Science(TM) units for me so I can understand?
Yes. Because they advance humanity. It is impossible, with our current resources, to pull people out of poverty. It is work like this that will one day give us the chance to pull humanity out of the gutter. Back when some crazy's were fooling around with dead human bodies, nobody cared. Now, we call them doctors and thank them for delivering us from the horrors of disease. Besides, don't you think there are better uses for the power used by your computer than posting on /.?
I always love morons who criticize things when they aren't part of their intended audience. Unless you're a gamer, you have no reason to be at those sites. If you are, those sites are invaluable. Quick, what's faster, the Radeon 8500 or the GeForce4 MX? I don't really care, my gaming is limited to some RPGs and the occasional CounterStrike match. However when I wanted to buy my brother a new graphics card, you bet he was happy that I did my research at those sites first. I used to frequent those sites a lot, back when I did a lot of 3D programming and gaming. I honestly needed the performance, and those sights provided a lot of good guidence, as well as useful ways to get cheap hardware to do cool things (overclocking Celeron 300A's and running them in SMP for example, or joining the bridges on Athlon XPs to get them running in SMP). As for the bit about advertisers, that's bull. The internet keeps these sites honest. A GeForce 4 TI 4600 really is a lot faster than a Radeon 8500. If a site says otherwise because of advertising, a second opinion (or third or fourth) is just a click away. In the future, it would be wise to provide specific proof before criticizing people of lacking journalistic integrity.
Um, disk isn't a bottleneck on my system. I do lots of C++ programming, and everything gets cached in RAM after the first build. Thus, my bottleneck becomes gcc ...err... CPU and memory bandwidth :) Same thing when I'm doing 3D rendering (which had better fit in RAM or else) or playing 3D games and whatnot. I thought moving to a 4200 RPM laptop hard drive was going to be bad, after my nice 7200 RPM IDE RAID. In truth, thanks to the Linux VM, I don't notice the difference after the first half hour of using the system. However, I did notice the big boost that came with moving from a PC100 memory system to a PC266, even just palying around with GUI wigets (resizing and whatnot).
Rich immigrants aren't willing to work for an unnaturally low cost. Thus, in this case, its not stealing the job, but taking it fairly.
If American engineers can't compete and they lose their jobs, well too bad, such are the costs of capitalism. And don't tell me these are real engineers with real families, because so are they, and they need the jobs more than we do.
Actually, if you had been keeping up, some guy had broken XFS up into nice tiny patches, while also converting it to use the generic I/O routines. In the end, there were four lines of changes to non-XFS code. There is an OS news article about this, with a link to the relavent lkml thread.
Just make sure you shut of sharing with other users :)
Example: IPsec. Not included in the standard kernel. In order to get it working, I'll have to patch, configure, and recompile kernels for half a dozen different machines. For handhelds running Linux, this will be even more of a chore.
>>>>>>
If you need IPsec, then you can certainly recompile your kernel. The thing is that external stuff like IPsec is entirely analagous to external stuff in Windows. For example, Windows prior to Win98 SE didn't come with any NAT capability. Installing an add-on like raspppoe took a good bit of work. IPsec isn't a part of the standard Linux installation. Thus, it is to be expected that installing it takes some extra work.
Yes, and that is one of the problems: rather than fixing the kernel, kernel developers just stick more and more drivers into the kernel source tree.
>>>>>>>>>>
There is no problem to fix. If the kernel developers wanted a standard ABI, they could have done it long ago. But they didn't for a reason. PC hardware changes too quickly for that to be feasible. Look at Windows and how long it takes them to support advances to PC hardware. Why? Because they have to be very careful that any changes they make don't break old binary drivers. Both Linux and Windows were designed at a time when drivers needed to know nothing about power management or ACPI or hotplugging. Now, drivers need to know these things. To support this, Windows has all sorts of strange interfaces (writing Windows drivers is a lot harder than writing UNIX drivers. I/O request packets and whatnot are a bitch). Linux has had to break interfaces, but the end result is much cleaner and more managable. As for adding more drivers, that's a good thing. The more drivers that are in the standard tree, the more support there is for hardware out of box.
I agree 100%. And the software they need that isn't working is the drivers and other kernel modules they need to get their hardware working and communicate with the rest of the world.
>>>>>>>>>>
Again, what hardware drivers are you talking about? If you're using a modern distribution, everything should be supported out of box. If it isn't, then consider it a piece of hardware that Linux doesn't support by default. Just as Windows doesn't support certain hardware out of box. Both take some work to get running.
If, after 10 years of hacking, it's not possible to provide a basic set of APIs for drivers, file systems, and other common kernel components, then the design is at fault. If not anything else, the Linux kernel could have two sets of APIs: stable and experimental. Other kernel architectures, involving message passing, RPC, or objects, also force people to think about this rather than keep changing things around haphazardly.
>>>>>>>>>>>>&g t;
Before you go faulting the design of the kernel, I'd ask you, what do you know that everyone hacking on the kernel doesn't. Face it, hardware changes, the goals of the OS change. Even now, if the driver API were frozen, it might be (for example) unsuitable for the NUMA machines that Linux is trying to target. Keeping the driver API fluid allows developers to fix stuff as needed, instead of being a slave to backwards compatibility. One thing an open source kernel gains over a closed-source one is more freedom with interfaces. GCC can keep breaking binary compatibility (and hence keep improving the ABI) because people can just recompile their software. Closed source OSs can't do that, and there is no reason for Linux to try to emulate that undesireable behavior.
Umm, the same thing happens with most OSs.
I don't know what "most" means, but there are certainly many ways of avoiding that problem. For example, if you write the kernel in something other than assembly or C/C++,
>>>>>>>>
You've immediatly lost all credibility right there. In the real world, people don't use sissy languages like Scheme to do OS programming. Its ASM and C, live with it.
it gets much harder to crash the kernel accidentally.
>>>>
And you lose all semblence of performance.
If you build a message passing kernel, you can transparently move drivers in and out of kernel address space, trading off performance and safety as needed.
>>>>>
If you're drivers are bothering you that much, you've got a problem. I've used some pretty flaky drivers in the past (NVIDIA's early kernel drivers) and I have yet to crash the kernel due to a driver problem. This is a dead horse. People long ago figured out that existing architectures were just not designed for microkernel systems, and that the saftey of a message passing interface did not justify the overhead required to give drivers access to the hardware. Even OS-X realized this, and put the whole kernel back in kernel-space, and replaced messaging with procedure calls.
It's only monolithic kernels written in an unsafe language that are this sensitive.
>>>>>>>>>
Which is basically all of them. And they are that way for a reason. Besides, the fact that you use the word monolithic identifies you as a throwback to the 1990's. There are no monolithic kernels anymore, they're all modular. They don't have the safety of microkernels, but have all the advantage of seperating out components, and that's always been the real win.
Don't get me wrong: Linux has been a reliable workhorse for many years, and the functionality in it is wonderful. But I think these issues are really becoming the biggest obstacle to its more widespread adoption and use on the desktop, and it's only going to get worse.
>>>>>>>>>
People don't go, "I don't use Linux because its not a microkernel written in Scheme," people say "I don't use Linux because it doesn't have the software I need." I mean what problem exactly are you trying to solve? Instability? Come on, not even MS claims that Linux is unstable. How about ease of use? Nope. As long as you're using an easy distro like Redhat or Mandrake, driver updates are a simple urpmi kernel-2.4.XX away. These days, there is very little mainstream hardware not supported in the stock kernel. On my Inspiron 8200 laptop, for example, every single gadget I have, from my USB mouse to my Pocket PC to my QuickCam is supported in the stock kernel. There is simply no reason to install outside drivers. And because there is no reason to do that, it makes no sense to limit the kernel developers just so the 3 people that distribute seperate drivers can have a stable ABI.
You didn't know what to expect. Now if they could have just gotten someone besides Kirsten Dunst...
>>>>>>>>>>>>&g t;
Say it ain't so! Kirsten Dunst was great. Not entirely true to the character, but she's Kirsten Dunst!
1.3.7 I wouldn't expect newer versions to be dramatically faster, most of the changes (read the changelogs) occured in userspace (Quartz and whatnot).
. tx t
http://clustermonkey.org/~laz/pbook/rob.lmbench
Face the music: there are not enough users on Linux to justify having any developers work on a port of, say, Photoshop. It would take millions of dollars to port, and nobody will buy it. Given that Linux has maybe 0.5% of the desktop, and that maybe 1% of that will ever buy software that costs more than $30, I doubt the expense is justified.
>>>>>>>>>>
Actually, given the number of development houses that are switching their workstation's to Linux, stuff like Photoshop is what has a market on Linux. What really won't sell on Linux is home user stuff, like Mavis Beacon's Typing Tutor.
Actually, Adobe's Acrobat is quite full-featured in the Linux version. It supports CoolType and everything.
Actually, Darwin is about half as fast as Linux on the same hardware for basic kernel operations (mmap, open, etc). And its subsystems (especially the VM) aren't anything to brag about in comparison to their counterparts in FreeBSD and NetBSD.
Because, empirically, kernel modules seem to end up being very dependent on kernel versions; if they weren't, distributions like Debian wouldn't ship with different collections of most kernel modules for each kernel, they would ship with one kernel module per package for each function/driver, without much of a notion of a "kernel version".
>>>>>>
That's more attributable to the fact that Linus doesn't want to freeze the driver API rather than any fault of the design. BeOS (and Windows!) both use dynamically loaded drivers into kernel space, and they work just fine (Windows' other design weirdness aside).
Another reason is that one bug in one kernel modules brings down the whole thing. That's unnecessary and makes driver development a huge pain.
>>>>>>>>
Umm, the same thing happens with most OSs. If it works for a 64-proc Solaris box, it sure as hell is good enough for my laptop! Given that OS X runs everything in kernel space, it isn't immune to this either.
Hey, I liked Con Air. It was a bit different from other movies, but as far as action movies go, it was pretty god. Not a bad story, either.
To answer your points:
1) Innovation only increases the cost of product for a limited period of time after the innovation. Once that period is over, market forces will bring selling prices back in line with production costs.
2) Monopolies are illegal, exactly because monopolies lead to prices that don't reflect production costs. Competition leads to lowering of production costs, which leads to lower prices.
3) Cartel's are also illegal.
4) Advertising can be looked as an increasing in production costs. It allows for some level of increase in price, but doesn't allow it to stay at levels vastly in excess of production costs.
Now, if you take a look at the actual market place, you'll see that this is generally true. Most products do not sell that far in excess of production costs. Cars, industrial machines, and even computers, at the large scale, have very low profit margins. Computer manufacturers, for example, make only a few percent profit per machine. Now, software on the other hand, has huge profit margins. Why? Because people don't understand it, and because its new. Once the novelty of software wears off, and it becomes commodity (and people figure out that software monopolies like Microsoft are just as bad as any steel or oil monopoly ever was) the cost of software will decrease to the point where it is somewhat in line with production costs.
Even if it is, who cares? Is Microsoft (the chief competitor to this groupware product) in any position to complain about unfair competition?
The idea that a few months of subsidized work could undermine a huge sector of the economy is a pretty new one.
>>>>>>>>>
Methinks that if a few months of subsidized work can undermine a huge sector of the economy, that sector of the economy should not have become as large as it is, and eventually would have crashed anyway. Don't forget, at some point in economics, there is reality involved. If software just isn't worth the price people are charging for it (and my guess is that the $40 billion in cash in Microsoft's bank indicates the cost of labor is far outstripped by the final selling costs) then eventually something will happen to stabilize the situation again.
Err, I have no such illusions. I expect the Hammer to be about 20-30% faster at a given clock than existing chips, which is somewhat optimistic, but entirely within the expectations for the chip. I want a 64-bit machine because there are some things in OS development that are more fun when you have 64-bits of address space. Things like single-address space operating systems and persistant virtual memory stores become feasible when with 64-bits of address space while they aren't so nice to implement with only 32 bits.
Because the sooner it comes out, the sooner I get to play with a 64-bit OS development on a machine that gets top performance and doesn't cost $20,000. That alone is reason enough for AMD to ship it sooner.
Um, who cares? Not everything has to help the masses. Some stuff is by hackers for hackers.