I have never seen good documentation written by a developer. Developers usually suck at writing, and think that every feature of their program is completely obvious. It is much better to have an experienced user write the documentation.
Unless it's very small volume manufacturing, manufacturers simply do not have the resources to test things. Electronics assembly is done completely automatically these days, and it's very expensive to individually test or adjust anything. Unless the equipment in question is very expensive (microprocessors, for instance), it's probably identical across brands and such.
If you use a PC to control a machine like that, you are an idiot and should not be let near such hardware. The Win3.1 machine probably just sets the parameters or something, the real control is usually done by a specialized controller or DSP that's not running any operating system, just the program itself.
Don't you think that a private developer would be even more likely to make an error? All important code has to get verified very extensively, so the panel of experts will be present in either case. You can't just trust somebody - everyone makes errors. For REALLY important projects, three separate teams will write different code for the same specification and it is then compared. I really doubt that someone would use a commercial package for something like that, anyway.
The thing is, there are lots of things in government that require embedded systems and don't require that level of confidence. That's where Linux and the proprietary vendors come in. And that's where Linux has the potential to eat their lunch.
If it was an open standard, there would be no need to reverse-engineer it.
But it isn't- it is a proprietary standard that Apple reserves the right to license to others and I see no basis for the legal system to NOT uphold this right.
Tell that to the developers of Samba (which, incidentally, OS X uses to access Windows file sharing). I am sure they would be falling over themselves to license SMB from Microsoft.
"Hats" with 96-electrodes in them are capable of collecting data that can do crazy things, like predict the occurrence of seizures up to ten minutes in advance.
I didn't mean to say EEG was useless. Obviously, it has its uses, and this is one where it shines. Read the post I was originally replying to, and I think you will agree that EEG is useless for most things that are mentioned there (controlling your computer, finding "vulnerabilities" in the brain, all in the Open-source tradition).
As far as implanted multielectrode arrays providing interesting data, I guess "interesting" is subjective, but they can record individual action potentials from hundreds of neurons with sub-millisecond time resolution, and I'm not sure what more you could ask for than that, aside from increasing the number of neurons recorded.
The thing is, it's like recording from 10 random wires on your computer's motherboard. You might see something interesting, but it won't give you the ability to do complex things with your computer. It's very useful if you are trying to map areas of the brain, but it doesn't let you read someone's thoughts. You have to read my reply in the context of the original post to see my point.
That's pretty well known. I have a book from the early 80s with a simple analog circuit that's supposed to do something similar (control the direction of a toy train, IIRC). This has nothing to do with EEG, it's just muscle signals. Your brain sends small signals to the muscles when you think of moving them. If you amplify them, you can use these signals.
Interesting? Maybe. Useful? Not for an average person. You have to attach lots of electrodes that are intrusive, uncomfortable, possibly expensive, and not very reliable. It's also very difficult to get any real precision, because you are just seeing some average signal for a huge group of muscles.
I am not saying it's useless. It's very useful for research. The problem is, research is not something easy or cheap. My point was that it's not something an average Slashdotter can do anything interesting with. If you have sufficient knowledge to actually make sense of the EEG signal, chances are good you will not need to build a DIY EEG machine (since your grants will pay for one easily).
Seriously. A carpenter can't sell the abstract concept of a table. He/she has to sell a physical table. Until very recently, a carpenter had no way of stopping someone from copying the design of his table and selling it for less.
Also, I fail to see how it is someone's right to make money from something. If I choose to build tables and give them away for free (and thus destroy someone's business model), it's my right to do so. Software is something that requires a high initial investment and has zero unit cost. If I am willing to donate the investment, and make the software free, it's my choice and you have no right to whine. Obviously, really large and complex projects generally aren't built by one person, so it's self regulating.
WTF is it with Apple zealots these days? You do know that these provisions of the EULA are not legally enforceable? The courts have consistently held up the viewpoint that if you reverse-engineer for interoperability, it's perfectly legal to do so.
Whatever. As someone who has some experience with neurobiology, it is very unlikely you can actually get anything useful from EEG signals. They are totally useless for exploring the brain's functions. Sorry to burst your bubble, but even microelectrode arrays implanted directly into the brain don't provide anything too interesting. At best, you might see a different firing pattern based on external stimulus, and even that is rare.
Almost any laser can damage your vision. Even a laser pointer-class laser is several times brighter than the sun when focused. However, it won't damage your eyes unless you act like a total idiot and stare into it. In any case, putting lots of BS or non-BS warnings on a DIY project is commonly referred to as "covering your ass" and is done by many people, mainly to avoid getting sued.
Running 5V around the whole house would be much more inefficient than putting a transformer inside each bulb. It would also be expensive, and a fire hazard. Think about it: running 1kW of power through a 5V system would require 200 amps of current! That would require welding cable-sized wires.
Although you can handle more simultaneous small requests faster with more drives
That's exactly the point. What do you think limits the bandwidth of, say, a database?
you can only transfer the data from those requests at 320mb/s max
I am pretty sure fiber channel is faster than that. That's what fast arrays are hooked up with, anyway. These are generally independent boxes, with their own highly sophisticated and intelligent controller and a really fat pipe.
And there are always those who wants lots of larger drives.
If you have a database that takes up 500GB, you don't need 15 terabytes. You do need reliability and extremely high throughput and low latency. If the 500GB is all on one drive, you will be limited by the speed with which that drive can move its head. If it's spread across 30 smaller disks in a RAID configuration, your latency will be greatly reduced and the throughput will be greatly increased.
If you are actually going to use those drives in a server (guess where they are used, for the most part), you need them to be fast. An array with fewer larger drives is much slower than one with lots of small drives.
I am pretty sure it's still a scam. Quite possibly they are a real company. But the thing is, you are not going to get jack shit (no pun intended) for your $100. They make you wait 90 days before it ships (so your cc company won't be able to do anything about it). Their website has no physical address. They have logos of various companies on their website -- obviously an attempt to make them look legitimate. Their website has no pictures of the executives -- strange. If this is not a scam, then the company is deliberately making it look like one.
Kernel fork so that companies can have a stable branch that they can trust, and just cherry pick new things from the main tree as and when they want
Actually, companies don't want that. Have you ever looked at the redhat kernel? Or the Mandrake kernel? There is a shitload of backported features. Do you think this leads to stability? Do you think kernel developers can support custom-hacked kernels? This change is a great thing, because the kernel developers have been out of touch with the needs of end-users for quite a long time.
OR - vendors like Oracle, Sybase, IBM, etc. only supporting one or at most two distributions
That's already the case. If you want to install Oracle, you better have Redhat version something or other. Sure, it will work on any version, but don't expect them to help you when it breaks.
Each real distro already adds tons of patches to stabilize and add features. This already results in really nasty forking. At least with the new model, those patches will quickly get merged into the kernel, which is not the case with backports.
The expectation that the vanilla kernel was ever stable needs to DIE. That has not been the case since 2.2. 2.4 was buggy as hell until about 2.4.19, 2.6 is still buggy as hell. The distro makers are already having to stabilize the kernels AND backport features on top of that. This will, in all likelihood, actually result in more stable kernels from the distributors, because they will spend less time on backporting stuff.
Just because you are a bad driver doesn't mean everyone else is. Good drivers can drive for decades without a single accident. It's just that it's too easy to get a driver's license in the US.
Slashdot Mirror
I have never seen good documentation written by a developer. Developers usually suck at writing, and think that every feature of their program is completely obvious. It is much better to have an experienced user write the documentation.
Unless it's very small volume manufacturing, manufacturers simply do not have the resources to test things. Electronics assembly is done completely automatically these days, and it's very expensive to individually test or adjust anything. Unless the equipment in question is very expensive (microprocessors, for instance), it's probably identical across brands and such.
If you use a PC to control a machine like that, you are an idiot and should not be let near such hardware. The Win3.1 machine probably just sets the parameters or something, the real control is usually done by a specialized controller or DSP that's not running any operating system, just the program itself.
Don't you think that a private developer would be even more likely to make an error? All important code has to get verified very extensively, so the panel of experts will be present in either case. You can't just trust somebody - everyone makes errors. For REALLY important projects, three separate teams will write different code for the same specification and it is then compared. I really doubt that someone would use a commercial package for something like that, anyway.
The thing is, there are lots of things in government that require embedded systems and don't require that level of confidence. That's where Linux and the proprietary vendors come in. And that's where Linux has the potential to eat their lunch.
Maybe, if Fairplay were an open standard.
If it was an open standard, there would be no need to reverse-engineer it.
But it isn't- it is a proprietary standard that Apple reserves the right to license to others and I see no basis for the legal system to NOT uphold this right.
Tell that to the developers of Samba (which, incidentally, OS X uses to access Windows file sharing). I am sure they would be falling over themselves to license SMB from Microsoft.
Sure. But you won't be playing Doom 3 with it.
"Hats" with 96-electrodes in them are capable of collecting data that can do crazy things, like predict the occurrence of seizures up to ten minutes in advance.
I didn't mean to say EEG was useless. Obviously, it has its uses, and this is one where it shines. Read the post I was originally replying to, and I think you will agree that EEG is useless for most things that are mentioned there (controlling your computer, finding "vulnerabilities" in the brain, all in the Open-source tradition).
As far as implanted multielectrode arrays providing interesting data, I guess "interesting" is subjective, but they can record individual action potentials from hundreds of neurons with sub-millisecond time resolution, and I'm not sure what more you could ask for than that, aside from increasing the number of neurons recorded.
The thing is, it's like recording from 10 random wires on your computer's motherboard. You might see something interesting, but it won't give you the ability to do complex things with your computer. It's very useful if you are trying to map areas of the brain, but it doesn't let you read someone's thoughts. You have to read my reply in the context of the original post to see my point.
That's pretty well known. I have a book from the early 80s with a simple analog circuit that's supposed to do something similar (control the direction of a toy train, IIRC). This has nothing to do with EEG, it's just muscle signals. Your brain sends small signals to the muscles when you think of moving them. If you amplify them, you can use these signals.
Interesting? Maybe. Useful? Not for an average person. You have to attach lots of electrodes that are intrusive, uncomfortable, possibly expensive, and not very reliable. It's also very difficult to get any real precision, because you are just seeing some average signal for a huge group of muscles.
I am not saying it's useless. It's very useful for research. The problem is, research is not something easy or cheap. My point was that it's not something an average Slashdotter can do anything interesting with. If you have sufficient knowledge to actually make sense of the EEG signal, chances are good you will not need to build a DIY EEG machine (since your grants will pay for one easily).
Seriously. A carpenter can't sell the abstract concept of a table. He/she has to sell a physical table. Until very recently, a carpenter had no way of stopping someone from copying the design of his table and selling it for less.
Also, I fail to see how it is someone's right to make money from something. If I choose to build tables and give them away for free (and thus destroy someone's business model), it's my right to do so. Software is something that requires a high initial investment and has zero unit cost. If I am willing to donate the investment, and make the software free, it's my choice and you have no right to whine. Obviously, really large and complex projects generally aren't built by one person, so it's self regulating.
BULL shit. Off to jail, fuckers.
WTF is it with Apple zealots these days? You do know that these provisions of the EULA are not legally enforceable? The courts have consistently held up the viewpoint that if you reverse-engineer for interoperability, it's perfectly legal to do so.
Whatever. As someone who has some experience with neurobiology, it is very unlikely you can actually get anything useful from EEG signals. They are totally useless for exploring the brain's functions. Sorry to burst your bubble, but even microelectrode arrays implanted directly into the brain don't provide anything too interesting. At best, you might see a different firing pattern based on external stimulus, and even that is rare.
Almost any laser can damage your vision. Even a laser pointer-class laser is several times brighter than the sun when focused. However, it won't damage your eyes unless you act like a total idiot and stare into it. In any case, putting lots of BS or non-BS warnings on a DIY project is commonly referred to as "covering your ass" and is done by many people, mainly to avoid getting sued.
Running 5V around the whole house would be much more inefficient than putting a transformer inside each bulb. It would also be expensive, and a fire hazard. Think about it: running 1kW of power through a 5V system would require 200 amps of current! That would require welding cable-sized wires.
Although you can handle more simultaneous small requests faster with more drives
That's exactly the point. What do you think limits the bandwidth of, say, a database?
you can only transfer the data from those requests at 320mb/s max
I am pretty sure fiber channel is faster than that. That's what fast arrays are hooked up with, anyway. These are generally independent boxes, with their own highly sophisticated and intelligent controller and a really fat pipe.
And there are always those who wants lots of larger drives.
If you have a database that takes up 500GB, you don't need 15 terabytes. You do need reliability and extremely high throughput and low latency. If the 500GB is all on one drive, you will be limited by the speed with which that drive can move its head. If it's spread across 30 smaller disks in a RAID configuration, your latency will be greatly reduced and the throughput will be greatly increased.
If you are actually going to use those drives in a server (guess where they are used, for the most part), you need them to be fast. An array with fewer larger drives is much slower than one with lots of small drives.
The fact that you are an idiot does not, in any way, discredit the theory of evolution.
I am pretty sure it's still a scam. Quite possibly they are a real company. But the thing is, you are not going to get jack shit (no pun intended) for your $100. They make you wait 90 days before it ships (so your cc company won't be able to do anything about it). Their website has no physical address. They have logos of various companies on their website -- obviously an attempt to make them look legitimate. Their website has no pictures of the executives -- strange. If this is not a scam, then the company is deliberately making it look like one.
Kernel fork so that companies can have a stable branch that they can trust, and just cherry pick new things from the main tree as and when they want
Actually, companies don't want that. Have you ever looked at the redhat kernel? Or the Mandrake kernel? There is a shitload of backported features. Do you think this leads to stability? Do you think kernel developers can support custom-hacked kernels? This change is a great thing, because the kernel developers have been out of touch with the needs of end-users for quite a long time.
OR - vendors like Oracle, Sybase, IBM, etc. only supporting one or at most two distributions
That's already the case. If you want to install Oracle, you better have Redhat version something or other. Sure, it will work on any version, but don't expect them to help you when it breaks.
Each real distro already adds tons of patches to stabilize and add features. This already results in really nasty forking. At least with the new model, those patches will quickly get merged into the kernel, which is not the case with backports.
The expectation that the vanilla kernel was ever stable needs to DIE. That has not been the case since 2.2. 2.4 was buggy as hell until about 2.4.19, 2.6 is still buggy as hell. The distro makers are already having to stabilize the kernels AND backport features on top of that. This will, in all likelihood, actually result in more stable kernels from the distributors, because they will spend less time on backporting stuff.
Also, if you buy the hardware now, you'll be paying the MS tax now, and then you'll have to pay for a windows upgrade later.
Apple users have no problem paying $129 a year for minor OS upgrades. Microsoft's upgrades are $30 cheaper, and occur less frequently.
Very handy for hospitals whose equipment can potentially be sensitive to the high interference caused by cell phones.
Ever see a doctor WITHOUT a cell phone? Most of them carry one.
Just because you are a bad driver doesn't mean everyone else is. Good drivers can drive for decades without a single accident. It's just that it's too easy to get a driver's license in the US.
It's not a LOT quieter, it's only 30% quieter. Human hearing is logarithmic, that's why the decibel scale is logarithmic.