A device like his, properly done, should cost no more than $30 done in market quantitites (10k+). Even a prototype, properly done, would cost less than $100 in materials, easy. How the heck did he spend $300 on his thing I just don't know. For $300 you could build a 16 channel EEG logger...
EOG can be done right, if you know what you're doing. The way he is doing it -- it barely works. He cut all the corners that there were to be cut. The quoted cost ($300 USD) is pretty silly for what he has done.
For noise mitigation and ease of use, you need a system that is physically small (forget any long wires) and uses integrated, reusable electrodes. It needs to be no harder to put on than eyeglasses. This seems like an obvious requirement. Who has time to play with electrodes?
This means the following (BTDT): 1. The electrodes are machined as rods with rounded ends out of titanium (lightweight!) or stainless steel. The inside can be drilled out to further reduce weight. 2. The electrodes are mounted on a PC board that straddles the nose bridge. They are on the inside corners the eye, aiding in keeping the system compact. 3. The PC board extends above the nose bridge and has slots for a flat stretchable "headband" that pushes it against your lower forehead. You may need an adapter piece (a plastic shim) to keep it at the right angle, this fitting is done once per person -- until one changes the shape of one's skull, that is. 4. The PC board holds the differential preamp and a low power CPU that does A/D conversion and transmission over IR. The battery (two AAAs) can be on the other side of the head, on the strap.
Ideally, you'd fit a two-cell charger on board (it's one more chip), with provision for charging the two AAA NiMH cells on the strap via, say, USB. Then you don't have to worry about having to mess with replacing batteries, and the computer is the power supply for charging -- it's there already, might as well use the standby power supply for something:)
If you have access to a lathe and scrap metal, you can keep the cost of the entire thing under $100 in materials, even in quantity one, even with a custom PC board.
Since you are, presumably, in front of your computer, using IR is the simplest, lowest cost means of one way communication. On a Mac you have built-in IR receiver so you need no extra hardware on the receiving end.
His is a nifty project, but done with little attention to detail.
Yep. And that's what he said: trademark owners are required to protect their IP rights. Patent holders' won't risk "losing" anything if they don't fight every patent violation they learn about.
For patents, the filing date is what counts in this scenario. The one to file first "wins". There's more to it, IIRC, as you can claim priority on public disclosures and foreign patents. So if you publish something in a scientific journal, say, you have a year to file a patent for it, and your invention is protected retroactively since the publication date in the journal. Someone who knows U.S. patent law better feel free to chime in with corrections, I'm not 100% sure about it. I'll ask a patent lawyer at work to see if he has anything to add to that.
Forgot to add: this was about very local-in-scope laws -- municipal, no less. So the assertion that somehow only federal laws are exempt from copyright is false in light of this.
The issue in this en banc case is the extent to which a private organization may assert copyright protection for its model codes, after the models have been adopted by a legislative body and become "the law". Specifically, may a code-writing organization prevent a website operator from posting the text of a model code where the code is identified simply as the building code of a city that enacted the model code as law? Our short answer is that as law, the model codes enter the public domain and are not subject to the copyright holder's exclusive prerogatives. As model codes, however, the organization's works retain their protected status.
As long as you treat the codes as part of law, they are public domain. Easy. So, if a code is adopted with any changes (even reformatting), only the changed version as adopted into law is in public domain. The original "model code", if any, is still subject to copyright protection.
Well, if you adopt such standards then they become law, plain and simple. Alas, I messed up my link: this is bulk resource and they have various building codes for your perusal. No need to spend a single cent for it. I think it's unconscionable that laws are a subject of copyright protection, just as it is unconscionable that the Queen holds crown copyright of King James Bible (in the U.K.). Both are fucked up IMHO (yep, I'm strongly opinionated on that front). If it's a law, it should be accessible online in an unencumbered fashion. Make it a U.S. constitutional amendment, for all I care.
I believe that the U.S. courts repeatedly held that legislation is exempt from copyright protection. So, even if the NEC is protected by copyright as its printed in the NFPA publications, if you get a copy from your local legislation, you can do with it as you please. Thus bulk resource. Alas, the laws in all of Europe are in complete opposition to this: they have plenty of laws that incorporate fairly expensive standards, and the courts have held that it's OK. Just E.U. legislation (so we exclude member countries' laws!) includes by reference standards that would cost you tens of thousands of Euros to obtain!
Yeah, I was just wondering: so we have a sterile bag that promotes cell growth (of certain cells, at least). Now, how does that fix contamination of the thing you're going to culture? That's not to be trivially dismissed, I'd imagine that contamination of the "seed" material is just as big of an issue as the contamination of the growth medium.
Many people are not upgrading because there's no need to, and because application support is lagging. It has been only fairly recently (last year or so) that some macports started working correctly on snow leopard. There are still ports that do not work when compiled for 64 bits, so I still compile macports for 32 bits. The reason to update to 10.6, for me, was speed and stability. Those were killer features. Lion doesn't have anything that would be a killer feature, not for me at least.
I don't think that there is any concrete dam out there that's not designed to survive with passive safety only. That means it can get full and spill over the top and nothing bad will happen. Eventually, if you let it spill over for many years, the top would erode away, perhaps facilitating crack initiation, but this is nothing immediate. With earthen dams it's a different story, but there isn't too many of those that hold enough water to cause large scale damage. By large scale I mean, say, 10% of country's population would be affected.
I still hold that if you closed every flow control that can be closed on all of the U.S. dams, there'd be no major emergency resulting from that. Of course whatever secondary function the equipment on the dams may have would likely be compromised (say power generation), but that's, like, duh.
I don't think it would be a factor, but it's easy to test for so it's silly not to. It would end all arguments. If you suspect creep, you take a couple samples (here: magazines loaded with mechanical blanks), and test them under different temperatures. It's then real easy to see if you have creep or not. I'd suspect that we're dealing with a folk tale here anyway (the loaded mag story). It's too far fetched.
What I stated was hopefully an accurate reflection of reality. Our way of life is not really compatible with large scale disasters. Jost look at how bad it was in Japan, and that was just in one coastal area with the rest of the country being able to function more-or-less normally.
You may have more luck on township/county roads. If I had to flee any large city, I'd head for those and stay the heck away from any popular, large thoroughfare.
- Widespread failure of dam control systems that risk flooding large parts of the country.
Maybe if you lived in a small alpine country or in the Netherlands. There is nothing you could do to the dams in large countries like U.S., China, Russia, India, Australia, etc. that would affect significant number of people.
If this would be a nationwide attack, presumably all the infrastructure would be affected to such a degree that the society as we know it would collapse anyway. You're right that individual suffering may be reduced in the immediate aftermath (hours and days), but in the following months those same people who were initially spared would die due to disease and lack of food. The cities, with their high density of population, would become unsustainable pretty much overnight. Folks living out there in the farmland would have the best chances.
I have not heard of fatigue, nor any theories used to model it, in high school, and I think it's OK to leave it out. There's plenty of things more fundamental than that to learn about before you enter college. Fatigue is pretty much an engineering subject, and its detailed study would include theory of fracture, and that's not exactly an afternoon tea subject (nor a high school one).
Oh, all the half-truths in those posts are just saddening. Why can't anyone get it completely right, at least in the realm of first-order effects?
Fatigue is, by definition, when you have cyclic loads. Obviously a spring left loaded and left alone doesn't experience any more cyclic loading than it would without being loaded at all. So it's nothing to do with fatigue.
What they may be talking about is creep -- the relaxation of stresses in material. This process is highly temperature dependent, so if you think you're dealing with creep, you immediately set up multiple specimens at different temperatures. If whatever effect you observe (here: loss of spring tension) accelerates with higher temperatures, then you can be 99% sure it's creep pretty much without having to do any other testing at all. If, though, there is no change with temperature, then you can safely exclude creep.
This gave me an idea: I've searched for the word temperature in the discussions you linked to. And hey, there was ONE post in all the misinformed noise that came from someone who had a bit of a clue. I'll cite it:
The phenomena you guys are talking about is called creep, where a material slowly loses strength while elastically deformed. Elastic deformation is when a material is under enough stress to deform it, but when the stress is removed the material returns to it's original shape. Springs compressed within their limits are elastically deformed. Creep is mostly a function of temperature, and typically isn't an issue in environments below half the melting point of the stressed material.
The person quoted wasn't fully correct either. Creep has got nothing to do with loss of strength, at least it's not a first-order effect during creep. A spring that experiences creep is not becoming any less strong, its internal stresses merely become relaxed (they go down), and thus the zero length of the spring is moving. This makes the spring act, over time, as if you had been squeezing it less and less: the force at the end of the spring goes down over time. That's all.
This is the silliest thing I keep hearing. There's nothing about gasoline that causes it to go bad over time, except if there is VERY significant evaporation going on. If you have a hermetically closed tank, you're good. Even with evaporation, I've had a 1993 Camry sit for about 3 years without being refueled, it'd be started up maybe twice a year but there was nothing bad about the gasoline in the tank. The engine ran, to my ear, as well as it ever did. It'd start right up. I had kept the battery hooked up, and a standby (maintenance) charger connected to keep it ready.
If you have two cycle engines that need oil mixed with gasoline, then all you have to do is make sure the tank is full when you put them away, and that the level hasn't gone down to more than 50% of tank full. I've found a lawnmower that was sitting in a shed somewhere long enough (at least 5 years) that it appeared that 2/3 of the gas has evaporated from the tank. It still started up after about a minute of working on it.
There is a lot of counterfeiting going on in the passive component market. Most electronic equipment is brought down by passive components failing. There were few chip fiascos, and those usually didn't even have anything to do with the chip but with packaging and mounting (say nVidia gpus). Most of the electronics I repair have problems with passive components. Shorted out SMT capacitors, dried out electrolytics, a broken track, a poor connector somewhere. Sometimes a failing passive component can take semiconductors with it, especially power semiconductors. I'd say if you're repairing stuff, you'll be better than 50% successful if all you have for replacement parts is passives and wire. With modern SMT assemblies, we're often talking of parts that cost lestt than $5 to buy qty 1. I've had a laptop 3 years ago where a $0.20 ceramic capacitor shorted out and would prevent the machine from running. Plenty of oscilloscopes and other T&M equipment with bad tantalums -- too many to count. All cost on the order of $1.
Moore's law only talks about a number of transistors, not about how performant a chip is! If your design scales to multiple threads/cores, and can utilize simpler and faster cores, then you can stay with a fixed number of transistors. Current architectures are not an end-all be-all, yaknow.
I don't think that, necessarily going below 50nm is all that great, at least not for a company like Apple where their custom chips are for the portable market. Small geometries increase leakage! It's probably better to optimize existing designs, or even do major redesigns or paradigm shifts with existing technologies. Say if Apple wanted to add a bunch of micro-threads to iOS and the hardware it runs on -- small but fast threads that have small memory (say 64kB) but do things very quickly. They could easily do that without going to a smaller geometry, because such microthread CPUs don't take too much silicon. That's but an example.
Slashdot Mirror
A device like his, properly done, should cost no more than $30 done in market quantitites (10k+). Even a prototype, properly done, would cost less than $100 in materials, easy. How the heck did he spend $300 on his thing I just don't know. For $300 you could build a 16 channel EEG logger...
EOG can be done right, if you know what you're doing. The way he is doing it -- it barely works. He cut all the corners that there were to be cut. The quoted cost ($300 USD) is pretty silly for what he has done.
For noise mitigation and ease of use, you need a system that is physically small (forget any long wires) and uses integrated, reusable electrodes. It needs to be no harder to put on than eyeglasses. This seems like an obvious requirement. Who has time to play with electrodes?
This means the following (BTDT):
1. The electrodes are machined as rods with rounded ends out of titanium (lightweight!) or stainless steel. The inside can be drilled out to further reduce weight.
2. The electrodes are mounted on a PC board that straddles the nose bridge. They are on the inside corners the eye, aiding in keeping the system compact.
3. The PC board extends above the nose bridge and has slots for a flat stretchable "headband" that pushes it against your lower forehead. You may need an adapter piece (a plastic shim) to keep it at the right angle, this fitting is done once per person -- until one changes the shape of one's skull, that is.
4. The PC board holds the differential preamp and a low power CPU that does A/D conversion and transmission over IR. The battery (two AAAs) can be on the other side of the head, on the strap.
Ideally, you'd fit a two-cell charger on board (it's one more chip), with provision for charging the two AAA NiMH cells on the strap via, say, USB. Then you don't have to worry about having to mess with replacing batteries, and the computer is the power supply for charging -- it's there already, might as well use the standby power supply for something :)
If you have access to a lathe and scrap metal, you can keep the cost of the entire thing under $100 in materials, even in quantity one, even with a custom PC board.
Since you are, presumably, in front of your computer, using IR is the simplest, lowest cost means of one way communication. On a Mac you have built-in IR receiver so you need no extra hardware on the receiving end.
His is a nifty project, but done with little attention to detail.
Yep. And that's what he said: trademark owners are required to protect their IP rights. Patent holders' won't risk "losing" anything if they don't fight every patent violation they learn about.
For patents, the filing date is what counts in this scenario. The one to file first "wins". There's more to it, IIRC, as you can claim priority on public disclosures and foreign patents. So if you publish something in a scientific journal, say, you have a year to file a patent for it, and your invention is protected retroactively since the publication date in the journal. Someone who knows U.S. patent law better feel free to chime in with corrections, I'm not 100% sure about it. I'll ask a patent lawyer at work to see if he has anything to add to that.
Forgot to add: this was about very local-in-scope laws -- municipal, no less. So the assertion that somehow only federal laws are exempt from copyright is false in light of this.
Thanks. One tries, one fails, one occasionally succeeds :)
Nope. See
293 F.3d 79
The issue in this en banc case is the extent to which a private organization may assert copyright protection for its model codes, after the models have been adopted by a legislative body and become "the law". Specifically, may a code-writing organization prevent a website operator from posting the text of a model code where the code is identified simply as the building code of a city that enacted the model code as law? Our short answer is that as law, the model codes enter the public domain and are not subject to the copyright holder's exclusive prerogatives. As model codes, however, the organization's works retain their protected status.
As long as you treat the codes as part of law, they are public domain. Easy. So, if a code is adopted with any changes (even reformatting), only the changed version as adopted into law is in public domain. The original "model code", if any, is still subject to copyright protection.
Well, if you adopt such standards then they become law, plain and simple. Alas, I messed up my link: this is bulk resource and they have various building codes for your perusal. No need to spend a single cent for it. I think it's unconscionable that laws are a subject of copyright protection, just as it is unconscionable that the Queen holds crown copyright of King James Bible (in the U.K.). Both are fucked up IMHO (yep, I'm strongly opinionated on that front). If it's a law, it should be accessible online in an unencumbered fashion. Make it a U.S. constitutional amendment, for all I care.
I believe that the U.S. courts repeatedly held that legislation is exempt from copyright protection. So, even if the NEC is protected by copyright as its printed in the NFPA publications, if you get a copy from your local legislation, you can do with it as you please. Thus bulk resource. Alas, the laws in all of Europe are in complete opposition to this: they have plenty of laws that incorporate fairly expensive standards, and the courts have held that it's OK. Just E.U. legislation (so we exclude member countries' laws!) includes by reference standards that would cost you tens of thousands of Euros to obtain!
Yeah, I was just wondering: so we have a sterile bag that promotes cell growth (of certain cells, at least). Now, how does that fix contamination of the thing you're going to culture? That's not to be trivially dismissed, I'd imagine that contamination of the "seed" material is just as big of an issue as the contamination of the growth medium.
Many people are not upgrading because there's no need to, and because application support is lagging. It has been only fairly recently (last year or so) that some macports started working correctly on snow leopard. There are still ports that do not work when compiled for 64 bits, so I still compile macports for 32 bits. The reason to update to 10.6, for me, was speed and stability. Those were killer features. Lion doesn't have anything that would be a killer feature, not for me at least.
Gee whiz, it's not hard to put an Apple label on a beige box ;)
I don't think that there is any concrete dam out there that's not designed to survive with passive safety only. That means it can get full and spill over the top and nothing bad will happen. Eventually, if you let it spill over for many years, the top would erode away, perhaps facilitating crack initiation, but this is nothing immediate. With earthen dams it's a different story, but there isn't too many of those that hold enough water to cause large scale damage. By large scale I mean, say, 10% of country's population would be affected.
I still hold that if you closed every flow control that can be closed on all of the U.S. dams, there'd be no major emergency resulting from that. Of course whatever secondary function the equipment on the dams may have would likely be compromised (say power generation), but that's, like, duh.
I don't think it would be a factor, but it's easy to test for so it's silly not to. It would end all arguments. If you suspect creep, you take a couple samples (here: magazines loaded with mechanical blanks), and test them under different temperatures. It's then real easy to see if you have creep or not. I'd suspect that we're dealing with a folk tale here anyway (the loaded mag story). It's too far fetched.
What I stated was hopefully an accurate reflection of reality. Our way of life is not really compatible with large scale disasters. Jost look at how bad it was in Japan, and that was just in one coastal area with the rest of the country being able to function more-or-less normally.
You may have more luck on township/county roads. If I had to flee any large city, I'd head for those and stay the heck away from any popular, large thoroughfare.
- Widespread failure of dam control systems that risk flooding large parts of the country.
Maybe if you lived in a small alpine country or in the Netherlands. There is nothing you could do to the dams in large countries like U.S., China, Russia, India, Australia, etc. that would affect significant number of people.
If this would be a nationwide attack, presumably all the infrastructure would be affected to such a degree that the society as we know it would collapse anyway. You're right that individual suffering may be reduced in the immediate aftermath (hours and days), but in the following months those same people who were initially spared would die due to disease and lack of food. The cities, with their high density of population, would become unsustainable pretty much overnight. Folks living out there in the farmland would have the best chances.
You'd need to look for a spike in data that has granularity of months, not years.
I have not heard of fatigue, nor any theories used to model it, in high school, and I think it's OK to leave it out. There's plenty of things more fundamental than that to learn about before you enter college. Fatigue is pretty much an engineering subject, and its detailed study would include theory of fracture, and that's not exactly an afternoon tea subject (nor a high school one).
Oh, all the half-truths in those posts are just saddening. Why can't anyone get it completely right, at least in the realm of first-order effects?
Fatigue is, by definition, when you have cyclic loads. Obviously a spring left loaded and left alone doesn't experience any more cyclic loading than it would without being loaded at all. So it's nothing to do with fatigue.
What they may be talking about is creep -- the relaxation of stresses in material. This process is highly temperature dependent, so if you think you're dealing with creep, you immediately set up multiple specimens at different temperatures. If whatever effect you observe (here: loss of spring tension) accelerates with higher temperatures, then you can be 99% sure it's creep pretty much without having to do any other testing at all. If, though, there is no change with temperature, then you can safely exclude creep.
This gave me an idea: I've searched for the word temperature in the discussions you linked to. And hey, there was ONE post in all the misinformed noise that came from someone who had a bit of a clue. I'll cite it:
The phenomena you guys are talking about is called creep, where a material slowly loses strength while elastically deformed. Elastic deformation is when a material is under enough stress to deform it, but when the stress is removed the material returns to it's original shape. Springs compressed within their limits are elastically deformed. Creep is mostly a function of temperature, and typically isn't an issue in environments below half the melting point of the stressed material.
The person quoted wasn't fully correct either. Creep has got nothing to do with loss of strength, at least it's not a first-order effect during creep. A spring that experiences creep is not becoming any less strong, its internal stresses merely become relaxed (they go down), and thus the zero length of the spring is moving. This makes the spring act, over time, as if you had been squeezing it less and less: the force at the end of the spring goes down over time. That's all.
This is the silliest thing I keep hearing. There's nothing about gasoline that causes it to go bad over time, except if there is VERY significant evaporation going on. If you have a hermetically closed tank, you're good. Even with evaporation, I've had a 1993 Camry sit for about 3 years without being refueled, it'd be started up maybe twice a year but there was nothing bad about the gasoline in the tank. The engine ran, to my ear, as well as it ever did. It'd start right up. I had kept the battery hooked up, and a standby (maintenance) charger connected to keep it ready.
If you have two cycle engines that need oil mixed with gasoline, then all you have to do is make sure the tank is full when you put them away, and that the level hasn't gone down to more than 50% of tank full. I've found a lawnmower that was sitting in a shed somewhere long enough (at least 5 years) that it appeared that 2/3 of the gas has evaporated from the tank. It still started up after about a minute of working on it.
There is a lot of counterfeiting going on in the passive component market. Most electronic equipment is brought down by passive components failing. There were few chip fiascos, and those usually didn't even have anything to do with the chip but with packaging and mounting (say nVidia gpus). Most of the electronics I repair have problems with passive components. Shorted out SMT capacitors, dried out electrolytics, a broken track, a poor connector somewhere. Sometimes a failing passive component can take semiconductors with it, especially power semiconductors. I'd say if you're repairing stuff, you'll be better than 50% successful if all you have for replacement parts is passives and wire. With modern SMT assemblies, we're often talking of parts that cost lestt than $5 to buy qty 1. I've had a laptop 3 years ago where a $0.20 ceramic capacitor shorted out and would prevent the machine from running. Plenty of oscilloscopes and other T&M equipment with bad tantalums -- too many to count. All cost on the order of $1.
Moore's law only talks about a number of transistors, not about how performant a chip is! If your design scales to multiple threads/cores, and can utilize simpler and faster cores, then you can stay with a fixed number of transistors. Current architectures are not an end-all be-all, yaknow.
I don't think that, necessarily going below 50nm is all that great, at least not for a company like Apple where their custom chips are for the portable market. Small geometries increase leakage! It's probably better to optimize existing designs, or even do major redesigns or paradigm shifts with existing technologies. Say if Apple wanted to add a bunch of micro-threads to iOS and the hardware it runs on -- small but fast threads that have small memory (say 64kB) but do things very quickly. They could easily do that without going to a smaller geometry, because such microthread CPUs don't take too much silicon. That's but an example.