Once upon a time, supercomputers were bunches of general-purpose cpu's, and you made them faster by connecting up more of them.
Now people have realized that massively parallel special purpose chips (like Cell and, even more so, GPU's) can be used to do general-purpose computing, and have started to add those to clusters. But those chips have a lower bandwidth:flops ratio than the x86 etc. CPU's that have been historically used; the gap between a computer's "peak" FLOPS (on an ideal job with no communication requirements to either other nodes or to memory) and the performance it actually achieves is wider using something like CUDA than on a standard supercomputer. CUDA machines are so bandwidth-limited that people use rather hairbrained data compression schemes to move data from place to place, just because all the nodes have extra compute power lying around anyway, and the bottleneck is in communication. (The example that comes to mind is sending the coefficients of the eight generators of an SU(3) matrix rather than just sending the eighteen floats that make up the damn matrix. It's a lot of work to reassemble, relatively speaking, but it's worth it to avoid sending a few bits down the wire.)
CUDA is wonderful, and my field at least (lattice QCD) is falling over itself trying to port stuff to it. Even though it falls far short of its theoretical FLOPS, it's still a hell of a lot faster than a supercomputer made of Opterons. But we shouldn't fool ourselves into thinking that you can accurately measure computer speed now by looking at peak FLOPS. It makes the CUDA/Cell machines look better than they really are.
There is a/huge/ spectrum of different chemicals to kill different sorts of pests, used in different ways, and with different mechanisms of action. Saying that "pesticides" affect nitrogen fixation is an overbroad statement.
While I agree with the sentiment of your post, you misrepresent the neonicotinoids. Nicotine evolved as a natural insecticide; it's only logical that we use it too if we want to kill insects. (When insects should be killed, of course, is not the question here.)
There are pretty stringent controls on how recently food can be sprayed before it's harvested; I'd worry far more about environmental degradation from the stuff than harm to food consumers.
The issue is that there are three points of view: strong government controls (a la Europe), weak government controls (what you're calling "capitalism"), and fascism. We have the latter right now.
My father is an entomologist for a university Extension Service. For those who don't know (non-Americans), the Extension Services are outreach arms of the universities set up to provide advice to the public. His main job is to advise farmers on pest control measures for crops, mostly cotton; the advice is often "if you spray to kill pest A, you'll also kill predator B, which eats pest C which is resistant to insecticide, and C will eat your cotton. So don't do anything and put up with A, they won't eat that much."
Many of the meetings are sponsored by chemical companies. There are responsible uses of insecticides; used wisely, some insecticides can provide a cost-effective way to increase yields with very minimal long-term environmental harm. But the chemical companies are corrupt as hell. They try to bribe the scientists with lavish gifts to publish studies that favor their products, and encourage farmers (and scientists) to use too much insecticide, or use it when it's not really appropriate. It's sham science done for the sake of greed, and it is disgusting.
On the flip side, there are "studies" that show environmental harm where there really isn't any -- either by misguided "everything must be grown organically" types, or by people pushing back against the chemical company propaganda.
It's hard to tell a damn thing from "studies" on this sort of thing, because everyone is so busy grinding axes that who's right and who's wrong gets completely lost. This makes me, as a scientist in another field where there is far less of that, rather angry.
What is not right is that Comcast is threatening to block ONE PARTICULAR SERVICE if they don't get paid.
Saying "Pay us or we will no longer carry your traffic" is business. Saying "Pay us or we will sort through all the packets and dump only those that come from Netflix" is unacceptable.
GP argued (correctly) that even when no free market exists, like with ISP's, Americans will believe that since they live in the Land of the Free then there has to be a Free Market (tm), and Capitalism will sort everything out, and for anything else to happen would be Socialism.
The anti-war crowd is the group that says "America is strong enough to survive without spending huge amounts of money and violating our principles doing something that won't make us safer in any significant way." This is the courageous position.
The pro-war crowd is the group that says "We have to invade Iraq to keep ourselves safe!", and demands an enormous military action for no real reason.
The brave person lets a wasp land on his nose, knowing he has nothing to fear from it; the coward sprays it with Raid.
How many billions of dollars did that cost the US? What else could have been done with those billions that would have done more to advance the interests of the American people and/or the cause of liberty around the world?
Withdrawal would have been possible with or without the surge. It was possible in 2005, 2006, and 2007. Maybe not the withdrawal you wanted under the conditions you want, but it was possible and feasible all along.
I'm in a big city with some coyotes, and tons of coyotes all over the desert outside. The people who live near the desert sometimes get their little fluffy critters eaten, but -- by what? It's either coyotes, or bobcats, or the giant owls that will eat pretty much anything smaller than them. Some of them probably run afoul of rattlesnakes, too, which we have about a zillion of.
Coyotes, along with bobcats and owls and snakes, are part of nature. They eat stuff, and if you keep little stupid creatures outside they might eat your little stupid creatures. None of them are going to eat *you*, or your kids, unless you tie them to a tree at night with a sign that says FREE DOG CHOW.
Coyotes are natural animals. They can live pretty much anywhere. I live in a city of a million people and we hear them in the city all the time, and nobody really minds. They mind their own business just like they have for millions of years.
The fact that they do what they always do and we appreciate it is no big deal. People put up roosting boxes for owls to control rodent populations too.
The 4/3 digital cameras are ordinary SLR's, with the the flippy mirror and the pentaprism. You're thinking of Micro 4/3, which is the different no-mirror system.
And that's the problem with their law. They need to be far more afraid of the gal with a Canon SX10 hidden under her hijab than the DSLR user carrying around a 400mm lens.
This was true once, but the new non-SLR digitals actually do manual focus better than the SLR's. They have an electronic viewfinder that will magnify the bit of the image you select when you twist the focus ring, so you can get higher precision for manual focus than you can get from a ground glass viewfinder.
More photons do NOT mean better resolution. More photons landing on each pixel means a better per-pixel signal-to-noise ratio. But in daylight, it turns out that even the relatively tiny pixels of a compact camera sensor deliver a perfectly acceptable signal to noise ratio for many purposes, so nobody cares.
There are no "ultra small, ultra sensitive" CCD's. If it's small it's not that sensitive; the efficiency per size of the best compact camera sensors (like the one in the Panasonic LX3) are about the same as the best digital SLR sensors, and in the end the only thing that matters for sensitivity is the size.
Optical glass doesn't absorb many photons. Glass-air interfaces reflect them, which is why we have coatings, which are pretty damn good these days. There are SLR lenses with upwards of 20 elements that spit out good crisp beautiful images. (One of the best tele zooms ever made, the Olympus 35-100 f/2, has 21 elements.) They have all those elements to correct for aberrations, which are what really craps on your image quality. The reason pictures taken with conversion lenses often suck is because either the conversion lens introduces aberrations or magnifies aberrations already present in the main lens, but using a good conversion lens on a good lens can produce good pictures -- it has nothing at all to do with lenses absorbing light.
More photons doesn't necessarily mean better pictures. If you collect too few photons your picture will suck (because of noise, not a lack of resolution), but in daylight you can collect a lot of photons with even a small sensor. ISO 100 on a point-and-shoot is about as good as ISO 600-800 on a Four Thirds digital SLR, which is perfectly capable of making beautiful poster prints. (I have 16x20's from both in my apartment.) The key is that the point and shoot has a very good lens, which is the *real* limiter on image quality in good light for pretty much every camera on the market.
The issue is that the sensel density on something like a Panasonic FZ50 or Canon SX series is astronomically high, such that you can count the eyelashes on a tweetybird at twenty paces with a 100mm lens. Many of these lens/sensor combinations have so many pixels that they are diffraction limited wide open. They just plain don't make full frame sensors with that many pixels, or (many) full frame lenses with that much resolution (in lp/mm).
I have a perfectly good SLR sitting in the other room that has a CCD in it. It is a bog-standard SLR in all other respects -- flippy mirror, optical viewfinder, lens mount.
Slashdot Mirror
Once upon a time, supercomputers were bunches of general-purpose cpu's, and you made them faster by connecting up more of them.
Now people have realized that massively parallel special purpose chips (like Cell and, even more so, GPU's) can be used to do general-purpose computing, and have started to add those to clusters. But those chips have a lower bandwidth:flops ratio than the x86 etc. CPU's that have been historically used; the gap between a computer's "peak" FLOPS (on an ideal job with no communication requirements to either other nodes or to memory) and the performance it actually achieves is wider using something like CUDA than on a standard supercomputer. CUDA machines are so bandwidth-limited that people use rather hairbrained data compression schemes to move data from place to place, just because all the nodes have extra compute power lying around anyway, and the bottleneck is in communication. (The example that comes to mind is sending the coefficients of the eight generators of an SU(3) matrix rather than just sending the eighteen floats that make up the damn matrix. It's a lot of work to reassemble, relatively speaking, but it's worth it to avoid sending a few bits down the wire.)
CUDA is wonderful, and my field at least (lattice QCD) is falling over itself trying to port stuff to it. Even though it falls far short of its theoretical FLOPS, it's still a hell of a lot faster than a supercomputer made of Opterons. But we shouldn't fool ourselves into thinking that you can accurately measure computer speed now by looking at peak FLOPS. It makes the CUDA/Cell machines look better than they really are.
Which pesticides?
There is a /huge/ spectrum of different chemicals to kill different sorts of pests, used in different ways, and with different mechanisms of action. Saying that "pesticides" affect nitrogen fixation is an overbroad statement.
While I agree with the sentiment of your post, you misrepresent the neonicotinoids. Nicotine evolved as a natural insecticide; it's only logical that we use it too if we want to kill insects. (When insects should be killed, of course, is not the question here.)
There are pretty stringent controls on how recently food can be sprayed before it's harvested; I'd worry far more about environmental degradation from the stuff than harm to food consumers.
The issue is that there are three points of view: strong government controls (a la Europe), weak government controls (what you're calling "capitalism"), and fascism. We have the latter right now.
My father is an entomologist for a university Extension Service. For those who don't know (non-Americans), the Extension Services are outreach arms of the universities set up to provide advice to the public. His main job is to advise farmers on pest control measures for crops, mostly cotton; the advice is often "if you spray to kill pest A, you'll also kill predator B, which eats pest C which is resistant to insecticide, and C will eat your cotton. So don't do anything and put up with A, they won't eat that much."
Many of the meetings are sponsored by chemical companies. There are responsible uses of insecticides; used wisely, some insecticides can provide a cost-effective way to increase yields with very minimal long-term environmental harm. But the chemical companies are corrupt as hell. They try to bribe the scientists with lavish gifts to publish studies that favor their products, and encourage farmers (and scientists) to use too much insecticide, or use it when it's not really appropriate. It's sham science done for the sake of greed, and it is disgusting.
On the flip side, there are "studies" that show environmental harm where there really isn't any -- either by misguided "everything must be grown organically" types, or by people pushing back against the chemical company propaganda.
It's hard to tell a damn thing from "studies" on this sort of thing, because everyone is so busy grinding axes that who's right and who's wrong gets completely lost. This makes me, as a scientist in another field where there is far less of that, rather angry.
See New York Times Co. v. United States.
That's already been done.
There's a giant file called "Insurance" floating around on wikileaks that's about 4GB in size. It's encrypted, and nobody knows what's inside.
What is not right is that Comcast is threatening to block ONE PARTICULAR SERVICE if they don't get paid.
Saying "Pay us or we will no longer carry your traffic" is business. Saying "Pay us or we will sort through all the packets and dump only those that come from Netflix" is unacceptable.
You missed GP's point.
GP argued (correctly) that even when no free market exists, like with ISP's, Americans will believe that since they live in the Land of the Free then there has to be a Free Market (tm), and Capitalism will sort everything out, and for anything else to happen would be Socialism.
And, frankly, anyone reading the news knows perfectly well b) is true.
[citation needed]
No, you have it wrong.
The anti-war crowd is the group that says "America is strong enough to survive without spending huge amounts of money and violating our principles doing something that won't make us safer in any significant way." This is the courageous position.
The pro-war crowd is the group that says "We have to invade Iraq to keep ourselves safe!", and demands an enormous military action for no real reason.
The brave person lets a wasp land on his nose, knowing he has nothing to fear from it; the coward sprays it with Raid.
the successful "surge" strategy
How many billions of dollars did that cost the US? What else could have been done with those billions that would have done more to advance the interests of the American people and/or the cause of liberty around the world?
Withdrawal would have been possible with or without the surge. It was possible in 2005, 2006, and 2007. Maybe not the withdrawal you wanted under the conditions you want, but it was possible and feasible all along.
That system is still being used (successfully) to fund large sectors of the arts in this country.
South Africa has a rapist as president, for that matter.
The way taxes are supposed to work is that you pay a tax and get something for it.
Paying a tax to the IFPI (or whoever) and getting nothing in return isn't taxation; it's confiscation.
Mod parent up.
I'm in a big city with some coyotes, and tons of coyotes all over the desert outside. The people who live near the desert sometimes get their little fluffy critters eaten, but -- by what? It's either coyotes, or bobcats, or the giant owls that will eat pretty much anything smaller than them. Some of them probably run afoul of rattlesnakes, too, which we have about a zillion of.
Coyotes, along with bobcats and owls and snakes, are part of nature. They eat stuff, and if you keep little stupid creatures outside they might eat your little stupid creatures. None of them are going to eat *you*, or your kids, unless you tie them to a tree at night with a sign that says FREE DOG CHOW.
This.
Coyotes are natural animals. They can live pretty much anywhere. I live in a city of a million people and we hear them in the city all the time, and nobody really minds. They mind their own business just like they have for millions of years.
The fact that they do what they always do and we appreciate it is no big deal. People put up roosting boxes for owls to control rodent populations too.
The 4/3 digital cameras are ordinary SLR's, with the the flippy mirror and the pentaprism. You're thinking of Micro 4/3, which is the different no-mirror system.
And that's the problem with their law. They need to be far more afraid of the gal with a Canon SX10 hidden under her hijab than the DSLR user carrying around a 400mm lens.
This was true once, but the new non-SLR digitals actually do manual focus better than the SLR's. They have an electronic viewfinder that will magnify the bit of the image you select when you twist the focus ring, so you can get higher precision for manual focus than you can get from a ground glass viewfinder.
No, you won't. But you can get pretty good resolution in good light, which is what matters if they're worried about spies taking pictures in a desert.
You're condescending and wrong, actually.
More photons do NOT mean better resolution. More photons landing on each pixel means a better per-pixel signal-to-noise ratio. But in daylight, it turns out that even the relatively tiny pixels of a compact camera sensor deliver a perfectly acceptable signal to noise ratio for many purposes, so nobody cares.
There are no "ultra small, ultra sensitive" CCD's. If it's small it's not that sensitive; the efficiency per size of the best compact camera sensors (like the one in the Panasonic LX3) are about the same as the best digital SLR sensors, and in the end the only thing that matters for sensitivity is the size.
Optical glass doesn't absorb many photons. Glass-air interfaces reflect them, which is why we have coatings, which are pretty damn good these days. There are SLR lenses with upwards of 20 elements that spit out good crisp beautiful images. (One of the best tele zooms ever made, the Olympus 35-100 f/2, has 21 elements.) They have all those elements to correct for aberrations, which are what really craps on your image quality. The reason pictures taken with conversion lenses often suck is because either the conversion lens introduces aberrations or magnifies aberrations already present in the main lens, but using a good conversion lens on a good lens can produce good pictures -- it has nothing at all to do with lenses absorbing light.
More photons doesn't necessarily mean better pictures. If you collect too few photons your picture will suck (because of noise, not a lack of resolution), but in daylight you can collect a lot of photons with even a small sensor. ISO 100 on a point-and-shoot is about as good as ISO 600-800 on a Four Thirds digital SLR, which is perfectly capable of making beautiful poster prints. (I have 16x20's from both in my apartment.) The key is that the point and shoot has a very good lens, which is the *real* limiter on image quality in good light for pretty much every camera on the market.
Learn to physics before posting.
The issue is that the sensel density on something like a Panasonic FZ50 or Canon SX series is astronomically high, such that you can count the eyelashes on a tweetybird at twenty paces with a 100mm lens. Many of these lens/sensor combinations have so many pixels that they are diffraction limited wide open. They just plain don't make full frame sensors with that many pixels, or (many) full frame lenses with that much resolution (in lp/mm).
Why does SLR imply film?
I have a perfectly good SLR sitting in the other room that has a CCD in it. It is a bog-standard SLR in all other respects -- flippy mirror, optical viewfinder, lens mount.
"Making a scene" is not grounds for violation of legal rights or for retribution.
We are supposed to live in a country where saying "Fuck you, sir" to the police is legally protected.