It's not just France and Canda. France does generate 78.5% of its electricity (415Bkwh) from nuclear power, but the USA generates 21% of its electricity (808Bkwh) from nuclear power (second after coal). Canada produces only 13% of its power using nuclear plants (73 Bkwh). That may be higher regionally; Ontario has most of Canada's nuclear generation capacity (it's also the most populous province).
Nuclear is cleaner in air emissions than burning hydrocarbons (although natural gas in modern plants is also quite clean). However, the cost of nuclear generation is pretty high when you factor in all of the safety measures that we require in modern society.
It's not clear to me that nuclear is the way of the future, although I'm willing to keep an open mind and compare based on a realistic analysis of the costs and benefits. I think that nuclear may win out over coal in the short term, although fuel disposal is a concern. Alternatives such as wind are interesting, although they have capacity concerns and other uncertainties.
Yes, the CREDITS file is not comprehensive, and (at least in 1998), the copyright notices were somewhat better but also not definitive.
You might be interested in a (old, 1999) paper on how you can use the CREDITS file to try to figure out how the Linux kernel interacts by looking at how developers work on different parts of the system.
I think that the lack of definitiveness of the CREDITS file and copyright notices is very understandable given the way they are updated manually. It might be an argument for having meta-data associated with source code. This type of information is encoded in a revision control system if everyone uses is, but with open source, the author of some code fragment is not necessarily going to log into your cvs repository and edit the text. Someone else may be cloning the code and inserting it (with the author's GPL blessing). It would be nice if there were a standardized way to track these types of edits and code provenance. Might be especially useful in refuting SCO type FUD.
I haven't looked at what Linus is doing with enhanced copyright tracking; is that something that can be standardized/automated/shared with other projects?
The article didn't count oil that went into raising the turkeys (for working the land, pesticides, heat/light, transportation to the processing plant and so on). The process is certainly of use because it reduces the external fuel needs of the turkey operation. It's not clear to me, though, whether the process breaks even with these other costs taken into account. Does the process produce enough fuel to fully power all of the farms that generate the turkeys?
I'd also like to know more about the inputs to the system. 500 barrels of oil must take a lot of turkey guts, especially as they are mostly water (which is burned off during coking). According to a CFACT article, it takes 1 ton of turkey to make 640 pounds of oil, 100 pounds of gasses, and 60 pounds of solids. This is surprising to me, I thought more water would be present; they are getting 40% "useful" material out of the turkey waste. The top production is based on estimates of 200tons/day of turkey waste. How many farms feed into this turkey processing plant, and how many such plants are there?
Another issue is the cost per barrel. The FAQ has a bit more information, and it seems to suggest the price is still a bit high, although hopefully it can be brought down.
It seems this is an interesting conservation measure that can reduce the amount of consumption while doing something useful with waste byproducts. The waste reduction seems like the most useful benefit for now (perhaps the original motivation for ConAgra to get involved in the project, as the process "will dramatically reduce the charges it faces for disposing of waste").
I think it's too early, though, to start investing in turkey gut futures.
I have one of these sitting on my desk. It is useful for copying old articles from the library (the ones that are not available on-line). I've decided to keep all my research material digital (over a thousand documents go with me everywhere on my laptop). Being able to scan directly saves me the step of photocopying then scanning back home (and the whole point was to avoid killing more trees and lugging their carcasses back to my office).
The CapShare works quite well. It can easily scan and store three or four papers on one charge of the AA batteries. I find the quality is good if you can get the paper flat enough. For bound journals, it does about as well as a photocopier. You really need to cut the binding to get a great image, which is obviously not an option at the library. I have been using the scanner less and less these days as the University library has access to older and older articles available (thanks to the good efforts of ACM student volunteers and some publishers).
I see a future in which all academic papers are available digitally, so the need for the CapShare or portable scanner will wane in that regard. But, I do also use it for scanning handwritten notes, signed documents, and so on. Often, I want to scan these while I am far away from my desk. I'm no fan of over-featured cell-phones, but it is a pretty neat hack. I probably paid too much for the CapShare for the value I get(heady dot com days), but I would still be willing to pay extra for a cell phone that could scan documents. It's worth it to me even if it is moderately inconvenient and moderate quality.
For cold places, a more cost-efficient use of solar energy is for solar heating. Efficiency of around 70% can be achieved, giving low-grade heat which can be used to pre-heat air or water for a heating system. While a traditional heating system is still needed to bring the temperature up to desired levels, less fuel is needed to maintain the same temperature. At a GM battery plant in Oshawa, Ontario, such a system generated 455 kWh/m of solar energy per year, representing a contribution of 317 MWh annually. This is a savings of between $4,700 and $12,200 depending on the fuel used for traditional heating (1991 prices, CAD$).
Similar technology is being used in car-washes to pre-heat water. I wonder if the flexible material technology could be incorporated into a system that pre-heats in the winter, and in the summer generates electricity which could be used for cooling. The summer months have some of the highest electrical energy demands as air conditioners are running.
I would imagine that you might be able to reduce the heat incident on the building using this material in the summer. For example, instead of introducing the heated air into the building, you could vent it out the top using either convection or wind-powered turbines.
Solar heating seems like a great alternative. Our house has good southern exposure, and stays reasonably warm during the day even with the heat set low (10C). Right now it is about 19C while it is -22C outside (-7F).
Photo-voltaics may also be useful, but the parent comment's points are quite good explanations for why they aren't in widespread use. On the other hand, if the cost of installing a system is sufficiently low and the esthetic is good enough, this could be a reason for buildings like warehouses and so on to reduce their reliance on non-renewable energy.
This portion of the Visionics press release gives a bit more of a clue (though not much):
The application, developed specifically for a law enforcement agency, uses Visionics' FaceIt® ARGUS as the delivery platform for facial recognition capabilities and Wirehound's Birddog software on Java 2 Platform, Micro Edition (J2METM) technology-enabled mobile phone with a color display. The FaceIt ARGUS system automatically finds faces in a field of view and searches them against a mug shot database. Upon finding a match, the Birddog component generates a wireless alert to the phones used by mobile law enforcement officials, who are then able to verify the identity of the subject. The phones can store multiple images and are alerted when a new image arrives. Non-matched images are automatically discarded from system.
Other than buzz-words, it sounds like this is intended to check the identity of suspects who are in a custody situation, instead of just scanning everybody passing by. This could be helpful to police and to innocents, since the 'mothership' sends the mug shot to the phone for human verification of identification.
However, I think it might be easier in these situations to use fingerprints, and send the mug-shot to the phone. I guess that the difference is that COTS phones have the cameras to take pictures of faces, while fingerprinting would take special hardware. Presumably, a cruiser could carry a mobile fingerprint device, but it would take $$ to make a mobile version. If they could use COTS to the same effect, it could be a big win.
I'm pretty skeptical, though, having used a 'FaceIt' trial version. I was able to train it to my face, but only by presenting it with a lot of variations (lighting, angle, etc.). This training data isn't available with mug shots.
However, I could envision a case where a cop remembers a 'wanted' picture, pulls someone over because they look like the picture, then uses the phone to pull up the right picture for on-the-spot comparison. Might work. But I doubt we're there--yet.
I think most people are willing to pay a reasonable amount for content (after all, it costs $$ to create good content and provide the bandwith to make it available). Further, $$ provides an additional motivation to provide high quality content.
The problem as I see it is that it is too much of an irritation to actually pay for anything while you are surfing content. Can you imagine anyone plugging in a nickel every time they flipped stations on a TV? Or paying for unlimited use of a single station? I am unwilling to pay for any content on the web, since there is always lots out there that is 'close enough' in quality to what I need so there's no need to pay for the best stuff. The adequate is the enemy of the excellent here.
If these costs could be incorporated into your ISP bill at a reasonable rate (maybe $1-$5 per month based on usage) and distributed based on your usage, then sites you enjoy could be rewarded for generating content and providing the bandwidth for it.
Of course, such a scheme would need to be designed carefully to avoid privacy concerns, but I think this could be done by aggregating to the ISP level. It would be completely unacceptable to have site usage tracked back to individuals.
This model moves the net closer to its real economic model, which is really based on content, not respect as suggested. I respect Linus quite a bit, but I don't tend to consume much content generated by him (not counting kernel sources).
On the other hand, I consume quite a bit of content from Slashdot, although my respect level is generally not as high.
California is proposing a law requiring companies to take back their products.
To me, this seems like the right way to go, provided that the companies don't just ship the product out of country where it becomes someone else's problem.
If we force manufacturers to charge for the full cost of technology, instead of subsidizing them as tax-payers, then they will tend to develop interesting ways to reduce the cost of recycling. This also lets consumers integrate the price of disposal into the purchasing decision, rewarding companies that have cleaner products.
However, if you want that way 31337 toaster with embedded, overclocked, uranium cooled processor, then you are welcome to it -- provided that you pay for the full cost, including its safe recycling.
I think it is reasonable for a party to the conversation to submit a digital copy of what they received -- the participants should know that the recipient has a durable copy that they could use to show others.
While I don't have a problem with using this evidence from that point of view, I'm concerned about the ability to prove that the submitted evidence actually came from a particular sender. With a voice recording, we have bio-metric information encoded which may attest to the authenticity. What can we use for email or IM? I suppose the law enforcement will be relying on the logs of the ISP (both sender and receiver) to verify that a message was sent. How can they verify the content of the message wasn't altered by the recipient?
I assume that most ISP email logs, at least, do not maintain a copy of the content. I suppose digital signatures could be used for non-repudiation, but only if the sender signs the message.
In this particular example, the nude picture of the sender could be used as proof, but only if he did not make it available elsewhere on the net (otherwise, a malicious person could have gotten it). I doubt that is the case here due to the corroborating evidence, but I'm a little uncomfortable with leaving it up to judges or jury's to judge the credibility of such digital evidence.
Slashdot Mirror
Nuclear is cleaner in air emissions than burning hydrocarbons (although natural gas in modern plants is also quite clean). However, the cost of nuclear generation is pretty high when you factor in all of the safety measures that we require in modern society.
It's not clear to me that nuclear is the way of the future, although I'm willing to keep an open mind and compare based on a realistic analysis of the costs and benefits. I think that nuclear may win out over coal in the short term, although fuel disposal is a concern. Alternatives such as wind are interesting, although they have capacity concerns and other uncertainties.
Perhaps you would notice the wires extending from the patient's abdomen.
You might be interested in a (old, 1999) paper on how you can use the CREDITS file to try to figure out how the Linux kernel interacts by looking at how developers work on different parts of the system.
I think that the lack of definitiveness of the CREDITS file and copyright notices is very understandable given the way they are updated manually. It might be an argument for having meta-data associated with source code. This type of information is encoded in a revision control system if everyone uses is, but with open source, the author of some code fragment is not necessarily going to log into your cvs repository and edit the text. Someone else may be cloning the code and inserting it (with the author's GPL blessing). It would be nice if there were a standardized way to track these types of edits and code provenance. Might be especially useful in refuting SCO type FUD.
I haven't looked at what Linus is doing with enhanced copyright tracking; is that something that can be standardized/automated/shared with other projects?
I'd also like to know more about the inputs to the system. 500 barrels of oil must take a lot of turkey guts, especially as they are mostly water (which is burned off during coking). According to a CFACT article, it takes 1 ton of turkey to make 640 pounds of oil, 100 pounds of gasses, and 60 pounds of solids. This is surprising to me, I thought more water would be present; they are getting 40% "useful" material out of the turkey waste. The top production is based on estimates of 200tons/day of turkey waste. How many farms feed into this turkey processing plant, and how many such plants are there?
Another issue is the cost per barrel. The FAQ has a bit more information, and it seems to suggest the price is still a bit high, although hopefully it can be brought down.
It seems this is an interesting conservation measure that can reduce the amount of consumption while doing something useful with waste byproducts. The waste reduction seems like the most useful benefit for now (perhaps the original motivation for ConAgra to get involved in the project, as the process "will dramatically reduce the charges it faces for disposing of waste"). I think it's too early, though, to start investing in turkey gut futures.
The CapShare works quite well. It can easily scan and store three or four papers on one charge of the AA batteries. I find the quality is good if you can get the paper flat enough. For bound journals, it does about as well as a photocopier. You really need to cut the binding to get a great image, which is obviously not an option at the library. I have been using the scanner less and less these days as the University library has access to older and older articles available (thanks to the good efforts of ACM student volunteers and some publishers).
I see a future in which all academic papers are available digitally, so the need for the CapShare or portable scanner will wane in that regard. But, I do also use it for scanning handwritten notes, signed documents, and so on. Often, I want to scan these while I am far away from my desk. I'm no fan of over-featured cell-phones, but it is a pretty neat hack. I probably paid too much for the CapShare for the value I get(heady dot com days), but I would still be willing to pay extra for a cell phone that could scan documents. It's worth it to me even if it is moderately inconvenient and moderate quality.
Similar technology is being used in car-washes to pre-heat water. I wonder if the flexible material technology could be incorporated into a system that pre-heats in the winter, and in the summer generates electricity which could be used for cooling. The summer months have some of the highest electrical energy demands as air conditioners are running.
I would imagine that you might be able to reduce the heat incident on the building using this material in the summer. For example, instead of introducing the heated air into the building, you could vent it out the top using either convection or wind-powered turbines.
Solar heating seems like a great alternative. Our house has good southern exposure, and stays reasonably warm during the day even with the heat set low (10C). Right now it is about 19C while it is -22C outside (-7F).
Photo-voltaics may also be useful, but the parent comment's points are quite good explanations for why they aren't in widespread use. On the other hand, if the cost of installing a system is sufficiently low and the esthetic is good enough, this could be a reason for buildings like warehouses and so on to reduce their reliance on non-renewable energy.
I'm pretty skeptical, though, having used a 'FaceIt' trial version. I was able to train it to my face, but only by presenting it with a lot of variations (lighting, angle, etc.). This training data isn't available with mug shots.
However, I could envision a case where a cop remembers a 'wanted' picture, pulls someone over because they look like the picture, then uses the phone to pull up the right picture for on-the-spot comparison. Might work. But I doubt we're there--yet.
I think most people are willing to pay a reasonable amount for content (after all, it costs $$ to create good content and provide the bandwith to make it available). Further, $$ provides an additional motivation to provide high quality content.
The problem as I see it is that it is too much of an irritation to actually pay for anything while you are surfing content. Can you imagine anyone plugging in a nickel every time they flipped stations on a TV? Or paying for unlimited use of a single station? I am unwilling to pay for any content on the web, since there is always lots out there that is 'close enough' in quality to what I need so there's no need to pay for the best stuff. The adequate is the enemy of the excellent here.
If these costs could be incorporated into your ISP bill at a reasonable rate (maybe $1-$5 per month based on usage) and distributed based on your usage, then sites you enjoy could be rewarded for generating content and providing the bandwidth for it.
Of course, such a scheme would need to be designed carefully to avoid privacy concerns, but I think this could be done by aggregating to the ISP level. It would be completely unacceptable to have site usage tracked back to individuals.
This model moves the net closer to its real economic model, which is really based on content, not respect as suggested. I respect Linus quite a bit, but I don't tend to consume much content generated by him (not counting kernel sources). On the other hand, I consume quite a bit of content from Slashdot, although my respect level is generally not as high.
To me, this seems like the right way to go, provided that the companies don't just ship the product out of country where it becomes someone else's problem.
If we force manufacturers to charge for the full cost of technology, instead of subsidizing them as tax-payers, then they will tend to develop interesting ways to reduce the cost of recycling. This also lets consumers integrate the price of disposal into the purchasing decision, rewarding companies that have cleaner products.
However, if you want that way 31337 toaster with embedded, overclocked, uranium cooled processor, then you are welcome to it -- provided that you pay for the full cost, including its safe recycling.
I think it is reasonable for a party to the conversation to submit a digital copy of what they received -- the participants should know that the recipient has a durable copy that they could use to show others. While I don't have a problem with using this evidence from that point of view, I'm concerned about the ability to prove that the submitted evidence actually came from a particular sender. With a voice recording, we have bio-metric information encoded which may attest to the authenticity. What can we use for email or IM? I suppose the law enforcement will be relying on the logs of the ISP (both sender and receiver) to verify that a message was sent. How can they verify the content of the message wasn't altered by the recipient? I assume that most ISP email logs, at least, do not maintain a copy of the content. I suppose digital signatures could be used for non-repudiation, but only if the sender signs the message. In this particular example, the nude picture of the sender could be used as proof, but only if he did not make it available elsewhere on the net (otherwise, a malicious person could have gotten it). I doubt that is the case here due to the corroborating evidence, but I'm a little uncomfortable with leaving it up to judges or jury's to judge the credibility of such digital evidence.