It seems that FEMA only uses a limited set of frequencies. Why not install notch filters at select access points and design the broadband to only use the remaining bandwidth (either in frequency space or via notch-resistant error correction protocols in the physical layer). The same could be done for ham radio users -- bandpass filtering outside the traditional X-meter bands used by SW radio operators.
Broadband use of powerlines does not have to create a broadband noise source.
Since any chemical reaction not completed results in some isotopic enrichment one might enrich U235 by, feeding the dissolved Uranium oxide produced by Thiobacillus Ferrooxidans from raw ore to the anaerobic Desulfovibrio ferrireducens where it would reprecipitate. Then feed the precipitated uranium oxide back to thiobacillus ferrooxidans to produce more uranium liquor to feed to desulfovibrio ferrireducens forming cascaded stages which would gradually enrich the U235 until it was useful for fuel rods etc.
WHOA DUDE! That's a very very interesting idea. AFAIK, isotopic enrichment is very common in biological systems. In fact, isotopic ratios are used to determine if the creature was a herbivore or carnivore (see this for more info). My understanding is that the differences in nuclear weight of different isotopes change the reaction kinetics and can interfere with enzyme dynamics. Since U-238 is not much different from U-235 in mass (compared to the differences among Iron or Oxygen isotopes) the enrichment might be modest. On the other hand, I suspect that one could artificailly evolve the bacteria for their enrichment prowess.
You have suggested a very interesting approach to what is often a very hard problem. Like you, I smell profit, but it may not come from uranium enrichment, but from carbon enrichment. Apparently, isotopically pure Carbon-12 should create diamonds (think semiconductors) with twice the thermal conductivity of mixed carbon diamonds. If biological enrichment were cheap enough, one could become a monopolist supplier of C-12 for high performance semiconductor manufacturing.
I'm surprised that more libraries don't have a library of CDs of various Linux distros and larger open source packages. I also wonder if some chariable OSS-minded soul could donate a pre-configured tighty locked low-end PC and CD-burner to a local library. With used PCs being so cheap, a basic setup (with a 100 GB IDE HD) would be under $200. Either the donor or librarians could make a set of for-checkout CDs or library patrons could make their own CD bundles (paying a nominal fee for media or bringing in their own media).
Do we need an open source project to create a simple locked linux library distro and easy-to-use CD maker?
Re:Broadcaster/Advertizer hacks for this
on
Smart Billboards
·
· Score: 1
Broadcasters wouldn't want to do this because the ads on the billboards would compete with the ads they play on their stations. The station doesn't own the billboard, Mobiltrak (or some advertising agency) does.
Good point. But if Mobiltrak data is used for radio station listener rankings, then hacking is good for broadcasters. If a broadcaster (call them K-HACK) looks popular (even by artificial means), then it can charge a higher rate for all of its ads. In fact, the data will inevitably bias advertizer's choice of radio stations, too. If an advertizer finds they are paying for 24 hours of impressions on a billboard because everyone seems to be listening to the K-HACK station, then that advertizer is likely buy radio ads on K-HACK.
If the advertising agency charges per impression, then advertisers also would not want to hack the system in this way.
Excellent point! Under most pay-per-impression rate schedules, the advertizer wants a fair system. But your point also suggests that Mobiltrak has an incentive to hack its own system - hacking it to bias results toward the radio station that maximizes display of ads with the highest negotiated per-impression price.
Broadcaster/Advertizer hacks for this
on
Smart Billboards
·
· Score: 5, Interesting
It would seem that a particular radio station or advertizer could hack the system and bias the billboard by leaving a box of cheap battery powered FM radios by the side of the road (or a weak FM transmitter). With all those radios tuned to the same station it would fool the billboard into thinking that the cars where tuned to that station. Thus the billboard would leave the same ad up and log high ratings for the station.
Find people with longer circadian cycles
on
Living on Mars Time
·
· Score: 4, Interesting
Not everyone has a body clock that runs on an exact 24 hour cycle. Some people's circadian rhythms run as fast as 23 hours/cycle, some as slow as 25 hrs/cycle. JPL could test its employees for their natural cycle. A few days in a sleep test chamber quickly show which people tend to get up earlier and earlier each day vs. those that get up later and later. Then, they could selectively use people whose body clock matches that of Mars. Of course, I would still pity the families of the people that are on Mars time.
The reminds me of an Artificial Life example. A scientist was simulating evolution using some little virtual animals. Built into the simulation was an energy cost for moving -- the virtual creatures were supposed to evolve efficient strategies for finding food and mates. The creatures quickly evolved the ability to move backwards becuase this was counted as negative movement which meant negative cost, so that actually gave the creatures more energy.
For every clever policy created by some scientist, game designer, economist, corporate manager, or clever politician, there is an even more clever counterstrategy that someone is bound to discover.
Distributed Genome Variability Analysis
on
Home DNA Sequencing
·
· Score: 4, Interesting
It would seem that products like this one (or maybe slightly more professional versions) would eventually support distributed human genome sequencing efforts by individuals. More data on the DNA sequences of more people would help scientists, biomed, and pharma types understand the genetic variability of people.
I guess the next frontier is Sequencing@Home with people bragging about how many of their own base pairs or chromosomes they have sequenced.
The point of the whole article is that exactly these sorts of rules dumb down presentations, making it difficult to present complex ideas.
Good point, I guess I did not explain what I meant by readability analysis. In the field of text document analysis, the entire point of readability analysis is not necessarily to get the highest (or lowest) possible score. For text documents, the readability score is often equated to the required school grade reading level of the audience. Thus, too low a grade level score means the text is dumbed down, whereas too high a number means the text is excessively complicated. Therefore the goal is to match the readability to the audience, not maximize a readability score.
The original articles illustrate this problem nicely. On the one hand and as you pointed out, Tufte complains that Powerpoint generally leads to dumbed-down slides. He finds that Powerpoint users create slides with low information content. On the other hand, the Columbia Accident Investigation Board cited excessive complexity in the Powerpoint slides associated with the accident. Clearly, most Powerpoint users author slides at too low a readability grade level and the Columbia engineers authored slides at too high a readability grade level.
Analysis tools could help slide creators to author at the right level of complexity and avoid common mistakes that obscure the content.
Although the article did a good job of discussing flight technology, they did not say enough on the market forces that might drive different scenarios. Its not clear whether Boeing's vision of direct point-to-point travel or Airbus's visions of mass-transit hub-and-spoke will be the future of air travel. On the one hand, the decline in business travel hurts the economics of offering quick direct flights to everywhere while new technologies like free flight aid point-to-point travel. On the other hand, its not clear whether people will tolerate multiple connections and long boarding processes required for larger aircraft like the A380.
I don't see how convergence is possible because different consumer groups seek different functionalities in the device. Some people want a gaming platform, others want a DVR, some need a single-person home computer, others need a family home theater, etc.
Any device that tries to be all things to all people will fail. The device will be forced to include too much expensive hardware. Each feature and port (a hot GPU, large HD, DVI, VGA, NTSC, DOCSIS cable, ethernet) adds cost and complexity that not everyone wants to pay for. And the device will be unprofitable -- how can a company make up its losses on the bloated hardware if a large fraction of people don't buy gaming software or pay for monthly services?
Dreams of a convergent device assume convergent usage patterns and that will never happen.
Dups provide a chance to post additional insights that emerge from the original story. I find that reading all the +5 comments from the first posting of the story provides more food for thought once the dup appears.
I wonder if someone is goign to make a killing by selling little RFID chip & reader detectors. Richard Stallman suggested RFID detectors and destroyers as a challenge for privacy adocates. Perhaps clothing with conductive/dissapative threads will be the next fashion trend (just don't count on your cellphone ringing if its inside your pocket;) ).
It seems most of us can agree that PowerPoint makes it too easy to make bad, form-over-function slide presentations. But why not produce tools that help the author check the readability/confusability of the slides. This scoring system could work on the slide-pack level or on the slide level. I can also see ways in which the scoring system could provide advice on correcting the problem.
I see the scoring system as checking the following 6 dimensions of readability. It should probably score each dimension separately because a bad score on each of the different dimensions yields a different recommendation for correction.
1. Legibility analysis: We've all sees slides that use illegible 10 point (or smaller) lettering. Sometimes small type is justified (e.g., for a necessarily complex data table that will be handed out to the audience) but usually it is bad.
2. Contrast Analysis: Yellow text on a white background is bad! Using purple and maroon to color-code two data lines is bad. A simple analysis of the colors in a slide would give one a contrast score and could even provide recommendations on how to move colors away from each other.
3. Object Density: Some slides are too dense. If we analyze the number of "features" on the slide or the ratio of information to whitespace, then we can give a density score. An appropriate score might vary -- I've noticed that German engineers (and myself) like information-dense slides.
4. Text-to-Graphics ratio: Slides with 100% words are bad and ones with 100% graphics (no words) are bad IMO. Scoring the ratio of words to graphic features might help people see if they are near the sweet-spot (whatever that is). The only problem with this dimension is that it is hard to assess the text-to-graphics ratio in information content terms -- one can add useless grpahics to a wordy slide and think that one has improved the text-to-graphics ratio.
5. Word Reuse: Repetition is good. If every slide uses different words, with no word overlap between slides, then audience comprehension will drop. This dimension can also catch terminology consistency problems -- such as when the presentation agenda slide uses different words than the slide titles for the respective sections.
6. Jargon Use: By scoring the slides against word-frequency data, we can detect the use of too many rare words in the presentation. There might be different word-frequency datasets for basic English, college English, mechanical engineering, medical research, etc. that lets the presentor see if their slides are right for the audience.
I'm sure that others might suggest other dimensions.
The two most common threats to consumers who would use the system would seem to be:
1. Charge Theft: the thief charges your card by bringing a payment terminal near you. This depends on the security of the payment terminals. If the credit card processing system authenticates the terminal, then it would be hard for the thief to use the terminal to get the money. Even if the thief steals a terminal, the only thing that would happen is that the money would go to the retailer where the thief obtained the payment terminal. The real threat comes from a home-made or modded terminal. But this approach also requires a break in to the credit card processor to hack a record for the hacked terminal to ensure that charges to that terminal goes to a destination of the thief's choosing.
2. Card Theft: the thief remotely steals a person's card. This seems highly unlikely. The card would need to provide enough data in a reasonable number of monitored transactions to enable the thief to deduce how the card would respond to any future transaction. I would assume that the system would use a highly encrypted challenge-response system that would make it hard to reverse engineer the parameters for the response from a reasonable number of data points. But if someone hacks or steals the algorithm that is used to create the cards, then all bets are off.
It seems like the system could be secure if the encryption is sufficiently good and the data terminals are well controlled.
The Octopus card is widely used in Hong Kong. Its a stored value card, so its anonymous. It started life in the MTR (the local mass transit system) and has since expanded to convenience stores, Macdonalds, Starbucks, etc.
Walmart gets a very large portion of their stuff from China. It would probably cause prices to rise on darn near everything in the US. That would not necessarily be a bad thing. Probably help out domestic manufacturing.
The impact of war on this country would be none too pleasant. Goods shortages mean price rises means inflation means rising interest rates. The war would also disrupt Chinese purchases of American debt -- further driving up interest rates.
With higher interest rates would come debt problems for many people (who's ready for 24% credit card interest rates and 8-12% mortgage rates?). Rising interest rates would also kill the affordability of housing. Housing prices would drop and a bunch of people would find that they own more on their house than the house is worth. I doubt that domestic manufacturers would be able to pick up the slack because they woudl not be able to afford to borrow the money needed to invest in equipment and people.
Its a global economy and this war would not be pretty for anyone (and I won't even go into the possibility of direct U.S. intervention in any attack on Taiwan).
It would seem that if you have a clear line of sight to multiple APs, then you could combine them and have more bandwidth than a single AP-channel connection would provide. A ganged, multi-AP setup would use directional antennas to talk to each AP without collisions. Such a setup would also help if one AP were overloaded or down for some reason. The only problem would be if all the APs you talk to were routed through some narrow pipe somewhere in the network.
There is a strange and subtle distinction, in my mind... There is a sense in which the conventional photographic negative holds a physical trace of the actual "light event," if you will, that created it. The digital does not. As soon as those photons are converted to information all trace of the original physical event is lost. The light data becomes completely ephemeral, completely fungible. But in pragmatic reality this is pretty much a philosophical point. In the end the reliability of an image can only depend on documentation of method and faith in the source. The transition from film to digital doesn't change that.
Interesting. At one level I see your point. A digital camera slices and dices the physical event into a serialized set of pixels (its as if you took film and passed it through a carefully designed cross-cut paper shredder). This slicing and dicing emphasizes the fragility of the image -- one stray cosmic ray in RAM or a bit of dust on the disk's directory and the image is lost. A file on a harddisk seems so much less real than a developed frame of film.
At another level, digital is actually more objective than film at capturing that "light event." On the one hand, a CCD imager is like a geometrically regular array of little photon counters. The output of a digital camera represents a quantitative sample measurement of the photons that fell in each of those little geometrically arranged areas. On the other hand, film operates much more stochastically -- the random chances that enough photons all hit each randomly distributed film grain defines the chance of each grain developing (at least 3 or 4 photons must hit a grain to make it develop). Each grain is a stochastic event at a stochastic location. Only in the aggregate, with thousands of these randomly energized and developed grains, do we perceive the developed latent image. Looked at from this perspective, digital seems more real and film seems more capricious.
Ultimately, though, you are right. It is faith in the process that determines whether film or digital images are "real." Both can be manipulated.
No reason to be a pretentious twat (unless that's your particular superiority complex). Some people want realism. Some people want fantastic exageration.
Sorry . I do tend to get worked up when people stress the supposed "realism" of photographs. So many of a photographer's techniques are about manipulating the image with perspective, composition, framing, exposure, and lighting that happen before pressing the shutter, developing the film, or getting into the darkroom. You are right that photography, like painting, spans a spectrum of styles. Each style has an associated "permitted" set of image transformations. To each his own.
My photography teacher saw this issue in terms of some styles making it too easy to create a visually interesting image. He prefered B/W because he felt that color was too easy. In his opinion, and perhaps the opinion of Adams, B/W reduced the image to its bare essentials and forced the photographer to make good images rather than rely of a flash of color.
It seems that one can detect meteors by radio. Meteors create transient ionized trails in the ionosphere that reflect radio waves that would not otherwise bounce of those ionosphere (HF and high-frequency shortwave signals). It even sem that there is software such as this tool for PC-based analysis of radio signals that can help one detect meteors and measure the intensity of the shower. Has anyone ever tried this?
Anybody who has ever gone beyond darkroom 101 knows that the best photographers do some of their best work with subtle manipulations in the dark room. Adams' zone system is all about remapping the intensities in the original scene onto a pleasing span of whites to blacks in the print. Adams himself said that "Dodging and burning are steps to take care of mistakes God made in establishing tonal relationships." Digital cameras and image manipulation programs only pickup where the relatively crude processes in the darkroom leave off.
Anyone who claims that photography is about objectively and accurately portraying the real scene knows very little about the nonlinear properties of human vision, film, and image reproduction systems and they know even less about art.
Coase's theory of the relationship between information gathering costs and organization size is interesting, but not the most interesting impact of the internet on politics. One side effect of low-cost high-speed information gathering (and distribution) systems is that the competing parties can adjust their offers to voters using a much more rapid feedback cycle. Intensive use of polls, focus groups, trial balloons, e-mail, etc. let candidates fine tune their message like never before.
The two party system engenders a careful political calculus of stepping just far enough over the middle to steal an opponent's votes without alienating the extremists in the party. The democrats will try to appear just far enough right of center and the republicans will try to appear just far enough left of center to win. Everyone is shooting for the same 50.1% of the electoral votes and has the information gathering systems and information distribution systems to get it.
Unless one side achieves a huge advantage through external events (e.g., Dean wills if the economy tanks, weather disrupts voting in a key state, etc.) this will mean more close elections that reveal the statistical inaccuracies of our voting systems. It won't surpirse me if the Supreme Court will again decide the outcome of a presidential election in the near future.
What the hell kind of laptop are you running that consumes 50W!?
I agree that most current laptops don't consume 50W (your figure of 16 W is more representative). But it seems that laptop power consumption is on the rise, especially with the advent of "desktop replacement machines". It also seems that one rationale for fuel cells would be to provide larger power budgets for smaller devices (to support 64-bit processors, larger displays, and watt-guzzling graphics chips for gaming). Thus, the 50W is an extrapolation based on both the trend of increasing laptop power consumption and the seeming attractiveness of designing for increased laptop power budgets.
Fuel cells would seem to provide the power needed to bring near-desktop performance to the laptop. Only after the sale, would the poor consumer discover that they must buy and schlep fuel cartridges on every trip.
Slashdot Mirror
It seems that FEMA only uses a limited set of frequencies. Why not install notch filters at select access points and design the broadband to only use the remaining bandwidth (either in frequency space or via notch-resistant error correction protocols in the physical layer). The same could be done for ham radio users -- bandpass filtering outside the traditional X-meter bands used by SW radio operators.
Broadband use of powerlines does not have to create a broadband noise source.
Since any chemical reaction not completed results in some isotopic enrichment one might enrich U235 by, feeding the dissolved Uranium oxide produced by Thiobacillus Ferrooxidans from raw ore to the anaerobic Desulfovibrio ferrireducens where it would reprecipitate. Then feed the precipitated uranium oxide back to thiobacillus ferrooxidans to produce more uranium liquor to feed to desulfovibrio ferrireducens forming cascaded stages which would gradually enrich the U235 until it was useful for fuel rods etc.
WHOA DUDE! That's a very very interesting idea. AFAIK, isotopic enrichment is very common in biological systems. In fact, isotopic ratios are used to determine if the creature was a herbivore or carnivore (see this for more info). My understanding is that the differences in nuclear weight of different isotopes change the reaction kinetics and can interfere with enzyme dynamics. Since U-238 is not much different from U-235 in mass (compared to the differences among Iron or Oxygen isotopes) the enrichment might be modest. On the other hand, I suspect that one could artificailly evolve the bacteria for their enrichment prowess.
You have suggested a very interesting approach to what is often a very hard problem. Like you, I smell profit, but it may not come from uranium enrichment, but from carbon enrichment. Apparently, isotopically pure Carbon-12 should create diamonds (think semiconductors) with twice the thermal conductivity of mixed carbon diamonds. If biological enrichment were cheap enough, one could become a monopolist supplier of C-12 for high performance semiconductor manufacturing.
I'm surprised that more libraries don't have a library of CDs of various Linux distros and larger open source packages. I also wonder if some chariable OSS-minded soul could donate a pre-configured tighty locked low-end PC and CD-burner to a local library. With used PCs being so cheap, a basic setup (with a 100 GB IDE HD) would be under $200. Either the donor or librarians could make a set of for-checkout CDs or library patrons could make their own CD bundles (paying a nominal fee for media or bringing in their own media).
Do we need an open source project to create a simple locked linux library distro and easy-to-use CD maker?
Broadcasters wouldn't want to do this because the ads on the billboards would compete with the ads they play on their stations. The station doesn't own the billboard, Mobiltrak (or some advertising agency) does.
Good point. But if Mobiltrak data is used for radio station listener rankings, then hacking is good for broadcasters. If a broadcaster (call them K-HACK) looks popular (even by artificial means), then it can charge a higher rate for all of its ads. In fact, the data will inevitably bias advertizer's choice of radio stations, too. If an advertizer finds they are paying for 24 hours of impressions on a billboard because everyone seems to be listening to the K-HACK station, then that advertizer is likely buy radio ads on K-HACK.
If the advertising agency charges per impression, then advertisers also would not want to hack the system in this way.
Excellent point! Under most pay-per-impression rate schedules, the advertizer wants a fair system. But your point also suggests that Mobiltrak has an incentive to hack its own system - hacking it to bias results toward the radio station that maximizes display of ads with the highest negotiated per-impression price.
It would seem that a particular radio station or advertizer could hack the system and bias the billboard by leaving a box of cheap battery powered FM radios by the side of the road (or a weak FM transmitter). With all those radios tuned to the same station it would fool the billboard into thinking that the cars where tuned to that station. Thus the billboard would leave the same ad up and log high ratings for the station.
Not everyone has a body clock that runs on an exact 24 hour cycle. Some people's circadian rhythms run as fast as 23 hours/cycle, some as slow as 25 hrs/cycle. JPL could test its employees for their natural cycle. A few days in a sleep test chamber quickly show which people tend to get up earlier and earlier each day vs. those that get up later and later. Then, they could selectively use people whose body clock matches that of Mars. Of course, I would still pity the families of the people that are on Mars time.
The reminds me of an Artificial Life example. A scientist was simulating evolution using some little virtual animals. Built into the simulation was an energy cost for moving -- the virtual creatures were supposed to evolve efficient strategies for finding food and mates. The creatures quickly evolved the ability to move backwards becuase this was counted as negative movement which meant negative cost, so that actually gave the creatures more energy.
For every clever policy created by some scientist, game designer, economist, corporate manager, or clever politician, there is an even more clever counterstrategy that someone is bound to discover.
It would seem that products like this one (or maybe slightly more professional versions) would eventually support distributed human genome sequencing efforts by individuals. More data on the DNA sequences of more people would help scientists, biomed, and pharma types understand the genetic variability of people.
I guess the next frontier is Sequencing@Home with people bragging about how many of their own base pairs or chromosomes they have sequenced.
The point of the whole article is that exactly these sorts of rules dumb down presentations, making it difficult to present complex ideas.
Good point, I guess I did not explain what I meant by readability analysis. In the field of text document analysis, the entire point of readability analysis is not necessarily to get the highest (or lowest) possible score. For text documents, the readability score is often equated to the required school grade reading level of the audience. Thus, too low a grade level score means the text is dumbed down, whereas too high a number means the text is excessively complicated. Therefore the goal is to match the readability to the audience, not maximize a readability score.
The original articles illustrate this problem nicely. On the one hand and as you pointed out, Tufte complains that Powerpoint generally leads to dumbed-down slides. He finds that Powerpoint users create slides with low information content. On the other hand, the Columbia Accident Investigation Board cited excessive complexity in the Powerpoint slides associated with the accident. Clearly, most Powerpoint users author slides at too low a readability grade level and the Columbia engineers authored slides at too high a readability grade level.
Analysis tools could help slide creators to author at the right level of complexity and avoid common mistakes that obscure the content.
Although the article did a good job of discussing flight technology, they did not say enough on the market forces that might drive different scenarios. Its not clear whether Boeing's vision of direct point-to-point travel or Airbus's visions of mass-transit hub-and-spoke will be the future of air travel. On the one hand, the decline in business travel hurts the economics of offering quick direct flights to everywhere while new technologies like free flight aid point-to-point travel. On the other hand, its not clear whether people will tolerate multiple connections and long boarding processes required for larger aircraft like the A380.
I don't see how convergence is possible because different consumer groups seek different functionalities in the device. Some people want a gaming platform, others want a DVR, some need a single-person home computer, others need a family home theater, etc.
Any device that tries to be all things to all people will fail. The device will be forced to include too much expensive hardware. Each feature and port (a hot GPU, large HD, DVI, VGA, NTSC, DOCSIS cable, ethernet) adds cost and complexity that not everyone wants to pay for. And the device will be unprofitable -- how can a company make up its losses on the bloated hardware if a large fraction of people don't buy gaming software or pay for monthly services?
Dreams of a convergent device assume convergent usage patterns and that will never happen.
Dups provide a chance to post additional insights that emerge from the original story. I find that reading all the +5 comments from the first posting of the story provides more food for thought once the dup appears.
I wonder if someone is goign to make a killing by selling little RFID chip & reader detectors. Richard Stallman suggested RFID detectors and destroyers as a challenge for privacy adocates. Perhaps clothing with conductive/dissapative threads will be the next fashion trend (just don't count on your cellphone ringing if its inside your pocket ;) ).
It seems most of us can agree that PowerPoint makes it too easy to make bad, form-over-function slide presentations. But why not produce tools that help the author check the readability/confusability of the slides. This scoring system could work on the slide-pack level or on the slide level. I can also see ways in which the scoring system could provide advice on correcting the problem.
I see the scoring system as checking the following 6 dimensions of readability. It should probably score each dimension separately because a bad score on each of the different dimensions yields a different recommendation for correction.
1. Legibility analysis: We've all sees slides that use illegible 10 point (or smaller) lettering. Sometimes small type is justified (e.g., for a necessarily complex data table that will be handed out to the audience) but usually it is bad.
2. Contrast Analysis: Yellow text on a white background is bad! Using purple and maroon to color-code two data lines is bad. A simple analysis of the colors in a slide would give one a contrast score and could even provide recommendations on how to move colors away from each other.
3. Object Density: Some slides are too dense. If we analyze the number of "features" on the slide or the ratio of information to whitespace, then we can give a density score. An appropriate score might vary -- I've noticed that German engineers (and myself) like information-dense slides.
4. Text-to-Graphics ratio: Slides with 100% words are bad and ones with 100% graphics (no words) are bad IMO. Scoring the ratio of words to graphic features might help people see if they are near the sweet-spot (whatever that is). The only problem with this dimension is that it is hard to assess the text-to-graphics ratio in information content terms -- one can add useless grpahics to a wordy slide and think that one has improved the text-to-graphics ratio.
5. Word Reuse: Repetition is good. If every slide uses different words, with no word overlap between slides, then audience comprehension will drop. This dimension can also catch terminology consistency problems -- such as when the presentation agenda slide uses different words than the slide titles for the respective sections.
6. Jargon Use: By scoring the slides against word-frequency data, we can detect the use of too many rare words in the presentation. There might be different word-frequency datasets for basic English, college English, mechanical engineering, medical research, etc. that lets the presentor see if their slides are right for the audience.
I'm sure that others might suggest other dimensions.
The two most common threats to consumers who would use the system would seem to be:
1. Charge Theft: the thief charges your card by bringing a payment terminal near you. This depends on the security of the payment terminals. If the credit card processing system authenticates the terminal, then it would be hard for the thief to use the terminal to get the money. Even if the thief steals a terminal, the only thing that would happen is that the money would go to the retailer where the thief obtained the payment terminal. The real threat comes from a home-made or modded terminal. But this approach also requires a break in to the credit card processor to hack a record for the hacked terminal to ensure that charges to that terminal goes to a destination of the thief's choosing.
2. Card Theft: the thief remotely steals a person's card. This seems highly unlikely. The card would need to provide enough data in a reasonable number of monitored transactions to enable the thief to deduce how the card would respond to any future transaction. I would assume that the system would use a highly encrypted challenge-response system that would make it hard to reverse engineer the parameters for the response from a reasonable number of data points. But if someone hacks or steals the algorithm that is used to create the cards, then all bets are off.
It seems like the system could be secure if the encryption is sufficiently good and the data terminals are well controlled.
The Octopus card is widely used in Hong Kong. Its a stored value card, so its anonymous. It started life in the MTR (the local mass transit system) and has since expanded to convenience stores, Macdonalds, Starbucks, etc.
This sugar eating bacteria battery looks more promising. Runs on sugar and has an 80% conversion efficiency.
Walmart gets a very large portion of their stuff from China. It would probably cause prices to rise on darn near everything in the US. That would not necessarily be a bad thing. Probably help out domestic manufacturing.
The impact of war on this country would be none too pleasant. Goods shortages mean price rises means inflation means rising interest rates. The war would also disrupt Chinese purchases of American debt -- further driving up interest rates.
With higher interest rates would come debt problems for many people (who's ready for 24% credit card interest rates and 8-12% mortgage rates?). Rising interest rates would also kill the affordability of housing. Housing prices would drop and a bunch of people would find that they own more on their house than the house is worth. I doubt that domestic manufacturers would be able to pick up the slack because they woudl not be able to afford to borrow the money needed to invest in equipment and people.
Its a global economy and this war would not be pretty for anyone (and I won't even go into the possibility of direct U.S. intervention in any attack on Taiwan).
It would seem that if you have a clear line of sight to multiple APs, then you could combine them and have more bandwidth than a single AP-channel connection would provide. A ganged, multi-AP setup would use directional antennas to talk to each AP without collisions. Such a setup would also help if one AP were overloaded or down for some reason. The only problem would be if all the APs you talk to were routed through some narrow pipe somewhere in the network.
There is a strange and subtle distinction, in my mind... There is a sense in which the conventional photographic negative holds a physical trace of the actual "light event," if you will, that created it. The digital does not. As soon as those photons are converted to information all trace of the original physical event is lost. The light data becomes completely ephemeral, completely fungible. But in pragmatic reality this is pretty much a philosophical point. In the end the reliability of an image can only depend on documentation of method and faith in the source. The transition from film to digital doesn't change that.
Interesting. At one level I see your point. A digital camera slices and dices the physical event into a serialized set of pixels (its as if you took film and passed it through a carefully designed cross-cut paper shredder). This slicing and dicing emphasizes the fragility of the image -- one stray cosmic ray in RAM or a bit of dust on the disk's directory and the image is lost. A file on a harddisk seems so much less real than a developed frame of film.
At another level, digital is actually more objective than film at capturing that "light event." On the one hand, a CCD imager is like a geometrically regular array of little photon counters. The output of a digital camera represents a quantitative sample measurement of the photons that fell in each of those little geometrically arranged areas. On the other hand, film operates much more stochastically -- the random chances that enough photons all hit each randomly distributed film grain defines the chance of each grain developing (at least 3 or 4 photons must hit a grain to make it develop). Each grain is a stochastic event at a stochastic location. Only in the aggregate, with thousands of these randomly energized and developed grains, do we perceive the developed latent image. Looked at from this perspective, digital seems more real and film seems more capricious.
Ultimately, though, you are right. It is faith in the process that determines whether film or digital images are "real." Both can be manipulated.
No reason to be a pretentious twat (unless that's your particular superiority complex). Some people want realism. Some people want fantastic exageration.
Sorry . I do tend to get worked up when people stress the supposed "realism" of photographs. So many of a photographer's techniques are about manipulating the image with perspective, composition, framing, exposure, and lighting that happen before pressing the shutter, developing the film, or getting into the darkroom. You are right that photography, like painting, spans a spectrum of styles. Each style has an associated "permitted" set of image transformations. To each his own.
My photography teacher saw this issue in terms of some styles making it too easy to create a visually interesting image. He prefered B/W because he felt that color was too easy. In his opinion, and perhaps the opinion of Adams, B/W reduced the image to its bare essentials and forced the photographer to make good images rather than rely of a flash of color.
It seems that one can detect meteors by radio. Meteors create transient ionized trails in the ionosphere that reflect radio waves that would not otherwise bounce of those ionosphere (HF and high-frequency shortwave signals). It even sem that there is software such as this tool for PC-based analysis of radio signals that can help one detect meteors and measure the intensity of the shower. Has anyone ever tried this?
Anybody who has ever gone beyond darkroom 101 knows that the best photographers do some of their best work with subtle manipulations in the dark room. Adams' zone system is all about remapping the intensities in the original scene onto a pleasing span of whites to blacks in the print. Adams himself said that "Dodging and burning are steps to take care of mistakes God made in establishing tonal relationships." Digital cameras and image manipulation programs only pickup where the relatively crude processes in the darkroom leave off.
Anyone who claims that photography is about objectively and accurately portraying the real scene knows very little about the nonlinear properties of human vision, film, and image reproduction systems and they know even less about art.
Coase's theory of the relationship between information gathering costs and organization size is interesting, but not the most interesting impact of the internet on politics. One side effect of low-cost high-speed information gathering (and distribution) systems is that the competing parties can adjust their offers to voters using a much more rapid feedback cycle. Intensive use of polls, focus groups, trial balloons, e-mail, etc. let candidates fine tune their message like never before.
The two party system engenders a careful political calculus of stepping just far enough over the middle to steal an opponent's votes without alienating the extremists in the party. The democrats will try to appear just far enough right of center and the republicans will try to appear just far enough left of center to win. Everyone is shooting for the same 50.1% of the electoral votes and has the information gathering systems and information distribution systems to get it.
Unless one side achieves a huge advantage through external events (e.g., Dean wills if the economy tanks, weather disrupts voting in a key state, etc.) this will mean more close elections that reveal the statistical inaccuracies of our voting systems. It won't surpirse me if the Supreme Court will again decide the outcome of a presidential election in the near future.
What the hell kind of laptop are you running that consumes 50W!?
I agree that most current laptops don't consume 50W (your figure of 16 W is more representative). But it seems that laptop power consumption is on the rise, especially with the advent of "desktop replacement machines". It also seems that one rationale for fuel cells would be to provide larger power budgets for smaller devices (to support 64-bit processors, larger displays, and watt-guzzling graphics chips for gaming). Thus, the 50W is an extrapolation based on both the trend of increasing laptop power consumption and the seeming attractiveness of designing for increased laptop power budgets.
Fuel cells would seem to provide the power needed to bring near-desktop performance to the laptop. Only after the sale, would the poor consumer discover that they must buy and schlep fuel cartridges on every trip.