How about a driver for Linux on the PlayStation 3, which would let the PS3 RSX chip actually work for Linux apps?
I know reading the article is considered some sort of terrible burden around here, but if you're going to ask a specific question, you might as well see if the interviewer already asked it for you:
4) Do you expect this situation [x86-only support] to change in case that Sony allows Linux to access the RSX chip on its PowerPC-based PlayStation 3? We have no current plans to port our Linux driver to PS3.
"In the meantime, their children are outside getting burnt without sunscreen."
With recent studies on vitamin D showing a strong link to preventing the major cancers, putting tons of sunblock on someone seems to increase their overall risk of cancer by inhibiting the vitamin's production. If you are not in an equatorial region and/or don't spend much time outside (e.g. most Americans) then you should think twice about slathering on the 45 SPF.
OTOH, I wonder if the reason regenerative braking isn't as big a deal as I thought comes from a fundamental limitation in the scheme, or from efficiency losses in the regenerating/storage/discharge mechanisms in place. If the latter, then it might still be a huge advantage if the efficiency of the system can be improved.
There's definitely room for improving the Prius implementation. After all, it still has brake pads:-). Regeneration only provides a certain amount of braking power... Beyond that, they chose the safety, reliability, and familiarity of a standard brake pad system. It might be because of the stated design goal of having it feel like a normal car and not require any change in driving habits, or there might actually be a limit to how much braking power a system like that can even apply.
I'm guessing what regeneration the Prius does do is further limited by conversion inefficiencies, although I really have no idea how to estimate that or how much better it could get. Like you said, it's a very interesting topic.
if the primary gain is the efficiency of the ICE, then there may well be other ways to optimize that performance without incurring the significant mass penalty of the additional motor and battery pack. I realize constant velocity transmissions have been tricky at best, but progress in that field could just as easily yield the same efficiency gains as hybrids.
Certainly. As long as the hybrid system is still a "gasoline in, kinetic energy out" black box, it's still just an evolutionary technology. There could probably be other evolutions to increase the ICE efficiency to something comparable. For a technology to be truly "unbeatable" by an evolution of current methods, it has to be revolutionary... for example, it could change the external source of stored energy, such as plug-in electric or hydrogen-stored power.
The Prius shows how many KwH are regenerated through regenerative braking as you drive. I've done some back-of-the-envelope calculations based on what I've seen under city driving conditions, and found that it only adds a couple percent to my efficiency. Maybe 5% at most, although I'd guess closer to 2-3%. It's definitely not the main efficiency win of hybrids, although before I calculated it I assumed the same thing as you!
Here's some other possible reasons why my Prius got 47mpg on my last tank:
A smaller engine is more efficient. Ordinarily this is a tradeoff with power (c.f. Geo Metro), but the battery in a Prius makes up for that.
Shutting off the engine when slowing or stopped eliminates wasted gas that's just spinning the engine otherwise. It can even be shut off when cruising at lower speeds (~30mph) or driving in parking lots or traffic jams (e.g. stop-n-go up to 10mph). I've had it shut itself off for over 10 minutes before.
Internal combustion engines are more efficient at a "sweet spot." Cruising (such as on a freeway) is well below that sweet spot, and accelerating hard is above. With a hybrid, the battery assist allows the engine to run softer during hard acceleration, and then the battery is charged by running the engine a little extra at cruising speeds. In both cases, the internal combustion engine is operating closer to its peak efficiency than it would otherwise. The battery allows it to smooth out the load, like how a water tower smooths a city's water demand over the course of a day.
Note: I am not an engineer, and would welcome corrections to any of my ideas.
As other people have noted, "terroristic" is a well-defined legalese term. If you're going to blame some dialect for growing out of control, it's the extremely precise demands of law that cause it, rather than the typical American bastardization of English. (And yes, I am American.)
From uslegal.com: "It may mean an offense against property or involving danger to another person that may include but is not limited to recklessly endangering another person, harrassment, stalking, ethnic intimidation, and criminal mischief."
Here's an example from the article you linked: "[O]ne cannot get a job without work experience, but one cannot gain experience without a job."
Here's the current situation: One cannot prove one's innocence to the principal without giving trusted evidence, but one cannot give trusted evidence without being considered innocent by the principal.
It's parallel to th example I always think of for Catch-22: you need a permit to get into a secure building, but the only office where you can apply to receive the permit is inside the building.
The problem is that there are better extensions for the same thing that don't require you to go through a 3rd-party website proxy, which is presumably what makes the downloads so slow. This also means your downloads aren't tracked by said 3rd-party. (You could even just use http://www.kissyoutube.com/ for the same thing.) They should have had an "alternatives" section listed for each extension.
I frequently watch 1.4gb 2hr movies full screen on my computer without any noticeable loss in quality over a DVD version. The ones I watch are typically encoded with an MPEG-4 variant (Divx, Xvid)... not even a "next-gen" codec. That means your estimate of 500MB for a 45 minute standard definition show should actually look extremely good if properly encoded using H.264. (HD is another story, but if we're comparing to DVDs, that's already out of the picture.)
MPEG-2 really is quite shitty in comparison to any decent, modern video codec... that's why they need to crank the bitrate so high on video DVDs. The AC3 audio is also needlessly high-bitrate. Yvan256 (GP) was absolutely right to make fun of the OP for not understanding why video DVD box sets occupy way more discs than the data DVDs would, for nearly indistinguishable quality.
For an in-depth comparison of what various high-end codecs look like, check out doom9's 2005 Codec Shootout. You can compare high-resolution screen captures from various difficult scenes across multiple codecs and against the original source. You have to really zoom in to see some of the compression artifacts, and that's for a 700mb/120min video or a 1400mb/160min video (including a 128k audio track). Go up another 50-100% in bitrate to 500mb/45min, and add a year and a half of codec development since then, and try to claim that it can't be DVD quality.
It might not be a fixed number at all, in which case their estimate is as exact as they can be without going into explicit detail.
I'm guessing they will have a new process, executed every 6 months, which anonymizes all logs older than 18 months. How long would any given search remain non-anonymous under that approach? 18-24 months.
I always thought voting for some guy who might have ideas that might be more to your liking than some other guy's is far from 'democracy'.
You're right. It is far from democracy. That's the point. We do not have a democracy in the United States, nor are we supposed to. We have a Constitutional Republic. If we had a democracy, we could have a vote on whether to bulldoze your house and build a freeway, without buying the land from you first, and you would be SOL if it passed. That's tyranny of the majority, and it really sucks.
By having representative officials who are bound by the Constitution to respect people's inalienable rights to property and privacy, the idea is to prevent the abuses that would result from a straight democracy. (Whether it worked or not is another story.)
Using the sky or a white piece of paper may be interesting, but it probably won't give you anything you can use to calibrate the rest of your photos.
A better bet for isolating the noise your camera generates is to take completely black photos, using the lens cap and some extra covering (and a dark room) to make sure absolutely no light hits the sensors. This will let you make raw images of the "dark noise" and "bias noise" that your camera generates, and subtract those images from your real photos before doing any other processing.
"If you want accurate, buy the World Book or Encyclopedia Britannica encyclopedias."
Funny you should mention that. When Nature had a number of professors review articles from their (science-related) fields in WP and EB, they found that Wikipedia had fewer errors per word, as well as longer average articles. The professors picked terms in advance and then checked them in both encyclopedias, throwing out any where the EB had no article or the article length was "vastly different." (WP was not missing any of the tested articles that appeared in EB.)
Note: no encyclopedias are acceptable sources to cite in a research paper, regardless of accuracy. Your suggestion to only use it as a starting point is correct, but not for that reason.
I don't think the grandparent poster is trying to come up with a *stronger* hash, exactly. I don't think he expects a combination of three hashes to have a strength equal to the product of their strengths, e.g. MD5*SHA1*RIPEMD. Instead, he wants a hash which has the strength of Max(MD5, SHA1, RIPEMD). That way, if two of the three are broken but the other one is not, it is still as hard to find a collision as if he were just using the unbroken one.
Certainly if they are all broken, using more than one doesn't help at all. Also, if they are all of similar complexity (e.g. 2^80), using multiple parallel hashes wouldn't increase it more than a few powers. Using the math from your linked post, two 2^80 complexity ciphers with 160-bit keys might increase to 160*2^80 complexity, or about 2^87. It's nowhere close to the 2^160 that one might expect from having a 320-bit key, but the fault tolerance against any individual algorithm breaking is much higher.
You might ask why one wouldn't just use a true 320-bit key or higher, like SHA-512. There are two reasons why the GP's suggestion might be better: - A single 320-bit algorithm might be more vulnerable to researchers finding a single weakness that dramatically diminishes the security. (Yes, I know that flaws are usually found gradually, lowering the strength of an individual algorithm in steps rather than completely breaking it all at once, but a sudden full-break is even less likely to occur over multiple algorithms). - There might be more applications out there that support all 3 of MD5, SHA1 and RIPEMD than those that support your preferred higher-bit algorithm, which makes the GP's suggestion more compatible in the short-term.
No, it doesn't restrict their options. They require copyright assignment on any contributed code, so they can re-license it whenever they want, using whatever license they want.
This only restricts the options of other people who might want to fork the code.
"I'm sure there will be replies that say this or that OS had had that functionality for years, which is fine...there are some great OSes out there."
I'm not normally one of those people, and I understand that it doesn't have anything to do with the point you were making, but I was shocked by how closely most of those features matched what I do in GoboLinux/KDE. The uncanny similarity goes all the way down to having the folder for people's personal data be called/Users! The main one I'm missing is a filter-bar in my file browser. I'll have to see if there's a Konqueror plugin or something for that functionality, because I've used it in various apps like Amarok and Thunderbird, and it is very handy.
None of this takes away from your post, which compares XP to Vista. Actually, I want to thank you for writing that all out, since those are the kind of improvements I'm really interested in: the ones that actually make existing tasks on the computer easier to perform. This is the first time I've heard about most of them. I won't be buying Vista for myself, but it may be installed on my work machine within a few years, so it's nice to know the good side of what's coming.
OK, 3 reasons if you're paying for gas with cash. But note that gas stations already advertise prices to the thousandth of a dollar -- as far as I know, the US has never actually minted a mil -- and they already get rounded up to the nearest penny. I'm sure gas stations would be quite happy to round to the nearest nickel instead.
Since I got a Prius, I have been keeping a record of my gas purchases. It turns out the two gas stations I go to often (a Holiday and a BP) compute the total price based on their thousandths figure, and then round the total to the nearest penny whether up or down. I was also surprised to see myself getting a fractional penny credit half the time, but empirical evidence beats preconceived notions (yes, even on Slashdot). Thus, gas stations would be indifferent to the change.
Your actual points about using a nickel as the smallest physical piece of currency are all valid, and I would applaud such a move. I routinely throw pennies away (or in tip jars) rather than carry them home to my change jar, because it isn't even worth my time to do that. I would rather that stores simply not give them to me in change, so I don't have to deal with them. A few times at a local Chipotle they have given me a nickel instead of four pennies in change, which was a nice touch.
Your first impression is correct: an alternate tax like this wouldn't conjure up any money from thin air, nor should any proponent claim it does. However, it doesn't create any additional cost either, at least in total.
What it does is shift the cost from companies to companies which use a disproportionate amount of energy to make their profit, and people who use a disproportionate amount of energy in their daily living. The average company's tax burden increases by a similar amount that your salary does after income taxes aren't taken out of it anymore, which means that your after-tax salary and the prices of its product will be essentially unchanged. However, companies which are more energy-efficient will pay less, and companies which leave the computers on all night will pay more. This will inspire a 'race to the top' of efficiency on both personal and corporate levels.
Of course, if it's only applied per-kilowatt, it doesn't do much to boost clean energy sources. I'd like to see something closer to a pollution tax, where (e.g.) wind power would be exempt.
Guess which one works in MPlayer, Xine, VLC, ffplay, GStreamer, etc., and which doesn't?
Flash video works, and wmv doesn't consistently. Ooh, ooh, did I get it right? For me, yes.
I often have problems with recent versions of (or "protected" files encoded with) the Windows Media codec in both Xine and Mplayer. Flash Video, on the other hand (which can be grabbed as a.flv file from YouTube quite easily using the DownloadHelper extension for Firefox) plays just fine in Mplayer. Despite being a Linux user, I'd prefer the YouTube link. I'm sure.wmv is easier for you, and the point of my post was not to argue that you were wrong per se... just to remind you not to assume that everyone's else's experiences are the same as yours.
Of course, nearly ANY other major video format would be preferable: xvid, divx, mov, mpeg, h.263/264, x.264, theora, etc.
Slashdot needs to create a +10 "ridiculously informative" mod just for you. Making a convincing demonstration using your opponent's scare websites as a source should qualify you for some sort of medal.
10. Deep Thought - Hitchhikers Guide to the Galaxy (2005) The Deep Thought computer was created to answer one question. What is the ultimate answer to Life, the Universe, and Everything? As you might imagine, the question is a bit vague. So the computer begins to calculate the answer as it has been commanded. After 7.5 million years, the answer is 42. The amount of computer cycles to compute all possibilities is quite large. Calculating all the possibilities from a question is currently a busy project.
Interesting point. I think the researcher was on to some useful stuff, and you've identified a good one. It's too bad the VR sensationalism distracted from further inqueries like yours.
I don't really have any insights on why "textbook" memory would be less prone to false memories than "experience" memory, but I'd definitely be interested in further studies to tease them apart, find the differences in capability, and figure out ways to take advantage of that (e.g. in schools).
Thanks for the insightful post. If I weren't already in this discussion, I'd mod it up!:)
Let's get something straight: This was NOT a test of VR versus reality. One group interacted with a camera in VR, and the other group read about it with some pictures. (I don't have a problem with the study, but rather with bloggers who misinterpret it.)
As the post points out, could this be a serious problem for VR going forward?
No. It is going to create a problem for the concept of memories, which have always been volatile and unreliable, but for some reason are perceived as accurate fact-recall centers in our brain. Something has to force people to adjust how they think of "memories," and this suggests it might be VR.
Research into "flash-bulb" memories (e.g. "I can remember exactly where I was when I heard about the Challenger") has shown that people's confidence in their memory for small details is barely correlated with the amount of detail they actually recall correctly. Elizabeth Loftus's research into eyewitness accounts and false memories have already shown that it is possible to plant even completely false memories with a reasonable success rate, much less small differences in an otherwise real memory like whether a street sign in a video was a stop or a yield, or what specific features a digital camera has.
from the blog: "It wasn't tested, but I assume real experiences don't generate false memories to the same degree."
Actually, I would assume the opposite: allowing people to play with a real camera briefly would have the same effect.
Slashdot Mirror
I know reading the article is considered some sort of terrible burden around here, but if you're going to ask a specific question, you might as well see if the interviewer already asked it for you:
"In the meantime, their children are outside getting burnt without sunscreen."
With recent studies on vitamin D showing a strong link to preventing the major cancers, putting tons of sunblock on someone seems to increase their overall risk of cancer by inhibiting the vitamin's production. If you are not in an equatorial region and/or don't spend much time outside (e.g. most Americans) then you should think twice about slathering on the 45 SPF.
I'm guessing what regeneration the Prius does do is further limited by conversion inefficiencies, although I really have no idea how to estimate that or how much better it could get. Like you said, it's a very interesting topic.Certainly. As long as the hybrid system is still a "gasoline in, kinetic energy out" black box, it's still just an evolutionary technology. There could probably be other evolutions to increase the ICE efficiency to something comparable. For a technology to be truly "unbeatable" by an evolution of current methods, it has to be revolutionary... for example, it could change the external source of stored energy, such as plug-in electric or hydrogen-stored power.
Here's some other possible reasons why my Prius got 47mpg on my last tank:
Note: I am not an engineer, and would welcome corrections to any of my ideas.
As other people have noted, "terroristic" is a well-defined legalese term. If you're going to blame some dialect for growing out of control, it's the extremely precise demands of law that cause it, rather than the typical American bastardization of English. (And yes, I am American.)
From uslegal.com: "It may mean an offense against property or involving danger to another person that may include but is not limited to recklessly endangering another person, harrassment, stalking, ethnic intimidation, and criminal mischief."
Seems like it applies just fine.
Here's an example from the article you linked: "[O]ne cannot get a job without work experience, but one cannot gain experience without a job."
Here's the current situation: One cannot prove one's innocence to the principal without giving trusted evidence, but one cannot give trusted evidence without being considered innocent by the principal.
It's parallel to th example I always think of for Catch-22: you need a permit to get into a secure building, but the only office where you can apply to receive the permit is inside the building.
That one seems to be a justified inclusion.
The problem is that there are better extensions for the same thing that don't require you to go through a 3rd-party website proxy, which is presumably what makes the downloads so slow. This also means your downloads aren't tracked by said 3rd-party. (You could even just use http://www.kissyoutube.com/ for the same thing.) They should have had an "alternatives" section listed for each extension.
I frequently watch 1.4gb 2hr movies full screen on my computer without any noticeable loss in quality over a DVD version. The ones I watch are typically encoded with an MPEG-4 variant (Divx, Xvid)... not even a "next-gen" codec. That means your estimate of 500MB for a 45 minute standard definition show should actually look extremely good if properly encoded using H.264. (HD is another story, but if we're comparing to DVDs, that's already out of the picture.)
MPEG-2 really is quite shitty in comparison to any decent, modern video codec... that's why they need to crank the bitrate so high on video DVDs. The AC3 audio is also needlessly high-bitrate. Yvan256 (GP) was absolutely right to make fun of the OP for not understanding why video DVD box sets occupy way more discs than the data DVDs would, for nearly indistinguishable quality.
For an in-depth comparison of what various high-end codecs look like, check out doom9's 2005 Codec Shootout. You can compare high-resolution screen captures from various difficult scenes across multiple codecs and against the original source. You have to really zoom in to see some of the compression artifacts, and that's for a 700mb/120min video or a 1400mb/160min video (including a 128k audio track). Go up another 50-100% in bitrate to 500mb/45min, and add a year and a half of codec development since then, and try to claim that it can't be DVD quality.
It might not be a fixed number at all, in which case their estimate is as exact as they can be without going into explicit detail.
I'm guessing they will have a new process, executed every 6 months, which anonymizes all logs older than 18 months. How long would any given search remain non-anonymous under that approach? 18-24 months.
By having representative officials who are bound by the Constitution to respect people's inalienable rights to property and privacy, the idea is to prevent the abuses that would result from a straight democracy. (Whether it worked or not is another story.)
Using the sky or a white piece of paper may be interesting, but it probably won't give you anything you can use to calibrate the rest of your photos.
A better bet for isolating the noise your camera generates is to take completely black photos, using the lens cap and some extra covering (and a dark room) to make sure absolutely no light hits the sensors. This will let you make raw images of the "dark noise" and "bias noise" that your camera generates, and subtract those images from your real photos before doing any other processing.
Details of this method can be found here: http://photo.net/learn/dark_noise/.
"If you want accurate, buy the World Book or Encyclopedia Britannica encyclopedias."
Funny you should mention that. When Nature had a number of professors review articles from their (science-related) fields in WP and EB, they found that Wikipedia had fewer errors per word, as well as longer average articles. The professors picked terms in advance and then checked them in both encyclopedias, throwing out any where the EB had no article or the article length was "vastly different." (WP was not missing any of the tested articles that appeared in EB.)
See the Wikipedia page on External peer review for more.
Note: no encyclopedias are acceptable sources to cite in a research paper, regardless of accuracy. Your suggestion to only use it as a starting point is correct, but not for that reason.
I don't think the grandparent poster is trying to come up with a *stronger* hash, exactly. I don't think he expects a combination of three hashes to have a strength equal to the product of their strengths, e.g. MD5*SHA1*RIPEMD. Instead, he wants a hash which has the strength of Max(MD5, SHA1, RIPEMD). That way, if two of the three are broken but the other one is not, it is still as hard to find a collision as if he were just using the unbroken one.
Certainly if they are all broken, using more than one doesn't help at all. Also, if they are all of similar complexity (e.g. 2^80), using multiple parallel hashes wouldn't increase it more than a few powers. Using the math from your linked post, two 2^80 complexity ciphers with 160-bit keys might increase to 160*2^80 complexity, or about 2^87. It's nowhere close to the 2^160 that one might expect from having a 320-bit key, but the fault tolerance against any individual algorithm breaking is much higher.
You might ask why one wouldn't just use a true 320-bit key or higher, like SHA-512. There are two reasons why the GP's suggestion might be better:
- A single 320-bit algorithm might be more vulnerable to researchers finding a single weakness that dramatically diminishes the security. (Yes, I know that flaws are usually found gradually, lowering the strength of an individual algorithm in steps rather than completely breaking it all at once, but a sudden full-break is even less likely to occur over multiple algorithms).
- There might be more applications out there that support all 3 of MD5, SHA1 and RIPEMD than those that support your preferred higher-bit algorithm, which makes the GP's suggestion more compatible in the short-term.
No, it doesn't restrict their options. They require copyright assignment on any contributed code, so they can re-license it whenever they want, using whatever license they want.
This only restricts the options of other people who might want to fork the code.
C'mon, Altanar You're posting this now? It's not yet fixed.
(Yes, that's your own link. Read the discussion.)
"I'm sure there will be replies that say this or that OS had had that functionality for years, which is fine...there are some great OSes out there."
/Users! The main one I'm missing is a filter-bar in my file browser. I'll have to see if there's a Konqueror plugin or something for that functionality, because I've used it in various apps like Amarok and Thunderbird, and it is very handy.
I'm not normally one of those people, and I understand that it doesn't have anything to do with the point you were making, but I was shocked by how closely most of those features matched what I do in GoboLinux/KDE. The uncanny similarity goes all the way down to having the folder for people's personal data be called
None of this takes away from your post, which compares XP to Vista. Actually, I want to thank you for writing that all out, since those are the kind of improvements I'm really interested in: the ones that actually make existing tasks on the computer easier to perform. This is the first time I've heard about most of them. I won't be buying Vista for myself, but it may be installed on my work machine within a few years, so it's nice to know the good side of what's coming.
Since you're already at +5 funny, I'll give you your kudos here:
You win at the internet.
Your actual points about using a nickel as the smallest physical piece of currency are all valid, and I would applaud such a move. I routinely throw pennies away (or in tip jars) rather than carry them home to my change jar, because it isn't even worth my time to do that. I would rather that stores simply not give them to me in change, so I don't have to deal with them. A few times at a local Chipotle they have given me a nickel instead of four pennies in change, which was a nice touch.
Your first impression is correct: an alternate tax like this wouldn't conjure up any money from thin air, nor should any proponent claim it does. However, it doesn't create any additional cost either, at least in total.
What it does is shift the cost from companies to companies which use a disproportionate amount of energy to make their profit, and people who use a disproportionate amount of energy in their daily living. The average company's tax burden increases by a similar amount that your salary does after income taxes aren't taken out of it anymore, which means that your after-tax salary and the prices of its product will be essentially unchanged. However, companies which are more energy-efficient will pay less, and companies which leave the computers on all night will pay more. This will inspire a 'race to the top' of efficiency on both personal and corporate levels.
Of course, if it's only applied per-kilowatt, it doesn't do much to boost clean energy sources. I'd like to see something closer to a pollution tax, where (e.g.) wind power would be exempt.
Flash video works, and wmv doesn't consistently. Ooh, ooh, did I get it right? For me, yes.
I often have problems with recent versions of (or "protected" files encoded with) the Windows Media codec in both Xine and Mplayer. Flash Video, on the other hand (which can be grabbed as a
Of course, nearly ANY other major video format would be preferable: xvid, divx, mov, mpeg, h.263/264, x.264, theora, etc.
Slashdot needs to create a +10 "ridiculously informative" mod just for you. Making a convincing demonstration using your opponent's scare websites as a source should qualify you for some sort of medal.
Interesting point. I think the researcher was on to some useful stuff, and you've identified a good one. It's too bad the VR sensationalism distracted from further inqueries like yours.
:)
I don't really have any insights on why "textbook" memory would be less prone to false memories than "experience" memory, but I'd definitely be interested in further studies to tease them apart, find the differences in capability, and figure out ways to take advantage of that (e.g. in schools).
Thanks for the insightful post. If I weren't already in this discussion, I'd mod it up!
"(although they would have to create their own LED bar)."
The fact that you can use a pair of candles as a sensor bar suggests that cheap third-party wireless ones shouldn't take long to hit the market.
No. It is going to create a problem for the concept of memories, which have always been volatile and unreliable, but for some reason are perceived as accurate fact-recall centers in our brain. Something has to force people to adjust how they think of "memories," and this suggests it might be VR.
Research into "flash-bulb" memories (e.g. "I can remember exactly where I was when I heard about the Challenger") has shown that people's confidence in their memory for small details is barely correlated with the amount of detail they actually recall correctly. Elizabeth Loftus's research into eyewitness accounts and false memories have already shown that it is possible to plant even completely false memories with a reasonable success rate, much less small differences in an otherwise real memory like whether a street sign in a video was a stop or a yield, or what specific features a digital camera has.
from the blog: "It wasn't tested, but I assume real experiences don't generate false memories to the same degree."
Actually, I would assume the opposite: allowing people to play with a real camera briefly would have the same effect.