One part of Contact that I particularly like is the "hacking" of the signal from Vega. If I were to explain how things are reverse engineered and why, this is a very good explaination. In this case, it is with something that is specifically designed to be reverse engineered (as opposed to deliberate obfuscation in the design), but the principles still hold here. The TV signal processing was a little bizzare and super-convienient that they happened to have a flexable video processor on-hand for such an event. I can't even imagine what it would take to make a "generic" analog video signal processor and be able to make it work, even if all you had to deal with was all known broadcast formats for television (NTSC, PAL, SECAM, Farnsworth's original broadcast format, etc.) What an alien might do would add considerably more complexity to the issue, like scanning vertically instead.
Some parts of the movie were unbelieveable, and how Dr. Arroway was rolled by the White House and accepted their authority to shut up about the whole thing when it was clearly a civilian research project, particularly with their anti-government/anti-establishment leanings with the research group before hand. When a similar message (originating from Earth, but intended for "alien" audiences) was posted here on/., it was cracked within an hour and fully decyphered within three hours, so why would they not send the message to practically every geek news site or bureau?
The problem is though, with the 1st Ammendment of the U.S. Constitution, it is specifically written and interpreted as protecting political speech, and Nazi propaganda clearly fits within that definition. That Germany and France have specific exceptions to the Nazi party in their constitutions is besides the point.
Similar arguments were made back in the 1950's and 1960's regarding Communist literature being distributed on college campii. Unfortunately, for good or ill, the Nazi party is recognized as a political party in America (together with the American Fisherman's Party, the Nudist Party, and a couple other fringe groups) and in the eyes of American law there is no difference between the Nazis and Republicans or Democrats other than sheer numbers of party members and votes collected in November every two years.
I'm not saying that I support the Nazi Party of America (or related groups like the Ku Klux Klan), but political speech no matter how bizzare is a protected right.
I would have to agree though, that if France wants to block internet traffic going into their country because it offends their censors, they should be free to do so. This is no different than France banning the importation of a magazine that is full of Nazi propaganda getting confiscated by French customs when you land at Charles de Gaulle International Airport in Paris.
The trick, however, is trying to identify that "offending" internet traffic. That a French citizen could obtain the same thing through an e-gold account and some traffic on Freenet also needs to be taken into account when trying to set up such a censor system. I don't think it would be technically possible to block any forms of internet traffic unless you block all internet traffic, which essentially cuts off that group (like France) from the rest of the world if they try this approach.
I think you miss the point here. This is an American company doing business principly in America. The memorabilia in question (Nazi flags, ect.) are being offered by Americans.
The question comes up if the fact that this site can be viewed in France, and bid upon by French citizens because there is no way to identify them as such, should French law apply to a company doing business essentially in America?
This goes way beyond if France is a soverign entity. France could always do the same thing that is done in China, and ban certain websites that don't fit their criteria of public decency laws. Should E-bay and in this case Yahoo also be banned in France? Where do national laws end and what should be done when international commerce is being effected and who has the authority to regulate that commerce?
In the USA, there is a long standing precedence regarding interstate commerce being regulated by Federal courts, and I would imagine if this were a dispute between somebody in France and Germany that similar legal mechanisms are in place within the context of the EU.
IMHO, this is where France seems to thing that the USA is a member of the EU, and clearly it is not. That is the whole issue, and unfortunately the government of the USA has a tendancy of thinking in trade talks that it has one vote against the 15+ votes of the EU.
I would sum this whole issue up in one "bumper sticker quote":
Texas: Bigger Than France
And that is an American attitude. If they can sell more products in Texas than all of France, why worry about France? Especially when the Texas National Guard could by itself invade and defeat France militarily.
How is a ruby dagger any different from an obsidian knife? That is one of the oldest weapon implements ever made by mankind, so if you are talking about something new, this is not it. And one of those knifes could clearly be deadly in a cockpit situation. With a leather body, handle, and sheath, it could go totally undetected in that regard.
Still, having something with the strength of aluminum but with high transparency would be useful for many applications. Ships hulls (portals), including in submarines as well as in space would be quite useful. The holding tanks (aka Star Trek IV) might be interesting as well, although how this fits with transporter technology is yet another subject of discussion.
Most of the European settlements and much of the Eastern U.S.A. was built around water transportation, while most of the Western USA was built around railroad transportation systems, and later internal combustion transport systems (like the Interstate Highway system... that has been absorbed into the growth of the western USA...it is still almost an artificial construct in the eastern USA except in suburbia). You can see some of the rail routes in the lights, especially in areas through the Rocky Mountains, where the grid pattern of the mid-west breaks down. This is true even for areas where the rail lines are no longer in service, but the old rail lines were necessary for the creation of some of these towns.
What happened to passenger rail transportation in the western USA is also a subject of piles of commentaries that I won't get into, but it pretty much is for hobby rail lines with tourists and the few people who are paranoid about traveling by air...and even then you have to pay a healthy premium to travel by rail. Amtrack got hammered after 9/11, but business has since gone back to half-empty trains that are heavily subsidized and still overpriced and only run during months of a blue moon. It wasn't always that way. The train station where I live was turned into a restaurant, and that only because there was a historical easment to keep it from getting demolished. It hasn't been used for passenger travel since WWII. This is very typical throughout most of the western USA, and even much of the USA in general. Even the major train terminals are pretty much gutted and only kept for historical reasons, usually by a government entity of some sort.
This does, however, make it a nice "skin" to wrap around a sphere when doing 3D modeling. Any other projection would be a nightmare in trying to work with, except maybe a couple of polar projection maps. Even then, this gives virtually no distortion on the equator, and the poles are essentially irrelevant in this situation, unless you are trying to get the lights of the antarctic research station on the south pole.
I happen to like this numbering system. If you give me an address in Salt Lake City (or most of Salt Lake County as well), I can usually find the location within 3 or 4 blocks even if I don't have a map and have never been there before. The problem comes when somebody decides to build a subdivision in a new area and call it something more generic like Elm Street or Oak Street. When that happens, you are totally lost and even the locals at a gas station are clueless trying to help you out. This system is used throughout most of Utah, south eastern Idaho, and some of the "Mormon" areas of Arizona... at least parts that were settled by Mormon pioneers. If you are from Arizona you know what I mean.
BTW, the second part of the address that you were mentioned, like 1900 East, is the name of the street. It could be called 1900 East Street, but because all streets are called streets (not streets and avenues), for shorthand they call it simply 1900 East. Roads going east and west (hence are north and south of the center of town) are named 400 North or 600 South.
If you talk with locals it gets even more confusing yet, because when the cities in Utah were originally set up they didn't county by hundreds like is currently designated. Instead they simply called the street 1st to the east from the center of town 1st East. The next one was 2nd East, ect. that way 1900 East Street is sometimes called simply 19th East, or a famous place in Salt Lake County, 49th Street Galleria (since renamed) was on 4900 South. The 100's designation do help so a street that would be halfway between what should be 400 North and 500 North would be 450 North. That gets rid of really crazy street names (I've seen them) like 5-3/4er Street.
Because of the influence of people from outside of Utah who simply don't get it, some streets are getting names again, like Dr. Martin Luther King Jr. Avenue or Freedom Blvd. These usually also have an address like 600 South or 400 East, and the post office will deliver to either street name. And you think it is easier to find a place with that kind of address?
BTW, many mid-western cities also have somewhat similar street coordinates, where they have an addresses that are counted from the center of town, and the streets are labeled 1st Street, 2nd Street, etc. and then have a city quadrant like NW, SW, NE or SE. Streets go north & south and avenues go east & west. Or the opposite, but consistant throughout the city. The problem you come across is that sometimes you will have eight different 11th Avenues, each having a sub-range of addresses and broken up because of rivers, freeways, shopping malls, or other natural & artificial features. Stil, it isn't hard to find a place like 1354 15th Avenue SW.
The author is trying to describe the effects of acceleration, and specifically the acceleration due to this phenomona. He would have been better off trying to describe the gravitational effects of being on Phobos in comparison, but most people don't even understand that very real situation.
The size is not the size of an apple, but the gravity so weak that an object like an apple would take forever (seemingly) to accelerate due to that force. It makes sense to me, but it could have been better written, particularly if you got the apple mixed up as the size of the effect.
Actually, a photon has mass. Remember Einstein's famous equation?
E = mc^2
This mean, in a grossly simplified and totally non-scientific viewpoint, that mass and energy are simply forms of each other. Normally when you are talking about nuclear reactions you are more concerned about the transformation of mass into energy, but the opposite also occurs. Simple things like the production of electron/anti-electron pairs due to x-rays at a certain frequency that happens due to the energy of a photon being at the same energy level. Other bizzare things happen with elementary particles as well.
Getting back to photons having mass. Since energy by itself has mass, it can also be affected by mass. The space-time geometry descriptions are another valid model, but simply describes this basic fact about how gravity affects mass/energy.
Indeed, if you shine a bunch of photons into a very concentrated area simultaneously, you could in theory be able to form a "black hole" that is created purely out of photons, not normal matter. This is called a "kugelblitz", and behaves in some strange ways somewhat different from a normal black hole. This is theoretical anyway and if a normal black hole is still being debated regarding their existance, one of these guys is way out in terms of if it even exists.
So yeah, a photon doesn't have "rest mass", but indeed it does have mass, and an amount that you can measure.
E = hv (h is Plank's Constant and v is actually the greek letter nu, or the frequency of the light energy).
hv = mc^2
hv --- = m c^2
That is at least one way to calculate the actual amount of mass that a single photon would have at a particular frequency. It is pretty damn small for visible light frequencies, but still not zero. Photons don't stay still anyway, so talking about rest mass of a photon is kinda silly.
While the business model may be to recycle the cameras and charge for printing, I think they got the costs down fairly well here, with the cost of goods being quite a bit less than $20. They will make a small profit just from selling many of these units in volume.
I would actually hope they embrace hackers... this is not the group that they are targeting anyway, but instead people like my mother that don't want to hassle with setting up a printer, loading paper, replacing/refilling ink cartrages, geting drivers installed for file transfer, trying to find a file manager that could read the data from the camera, etc. Instead, you can grab the camera, point, click, and return the camera for prints of your daughter's wedding while you buy groceries and have the prints waiting for you when you get to the checkout counter.
I think there is a market for this sort of thing, and the cameras are cheap enough to make it worthwhile to folks who don't want the up front cost of settting up a computer system. It is not like photo processing shops won't help you with more expensive cameras either, but the market for really cheap cameras otherwise just isn't there yet.
Encode that information? Actually, yes you could. This is something, BTW, that the Kodak Photo Studio (*.psd files) do very well, as they encode the pixel data points as floating point numbers rather than integers.
The problem you are describing is due to the fact that display devices have a limited dynamic range, usually due to costs involved with trying to get the light emitters to produce light at their maximum intensity. There are usually limits due to one color or another. With monitors and color television sets, trying to get a bright red was a struggle for quite some time. With LED systems red and green LEDs have been around for quite some time so they are comparatively cheap compared to blue LEDs, and the maximum intensity with LED systems is driven by the quality of the blue elements. Since the pigment of blue LEDs also has a shorter life than the green or red LEDs, the displays of this nature don't last as long either. "White" LEDs are usually a combination of pigments that include red, green, and blue elements, and even there the blue fades quicker, making LED flashlights over time turn yellow.
Actually, I've worked with displays that have sub pixels with individual lightbulbs, like you've mentioned. Using discrete lightbulbs have their own characteristics that are somewhat different from monitor or LED phosphors due to the fact that you can treat the light coming from them as a black body curve, while phosphors generally emit light using quantum state transitions (going back to ye olde chemistry and elemental photon quantum state transition levels). They are in practice pretty much the same thing. With the light bulbs, however, you get a shifting spectral peak with temperature, and since the brightness was controlled by time limiting the period per power cycle that the bulb was on, you had to integrate the total power in an AC power cycle to determine just how many photons (roughly) were emitted, and then further mess stuff up by trying to gamma correct the figures to discrete power levels per pixel to give a good approximation on balancing colors. Colors were added by putting a colored filter in front of the bulb.
Oh, and the size of these discrete subpixel displays? About 100' by 30', and sucking up about 1 MW of power. A good thing I didn't have to pay the electric bill on those things. While these put out a whole lot of light, in reality the way you generated hues is no different than the way you do with a conventional computer monitor. BTW, it was a blast to play Quake on one of these monster displays.
Strictly speaks, yeah, you can't derive luminance directly from a color triplet, but you can get pretty close, and it is possible to do an RGB to Y conversion (the luminance portion of the XYZ coordinate system, also used for YUV or YCrCb color spaces as well, and the conversion for Y is the same for all three color spaces). For a quick & dirty conversion, I generally did 1/2 green + 1/4 red + 1/8 blue + 1/8 green (to make 1...green figures strongly with your perception of how bright something is) to come up with a good luminance value for B&W displays, as this conversion could be done with integer math and values like 1/2, 1/4, & 1/8 can be done with a shift instruction...multiplication wasn't even necessary. This made it nice with embedded systems as well. This rough approximation is pretty close to the CIE color point and even closer to what is used now for HDTV standards.
Re:Until I upgrade my eyes, why should I care?
on
RGB to become RGBCMY
·
· Score: 1
You are correct on all of these points so far. Also keep in mind that some people see not just three colors but a few see four, some see just two or in an extreme case just one (true color blindness). Blue-Green color blindness being the most common.
BTW, I am convinced that the "high quality" interior decorators are among those people with tetrachromaticity (being able to see with 4 colors). If there are people that see with 4 colors, I'm sure there are some with even broader ranges yet, even if it is just a very small number. Since I am color blind to those people, I can only remotely imagine what their world is like through their eyes. From what I understand, this extra color perception is some extra cones that perceive Orange directly, and has survival advantages because you can identify diseases easier, particularly in children. There was a writeup on this topic awhile back here on/.
In order to accurately depect a function using the six colors, it would have to be something like this:
HUE generate_hue(R,G,B,C,M,Y)
This is by definition a six-dimensional function. That you are confusing this dimensionality with our normal 3D world is a mistake. This is not just six vectors but total number of dimensions that would be needed to fully explain every potential point that could be generated with this system. By reducing this to a tri-value representation you are going to lose information. There is no way to reverse that sort of function, which is what I'm trying to point out.
If you define a color coordinate system as three coordinates (i.e. XYZ or RGB or YUV) you can't be able to get to all possible points available with a six-coordinate system. That is the whole point I'm trying to make here is that if you use six colors for display, you need to somehow define each and every potential color and have that encoded as the representation of that hue when used for data storage. By only recording three coordinates you will by definition lose information and lose hues, which makes it pointless to have more than three colors for display unless those other coordinates are there.
One way that image compressors work often, as you were alluding to earlier, is to transform the color coordinates to something like YUV, because you notice changes in brightness much more than the hue itself. Analog Color television does something like this as well, where most of the bandwidth is occupied with the luminance signal, and only a small portion is used to actually define the hues. That you don't see "banding" with color TV is only because it is an analog signal instead of a digital signal. Colors on television are very hard to get hues to display correctly, particularly reds. You can do the same thing with a six-coordniate system, but you would have to do a transformation through six values. You may only have to store some of those values in one or two bits, but it would still have to be there to be of any value in the overall system.
No, by reading between the lines, it is if you use functions in the database, like if you use function in the operating system to format a disk or retrieve a data file, you are technically in violation of what they consider to be "linking" to their software.
The real question is at what point does linking stop being linking. They point out that even connections that are based on raw TCP/IP, according to MySQL AB, are in violation of the GPL. In rough technicalities, if I write some software that scrapes data from a web page that uses an Apache webserver using MySQL for content, and mind you the web page scraper is written just for that specific website, I would be technically in violation of the GPL from their viewpoint. I would have to release that scraper under the GPL from this viewpoint.
That is where I think MySQL AB is overeaching on their reading of the GPL. In that last example, I may not even be aware of the fact that the web site is even using MySQL. It would also make using Linux or other GPL'd operating systems illegal (following this philosophy) when running propritary software. Since databases are moving into the operating system environment, this is going to be even more of an issue, although I don't thing MySQL is going to be the key component when that happens.
Actually, I think you have an idea here. While luminance can be derived from the rest of the colors, Reflectivity and Alpha blending is something that must be added seperately.
In encoding six colors, it give you only 48 bits, where 64 bits is more natural for computers to manipulate (just as 24 vs. 32 bits for a normal RGB is where alpha usually gets thrown in as the extra byte per pixel). What to do with that extra byte of data in addition to Alpha.
Re:Until I upgrade my eyes, why should I care?
on
RGB to become RGBCMY
·
· Score: 1
There is quite a bit more to explain here. First of all, while you might be able to see in three different wavelengths of light, others will see with a totally different set of three wavelengths of light, so what may look just fine to you will look off or even quite a bit different to another person. This is called genetic variations, and there is quite a bit of variance between different people.
Also, you need to remember that an RGB monitor is not really well tuned to specific freqnencies that your eye can perceive. I would dare say that for the most part it is pretty close, but it is off, and by having other colors to display it would "normalize" the display even more so your eyes would notice subtle color variations more to your normal experience with things "In Real Life(tm)". What drives monitor development is certain phosphors that emit light at a given frequency or a set of frequencies that are approximately close to one of the primary colors. While "popular" phosphors will come close to one of the major frequecies that your eye perceives best with one of its color receptors, that is also on average, and only an approximation.
You got the idea more or less correct when you mind perceives color as a ratio of your "red", "blue", and "green" color receptors. There is more, however, because these color receptors don't just see one specific frequency, but see along a whole range of frequencies that are close to the "peak" frequency of that receptor. In fact, there are sometimes more than one peak for a given color receptor across the whole EM visible light band, although there is clearly a dominant peak as well.
By adding extra colors to "smooth out" the color frequency distribution to make a given hue, or to give it several peaks in different places rather than just three peaks with a tri-color system, you can more faithfully reproduce the color. Dealing with non-human perception would require even more range, like how honey bees can see into the UV band, or some other animals being able to see into the infared band. You can see "chords" to form a hue, but this is quite limited as you can only see about 1 octave from red to blue.
This is not just precision loss due to conversion, this is a loss of information due to moving from a 6 dimensional representation to a 3 dimensional representation of a number. Information loss is going to occur when you do that, and you can't wave your hand around trying to pretend otherwise. Try to do a conversion of 3 dimensions to 2 as an exercise and get all three back somehow, like going to flatland and poping back to 3D world.
Yes, changing from one 3D representation to another 3D representation (ie RGB to YUV to XYZ) can be in theory kept free of loss, but this is not the same thing as converting for more than 3 colors to one of these 3D systems. This is raw information theory here.
While I would agree that your eye measures discrete "pixels" as well, the human eye is an analog conversion device, not digital. Frequency and intensity are measured by your eye through analog processes, and processed in your brain through an analog "computer", not a digital one. Furthermore, what one person perceives is slightly to substantially different than what somebody else will see, because their eyes respond to different frequencies, and the cones in their eyes can respond to more or fewer number of frequencies than yours. Even individual optic cones that are in theory receptive to the same color will respond to slight variations of a given frequency, which provides a slight "fuzzing" or blurring of the frequency range.
At best a restrictive color scheme will only work with a theoretical "typical human eye", and most people would still be able to notice a difference. Not only that, but producing materials that are dead on to emit or reflect light at precisely the peaks of human color perception are very difficult at best to manufacture. That means you need to make an encoding scheme for color that takes these other materials into account, and deals with people who have a wide range of varying color perceptions.
No, the issue is if MySQL AB has made certain that the code being distributed under commercial license is 100% certified by the developers as having copyright clearance for that commercial distribution. I would venture to say they have done exactly that, the GPL not withstanding. That is a big task, and what you are claiming is that somebody has added software to MySQL and not given permission to MySQL AB to redistribute it commercially, even though that software is being used in their product.
All I can say then is to "show me the money", or in this case the source code you question in their commercial distribution. $200 is paltry if it is your software, and the potential benefits including financial on your end if they are using your code would be quite substantial, just from a statutory viewpoint of somebody violating your copyright. I don't think you can find that code which is in violation.
No, I think you got that one wrong. The GPL allows you to redistrubte their code on your own website, flea market, subway station, commercial bookstore, etc. You can even distrubute a product that includes their code, as that is legal and encouraged by the GPL. You can even sell it for substantial profits, that you can even put directly into your own pocket, like $100 or $500 per CD if you can get buyers for it. None of that has to go to MySQL AB.
What MySQL AB is saying is that if you write software under a propritary license, and it somehow uses MySQL, you should purchase a commercial license of MySQL before you can sell that software that requires MySQL. That is why the argument regarding if the Linux kernel, which propritary software can and does link to accessing OS functions that are available only under the GPL, is brought up in comparison and the question begs to be asked at what point does MySQL AB really lose its argument with this linking requirement.
The GPL does not really specify this "right" that MySQL AB is claiming, and it doesn't really make sense to me. If I made changes to MySQL and relabeled MySQL as my own software, that would be a violation of the GPL. If instead I am installing MySQL or OpenOffice on a CD-ROM that also happens to include some propritary commercial software that I am selling for a profit, I don't see how MySQL AB should be demanding profits via the GPL from this propritary distrubution.
I can build a distribution of Linux and sell it for $10,000. That nobody may buy it is another story, but I can certainly try to sell it for that price, and I am not compelled to even place this distro on the internet (contrary to some people's viewpoints on the subject). You just have to make sure that all GPL'd software also includes the source code if you distribute the binaries with it.
I think you would be very hard pressed to find some source code in the commercial version that was "contributed" by somebody without a specific copyright waiver. I would say that MySQL is in much more sound legal water than Linux in general, specifically the Linux kernel, which obviously has some legal questions, the litigous idiots in Lindon, UT notwithstanding. Most of the code has been contributed by employees of MySQLAB.
You simply can't gain information once lost from a color conversion.
If you are talking about visual compression algorithms that are lossy, like JPEG, I would have to agree that often they do some sort of conversion like you are suggesting. Lossless compression systems, on the other hand, do no such thing.
While data formats like PNG or GIF can compress data using a YUV color pallet, that is because the information was originally stored in that format, not because that was the original color space.
A YUV or YCrCb to RGB conversion, even floating point signed RGB will eventually lose some information and is difficult to convert back. Very seldom are they stored as even signed numbers because of raw data storage limitations....bits for a color in an individual pixel are quite precious and are not automatically increased arbitrarily unless there is a good reason to do it.
The point here is that if you have more three chromatic color for a given hue, a conversion to YUV or some other encoding system (like CIE XYZ) is going to lose some data along the way. I actually dare you to prove otherwise to me. While the shade of color reproduced with such a conversion might be close, it won't get the color back all of the way and will lose some information, and information that will be noticed by a good eye.
Re:Color is not a discrete phenomena!
on
RGB to become RGBCMY
·
· Score: 2, Insightful
The problem is, and I'll admit this, that different people percieve color differently. While there might be a model that you can call a "typical" human eye color gamut, you need to go to hard physics ultimately in order to pull out the other colors.
Photoshop experience and an artistic eye can pull out colors to make them more life-like and even treat the other three colors in a hex printing pallet like colors on an oil-based paint pallet, but in reality you can't obtain new information that isn't there unless it was encoded in the first place. You add a little bit of that information with a good photoediting piece of software like Photoshop. An RGB color space is fairly good, and a reasonable model, as is the "color wheel", but it is just one model that works reasonably well. There is a point that ultimately it breaks down, and that is the point I was trying to make earlier. That you can create 70%-80% of all of the colors in human experience makes them very useful models, especially as the remaining colors are seldom seen by most people, and there are many other issues involved with art like proportion, balance, and perspective that are just as important if not more important. That colors get pretty close means you can concentrate on the other issues instead.
Trying to explain the value of even an RGB system is quite difficult to those who are color blind and barely see two colors, or are purely monochromatic in their vision is particularly difficult. What is worse is that often they don't realize that they don't see all of these colors.
My background is more along trying to engineer systems that can accurately display and portray colors for most people, which is why I have gone more for a purely scientific viewpoint. Having to deal with more unusual color gamuts like a pure RG system (systems that only display red and green, due to costs to add blue to the display), and RGBW systems (where you have the normal RGB and add white for additional contrast... and you though CMYK was tough). I did some limited experimentation with violet LEDs and some very dull near infared LEDs as well. They give some colors that are quite interesting, and unfortunately I never had the chance to see a full display made up of these colors tied together with RGB LEDs, like is being suggested by the article mentioned as the parent article. While understanding the physiological issues regarding color perception (and we did deal with them), we had a much easier time dealing with color from a raw physics viewpoint when designing our systems, in part because we were working on a more physical system level. I had to also deal with the user interface and trying to come up with a color picker that would work with these sometimes unusual color spaces.
No, standard NTSC (and PAL and SECAM) use a YUV system. That is, a Luminance value followed by two other sub-bands that control red and blue that get subtracted out of the base luminance signal (the remaing color being green). This is called YUV, and is also why NTSC is sometimes called "Never The Same Color", because getting this to work on an analog signal is particularly difficult. This system was chosen because the "Y" part of the signal was backward compatable with B&W television sets, so those old sets could work even with the new color signals added to the television broadcast.
Digital Video broadcasts will change this, but both RGB and YUV signals are going to be used by broadcasters. YUV in part because they don't have to rework as much equipment. I have never heard of RGBK used in television.
There is a "black mask" to help separate the individual phosphors from each other, and to add definition from the seperate color guns in the back of the CRT. On cheaper televisions with crude phosphors, the electrons from one color tend to "bleed" into the other phosphors, and make the image look "fuzzy" when you view them on a TV screen. This only applies however to CRT-based television sets and not plasma screens or LCD monitors. The same also applies to quality vs. cheap CRT-based computer monitors, for the same reason. The black mask has nothing to do with the color space, but with the physical construction of the monitor.
This is a BS alert that is valid, other than the fact that if you are digitizing film pigments that you would pull out some subtle hues that could not normally be found with a typical RGB scanner/encoding system. If you are talking about some extra processing of something that was previously encoded with just RGB pixel information, you are absolutely correct. This is utter BS.
Where you can see a hexcolor data encoding system in use is if you film the original scene with a camera that can capture the full gammut. Until that happens it would be like taking old films from the early 1900s and digitizing them in full RGB color, even though it is mostly B&W film. The subtle differences in stock and perhaps choices to have a "colored" stock would make a difference from the perspective of the editor or director and your viewing experience. If you digitized that same file with just luminance values you would miss that important piece of the movie, but of course that is for purests of film culture, right?
Slashdot Mirror
One part of Contact that I particularly like is the "hacking" of the signal from Vega. If I were to explain how things are reverse engineered and why, this is a very good explaination. In this case, it is with something that is specifically designed to be reverse engineered (as opposed to deliberate obfuscation in the design), but the principles still hold here. The TV signal processing was a little bizzare and super-convienient that they happened to have a flexable video processor on-hand for such an event. I can't even imagine what it would take to make a "generic" analog video signal processor and be able to make it work, even if all you had to deal with was all known broadcast formats for television (NTSC, PAL, SECAM, Farnsworth's original broadcast format, etc.) What an alien might do would add considerably more complexity to the issue, like scanning vertically instead.
/., it was cracked within an hour and fully decyphered within three hours, so why would they not send the message to practically every geek news site or bureau?
Some parts of the movie were unbelieveable, and how Dr. Arroway was rolled by the White House and accepted their authority to shut up about the whole thing when it was clearly a civilian research project, particularly with their anti-government/anti-establishment leanings with the research group before hand. When a similar message (originating from Earth, but intended for "alien" audiences) was posted here on
The problem is though, with the 1st Ammendment of the U.S. Constitution, it is specifically written and interpreted as protecting political speech, and Nazi propaganda clearly fits within that definition. That Germany and France have specific exceptions to the Nazi party in their constitutions is besides the point.
Similar arguments were made back in the 1950's and 1960's regarding Communist literature being distributed on college campii. Unfortunately, for good or ill, the Nazi party is recognized as a political party in America (together with the American Fisherman's Party, the Nudist Party, and a couple other fringe groups) and in the eyes of American law there is no difference between the Nazis and Republicans or Democrats other than sheer numbers of party members and votes collected in November every two years.
I'm not saying that I support the Nazi Party of America (or related groups like the Ku Klux Klan), but political speech no matter how bizzare is a protected right.
I would have to agree though, that if France wants to block internet traffic going into their country because it offends their censors, they should be free to do so. This is no different than France banning the importation of a magazine that is full of Nazi propaganda getting confiscated by French customs when you land at Charles de Gaulle International Airport in Paris.
The trick, however, is trying to identify that "offending" internet traffic. That a French citizen could obtain the same thing through an e-gold account and some traffic on Freenet also needs to be taken into account when trying to set up such a censor system. I don't think it would be technically possible to block any forms of internet traffic unless you block all internet traffic, which essentially cuts off that group (like France) from the rest of the world if they try this approach.
I think you miss the point here. This is an American company doing business principly in America. The memorabilia in question (Nazi flags, ect.) are being offered by Americans.
The question comes up if the fact that this site can be viewed in France, and bid upon by French citizens because there is no way to identify them as such, should French law apply to a company doing business essentially in America?
This goes way beyond if France is a soverign entity. France could always do the same thing that is done in China, and ban certain websites that don't fit their criteria of public decency laws. Should E-bay and in this case Yahoo also be banned in France? Where do national laws end and what should be done when international commerce is being effected and who has the authority to regulate that commerce?
In the USA, there is a long standing precedence regarding interstate commerce being regulated by Federal courts, and I would imagine if this were a dispute between somebody in France and Germany that similar legal mechanisms are in place within the context of the EU.
IMHO, this is where France seems to thing that the USA is a member of the EU, and clearly it is not. That is the whole issue, and unfortunately the government of the USA has a tendancy of thinking in trade talks that it has one vote against the 15+ votes of the EU.
I would sum this whole issue up in one "bumper sticker quote":
Texas: Bigger Than France
And that is an American attitude. If they can sell more products in Texas than all of France, why worry about France? Especially when the Texas National Guard could by itself invade and defeat France militarily.
How is a ruby dagger any different from an obsidian knife? That is one of the oldest weapon implements ever made by mankind, so if you are talking about something new, this is not it. And one of those knifes could clearly be deadly in a cockpit situation. With a leather body, handle, and sheath, it could go totally undetected in that regard.
Still, having something with the strength of aluminum but with high transparency would be useful for many applications. Ships hulls (portals), including in submarines as well as in space would be quite useful. The holding tanks (aka Star Trek IV) might be interesting as well, although how this fits with transporter technology is yet another subject of discussion.
Most of the European settlements and much of the Eastern U.S.A. was built around water transportation, while most of the Western USA was built around railroad transportation systems, and later internal combustion transport systems (like the Interstate Highway system... that has been absorbed into the growth of the western USA...it is still almost an artificial construct in the eastern USA except in suburbia). You can see some of the rail routes in the lights, especially in areas through the Rocky Mountains, where the grid pattern of the mid-west breaks down. This is true even for areas where the rail lines are no longer in service, but the old rail lines were necessary for the creation of some of these towns.
What happened to passenger rail transportation in the western USA is also a subject of piles of commentaries that I won't get into, but it pretty much is for hobby rail lines with tourists and the few people who are paranoid about traveling by air...and even then you have to pay a healthy premium to travel by rail. Amtrack got hammered after 9/11, but business has since gone back to half-empty trains that are heavily subsidized and still overpriced and only run during months of a blue moon. It wasn't always that way. The train station where I live was turned into a restaurant, and that only because there was a historical easment to keep it from getting demolished. It hasn't been used for passenger travel since WWII. This is very typical throughout most of the western USA, and even much of the USA in general. Even the major train terminals are pretty much gutted and only kept for historical reasons, usually by a government entity of some sort.
This does, however, make it a nice "skin" to wrap around a sphere when doing 3D modeling. Any other projection would be a nightmare in trying to work with, except maybe a couple of polar projection maps. Even then, this gives virtually no distortion on the equator, and the poles are essentially irrelevant in this situation, unless you are trying to get the lights of the antarctic research station on the south pole.
I happen to like this numbering system. If you give me an address in Salt Lake City (or most of Salt Lake County as well), I can usually find the location within 3 or 4 blocks even if I don't have a map and have never been there before. The problem comes when somebody decides to build a subdivision in a new area and call it something more generic like Elm Street or Oak Street. When that happens, you are totally lost and even the locals at a gas station are clueless trying to help you out. This system is used throughout most of Utah, south eastern Idaho, and some of the "Mormon" areas of Arizona... at least parts that were settled by Mormon pioneers. If you are from Arizona you know what I mean.
BTW, the second part of the address that you were mentioned, like 1900 East, is the name of the street. It could be called 1900 East Street, but because all streets are called streets (not streets and avenues), for shorthand they call it simply 1900 East. Roads going east and west (hence are north and south of the center of town) are named 400 North or 600 South.
If you talk with locals it gets even more confusing yet, because when the cities in Utah were originally set up they didn't county by hundreds like is currently designated. Instead they simply called the street 1st to the east from the center of town 1st East. The next one was 2nd East, ect. that way 1900 East Street is sometimes called simply 19th East, or a famous place in Salt Lake County, 49th Street Galleria (since renamed) was on 4900 South. The 100's designation do help so a street that would be halfway between what should be 400 North and 500 North would be 450 North. That gets rid of really crazy street names (I've seen them) like 5-3/4er Street.
Because of the influence of people from outside of Utah who simply don't get it, some streets are getting names again, like Dr. Martin Luther King Jr. Avenue or Freedom Blvd. These usually also have an address like 600 South or 400 East, and the post office will deliver to either street name. And you think it is easier to find a place with that kind of address?
BTW, many mid-western cities also have somewhat similar street coordinates, where they have an addresses that are counted from the center of town, and the streets are labeled 1st Street, 2nd Street, etc. and then have a city quadrant like NW, SW, NE or SE. Streets go north & south and avenues go east & west. Or the opposite, but consistant throughout the city. The problem you come across is that sometimes you will have eight different 11th Avenues, each having a sub-range of addresses and broken up because of rivers, freeways, shopping malls, or other natural & artificial features. Stil, it isn't hard to find a place like 1354 15th Avenue SW.
The author is trying to describe the effects of acceleration, and specifically the acceleration due to this phenomona. He would have been better off trying to describe the gravitational effects of being on Phobos in comparison, but most people don't even understand that very real situation.
The size is not the size of an apple, but the gravity so weak that an object like an apple would take forever (seemingly) to accelerate due to that force. It makes sense to me, but it could have been better written, particularly if you got the apple mixed up as the size of the effect.
Actually, a photon has mass. Remember Einstein's famous equation?
E = mc^2
This mean, in a grossly simplified and totally non-scientific viewpoint, that mass and energy are simply forms of each other. Normally when you are talking about nuclear reactions you are more concerned about the transformation of mass into energy, but the opposite also occurs. Simple things like the production of electron/anti-electron pairs due to x-rays at a certain frequency that happens due to the energy of a photon being at the same energy level. Other bizzare things happen with elementary particles as well.
Getting back to photons having mass. Since energy by itself has mass, it can also be affected by mass. The space-time geometry descriptions are another valid model, but simply describes this basic fact about how gravity affects mass/energy.
Indeed, if you shine a bunch of photons into a very concentrated area simultaneously, you could in theory be able to form a "black hole" that is created purely out of photons, not normal matter. This is called a "kugelblitz", and behaves in some strange ways somewhat different from a normal black hole. This is theoretical anyway and if a normal black hole is still being debated regarding their existance, one of these guys is way out in terms of if it even exists.
So yeah, a photon doesn't have "rest mass", but indeed it does have mass, and an amount that you can measure.
E = hv (h is Plank's Constant and v is actually the greek letter nu, or the frequency of the light energy).
hv = mc^2
hv
--- = m
c^2
That is at least one way to calculate the actual amount of mass that a single photon would have at a particular frequency. It is pretty damn small for visible light frequencies, but still not zero. Photons don't stay still anyway, so talking about rest mass of a photon is kinda silly.
While the business model may be to recycle the cameras and charge for printing, I think they got the costs down fairly well here, with the cost of goods being quite a bit less than $20. They will make a small profit just from selling many of these units in volume.
I would actually hope they embrace hackers... this is not the group that they are targeting anyway, but instead people like my mother that don't want to hassle with setting up a printer, loading paper, replacing/refilling ink cartrages, geting drivers installed for file transfer, trying to find a file manager that could read the data from the camera, etc. Instead, you can grab the camera, point, click, and return the camera for prints of your daughter's wedding while you buy groceries and have the prints waiting for you when you get to the checkout counter.
I think there is a market for this sort of thing, and the cameras are cheap enough to make it worthwhile to folks who don't want the up front cost of settting up a computer system. It is not like photo processing shops won't help you with more expensive cameras either, but the market for really cheap cameras otherwise just isn't there yet.
Encode that information? Actually, yes you could. This is something, BTW, that the Kodak Photo Studio (*.psd files) do very well, as they encode the pixel data points as floating point numbers rather than integers.
The problem you are describing is due to the fact that display devices have a limited dynamic range, usually due to costs involved with trying to get the light emitters to produce light at their maximum intensity. There are usually limits due to one color or another. With monitors and color television sets, trying to get a bright red was a struggle for quite some time. With LED systems red and green LEDs have been around for quite some time so they are comparatively cheap compared to blue LEDs, and the maximum intensity with LED systems is driven by the quality of the blue elements. Since the pigment of blue LEDs also has a shorter life than the green or red LEDs, the displays of this nature don't last as long either. "White" LEDs are usually a combination of pigments that include red, green, and blue elements, and even there the blue fades quicker, making LED flashlights over time turn yellow.
Actually, I've worked with displays that have sub pixels with individual lightbulbs, like you've mentioned. Using discrete lightbulbs have their own characteristics that are somewhat different from monitor or LED phosphors due to the fact that you can treat the light coming from them as a black body curve, while phosphors generally emit light using quantum state transitions (going back to ye olde chemistry and elemental photon quantum state transition levels). They are in practice pretty much the same thing. With the light bulbs, however, you get a shifting spectral peak with temperature, and since the brightness was controlled by time limiting the period per power cycle that the bulb was on, you had to integrate the total power in an AC power cycle to determine just how many photons (roughly) were emitted, and then further mess stuff up by trying to gamma correct the figures to discrete power levels per pixel to give a good approximation on balancing colors. Colors were added by putting a colored filter in front of the bulb.
Oh, and the size of these discrete subpixel displays? About 100' by 30', and sucking up about 1 MW of power. A good thing I didn't have to pay the electric bill on those things. While these put out a whole lot of light, in reality the way you generated hues is no different than the way you do with a conventional computer monitor. BTW, it was a blast to play Quake on one of these monster displays.
Strictly speaks, yeah, you can't derive luminance directly from a color triplet, but you can get pretty close, and it is possible to do an RGB to Y conversion (the luminance portion of the XYZ coordinate system, also used for YUV or YCrCb color spaces as well, and the conversion for Y is the same for all three color spaces). For a quick & dirty conversion, I generally did 1/2 green + 1/4 red + 1/8 blue + 1/8 green (to make 1...green figures strongly with your perception of how bright something is) to come up with a good luminance value for B&W displays, as this conversion could be done with integer math and values like 1/2, 1/4, & 1/8 can be done with a shift instruction...multiplication wasn't even necessary. This made it nice with embedded systems as well. This rough approximation is pretty close to the CIE color point and even closer to what is used now for HDTV standards.
You are correct on all of these points so far. Also keep in mind that some people see not just three colors but a few see four, some see just two or in an extreme case just one (true color blindness). Blue-Green color blindness being the most common.
/.
BTW, I am convinced that the "high quality" interior decorators are among those people with tetrachromaticity (being able to see with 4 colors). If there are people that see with 4 colors, I'm sure there are some with even broader ranges yet, even if it is just a very small number. Since I am color blind to those people, I can only remotely imagine what their world is like through their eyes. From what I understand, this extra color perception is some extra cones that perceive Orange directly, and has survival advantages because you can identify diseases easier, particularly in children. There was a writeup on this topic awhile back here on
In order to accurately depect a function using the six colors, it would have to be something like this:
HUE generate_hue(R,G,B,C,M,Y)
This is by definition a six-dimensional function. That you are confusing this dimensionality with our normal 3D world is a mistake. This is not just six vectors but total number of dimensions that would be needed to fully explain every potential point that could be generated with this system. By reducing this to a tri-value representation you are going to lose information. There is no way to reverse that sort of function, which is what I'm trying to point out.
If you define a color coordinate system as three coordinates (i.e. XYZ or RGB or YUV) you can't be able to get to all possible points available with a six-coordinate system. That is the whole point I'm trying to make here is that if you use six colors for display, you need to somehow define each and every potential color and have that encoded as the representation of that hue when used for data storage. By only recording three coordinates you will by definition lose information and lose hues, which makes it pointless to have more than three colors for display unless those other coordinates are there.
One way that image compressors work often, as you were alluding to earlier, is to transform the color coordinates to something like YUV, because you notice changes in brightness much more than the hue itself. Analog Color television does something like this as well, where most of the bandwidth is occupied with the luminance signal, and only a small portion is used to actually define the hues. That you don't see "banding" with color TV is only because it is an analog signal instead of a digital signal. Colors on television are very hard to get hues to display correctly, particularly reds. You can do the same thing with a six-coordniate system, but you would have to do a transformation through six values. You may only have to store some of those values in one or two bits, but it would still have to be there to be of any value in the overall system.
No, by reading between the lines, it is if you use functions in the database, like if you use function in the operating system to format a disk or retrieve a data file, you are technically in violation of what they consider to be "linking" to their software.
The real question is at what point does linking stop being linking. They point out that even connections that are based on raw TCP/IP, according to MySQL AB, are in violation of the GPL. In rough technicalities, if I write some software that scrapes data from a web page that uses an Apache webserver using MySQL for content, and mind you the web page scraper is written just for that specific website, I would be technically in violation of the GPL from their viewpoint. I would have to release that scraper under the GPL from this viewpoint.
That is where I think MySQL AB is overeaching on their reading of the GPL. In that last example, I may not even be aware of the fact that the web site is even using MySQL. It would also make using Linux or other GPL'd operating systems illegal (following this philosophy) when running propritary software. Since databases are moving into the operating system environment, this is going to be even more of an issue, although I don't thing MySQL is going to be the key component when that happens.
Actually, I think you have an idea here. While luminance can be derived from the rest of the colors, Reflectivity and Alpha blending is something that must be added seperately.
In encoding six colors, it give you only 48 bits, where 64 bits is more natural for computers to manipulate (just as 24 vs. 32 bits for a normal RGB is where alpha usually gets thrown in as the extra byte per pixel). What to do with that extra byte of data in addition to Alpha.
There is quite a bit more to explain here. First of all, while you might be able to see in three different wavelengths of light, others will see with a totally different set of three wavelengths of light, so what may look just fine to you will look off or even quite a bit different to another person. This is called genetic variations, and there is quite a bit of variance between different people.
Also, you need to remember that an RGB monitor is not really well tuned to specific freqnencies that your eye can perceive. I would dare say that for the most part it is pretty close, but it is off, and by having other colors to display it would "normalize" the display even more so your eyes would notice subtle color variations more to your normal experience with things "In Real Life(tm)". What drives monitor development is certain phosphors that emit light at a given frequency or a set of frequencies that are approximately close to one of the primary colors. While "popular" phosphors will come close to one of the major frequecies that your eye perceives best with one of its color receptors, that is also on average, and only an approximation.
You got the idea more or less correct when you mind perceives color as a ratio of your "red", "blue", and "green" color receptors. There is more, however, because these color receptors don't just see one specific frequency, but see along a whole range of frequencies that are close to the "peak" frequency of that receptor. In fact, there are sometimes more than one peak for a given color receptor across the whole EM visible light band, although there is clearly a dominant peak as well.
By adding extra colors to "smooth out" the color frequency distribution to make a given hue, or to give it several peaks in different places rather than just three peaks with a tri-color system, you can more faithfully reproduce the color. Dealing with non-human perception would require even more range, like how honey bees can see into the UV band, or some other animals being able to see into the infared band. You can see "chords" to form a hue, but this is quite limited as you can only see about 1 octave from red to blue.
This is not just precision loss due to conversion, this is a loss of information due to moving from a 6 dimensional representation to a 3 dimensional representation of a number. Information loss is going to occur when you do that, and you can't wave your hand around trying to pretend otherwise. Try to do a conversion of 3 dimensions to 2 as an exercise and get all three back somehow, like going to flatland and poping back to 3D world.
Yes, changing from one 3D representation to another 3D representation (ie RGB to YUV to XYZ) can be in theory kept free of loss, but this is not the same thing as converting for more than 3 colors to one of these 3D systems. This is raw information theory here.
While I would agree that your eye measures discrete "pixels" as well, the human eye is an analog conversion device, not digital. Frequency and intensity are measured by your eye through analog processes, and processed in your brain through an analog "computer", not a digital one. Furthermore, what one person perceives is slightly to substantially different than what somebody else will see, because their eyes respond to different frequencies, and the cones in their eyes can respond to more or fewer number of frequencies than yours. Even individual optic cones that are in theory receptive to the same color will respond to slight variations of a given frequency, which provides a slight "fuzzing" or blurring of the frequency range.
At best a restrictive color scheme will only work with a theoretical "typical human eye", and most people would still be able to notice a difference. Not only that, but producing materials that are dead on to emit or reflect light at precisely the peaks of human color perception are very difficult at best to manufacture. That means you need to make an encoding scheme for color that takes these other materials into account, and deals with people who have a wide range of varying color perceptions.
No, the issue is if MySQL AB has made certain that the code being distributed under commercial license is 100% certified by the developers as having copyright clearance for that commercial distribution. I would venture to say they have done exactly that, the GPL not withstanding. That is a big task, and what you are claiming is that somebody has added software to MySQL and not given permission to MySQL AB to redistribute it commercially, even though that software is being used in their product.
All I can say then is to "show me the money", or in this case the source code you question in their commercial distribution. $200 is paltry if it is your software, and the potential benefits including financial on your end if they are using your code would be quite substantial, just from a statutory viewpoint of somebody violating your copyright. I don't think you can find that code which is in violation.
No, I think you got that one wrong. The GPL allows you to redistrubte their code on your own website, flea market, subway station, commercial bookstore, etc. You can even distrubute a product that includes their code, as that is legal and encouraged by the GPL. You can even sell it for substantial profits, that you can even put directly into your own pocket, like $100 or $500 per CD if you can get buyers for it. None of that has to go to MySQL AB.
What MySQL AB is saying is that if you write software under a propritary license, and it somehow uses MySQL, you should purchase a commercial license of MySQL before you can sell that software that requires MySQL. That is why the argument regarding if the Linux kernel, which propritary software can and does link to accessing OS functions that are available only under the GPL, is brought up in comparison and the question begs to be asked at what point does MySQL AB really lose its argument with this linking requirement.
The GPL does not really specify this "right" that MySQL AB is claiming, and it doesn't really make sense to me. If I made changes to MySQL and relabeled MySQL as my own software, that would be a violation of the GPL. If instead I am installing MySQL or OpenOffice on a CD-ROM that also happens to include some propritary commercial software that I am selling for a profit, I don't see how MySQL AB should be demanding profits via the GPL from this propritary distrubution.
I can build a distribution of Linux and sell it for $10,000. That nobody may buy it is another story, but I can certainly try to sell it for that price, and I am not compelled to even place this distro on the internet (contrary to some people's viewpoints on the subject). You just have to make sure that all GPL'd software also includes the source code if you distribute the binaries with it.
I think you would be very hard pressed to find some source code in the commercial version that was "contributed" by somebody without a specific copyright waiver. I would say that MySQL is in much more sound legal water than Linux in general, specifically the Linux kernel, which obviously has some legal questions, the litigous idiots in Lindon, UT notwithstanding. Most of the code has been contributed by employees of MySQLAB.
You simply can't gain information once lost from a color conversion.
If you are talking about visual compression algorithms that are lossy, like JPEG, I would have to agree that often they do some sort of conversion like you are suggesting. Lossless compression systems, on the other hand, do no such thing.
While data formats like PNG or GIF can compress data using a YUV color pallet, that is because the information was originally stored in that format, not because that was the original color space.
A YUV or YCrCb to RGB conversion, even floating point signed RGB will eventually lose some information and is difficult to convert back. Very seldom are they stored as even signed numbers because of raw data storage limitations....bits for a color in an individual pixel are quite precious and are not automatically increased arbitrarily unless there is a good reason to do it.
The point here is that if you have more three chromatic color for a given hue, a conversion to YUV or some other encoding system (like CIE XYZ) is going to lose some data along the way. I actually dare you to prove otherwise to me. While the shade of color reproduced with such a conversion might be close, it won't get the color back all of the way and will lose some information, and information that will be noticed by a good eye.
The problem is, and I'll admit this, that different people percieve color differently. While there might be a model that you can call a "typical" human eye color gamut, you need to go to hard physics ultimately in order to pull out the other colors.
Photoshop experience and an artistic eye can pull out colors to make them more life-like and even treat the other three colors in a hex printing pallet like colors on an oil-based paint pallet, but in reality you can't obtain new information that isn't there unless it was encoded in the first place. You add a little bit of that information with a good photoediting piece of software like Photoshop. An RGB color space is fairly good, and a reasonable model, as is the "color wheel", but it is just one model that works reasonably well. There is a point that ultimately it breaks down, and that is the point I was trying to make earlier. That you can create 70%-80% of all of the colors in human experience makes them very useful models, especially as the remaining colors are seldom seen by most people, and there are many other issues involved with art like proportion, balance, and perspective that are just as important if not more important. That colors get pretty close means you can concentrate on the other issues instead.
Trying to explain the value of even an RGB system is quite difficult to those who are color blind and barely see two colors, or are purely monochromatic in their vision is particularly difficult. What is worse is that often they don't realize that they don't see all of these colors.
My background is more along trying to engineer systems that can accurately display and portray colors for most people, which is why I have gone more for a purely scientific viewpoint. Having to deal with more unusual color gamuts like a pure RG system (systems that only display red and green, due to costs to add blue to the display), and RGBW systems (where you have the normal RGB and add white for additional contrast... and you though CMYK was tough). I did some limited experimentation with violet LEDs and some very dull near infared LEDs as well. They give some colors that are quite interesting, and unfortunately I never had the chance to see a full display made up of these colors tied together with RGB LEDs, like is being suggested by the article mentioned as the parent article. While understanding the physiological issues regarding color perception (and we did deal with them), we had a much easier time dealing with color from a raw physics viewpoint when designing our systems, in part because we were working on a more physical system level. I had to also deal with the user interface and trying to come up with a color picker that would work with these sometimes unusual color spaces.
No, standard NTSC (and PAL and SECAM) use a YUV system. That is, a Luminance value followed by two other sub-bands that control red and blue that get subtracted out of the base luminance signal (the remaing color being green). This is called YUV, and is also why NTSC is sometimes called "Never The Same Color", because getting this to work on an analog signal is particularly difficult. This system was chosen because the "Y" part of the signal was backward compatable with B&W television sets, so those old sets could work even with the new color signals added to the television broadcast.
Digital Video broadcasts will change this, but both RGB and YUV signals are going to be used by broadcasters. YUV in part because they don't have to rework as much equipment. I have never heard of RGBK used in television.
There is a "black mask" to help separate the individual phosphors from each other, and to add definition from the seperate color guns in the back of the CRT. On cheaper televisions with crude phosphors, the electrons from one color tend to "bleed" into the other phosphors, and make the image look "fuzzy" when you view them on a TV screen. This only applies however to CRT-based television sets and not plasma screens or LCD monitors. The same also applies to quality vs. cheap CRT-based computer monitors, for the same reason. The black mask has nothing to do with the color space, but with the physical construction of the monitor.
This is a BS alert that is valid, other than the fact that if you are digitizing film pigments that you would pull out some subtle hues that could not normally be found with a typical RGB scanner/encoding system. If you are talking about some extra processing of something that was previously encoded with just RGB pixel information, you are absolutely correct. This is utter BS.
Where you can see a hexcolor data encoding system in use is if you film the original scene with a camera that can capture the full gammut. Until that happens it would be like taking old films from the early 1900s and digitizing them in full RGB color, even though it is mostly B&W film. The subtle differences in stock and perhaps choices to have a "colored" stock would make a difference from the perspective of the editor or director and your viewing experience. If you digitized that same file with just luminance values you would miss that important piece of the movie, but of course that is for purests of film culture, right?