Being "pretty" doesn't qualify something as being "art." In fact, some really fantastic art is quite ugly and disturbing (Munch, Giger, de Goya, etc.) Art involves a relationship between artist and viewer. Art is human expression, human communication. These pictures may be aesthetically pleasing. They may even resemble some imagery used by artists. But without human expression, it's just data.
MRI is more versatile than you think. Although it is certainly not sophisticated enough to actually "read your mind" and find out what you're thinking, MRI is quite capable of measuring blood flow to find out if you're thinking with a certain part of your brain.
Flow or diffusion can alter the measured NMR relaxation times of any fluid. As a simple spin-echo example, suppose that the 90 degree pulse is delivered to the arterial blood in some voxel in the selected plane. By the time the echo pulse is detected, this particular cohort of blood protons will have flowed into (and perhaps beyond) the adjacent plane, and the blood originally in the voxel be replaced with fresh blood. The apparent relaxation time that determines the pixel brightness will therefore be different from that of static blood, and the shift in apparent relaxation time will increase with the flow rate. (Wolbarst, Physics of Radiology, p. 401)
By using pulse sequences with different parameters, one can choose to suppress or enhance blood flow effects. This principle is commonly used for doing angiography via MRI. They acquire a set of images using a protocol that suppresses motion effects, then acquire a set of images using another protocol that enhances motion effects, and compute the difference. It also forms the basis of functional MRI, which is what they're using for their marketing research. Functional MRI relies on the relationship between blood flow and metabolism. When a region of the brain is active, blood flow in that region is increased. fMRI provides a map of how active different regions of the brain are.
You are correct that there is indeed a lot of noise present. Your brain is always doing a lot of different things at once and there will be some portion of the signal that results from the experience of being in the MRI. However, statistical analysis can filter that out because the control group experiences those same things. In the end, they will get an answer to questions about whether or not their stimulus (the advertising) produced a response (emotion) and some quantification of how strong the response was. I doubt they'll be able to differentiate between emotions, but they don't really need to. They can ask the participant whether they felt positively or negatively about a particular advertisement. The value of the MRI data is that it quantifies how much of a response there was.
In my opinion, however, this is a terrible waste of magnet time that could have been spent doing research that would actually benefit humanity.
Nice idea, but really we shouldn't name children names like RTDSG4232342. If we're going to name people unique identifiers, we should at least follow established international standards. ISO 9834 defines an appropriate naming convention.
Actually, free speech rights are an issue here. This takes away the First Amendment right of the librarian to speak out against the unjust treatment of the patrons whose Fourth Amendment rights are being violated.
Since politicians and the media alike always interpret low voter turnout as merely apathy or laziness rather than active dissatisfaction with the total lack of real choices in our political system, one option is to vote but opt not to vote for either the Dems or the Pubs. If a suitable third party candidate exists, vote for them. If not, invent one. You always have the option of a write-in vote. Vote for your mom. Vote for grandpa. Vote for famous dead people. Granted, none of these people are going to win. On the other hand, the media and the politicians can't write you off as apathetic or lazy because you took the time to go to the polls and actively register your rejection of the available candidates. If enough disenfranchised people actually did this, the major party candidates might actually have to take a real interest in what voters think because to ignore those dissatisfied voters would be to invite viable independent candidates who would not ordinarily have ever chosen to enter the political arena to run for office and present a real challenge to the major parties.
Inevitably, someone's going to claim that I'm "throwing my vote away" and that I'm advocating a position that is irresponsible with respect to my civic duty as a citizen of the United States. My answer to that is that I firmly believe that neither of the two major political parties have any interest in serving the people. In fact, I think the Democrats and Republicans are merely two factions of the same political party, both exclusively serving the interests of the elite ruling class. They only differ in terms of which subset of the elite they cater to most. I see the choice posed by our current electoral system as fundamentally equivalent to the question, "Would you rather have Green Eggs & Ham on Tuesday or Thursday?" My answer to that question is, "I will not eat Green Eggs & Ham. I will not eat them, Sam I am."
I don't know if it was available in Win2k, but you are definitely correct about it being available in XP. MSVDM is part of the PowerToys collection. It's still not quite as nice as the virtual desktops under UNIX/Linux (I always liked being able to move the mouse to the edge of the screen and have the window manager automatically switch to the virtual desktop adjacent to the current one after a short preset delay... a feature that is not available in the Windows virtual desktop feature) but it is, nonetheless, a big improvement.
A better idea is to lobby your state legislature to assign electors in proportion to your state's popular vote, rather than winner-take-all.
It'll never happen. The fundamental problem is that it only works if all 50 states implement it, but it is against the interests of each individual state to do it. It's a prisoner's dilemma problem. If your state splits its electoral votes according to the popular vote, then your state's power is diminished relative to any state that maintains winner-take-all policy.
When will Microsoft learn? The many arbitrary and unjustified changes they make in the UI with every new release serve only to confuse and annoy their existing customers. How happy would you be if GM decided to move the gas and brake pedals in every new car design? It's ridiculous and it hurts productivity. And Microsoft wonders why people don't feel inclined to upgrade their OS? Here's a clue: people don't want to throw away their hard earned knowledge without substantial benefits from learning something new. This is especially true for system administrators who have already memorized the arcane, and fundamentally arbitrary, rituals involving clicking through 99 layers of tabbed dialogs and wizards to get access to the desired tidbit of configuration info. Want to configure <feature x>? Oops. Sorry. We moved that. Go fish.
It's not a myth at all. The human eye can't even distinguish a 1 bit difference in 256 grayscale on a properly calibrated, medical grade, 0.2 to 100 foot-Lambert gamut, 5 megapixel CRT under optimum viewing conditions (low ambient light). That same 1 bit difference wasn't visible on output from a properly calibrated AGFA DryStar laser film printer when viewed on a medical light box either (which probably has a greater luminance range than the monitor does, although I didn't bother using a photometer and densitometer to verify that). The JND (Just Noticeable Difference) turns out to be around 3 or 4 graylevels. One should not confuse the human eye's spatial resolution abilities with its luminance resolution abilities (which are not so good).
The cost and somewhat slower performance of Matlab and other languages designed for scientific computation is a poor reason to go running away to code Fortran instead. If you're doing anything even remotely complex, any money you save by doing things in Fortran instead of buying Matlab, RSI's IDL, Maple, Mathematica, or some other appropriate tool will be eaten up by the huge productivity losses incurred by using such a low level tool as Fortran. Reinventing the wheel is inherently expensive in terms of opportunity cost. Do yourself a big favor and spend the money on a very high level, mathematically intuitive, interactive tool that is appropriate for the job and some badass hardware to achieve the performance you need. You'll get the job done in a jiffy without major debugging headaches and be able to move on to do other more productive things with your time that will add value to the company.
>> Try taking your tube - filling it with a sponge like material and then adding the water. Not such a great design problem when you think about it.
>This one is actually on the books, because a few builders thought that a saturated medium would make for low-splash fuel tanks (when it was proposed, sloshing fuel was a major guidance problem, as most rockets at that time burned kerosene). However, the design failed miserably, for two reasons. First, it was very difficult to get the fuel out of the medium when you needed it to burn (something my experiment doesn't address, but that a liquid-fueled engine must do). Second, When the medium was subjected to the G-forces of launch, it would simply squash down to the bottom of the tank, which caused the top portion of the tank to be only liquid (back to the sloshing fuel problem) and putting severe stress on the bottom of the tank (which caused more than one tank rupture with resultant kaboom). So, it's a good idea, but I must send you back to the drawing board.
You could accomplish this by using a packed bed instead of a spongy medium. The difficulty would be designing a chemically inert (with respect to H2O2) packing to have minimal mass, minimal volume, rigidity, and an appropriate geometrical configuration to restrict side to side fluid flow while allowing top to bottom flow. Should be doable though.
Seriously, this is IMO one of Herbert's best novels. Written in 1982 and more relevant than ever now. Ask yourself the question, "What if Theodore Kaczynski had been a molecular biologist?" and you realize just how plausible the hellish world Herbert describes really is. If Huxley's "Brave New World", Orwell's "1984", and Kafka's "The Trial" gave you nightmares, better leave the night light on and keep a can of Lysol and a gas mask by the bed while you're reading this book.
Using an EEG to make statements about the effects of gaming on the brain is like trying to benchmark your computer's processing power with a voltmeter. Likewise, using self-selecting experimental groups is poor science. If this researcher was at all serious about proving a causal relationship between gaming and negative effects on the brain, why didn't he:
1) Select a random sample consisting only of healthy people who don't play games for more than X hours a week to participate in the study.
2) Randomly divide that sample into a control group (who don't modify their behavior) and several experimental groups. Take measurements of brain activity for all participants before any of them play games. Use standardized psychological tests and health history questionnaires to rule out participants who are abnormal psychologically or physiologically.
3) Choose an archetypal set of games representative of different game categories (puzzle, shoot-'em-up, etc.)
4) Administer no games to the control group, games from a single archetype to individual experimental groups, and a mixture of games from different archetypes to one of the experimental groups. Give each experimental group an equal amount of time playing games.
5) Take measurements of the control group at rest. Take measurements of the experimental groups prior to gaming, while they are gaming, and after gaming.
6) Use a better system of measurement than surveys and EEG. Neurofunctional MRI and standardized psychological tests would work nicely.
Additionally, I don't know anybody that just turns off the PC anymore. Even my grandmother and mother know that you shut down the computer, and one's running Win98, the other's running MacOS 9.
I wish that were true. I've seen quite the opposite. People's rebooting habits for PCs are their default behavior and extend to all computers (because most users don't understand the difference between a Wintel PC and other computers). Incidentally, this behavior comes from otherwise intelligent folks, the more savvy users. The least savvy users just call help desk anytime stuff goes wrong and they don't reboot unless you tell them to. The dangerous ones are the people that know a bit more and think they can fix their own problems. They call help desk after they've power-cycled their Sun workstation and corrupted the filesystem. This has happened enough times that I've taken to putting stickers on UNIX workstations next to the power button telling people NOT to reboot them, let the sysadmin do it.
IANAL but it seems to me that if this ruling holds, then he isn't continuing to do anything because the court defined the publication date as being when the web page was first put online. It would be no different than if he had published the code on paper or on a CDROM.
The guys down the hall from me in the dorms had the same last four digits as the local Domino's Pizza for their phone number. The campus exchange was 487 and off-campus was 482. They would routinely answer their phone, "Thank you for choosing Domino's Pizza. May I take your order?" People would give them their orders and he'd give them an estimated delivery time, etc. Later, the people would call him back asking why their pizza hadn't been delivered yet. He'd make up the most ridiculous bullshit story about why the driver hadn't delivered yet. One time, he had somebody call back several times over the course of about three hours wondering why they hadn't received their pizza. He apologized and told the guy he'd receive coupons for free pizza. The guy wanted to speak to a manager, so he'd pass the phone to somebody else who was hanging out in his room. It was hilarious!
Hmm..... when I think of marching orcs chanting in unison, I still hear that song from the original cartoon LOTR films, "Where there's a whip, there's a way. Where there's a whip, there's a way. We don't wanna go no more today, but the Lord Of The Lashes says nay, nay, nay. We're gonna march all day, all day, all day. Where there's a whip, there's a way."
No, that's why we need to bring back unregulated Chlorofluorocarbons! Bathe your box in "inert" high heat-capacity liquid!
CFCs were never popular for their heat capacities. They were popular because their boiling points were in the right range and because of their vaporization enthalpies.
That's all fine and dandy until your fish die because the water picked up toxic substances from the metal heat exchangers or your computer fries because algae growth clogged the tubes.
The reason why water is used is because of it's exceptional thermal characteristics. Water has a heat capacity (at atmospheric pressure) of roughtly 4.2 J/g K. It also has a relatively high heat of vaporization (2.26 kJ/kg) and a boiling point above your target operating temperature. It's cheap to replentish. You can easily obtain cheap pumps that are designed to pump it safely and quietly.
Some heat capacities of common substances for comparison:
methanol, 2.5 J/g K
ethanol, 2.4 J/g K
ethylene glycol (antifreeze), 2.4 J/g K
vegetable oil, 1.7 J/g K
mercury, 0.14 J/g K
air, 1.0 J/g K
Obviously, to make this all meaningful, you need to take into account the density of the fluid at the operating temperature and thermal conductivity into your calculations to translate it all into a volumetric flow rate. You also need to consider whether or not a phase change is to be involved (and if it is, how you're going to deal with the associated pressure). Many substances are unsuitable for safety & health reasons (methanol, ethanol, etc. are a fire hazard.... mercury and methanol are quite poisonous). When you get through looking it all over, water is often used because it works without all kinds of complications in the design of the system. Your main concern with water is making sure it doesn't leak onto your mobo. If you use a halothane refrigerant, you're probably going to have to deal with phase changes (thus requiring a noisy compressor) and you're going to have to make it leak-free anyway (or you'll be shelling out big $$$ to replace the expensive refrigerant you lose).
Slashdot Mirror
Being "pretty" doesn't qualify something as being "art." In fact, some really fantastic art is quite ugly and disturbing (Munch, Giger, de Goya, etc.) Art involves a relationship between artist and viewer. Art is human expression, human communication. These pictures may be aesthetically pleasing. They may even resemble some imagery used by artists. But without human expression, it's just data.
MRI is more versatile than you think. Although it is certainly not sophisticated enough to actually "read your mind" and find out what you're thinking, MRI is quite capable of measuring blood flow to find out if you're thinking with a certain part of your brain.
Flow or diffusion can alter the measured NMR relaxation times of any fluid. As a simple spin-echo example, suppose that the 90 degree pulse is delivered to the arterial blood in some voxel in the selected plane. By the time the echo pulse is detected, this particular cohort of blood protons will have flowed into (and perhaps beyond) the adjacent plane, and the blood originally in the voxel be replaced with fresh blood. The apparent relaxation time that determines the pixel brightness will therefore be different from that of static blood, and the shift in apparent relaxation time will increase with the flow rate. (Wolbarst, Physics of Radiology, p. 401)
By using pulse sequences with different parameters, one can choose to suppress or enhance blood flow effects. This principle is commonly used for doing angiography via MRI. They acquire a set of images using a protocol that suppresses motion effects, then acquire a set of images using another protocol that enhances motion effects, and compute the difference. It also forms the basis of functional MRI, which is what they're using for their marketing research. Functional MRI relies on the relationship between blood flow and metabolism. When a region of the brain is active, blood flow in that region is increased. fMRI provides a map of how active different regions of the brain are.
You are correct that there is indeed a lot of noise present. Your brain is always doing a lot of different things at once and there will be some portion of the signal that results from the experience of being in the MRI. However, statistical analysis can filter that out because the control group experiences those same things. In the end, they will get an answer to questions about whether or not their stimulus (the advertising) produced a response (emotion) and some quantification of how strong the response was. I doubt they'll be able to differentiate between emotions, but they don't really need to. They can ask the participant whether they felt positively or negatively about a particular advertisement. The value of the MRI data is that it quantifies how much of a response there was.
In my opinion, however, this is a terrible waste of magnet time that could have been spent doing research that would actually benefit humanity.
Nice idea, but really we shouldn't name children names like RTDSG4232342. If we're going to name people unique identifiers, we should at least follow established international standards. ISO 9834 defines an appropriate naming convention.
Actually, free speech rights are an issue here. This takes away the First Amendment right of the librarian to speak out against the unjust treatment of the patrons whose Fourth Amendment rights are being violated.
Since politicians and the media alike always interpret low voter turnout as merely apathy or laziness rather than active dissatisfaction with the total lack of real choices in our political system, one option is to vote but opt not to vote for either the Dems or the Pubs. If a suitable third party candidate exists, vote for them. If not, invent one. You always have the option of a write-in vote. Vote for your mom. Vote for grandpa. Vote for famous dead people. Granted, none of these people are going to win. On the other hand, the media and the politicians can't write you off as apathetic or lazy because you took the time to go to the polls and actively register your rejection of the available candidates. If enough disenfranchised people actually did this, the major party candidates might actually have to take a real interest in what voters think because to ignore those dissatisfied voters would be to invite viable independent candidates who would not ordinarily have ever chosen to enter the political arena to run for office and present a real challenge to the major parties.
Inevitably, someone's going to claim that I'm "throwing my vote away" and that I'm advocating a position that is irresponsible with respect to my civic duty as a citizen of the United States. My answer to that is that I firmly believe that neither of the two major political parties have any interest in serving the people. In fact, I think the Democrats and Republicans are merely two factions of the same political party, both exclusively serving the interests of the elite ruling class. They only differ in terms of which subset of the elite they cater to most. I see the choice posed by our current electoral system as fundamentally equivalent to the question, "Would you rather have Green Eggs & Ham on Tuesday or Thursday?" My answer to that question is, "I will not eat Green Eggs & Ham. I will not eat them, Sam I am."
I don't know if it was available in Win2k, but you are definitely correct about it being available in XP. MSVDM is part of the PowerToys collection. It's still not quite as nice as the virtual desktops under UNIX/Linux (I always liked being able to move the mouse to the edge of the screen and have the window manager automatically switch to the virtual desktop adjacent to the current one after a short preset delay... a feature that is not available in the Windows virtual desktop feature) but it is, nonetheless, a big improvement.
A better idea is to lobby your state legislature to assign electors in proportion to your state's popular vote, rather than winner-take-all.
It'll never happen. The fundamental problem is that it only works if all 50 states implement it, but it is against the interests of each individual state to do it. It's a prisoner's dilemma problem. If your state splits its electoral votes according to the popular vote, then your state's power is diminished relative to any state that maintains winner-take-all policy.
When will Microsoft learn? The many arbitrary and unjustified changes they make in the UI with every new release serve only to confuse and annoy their existing customers. How happy would you be if GM decided to move the gas and brake pedals in every new car design? It's ridiculous and it hurts productivity. And Microsoft wonders why people don't feel inclined to upgrade their OS? Here's a clue: people don't want to throw away their hard earned knowledge without substantial benefits from learning something new. This is especially true for system administrators who have already memorized the arcane, and fundamentally arbitrary, rituals involving clicking through 99 layers of tabbed dialogs and wizards to get access to the desired tidbit of configuration info. Want to configure <feature x>? Oops. Sorry. We moved that. Go fish.
It's not a myth at all. The human eye can't even distinguish a 1 bit difference in 256 grayscale on a properly calibrated, medical grade, 0.2 to 100 foot-Lambert gamut, 5 megapixel CRT under optimum viewing conditions (low ambient light). That same 1 bit difference wasn't visible on output from a properly calibrated AGFA DryStar laser film printer when viewed on a medical light box either (which probably has a greater luminance range than the monitor does, although I didn't bother using a photometer and densitometer to verify that). The JND (Just Noticeable Difference) turns out to be around 3 or 4 graylevels. One should not confuse the human eye's spatial resolution abilities with its luminance resolution abilities (which are not so good).
The cost and somewhat slower performance of Matlab and other languages designed for scientific computation is a poor reason to go running away to code Fortran instead. If you're doing anything even remotely complex, any money you save by doing things in Fortran instead of buying Matlab, RSI's IDL, Maple, Mathematica, or some other appropriate tool will be eaten up by the huge productivity losses incurred by using such a low level tool as Fortran. Reinventing the wheel is inherently expensive in terms of opportunity cost. Do yourself a big favor and spend the money on a very high level, mathematically intuitive, interactive tool that is appropriate for the job and some badass hardware to achieve the performance you need. You'll get the job done in a jiffy without major debugging headaches and be able to move on to do other more productive things with your time that will add value to the company.
An excellent online resource for graphics (particularly B splines, NURBS, etc.):
Intro to Computing with Geometry Class Notes
The Computer Image Watt & Policarpo
Advanced Animation and Rendering Techniques Watt & Watt
Introduction to Computer Graphics Foley et al
OpenGL SuperBible Wright & Sweet
>> Try taking your tube - filling it with a sponge like material and then adding the water. Not such a great design problem when you think about it.
>This one is actually on the books, because a few builders thought that a saturated medium would make for low-splash fuel tanks (when it was proposed, sloshing fuel was a major guidance problem, as most rockets at that time burned kerosene). However, the design failed miserably, for two reasons. First, it was very difficult to get the fuel out of the medium when you needed it to burn (something my experiment doesn't address, but that a liquid-fueled engine must do). Second, When the medium was subjected to the G-forces of launch, it would simply squash down to the bottom of the tank, which caused the top portion of the tank to be only liquid (back to the sloshing fuel problem) and putting severe stress on the bottom of the tank (which caused more than one tank rupture with resultant kaboom). So, it's a good idea, but I must send you back to the drawing board.
You could accomplish this by using a packed bed instead of a spongy medium. The difficulty would be designing a chemically inert (with respect to H2O2) packing to have minimal mass, minimal volume, rigidity, and an appropriate geometrical configuration to restrict side to side fluid flow while allowing top to bottom flow. Should be doable though.
Seriously, this is IMO one of Herbert's best novels. Written in 1982 and more relevant than ever now. Ask yourself the question, "What if Theodore Kaczynski had been a molecular biologist?" and you realize just how plausible the hellish world Herbert describes really is. If Huxley's "Brave New World", Orwell's "1984", and Kafka's "The Trial" gave you nightmares, better leave the night light on and keep a can of Lysol and a gas mask by the bed while you're reading this book.
Using an EEG to make statements about the effects of gaming on the brain is like trying to benchmark your computer's processing power with a voltmeter. Likewise, using self-selecting experimental groups is poor science. If this researcher was at all serious about proving a causal relationship between gaming and negative effects on the brain, why didn't he:
1) Select a random sample consisting only of healthy people who don't play games for more than X hours a week to participate in the study.
2) Randomly divide that sample into a control group (who don't modify their behavior) and several experimental groups. Take measurements of brain activity for all participants before any of them play games. Use standardized psychological tests and health history questionnaires to rule out participants who are abnormal psychologically or physiologically.
3) Choose an archetypal set of games representative of different game categories (puzzle, shoot-'em-up, etc.)
4) Administer no games to the control group, games from a single archetype to individual experimental groups, and a mixture of games from different archetypes to one of the experimental groups. Give each experimental group an equal amount of time playing games.
5) Take measurements of the control group at rest. Take measurements of the experimental groups prior to gaming, while they are gaming, and after gaming.
6) Use a better system of measurement than surveys and EEG. Neurofunctional MRI and standardized psychological tests would work nicely.
?
Agreed. I also wonder why he didn't include any HTML text either. The QWERTY layout is absolutely loathsome for typing text with HTML tags.
Additionally, I don't know anybody that just turns off the PC anymore. Even my grandmother and mother know that you shut down the computer, and one's running Win98, the other's running MacOS 9.
I wish that were true. I've seen quite the opposite. People's rebooting habits for PCs are their default behavior and extend to all computers (because most users don't understand the difference between a Wintel PC and other computers). Incidentally, this behavior comes from otherwise intelligent folks, the more savvy users. The least savvy users just call help desk anytime stuff goes wrong and they don't reboot unless you tell them to. The dangerous ones are the people that know a bit more and think they can fix their own problems. They call help desk after they've power-cycled their Sun workstation and corrupted the filesystem. This has happened enough times that I've taken to putting stickers on UNIX workstations next to the power button telling people NOT to reboot them, let the sysadmin do it.
IANAL but it seems to me that if this ruling holds, then he isn't continuing to do anything because the court defined the publication date as being when the web page was first put online. It would be no different than if he had published the code on paper or on a CDROM.
Yes. This was indeed Michigan Tech. Wadsworth Hall, to be specific. And, yes, I know it's off topic, but that never stopped me before. :)
The guys down the hall from me in the dorms had the same last four digits as the local Domino's Pizza for their phone number. The campus exchange was 487 and off-campus was 482. They would routinely answer their phone, "Thank you for choosing Domino's Pizza. May I take your order?" People would give them their orders and he'd give them an estimated delivery time, etc. Later, the people would call him back asking why their pizza hadn't been delivered yet. He'd make up the most ridiculous bullshit story about why the driver hadn't delivered yet. One time, he had somebody call back several times over the course of about three hours wondering why they hadn't received their pizza. He apologized and told the guy he'd receive coupons for free pizza. The guy wanted to speak to a manager, so he'd pass the phone to somebody else who was hanging out in his room. It was hilarious!
Hmm... I always though "Shock the Monkey" was about electroconvulsive therapy.
Hmm..... when I think of marching orcs chanting in unison, I still hear that song from the original cartoon LOTR films, "Where there's a whip, there's a way. Where there's a whip, there's a way. We don't wanna go no more today, but the Lord Of The Lashes says nay, nay, nay. We're gonna march all day, all day, all day. Where there's a whip, there's a way."
No, that's why we need to bring back unregulated Chlorofluorocarbons! Bathe your box in "inert" high heat-capacity liquid!
CFCs were never popular for their heat capacities. They were popular because their boiling points were in the right range and because of their vaporization enthalpies.
That's all fine and dandy until your fish die because the water picked up toxic substances from the metal heat exchangers or your computer fries because algae growth clogged the tubes.
The reason why water is used is because of it's exceptional thermal characteristics. Water has a heat capacity (at atmospheric pressure) of roughtly 4.2 J/g K. It also has a relatively high heat of vaporization (2.26 kJ/kg) and a boiling point above your target operating temperature. It's cheap to replentish. You can easily obtain cheap pumps that are designed to pump it safely and quietly.
Some heat capacities of common substances for comparison:
methanol, 2.5 J/g K
ethanol, 2.4 J/g K
ethylene glycol (antifreeze), 2.4 J/g K
vegetable oil, 1.7 J/g K
mercury, 0.14 J/g K
air, 1.0 J/g K
Obviously, to make this all meaningful, you need to take into account the density of the fluid at the operating temperature and thermal conductivity into your calculations to translate it all into a volumetric flow rate. You also need to consider whether or not a phase change is to be involved (and if it is, how you're going to deal with the associated pressure). Many substances are unsuitable for safety & health reasons (methanol, ethanol, etc. are a fire hazard.... mercury and methanol are quite poisonous). When you get through looking it all over, water is often used because it works without all kinds of complications in the design of the system. Your main concern with water is making sure it doesn't leak onto your mobo. If you use a halothane refrigerant, you're probably going to have to deal with phase changes (thus requiring a noisy compressor) and you're going to have to make it leak-free anyway (or you'll be shelling out big $$$ to replace the expensive refrigerant you lose).