A 400x radiation dose over 5 weeks does not necessarily a long term 8x dose make, and Slashdotters are absolutely correct to call this out.
That sort of comparison is exactly what this study is calling out, too. You know how that "8x background = evacuation" threshold was developed? By measuring the effects of a short-term dose just low enough to avoid acute radiation poisoning, and assuming it's equivalent to a low-level long-term dose that provides as much total radiation. (It's called the "linear no-threshold model". Look it up.) You're correct to say that 120 millisieverts over the course of 5 weeks (the study dose) isn't equivalent to 1200 millisieverts over the course of 40 years (a long-term 8x dose), but 1000 millisieverts over the course of 30 seconds (a dose that won't quite trigger acute radiation poisoning) isn't equivalent to it either.
Mice may not live long, but genetic damage will still take the as long to occur in mice as in humans. The fact that humans live longer means the damage has longer to accumulate.
So while 5 weeks may be "20 years" for a mouse's life span, it's still only a miniscule fraction of a human's lifetime in terms of how long damage has been accumulating.
Humans have *much* stronger anti-cancer mechanisms than mice. A mouse in its natural environment will take about three years to accumulate enough unrepaired DNA damage to get cancer; a human takes about 60 years. Five weeks of an elevated dose for a mouse is equivalent to about two years for a human, assuming the linear no-threshold model (the model that's the basis for public policy).
4'6" is the lower end of normal for women from Southeast Asia. If you're going for a "general guideline" formula, it needs to cover at least 99% of the population, everything from the 4'6" woman from SE Asia to the 6'7" average North American basketball player.
(BMI falls down for tall people, as well: the average NBA forward is moderately overweight if you go by BMI.)
The article is about people coming in UNDER weight, and I've never seen any false-positives on that end.
You're looking at the wrong end of the height scale. Someone who's 4'6" would be considered "underweight" at 76 pounds, which is actually a healthy weight for someone that short.
You still think you know more than the doctors in the WHO and AMA who publish these BMI figures?
Yes. I, at least, am aware that weight increases as the cube of height, not the square. Consequently, BMI tends to give numbers that are too high for tall people, and too low for short people. As an extreme example, many professional basketball players would be considered "obese" based on their BMI numbers, but "normal weight" based on body-fat percentage.
(Actually, because of changing body shape, it's around the 2.7th power for humans, but 3 is a better approximation to that than 2.)
Because when something goes wrong in the breaker box, it's easier to pull the negatives out of the file and look at them with a flashlight than it is to find a neighbor who will let you plug your computer in.
Takeoff isn't very stressful on an aircraft. Touchdown can be, but the stresses are concentrated in specific areas, typically the strongest parts of the aircraft, so you can repair or replace them if needed. Pressurization stresses every single joint and opening in the fuselage. You can inspect and repair individual problem spots, but eventually, the only practical option is to replace the entire aircraft.
My parents did something like this once when buying a car. My father did all the talking, while my mother sat there with a Consumer Reports article on car-buying in one hand and a pen in the other, very visibly checking off the high-pressure sales tactics as they were used. It took about five minutes for the salesman to cave and give them the price they wanted.
The fundamental difference between consumer and enterprise hard drives is the error-recovery strategy in the drive firmware.
A consumer-grade drive, upon getting an unrecoverable ECC error while reading a sector, will make repeated tries at recovering the data, spending seconds or even minutes at it -- good for a drive operating on its own, but the delay could cause a storage array controller to mark the entire drive as failed. An enterprise-grade drive assumes it's part of some sort of storage redundancy array and will give up quickly, letting the higher-level controller recover the data from the rest of the array -- much faster, but less reliable if the drive is operating on its own.
I have about 6 TB of data that I need to keep backed up.
The only way my data is in jeopardy is if my house burns down (takes out 3 local redundancy & backup solutions) on the same day that Amazon has critical failure.
Are you sure of this? With my Internet connection, a full restore of 6TB of data would take somewhat over five years, which is a pretty big window for an Amazon failure.
The launch may have been believable, but the landing was dead wrong. There are only two ways to land a flapping-wing aircraft if you've got legs instead of wheels, and he didn't use either of them.
Why would you assume that, or the same platter size?
Historical evidence, mostly. Standard consumer drives have run at 7200 RPM for the past decade, high-performance drives have run at 15k RPM for just as long, and the 3.5" platter diameter has been standard for almost twice that.
the evolution perspective (100,000? or is it 2,000,000 now?, it changes so frequently and everyone disagrees so much I can't keep track).
They disagree because they disagree on how to define "the first human".
Two million years ago sounds about right for the "divergence of the Homo genus from the Australopithecus genus" definition. 100,000 years is about right for the "final divergence of Homo Sapiens and Homo Neanderthalensis" and "anatomically modern human" definitions. There's also the "minimum cranial volume" definition, the "minimum forebrain size" defintion, the "bipedal gait" and "upright posture" definitions, two different tool use definitions (using found tools vs. using made tools), and probably a number of others.
Then you are doing it wrong because in all of the backup technologies I use, I would either simply clone the backup drive (hours at most) or restore from a backup like Time Machine, also taking just a few hours...
Assuming constant RPM, linear read speed increases as the square root of areal density (or roughly, capacity). The square root of 60 is 7.75, so a 60TB hard drive will take 7.75 times longer to fully read than a 1TB drive; if your 1TB drive can be cloned in "hours", then a 60TB drive can be cloned in 7.75*hours, which is probably "days".
You can get LEDs in any color balance you want now, including very warm color balances.
What does the spectrum look like? With a suitable choice of primaries, you can get any tristimulus value you want (so you can give a white object whatever color you desire), but if you're not careful with the full spectrum, you can get some bizarre effects with other colors.
As an example, someone at the college I went to cooked up a pigment that would look red when lit by a compact fluorescent, green under his LED desk lamp, and yellow in broad daylight.
I call horseshit too, but for a different reason. He is indeed making a derivative work, but it doesn't fall into any of the classes of protected derivative work (parody, critical commentary, etc.), which makes it a straight-up copyright violation. Any lawyer or judge who's ever so much as glanced at copyright law knows this.
My computers have multiple names. They've got a canonical name that never changes (I use animals), and then a bunch of CNAME records for the services provided. So, for example, my internet services server ("firebird", I was feeling exotic that day) also has the names "www", "ftp", and "mail". If, for example, I decide to split the mailserver off to a new computer, the only thing I need to change is the DNS records.
Hell, I'd be happy if they just re-introduced Rhetoric and Logic as required courses. That alone would knock out at least half of the garbage we have to put up with in both media and society...
I wouldn't be so sure of that. My uncle is a professor who teaches logic at a university where it is a required course; about a third of his students simply never "get it" no matter how many different techniques he tries or how much effort they put into it. Around 10% have trouble with simple logical inference of the form "If A is true then B is true. A is true, therefore ____"
So if all the images that are open in your web browsers all add up to 2 GB of uncompressed pixel data then the browser still shouldn't use hit the GB mark? I want my computer to be magical too.
Nothing magic about it. A full-screen browser window on a 30" monitor needs about 16MB of RAM. By decompressing and rendering only those portions of web pages that are visible, and by caching page data to disk rather than RAM, total memory usage won't be much larger than this.
This isn't just theory. Almost all web browsers prior to the early 2000s worked like this. They stopped because it's harder to code and introduces barely-perceptible delays in page rendering.
Fusion bombs have (almost) all been air-bursts or fully contained underground tests, and for good reason: the 100-kiloton Sedan nuclear excavation test of Operation Plowshare was responsible for 7% of all radioactive contamination from US nuclear testing.
Electrical damage to what? The pipes themselves? My super-duper-Japanese-electronic-ass-cleaning-toilet?
A thousand miles of gas pipeline can easily build up an electrical potential of tens or hundreds of thousands of volts. Any conductive object near the pipe is at risk of damage from electrical arcing. We're not talking about "minor voltage surge burns out circuits" here, we're talking "minor lightning strike melts shut-off valve".
Slashdot Mirror
That sort of comparison is exactly what this study is calling out, too. You know how that "8x background = evacuation" threshold was developed? By measuring the effects of a short-term dose just low enough to avoid acute radiation poisoning, and assuming it's equivalent to a low-level long-term dose that provides as much total radiation. (It's called the "linear no-threshold model". Look it up.) You're correct to say that 120 millisieverts over the course of 5 weeks (the study dose) isn't equivalent to 1200 millisieverts over the course of 40 years (a long-term 8x dose), but 1000 millisieverts over the course of 30 seconds (a dose that won't quite trigger acute radiation poisoning) isn't equivalent to it either.
Humans have *much* stronger anti-cancer mechanisms than mice. A mouse in its natural environment will take about three years to accumulate enough unrepaired DNA damage to get cancer; a human takes about 60 years. Five weeks of an elevated dose for a mouse is equivalent to about two years for a human, assuming the linear no-threshold model (the model that's the basis for public policy).
4'6" is the lower end of normal for women from Southeast Asia. If you're going for a "general guideline" formula, it needs to cover at least 99% of the population, everything from the 4'6" woman from SE Asia to the 6'7" average North American basketball player.
(BMI falls down for tall people, as well: the average NBA forward is moderately overweight if you go by BMI.)
You're looking at the wrong end of the height scale. Someone who's 4'6" would be considered "underweight" at 76 pounds, which is actually a healthy weight for someone that short.
Yes. I, at least, am aware that weight increases as the cube of height, not the square. Consequently, BMI tends to give numbers that are too high for tall people, and too low for short people. As an extreme example, many professional basketball players would be considered "obese" based on their BMI numbers, but "normal weight" based on body-fat percentage.
(Actually, because of changing body shape, it's around the 2.7th power for humans, but 3 is a better approximation to that than 2.)
Because when something goes wrong in the breaker box, it's easier to pull the negatives out of the file and look at them with a flashlight than it is to find a neighbor who will let you plug your computer in.
Takeoff isn't very stressful on an aircraft. Touchdown can be, but the stresses are concentrated in specific areas, typically the strongest parts of the aircraft, so you can repair or replace them if needed. Pressurization stresses every single joint and opening in the fuselage. You can inspect and repair individual problem spots, but eventually, the only practical option is to replace the entire aircraft.
My parents did something like this once when buying a car. My father did all the talking, while my mother sat there with a Consumer Reports article on car-buying in one hand and a pen in the other, very visibly checking off the high-pressure sales tactics as they were used. It took about five minutes for the salesman to cave and give them the price they wanted.
The fundamental difference between consumer and enterprise hard drives is the error-recovery strategy in the drive firmware.
A consumer-grade drive, upon getting an unrecoverable ECC error while reading a sector, will make repeated tries at recovering the data, spending seconds or even minutes at it -- good for a drive operating on its own, but the delay could cause a storage array controller to mark the entire drive as failed. An enterprise-grade drive assumes it's part of some sort of storage redundancy array and will give up quickly, letting the higher-level controller recover the data from the rest of the array -- much faster, but less reliable if the drive is operating on its own.
Are you sure of this? With my Internet connection, a full restore of 6TB of data would take somewhat over five years, which is a pretty big window for an Amazon failure.
At Ease had better integration with System 7 than Metro does with classic Windows.
The launch may have been believable, but the landing was dead wrong. There are only two ways to land a flapping-wing aircraft if you've got legs instead of wheels, and he didn't use either of them.
Historical evidence, mostly. Standard consumer drives have run at 7200 RPM for the past decade, high-performance drives have run at 15k RPM for just as long, and the 3.5" platter diameter has been standard for almost twice that.
They disagree because they disagree on how to define "the first human".
Two million years ago sounds about right for the "divergence of the Homo genus from the Australopithecus genus" definition. 100,000 years is about right for the "final divergence of Homo Sapiens and Homo Neanderthalensis" and "anatomically modern human" definitions. There's also the "minimum cranial volume" definition, the "minimum forebrain size" defintion, the "bipedal gait" and "upright posture" definitions, two different tool use definitions (using found tools vs. using made tools), and probably a number of others.
Assuming constant RPM, linear read speed increases as the square root of areal density (or roughly, capacity). The square root of 60 is 7.75, so a 60TB hard drive will take 7.75 times longer to fully read than a 1TB drive; if your 1TB drive can be cloned in "hours", then a 60TB drive can be cloned in 7.75*hours, which is probably "days".
What does the spectrum look like? With a suitable choice of primaries, you can get any tristimulus value you want (so you can give a white object whatever color you desire), but if you're not careful with the full spectrum, you can get some bizarre effects with other colors.
As an example, someone at the college I went to cooked up a pigment that would look red when lit by a compact fluorescent, green under his LED desk lamp, and yellow in broad daylight.
Do you consider 3200 kilometers a reasonable diameter for the helix?
I've got a machine that can perform every part of the process except the "remove dirty dishes from the table" step.
I call horseshit too, but for a different reason. He is indeed making a derivative work, but it doesn't fall into any of the classes of protected derivative work (parody, critical commentary, etc.), which makes it a straight-up copyright violation. Any lawyer or judge who's ever so much as glanced at copyright law knows this.
My computers have multiple names. They've got a canonical name that never changes (I use animals), and then a bunch of CNAME records for the services provided. So, for example, my internet services server ("firebird", I was feeling exotic that day) also has the names "www", "ftp", and "mail". If, for example, I decide to split the mailserver off to a new computer, the only thing I need to change is the DNS records.
I wouldn't be so sure of that. My uncle is a professor who teaches logic at a university where it is a required course; about a third of his students simply never "get it" no matter how many different techniques he tries or how much effort they put into it. Around 10% have trouble with simple logical inference of the form "If A is true then B is true. A is true, therefore ____"
I just bought 32GB for $1900. Memory is only cheap until you max out your computer's capacity, at which point you need to buy a whole new computer.
Nothing magic about it. A full-screen browser window on a 30" monitor needs about 16MB of RAM. By decompressing and rendering only those portions of web pages that are visible, and by caching page data to disk rather than RAM, total memory usage won't be much larger than this.
This isn't just theory. Almost all web browsers prior to the early 2000s worked like this. They stopped because it's harder to code and introduces barely-perceptible delays in page rendering.
Fusion bombs have (almost) all been air-bursts or fully contained underground tests, and for good reason: the 100-kiloton Sedan nuclear excavation test of Operation Plowshare was responsible for 7% of all radioactive contamination from US nuclear testing.
A thousand miles of gas pipeline can easily build up an electrical potential of tens or hundreds of thousands of volts. Any conductive object near the pipe is at risk of damage from electrical arcing. We're not talking about "minor voltage surge burns out circuits" here, we're talking "minor lightning strike melts shut-off valve".