That poll is not scientific (people who find the poll are more likely to be people with problems), and out of people who upgraded the success rate is 68%. It is listed as 35% because they count successful installs separately from upgrades. Checking the polls for previous releases, the numbers are pretty much the same as this one.
I still don't use a new Ubuntu release for at least a few weeks though. There is always a flood of package upgrades for a few weeks after a release.
I once was handed a hard drive that would work fine when first plugged in, and then fail after 10-20 minutes of use when it heated up, and then work again after cooling down. I set up a fan to blow over it and copied off all the data in several short bursts and told my friend to buy a new drive. I can see how lazy technicians at a chain might quickly give up when handed something like that.
It could be a case of selection bias. Hot, fast moving planets are probably easier to detect than slow, cold ones. I don't think Kepler has much fear that this will disprove his work, his equations are based on geometry and are definitely correct (or can be corrected by general relativity). A problem in the model for how fast stars eat planets is more likely.
You're thinking of radiated power, this is magnetically coupled power. The wavelength is set by the resonant frequency of the coil and load. You want the coils to be much shorter than the wavelength to avoid radiating power (which is wasted).
This technology makes the very short range technique used in transformers work over a medium range (on the order of the coil diameter), but it is not and probably can not be a long range technique (like lasers). I only say probably and not definitely because I have studied science history.
This is magnetic coupling taking advantage of the resonant frequency of the system and is strictly a local effect. Radio's are electromagnetic radiation, which is where electrical and magnetic fields sustain each other will travelling through space indefinitely. The energy in the oscillation of a radio wave does not decrease with distance, it just gets spread out over a larger area (or absorbed, but that isn't relevant to this).
Your argument is effectively, "prove physical relation, break rules of physics" You can't make it smaller without making it less efficient. The resonance effect they are using to get decent efficiency degrades quickly once the separation is greater than the coil diameters.
And yet, every developed country with single payer (government run) healthcare spends less of its GDP on health care and achieves better results. Amazing. The reason of course, is that there is no real competition now but there is lots of profit for insurance companies; and bad performance due to screwed up fee structures that encourage expensive procedures instead of cheap prevention.
Personally, I think FORTRAN should be taken out and shot. Godawful unmaintainable code is the norm , and there is nothing you can do in FORTRAN that you can't do in a cleaner environment like C++. People only still use FORTRAN because they are used to it, and it will not go away until all the old fucks die.
A lot of that has more to do with who wrote the code than the language. The new C++ analysis code coming out also tends to be "godawful". And FORTRAN95 supposedly has a lot of modern features, so new code can be reasonable. As the article you quoted points out:
Finally, there is no need to be a religious extremist; these days, procedures in different languages can be intermixed (which will only become easier). Different subsets of applications can be written in the most appropriate language, e.g. Java for the web, and all can be glued together in a PYTHON framework. On the other hand, it is also possible to misuse almost any language: C++ includes C, and Fortran still has the GO TO function.
Fortran has tons of libraries specialized to whatever scientific field you are working in, and is unavoidable in high energy physics especially. Of course, most of these can be wrapped in C and then used in whatever high level language you like.
I like matplotlib and scipy but I have lots of problems with memory not being freed when making and closing a lot of plots. It might be something I'm doing though, so I'm curious if the book discusses this.
Normally I am skeptical of claims about radio waves causing people harm, but 100MW/9 means Vatican Radio is using over 10 MEGAWATTS. I nearly soiled myself. Cancer is no surprise. Hell, I'm surprised there aren't birds dropping from the sky pre-cooked (yes I'm aware that due to the wavelength absorption in tissue is low). Instead of mutating children and spewing nonsense, they could be providing power for a few thousand poor villages.
How would it help Chinese workers is they are out of jobs???
It wouldn't. But slave labor and unemployment are not the only options.
How would it help the developed world if everything costs more???
You might see a resurgence of a local service industry repairing these more expensive components. So the chinese could get paid more to build higher quality parts, and we wouldn't need to buy a new keyboard every year, and there would be some moderately skilled service jobs in the developed world.
When people/companies/countries trade, then both of the trading partners are better off. Otherwise they wouldn't trade...
Trade is generally, though not always, good for society as a whole. It is always good for the traders. It is often not good for the labors, especially when one partner in the trade uses unethical labor practices to reduce prices.
QAM and commercial AM radio are not the same thing. QAM is used extensively for digital communications, especially in modems. It isn't as common is wireless communication, so far as I know.
Of course, if this was just a lame excuse to bash Rush Limbaugh, carry on.
I'm no Microsoft fan, but it sounds as if $59.25 to get a completely different commercial OS, XP, isn't an egregious fee when you purchased the crummy consumer version of the newer OS, Vista.
In order to purchase the XP 'downgrade', you also had to purchase Vista Business. So the actual cost over Vista Home was closer to $150 dollars. Linux, or no OS, was probably not available as an option, arguably because of Microsoft's unfair business practices.
That's because the TFS left out a very important detail. The laptop will have two processors.
It runs with two chips, one from ARM and one from Intel. The ARM chip, provides instant on booting and is much more power efficient, while the Intel chip provides the juice to run apps that require more computing power.
So Linux can do heavy lifting, but the ARM chipping running it can't.
You can verify the result. If you're familiar with some basic communication theory, consider one of the operands as the source, encode it with some redundancy (parity bits for instance), and pass it through the channel which is the 'probabilistic' processor and the other operand. The channel parameters will give you a probability model for the output which you can use to try to correct any errors that occurred. This situation is very similar to a regular communications system, except there you don't even know the original signal (if you did, you wouldn't need to communicate it).
That is so 1960s. Modern error correction algorithms are much more efficient. If the machine runs 7x faster and you use an error correcting algorithm that uses 2x as many ops you're still 3x ahead. And you could selectively apply the error correction depending on where you needd more or less accurracy. Even for financial applications, there is an upper bound to the accuracy required.
Slashdot Mirror
Am I missing something?
Unsurprisingly, yes. You can get a DVD in the mail.
I have a better idea. Canonical can release every six months and you can upgrade once a year.
I still don't use a new Ubuntu release for at least a few weeks though. There is always a flood of package upgrades for a few weeks after a release.
I once was handed a hard drive that would work fine when first plugged in, and then fail after 10-20 minutes of use when it heated up, and then work again after cooling down. I set up a fan to blow over it and copied off all the data in several short bursts and told my friend to buy a new drive. I can see how lazy technicians at a chain might quickly give up when handed something like that.
mmm ... planet. Tasty.
You're thinking of radiated power, this is magnetically coupled power. The wavelength is set by the resonant frequency of the coil and load. You want the coils to be much shorter than the wavelength to avoid radiating power (which is wasted).
This technology makes the very short range technique used in transformers work over a medium range (on the order of the coil diameter), but it is not and probably can not be a long range technique (like lasers). I only say probably and not definitely because I have studied science history.
This is magnetic coupling taking advantage of the resonant frequency of the system and is strictly a local effect. Radio's are electromagnetic radiation, which is where electrical and magnetic fields sustain each other will travelling through space indefinitely. The energy in the oscillation of a radio wave does not decrease with distance, it just gets spread out over a larger area (or absorbed, but that isn't relevant to this).
Your argument is effectively, "prove physical relation, break rules of physics" You can't make it smaller without making it less efficient. The resonance effect they are using to get decent efficiency degrades quickly once the separation is greater than the coil diameters.
And yet, every developed country with single payer (government run) healthcare spends less of its GDP on health care and achieves better results. Amazing. The reason of course, is that there is no real competition now but there is lots of profit for insurance companies; and bad performance due to screwed up fee structures that encourage expensive procedures instead of cheap prevention.
Personally, I think FORTRAN should be taken out and shot. Godawful unmaintainable code is the norm , and there is nothing you can do in FORTRAN that you can't do in a cleaner environment like C++. People only still use FORTRAN because they are used to it, and it will not go away until all the old fucks die.
A lot of that has more to do with who wrote the code than the language. The new C++ analysis code coming out also tends to be "godawful". And FORTRAN95 supposedly has a lot of modern features, so new code can be reasonable. As the article you quoted points out:
Finally, there is no need to be a religious extremist; these days, procedures in different languages can be intermixed (which will only become easier). Different subsets of applications can be written in the most appropriate language, e.g. Java for the web, and all can be glued together in a PYTHON framework. On the other hand, it is also possible to misuse almost any language: C++ includes C, and Fortran still has the GO TO function.
Fortran has tons of libraries specialized to whatever scientific field you are working in, and is unavoidable in high energy physics especially. Of course, most of these can be wrapped in C and then used in whatever high level language you like.
For the homeless to use on their shiny laptops? Libraries already have those materials, and are warm.
I bought these when they had 5x2GB for the same price, but it is still a decent deal. They're not fancy, but they do work well. http://www.newegg.com/Product/Product.aspx?Item=N82E16820326009 100 1GB flash drives delivered to your door for $412.
I like matplotlib and scipy but I have lots of problems with memory not being freed when making and closing a lot of plots. It might be something I'm doing though, so I'm curious if the book discusses this.
Normally I am skeptical of claims about radio waves causing people harm, but 100MW/9 means Vatican Radio is using over 10 MEGAWATTS. I nearly soiled myself. Cancer is no surprise. Hell, I'm surprised there aren't birds dropping from the sky pre-cooked (yes I'm aware that due to the wavelength absorption in tissue is low). Instead of mutating children and spewing nonsense, they could be providing power for a few thousand poor villages.
How would it help Chinese workers is they are out of jobs???
It wouldn't. But slave labor and unemployment are not the only options.
How would it help the developed world if everything costs more???
You might see a resurgence of a local service industry repairing these more expensive components. So the chinese could get paid more to build higher quality parts, and we wouldn't need to buy a new keyboard every year, and there would be some moderately skilled service jobs in the developed world.
When people/companies/countries trade, then both of the trading partners are better off. Otherwise they wouldn't trade...
Trade is generally, though not always, good for society as a whole. It is always good for the traders. It is often not good for the labors, especially when one partner in the trade uses unethical labor practices to reduce prices.
Thanks for the heads up, I was about to dismiss this as fluff. Instead it is something 10-20+ years off, but much more interesting.
QAM and commercial AM radio are not the same thing. QAM is used extensively for digital communications, especially in modems. It isn't as common is wireless communication, so far as I know. Of course, if this was just a lame excuse to bash Rush Limbaugh, carry on.
I'm no Microsoft fan, but it sounds as if $59.25 to get a completely different commercial OS, XP, isn't an egregious fee when you purchased the crummy consumer version of the newer OS, Vista.
In order to purchase the XP 'downgrade', you also had to purchase Vista Business. So the actual cost over Vista Home was closer to $150 dollars. Linux, or no OS, was probably not available as an option, arguably because of Microsoft's unfair business practices.
It runs with two chips, one from ARM and one from Intel. The ARM chip, provides instant on booting and is much more power efficient, while the Intel chip provides the juice to run apps that require more computing power.
So Linux can do heavy lifting, but the ARM chipping running it can't.
You can run your installed windows partition under a VM and avoid having to reboot at all.
You can verify the result. If you're familiar with some basic communication theory, consider one of the operands as the source, encode it with some redundancy (parity bits for instance), and pass it through the channel which is the 'probabilistic' processor and the other operand. The channel parameters will give you a probability model for the output which you can use to try to correct any errors that occurred. This situation is very similar to a regular communications system, except there you don't even know the original signal (if you did, you wouldn't need to communicate it).
That is so 1960s. Modern error correction algorithms are much more efficient. If the machine runs 7x faster and you use an error correcting algorithm that uses 2x as many ops you're still 3x ahead. And you could selectively apply the error correction depending on where you needd more or less accurracy. Even for financial applications, there is an upper bound to the accuracy required.
Isn't that similar to the model Selinux uses?