Fair enough, but none of this addresses the question of value for the money. Maybe that $11 million would be better spent on some other project--the public money will create jobs wherever its allocated, but the whole point of infrastructure projects is that they create long term economic activity beyond just the initial lay out.
One could also question some of the architectural decisions--building it on a diagonal, the extra width to accomodate trees? None of that appears to be necessary if you look at the pictures. From here it looks a lot like MS had plans for a $25 million dollar bridge and then "upgraded" to a $36 million dollar bridge when they saw the stimulus money...
Fractals, being objects with a well defined fractal dimension, are not required to be iteratively defined or discontinuous, they just have to satisfy the limit formula for fractal dimension. Brownian motion, for example.
If you want to do scientific/numerical programming, learn Matlab, Mathematica, Python + NumPy / Sage, IDL (for astro) and maybe even Perl (for biotech). If you want to do statistical reporting you need to learn R, SPSS, etc, and get a masters degree in Stat because no one will trust you otherwise.
If you want to stick to the "ordinary" corporate world (meaning, everything else), forget all that and focus on SQL, XML, web services and refine the skills in Java, C#, VB.net etc.
An event can be completely determined by past information, but if that information is not available to the observer, then the outcome of the event is a probability distribution over possible outcomes given the unknowns.
For example: suppose I have an urn containing two marbles. One is black, the other is white. I draw one marble, and I *don't* look at it and throw it away. Now I draw the second marble. What is the probability the second marble is white? 50%. Similarly, the conditional probability of a quantum observation is completely determined given that you know the entire history of the universe (unless free will exists in which case there is "something else" still not accounted for).
The new Microchip PIC32 series has a 32-bit RISC core, single instruction per cycle at 80 Mhz and its nominally the same price as the older 8-bit micros (i.e. within a dollar or two). Moore's law applies to microprocessors, and 8-bit is on the way out.
There are many different standards for "accuracy", including "repeatability" and "flat", which are not the same. Headphones are superior to speakers for repeatability, when measured at the ear canal entrance. But they are not "flat" because they include a built-in simulated "free field response" HRTF that modifies the signal (at least, all consumer-market 'phones include this filter), plus some other geometric design issues.
With some work it is possible to get loudspeakers to give a flat response at a fixed reference listening position, but given two individuals it is impossible to guarantee that they will hear the "same" thing at that spot since there is no control over the HRTF--so, the repeatability isn't really there.
Also there is a difference between listening for artifacts (e.g. compression artifacts) and listening for mastering. Usually headphones are preferred for the former, but for mastering people usually prefer loudspeakers.
Yes, there is a reason, which is that they would sound terrible if they had a flat frequency response and nobody would buy them.
So, why is that: well, the "natural" way we hear sounds isn't "accurate" in the sense that not all frequency transduce with the same efficiency. The sound is modified by the geometry of your head and ears, also called the "head related transfer function" or HRTF for short. The HRTF is direction-dependent, it is also person-dependent as no two people have exactly the same head. Your auditory system understands your HRTF at a subconscious level and "factors it out" in determining the direction of sound and so on (for example sounds at higher elevation tend to have a bias towards higher frequency content created by the ear pinnae).
Now, headphones include a filter that applies a "simulated HRTF" that places the sound approximately directly "in front" of the listener. If they didn't include this, the sound would be very strange.
The downside to this is that the headphones' HRTF isn't individualized to your own head, and it can't be changed, and its exact specification varies from one model to another quite a lot. Usually the companies don't say exactly how the filter is constructed, and it requires some very fancy equipment (like dummy-heads and so on) to measure the headphone response accurately enough to make an inverse filter. The Sennheisser HD580 is one model (no longer in production) that we have some fairly extensive data for, and that is why it is still the standard for most auditory psychophysics research.
Loudspeakers on the other hand (in particular, reference loudspeakers for mastering) are actually designed to have a flat frequency response. Getting a good listening room isn't easy either, but if you work with a measurement microphone it is possible to check the results pretty easily.
On the subject of bass response, the impedance of air in the ear canal when closed off by the headphone is much much lower than the impedance of the driver in open air, which is why phones can deliver a quite good bass response with a very small driver.
Euphemism for "you're out of a job" (perhaps not the best one). i.e. so broke you have to buy ramen on credit, and so lacking in skills that you'll need to start fabricating a resume on linked in.
Apple did in fact close the source to some areas of Darwin. For example, certain low-level hardware drivers where the vendor didn't want the code to be published.
Why wouldn't these reports be available under FOIA? Considering that its "nominally public domain" already, what exemption would it fall under to bar a request?
But the communication bus between those brains is fantastically inefficient. It takes quite realistically dozens of years to transmit any idea of significant complexity.
Agreed. This could be solved by a semantic documentation generator. Input something like , run it through a processor and have it output either terminal instructions, a sequence of screen instructions, or a clickable automation script.
- Restoration of science to its proper place (a cheer went up in the crowd here at UC Berkeley) - "Non-believers" given equal mention in the laundry list of religions.
While its true they sold the bailout as a "loan", they also specifically stated that there was no guarantee taxpayers would get their money back.
Which is idiotic. Common practice is to use collateral to back a loan. The bailout operators refuse to take sufficient stock in companies to assure the collateral value. They also refuse to actually audit any of the companies they are bailing out to determine how much their stock is actually worth.
In other words, its either criminal mismanagement or a more likely, a giant scam. And tacking the the executive compensation terms on there is total PR stunt.
Anyone who has coded for microprocessors knows that silicon fabs contain bugs also. It is common practice to publish an "errata" to go with the datasheet that lists all the known problems with each revision of a processor. Sometimes the bugs have software workarounds, sometimes not and certain features will simply be unusable.
The reasons for hardware bugs are also quite similar--increasing complexity, and design tools that let you drop in large "prefab" modules.
Slashdot Mirror
Fair enough, but none of this addresses the question of value for the money. Maybe that $11 million would be better spent on some other project--the public money will create jobs wherever its allocated, but the whole point of infrastructure projects is that they create long term economic activity beyond just the initial lay out.
One could also question some of the architectural decisions--building it on a diagonal, the extra width to accomodate trees? None of that appears to be necessary if you look at the pictures. From here it looks a lot like MS had plans for a $25 million dollar bridge and then "upgraded" to a $36 million dollar bridge when they saw the stimulus money...
Fractals, being objects with a well defined fractal dimension, are not required to be iteratively defined or discontinuous, they just have to satisfy the limit formula for fractal dimension. Brownian motion, for example.
Maybe we should ban living in climates where the daily outside temp is hot enough to kill you.
Except that a heart attack is often confused, sometimes even by doctors, with an inflamation of the shoulder.
Ironically, this sort of tech is most likely to be picked up first by people who do extreme sports like mountaineering, high intensity training, etc.
If you want to do scientific/numerical programming, learn Matlab, Mathematica, Python + NumPy / Sage, IDL (for astro) and maybe even Perl (for biotech). If you want to do statistical reporting you need to learn R, SPSS, etc, and get a masters degree in Stat because no one will trust you otherwise.
If you want to stick to the "ordinary" corporate world (meaning, everything else), forget all that and focus on SQL, XML, web services and refine the skills in Java, C#, VB.net etc.
An event can be completely determined by past information, but if that information is not available to the observer, then the outcome of the event is a probability distribution over possible outcomes given the unknowns.
For example: suppose I have an urn containing two marbles. One is black, the other is white. I draw one marble, and I *don't* look at it and throw it away. Now I draw the second marble. What is the probability the second marble is white? 50%. Similarly, the conditional probability of a quantum observation is completely determined given that you know the entire history of the universe (unless free will exists in which case there is "something else" still not accounted for).
The new Microchip PIC32 series has a 32-bit RISC core, single instruction per cycle at 80 Mhz and its nominally the same price as the older 8-bit micros (i.e. within a dollar or two). Moore's law applies to microprocessors, and 8-bit is on the way out.
HD650 has a detachable cable. Maybe they've finally learned?
There are many different standards for "accuracy", including "repeatability" and "flat", which are not the same. Headphones are superior to speakers for repeatability, when measured at the ear canal entrance. But they are not "flat" because they include a built-in simulated "free field response" HRTF that modifies the signal (at least, all consumer-market 'phones include this filter), plus some other geometric design issues.
With some work it is possible to get loudspeakers to give a flat response at a fixed reference listening position, but given two individuals it is impossible to guarantee that they will hear the "same" thing at that spot since there is no control over the HRTF--so, the repeatability isn't really there.
Also there is a difference between listening for artifacts (e.g. compression artifacts) and listening for mastering. Usually headphones are preferred for the former, but for mastering people usually prefer loudspeakers.
BTW I use the HD650 also, they are awesome.
Yes, there is a reason, which is that they would sound terrible if they had a flat frequency response and nobody would buy them.
So, why is that: well, the "natural" way we hear sounds isn't "accurate" in the sense that not all frequency transduce with the same efficiency. The sound is modified by the geometry of your head and ears, also called the "head related transfer function" or HRTF for short. The HRTF is direction-dependent, it is also person-dependent as no two people have exactly the same head. Your auditory system understands your HRTF at a subconscious level and "factors it out" in determining the direction of sound and so on (for example sounds at higher elevation tend to have a bias towards higher frequency content created by the ear pinnae).
Now, headphones include a filter that applies a "simulated HRTF" that places the sound approximately directly "in front" of the listener. If they didn't include this, the sound would be very strange.
The downside to this is that the headphones' HRTF isn't individualized to your own head, and it can't be changed, and its exact specification varies from one model to another quite a lot. Usually the companies don't say exactly how the filter is constructed, and it requires some very fancy equipment (like dummy-heads and so on) to measure the headphone response accurately enough to make an inverse filter. The Sennheisser HD580 is one model (no longer in production) that we have some fairly extensive data for, and that is why it is still the standard for most auditory psychophysics research.
Loudspeakers on the other hand (in particular, reference loudspeakers for mastering) are actually designed to have a flat frequency response. Getting a good listening room isn't easy either, but if you work with a measurement microphone it is possible to check the results pretty easily.
On the subject of bass response, the impedance of air in the ear canal when closed off by the headphone is much much lower than the impedance of the driver in open air, which is why phones can deliver a quite good bass response with a very small driver.
Euphemism for "you're out of a job" (perhaps not the best one). i.e. so broke you have to buy ramen on credit, and so lacking in skills that you'll need to start fabricating a resume on linked in.
Don't worry, the economy sucks and they are expensive. Hence, available.
Apple deserves criticism in many areas. So does MSFT, but that doesn't mean they can't look at their competitors with a critical eye.
Apple did in fact close the source to some areas of Darwin. For example, certain low-level hardware drivers where the vendor didn't want the code to be published.
Why wouldn't these reports be available under FOIA? Considering that its "nominally public domain" already, what exemption would it fall under to bar a request?
But the communication bus between those brains is fantastically inefficient. It takes quite realistically dozens of years to transmit any idea of significant complexity.
How do you do the torture test? Is there a good tool for this?
The technical term for this feature is Time-limited Error Recovery (at least on WD drives, other brands might call it something else).
http://en.wikipedia.org/wiki/Time-Limited_Error_Recovery
Agreed. This could be solved by a semantic documentation generator. Input something like , run it through a processor and have it output either terminal instructions, a sequence of screen instructions, or a clickable automation script.
News for Nerds from the Obama speech:
- Restoration of science to its proper place (a cheer went up in the crowd here at UC Berkeley)
- "Non-believers" given equal mention in the laundry list of religions.
Mod parent up. The complex dynamics of economics around technological innovation have caused countless good ideas to be buried.
While its true they sold the bailout as a "loan", they also specifically stated that there was no guarantee taxpayers would get their money back.
Which is idiotic. Common practice is to use collateral to back a loan. The bailout operators refuse to take sufficient stock in companies to assure the collateral value. They also refuse to actually audit any of the companies they are bailing out to determine how much their stock is actually worth.
In other words, its either criminal mismanagement or a more likely, a giant scam. And tacking the the executive compensation terms on there is total PR stunt.
Anyone who has coded for microprocessors knows that silicon fabs contain bugs also. It is common practice to publish an "errata" to go with the datasheet that lists all the known problems with each revision of a processor. Sometimes the bugs have software workarounds, sometimes not and certain features will simply be unusable.
The reasons for hardware bugs are also quite similar--increasing complexity, and design tools that let you drop in large "prefab" modules.
Complete lack of conventional role separation is the norm at Google.
FWIW user-submitted material tends to produce low quality results anyways... not terribly surprising that they came up with this junk.