If you kept all the violent and really evil, sick, twisted people locked up, who would we have for lawyers? Politicians? Accountants? CEOs?!
The Huge Manatee!
Please, please, think of the scumbags!
(Seriously, I don't have any objection to people who need to be locked up being locked up. I do have an objection to that being the sole purpose of the legal system - people can change and it's fair to give them the means and opportunity even if they stay incarcerated. It also seems reasonable to give those who will be released the motive to try an alternative. Punishment alone never works and resentment is a great way to encourage people to think of other ways of being crappy to others.)
I regard the death penalty as somewhat childish and immature. "If X can't be alive, then... then... Neither Can Yoooooo! So nyah!" The idea that it gives closure to anything seemed to get a kick in the nuts with the Beltway Sniper's execution. If you don't get closure when the other person doesn't cry, then I'm not sure it's "closure" you're looking for. Try looking up "schoolyard bully".
I'm also not keen on the way a lot of these trials are handled, especially the insanity stuff. A person being insane doesn't alter whether or not they did something, it merely alters their culpability. That should be obvious.
Ergo, it follows that insanity should not be a plea in the trial phase but confined strictly to that phase which deals with culpability, the sentencing.
However, I also disagree with this idea that there are two options - total all-out criminal insanity and total all-out sanity. For a start, it doesn't leave you with anywhere to put lawyers or politicians.
I would far prefer to see a system in which sanity is regarded as a sliding scale and where sentencing allows the judge to split the time between punishment, treatment and rehabilitation (as and where appropriate) according to what produces the best outcome overall, rather than according to what gives the weenies in the press box a vicarious thrill.
Obviously, if a person is going to be incarcerated forever, then rehabilitation to the point where the person would be safe outside is not terribly useful. On the other hand, it seems reasonable to assume that having them stew, rebel and resent is both less cost-effective and less mature than encouraging them to make effective use of their abilities.
Just because someone is sealed off from society doesn't mean society can't benefit from their mind - there's probably plenty of intellectuals and artists behind bars.
Ian Brady is probably one of the craziest crazies to be in Broadmoor, but his book on the way serial killers think, feel and act should certainly be at least browsed by psychiatrists and detectives for insights no rational mind could ever have produced. No matter how little value it really is, the chances are really good that it'll do more good than the British Police's DNA database and CCTV camera system.
I'd rather let a hundred cold-blooded killers live in jail and receive at least some respect as a person if it meant that just one of those hundred produced a masterpiece of art or a book that had significance than have all hundred die purely for the viewing pleasure of Weekend Warriors.
In a hundred years time, which makes the difference? Something that might only rarely advance humanity - but when it does, advance it a lot - or something that provides a momentary mental orgasm for a bunch of f'ed-up "witnesses" and some losers outside and that's it?
I don't see why I should pay taxes for someone getting off on watching another die, when I could be paying taxes to give those in prison a chance to do something positive and worthwhile.
Pinhole cameras work great. Admittedly, they don't use cheap plastic lenses in those, but since I dissed the cheap plastic lenses elsewhere, I've got that covered.
Weren't casio watches considered evidence of involvement in terrorism in Afghanistan? And yours works off nuclear stuff, too! Definitely dodgy.
(On the other hand, that is one seriously cool watch. Probably not one I could afford right now, but with sales of anything being so bad this year, you never know.)
Out-nerded you, then. It was temporarily repaired in Logopolis, although the TARDIS only actually changed shape in one story quite a bit later than that and never changed again.
Come to think of it, it was invisible in The Invasion and therefore might have been any shape at all prior to rematerializing as a police box at the very end.
The Sensorites were able to remove the locking mechanism, resulting in a change to the outward appearance. Thus, even then, the chameleon circuit wasn't so stuck that it couldn't adjust the shape accordingly.
For that matter, although The Doctor has claimed the TARDIS can't change shape, it has not retained the same police box exterior, showing that some chameleon capability exists.
My personal suspicion is that the First Doctor forgot how to change the shape and that later Doctors kept the basic design as a cross between a trademark and a banner.
Heh. First, I'm increasingly skeptical of New Scientist. It went downhill after scrapping the Arianne column. Second, computers can't generate accurate information that isn't there, although they CAN synthesize data values that would be valid based on known information.
Using sound as an example, since you know the general shape of a sine wave, it is perfectly valid to non-linearly interpolate a digital recording to produce an analog waveform that would have produced an identical digital recording. What you cannot do is know if this was the actual original waveform.
If you're really clever, you may even be able to spot relationships between different components of the digital recording. For example, you know that any waveform can be split into simpler waveforms (fourier analysis) and you know that some component waves should relate to specific other component waves (many natural sounds will have harmonics, there may be echoes, and so on). This might allow you to identify some reconstructions as more probable than others. There may be other characteristics and meta-information you can use to refine the process further.
Now, I'll grant the possibility of some very, very good reconstructions from a good image. I'll even grant that a sophisticated enough camera may be able to capture meta-data that could be used to produce a brilliant reconstruction.
What I can't possibly accept, though, is a good reconstruction from a lossy capture at low dynamic range and low resolution and an even lossier image compression into JPEG or other lossy image format. Passable, perhaps, but not good.
Now, if someone could build a camera for a phone that could store images in OpenEXR, -and- have the sensor good enough that using an HDR format made sense, -and- have enough memory on the phone that you could store an image of a few megapixels res, THEN I could accept just about any reconstruction claim the inventor cared to make for the images. Mind you, why would you need to reconstruct anything if the camera is doing all that already?!
If all you measure is the impact of a photon of a very specific energy, the event happens or it doesn't. There's not a whole lot of alternative states.
And, no, there aren't going to be a whole lot of analog effects that interfere with the light-sensitive component. You can't have half an electron jump (it's quantized) and you can't have a fraction of a count (it increments by one or it doesn't). Photon counters are remarkably reliable and very accurate. They're also horribly slow (you accumulate state over a long time) which makes them brilliant for astronomy but useless for even still pictures if the scene has any moving component.
Analog effects WILL affect the analog component, but as I clearly differentiate, you're being an idiot if you confuse the two.
Hard radiation can always mangle solid-state circuits, sure, but I wasn't really considering those to be a typical scenario.
Thermal problems really won't affect a photon counter, as that just alters the movement of atoms. If the camera is exposed to enough heat that the electrons migrate, and it's a point-and-shoot (which basically means plastic lens and cardboard or plastic case), the camera is going to have more problems than a spurious value.
(Remember, we are talking point-and-shoot vs camera phone here, not the Hubble Telescope.)
Reference voltages won't affect the count, because you don't use any. Reference voltages will affect the analog circuits, though, but again those are considered separately.
Quantum effects shouldn't be a problem, since you are not measuring anything (you are only counting) and a pixel isn't on the same order of magnitude as the wavelength.
Remind me, is there anything I'm missing on the list?
Now THAT is a valid argument. If the phone camera does everything you want and need for your general use, then it doesn't matter if a PAS camera is better in any technical sense. I'm still not convinced you really get 5MP, even if the box says you do, but if it gives the results you want then who really cares?
Even DACs are digital devices that produce analog signals, but ANY device based on the photoelectronic effect at a fixed frequency is a digital device (it's either on or off). The analog signal generated is dependent on how many photons are collected in a given interval of time.
God, has Slashdot descended to such depths in only a few decades that people haven't looked up the alternatives to CCDs?
I am, and I'll accept that you are, but the vast majority of people have a really suspect intellectual capacity, a very shaky grasp on reality (I hear some people even believe politicians, accountants and/or Scientologists) and an attention-span of a 3 year old. On a good day.
When you consider that Oprah Winfrey is considered to be the height of intellectualism on US television and Coast-to-Coast AM has more credible stories than many of the popular news outlets, it's clear that the species has some serious shortcomings.
On the other hand, Neolithic people had perfectly functional and valuable neurosurgeons, had tools with better-than-millimetre precision, were capable of large-scale transatlantic sea voyages and were building some very good echo chambers.
So when you compare modern humanity (by which you realize you're including Palin supporters, rednecks, New York taxi drivers, Frank Bruno, Hulk Hogan and Terry Wogan) with Neolithic and Paleolithic people, I'm really not convinced humans are advanced as you think.
A decent analog watch has superbly-crafted components and therefore deserves the admiration. A decent digital watch needs something marginally better than a 555 timer, but only marginally.
That's not to say I wouldn't gave respect for a digital watch with the accuracy and drift of a scientific high-precision timer. I'd consider it adequately impressive for a digital watch to have a drift of a few nanoseconds a day.
(I'm judging analog differently than digital in the resolution, but really about the same in terms of the effort required to produce a system of a given standard because that's the only measure of craftsmanship you can ever really have.)
What you really want is a James Bond wristwatch with 0.5T magnetic field generator whenever you turn the dial. Saves on having to spend 5 seconds looking at the bloody time when the server room explodes and the users' machines go into meltdown.
Maybe it's the Asperger's in me, but after getting years of verbal grief for being different, I have no problem with neurotypicals begging. Actually, I don't think that is the Asperger's. I think I might just have become more cynical and more anti-social in my old age. (And gerroff the lawn!)
Well, there's that, but also bear in mind that cameras can afford to put a bit more power into the electronics, so that JPEG compression can be of higher quality.
Doubling the number of pixels on the CCD but more than halving the amount of retrievable data stored will give you a net loss of quality. High-res CCDs are relatively cheap and since the phones don't advertise the resolution of the image as stored, it's a great marketing ploy.
It's next to impossible. Phones need to be very small, lightweight and damage-resistant, the electronics need to be exceedingly low-power and the electronics for the camera and the electronics for the radio transceiver can't conflict.
That last requirement means is you use digital devices that produce analogue signals, the resolution on the ADC has to be so crappy that the RFI from the radio doesn't screw up the picture AND the voltage changes when a call is picked up or an alarm goes off or what have you can't throw the ADC.
The low-power means no fancy, power-hungry logic, the software zoom and other floating-point logic won't be terribly high precision, and the image compression algorithm will need to be light on the quality.
The size and damage-resistance impacts what sort of lens you can use, how rigid the structure has to be, how much the user can just seriously screw up the device before the image quality drops. Even for a disposable standalone camera, it's practical to put in some quite acceptable optics.
Even when such devices are of a size comparable to that OF the phone, you've got to remember that the camera is sans radio (or radios, for phones that have bluetooth and/or wifi and/or AM/FM tuners as well as the standard phone radio), sans keyboard, sans quite a bit of space-hungry stuff that phones either need or have as "features".
Talking of sci-fi, phone booths can't become obsolete! They're needed for "The Duelling Machine" and also for The Doctor's TARDIS! I'm pretty sure they'll be needed when the Triffids attack, too, and they could be handy when "A for Andromeda" happens.
Inherent in the idea that Firefox prohibit malicious code breaking into an extension and causing it to reformat/dev/hda1 is the idea that the extension itself cannot reformat/dev/hda1, as it is impossible for Firefox to know what will or will not cause an extension to do something maliciously.
Indeed, if an extension can do anything, it is possible to write an extension that allows an external program to control your computer with the privileges of Firefox AND (this is the important part) it is ALSO possible for a buggy extension to give such control of your computer with those same privileges.
If an extension can do absolutely anything, without any restriction, so can malicious code. The compromise suggestion was to have privileged operations in a distinct process. An extension overall can then still do anything but a given extension component cannot. (You didn't comment on that - not see it or you prefer to find something bad to say?)
Compartmentalizing extensions would, yes, make life much harder for extension writers. It would force them to be disciplined, code properly and not incorporate stupid security flaws. This would eliminate a lot of stupid coders, but I'm having a hard time seeing that as a bad thing.
If that was unacceptable, if you had to have the bad coders, then is sandboxing really so terrible? I honestly can't think of too many things Java application writers can't do, but Java is sandboxed. Forth programmers don't seem to have too many problems either - it's a very popular language for developing very low-level code like BIOSes - and yet also runs entirely in a compartmentalized virtual machine.
Hell, one could argue that any Trusted OS (ie: an OS that runs on top of a very thin security OS that provides all the operations that have security implications) is essentially an OS running in a sandbox. It might be harder to run the latest *ix games or applications under Trusted Solaris than it is under Linux, but unless you can name something you CANNOT run AT ALL, I'm inclined to believe that the limitation you speak of simply does not exist.
So if Firefox sandboxed extensions then it might need to provide some extra functionality via extensions to the existing API. Doesn't sound too horrible and it's certainly not fatal to developers.
So we definitely have multiple ways of improving security without preventing extension writers from doing what they want -- the only thing improving security would impose is HOW extension writers did things. Again, is that a bad thing?
I would rather see the death of bad code than see the death of Firefox because it got a reputation for being worse than IE on security. Particularly if the reputation was not due to Firefox per se but because extension writers were drunk or lobotomized at the time.
There's really no excuse for Firefox to allow at least some of the more common security flaws - or at least allowing those flaws to cause problems.
First, sandboxing of extensions should limit what problems can be caused.
Second, a lot of errors are caused by the overflowing of buffers - a problem that could be limited by the use of stretchy buffers or bounds-checking malloc implementations. Or not allowing direct access to the heap.
Third, Firefox (and indeed all programs) should run on the principle of least privilege. Where some specific subset of program functionality requires significantly greater privilege than the rest, run the subset as a different thread or process at a different level of privilege. By extension (bad pun, I know), extensions could also be run as a different thread or process with even fewer rights. (OS' that don't allow programs to shed rights might be a problem, though.)
Oh, it's practically a given that the non-linearity won't be the same at all scales. But complex behaviour can be produced by very simple non-linear systems - the Mandelbrot Set being the best-known example, so the presence of non-linearity merely creates a problem of practical computability rather than a problem of mathematical computability.
(Remember, to be computable in the mathematical sense, the algorithm has to complete in finite time. Which can mean anywhere from a few picoseconds to an hour after the heat-death of the Universe. To be practical, though, the model must produce results within the time the results are useful. To be commercially practical, it also has to produce results faster than other methods of getting those results.)
So we're looking at nested non-linear systems, no matter what starting point we're using.
Let's start with a bottom-up approach.
In the biological world, each cell has multiple mechanisms running in parallel where each mechanism is non-linear. The cell itself is a non-linear construct of these. There are different types of interconnect and these are also non-linear, so any network of cells is a non-linear construction of non-linear components. The brain has topological constraints, but unless there's grounds for believing those constraints to fundamentally alter the maths, the maths should be independent of implementation details.
This says we're looking at a nesting 3 deep. So we're looking at a chaotic system in which potentially all of the parameters are themselves chaotic systems in which potentially all of the parameters of that are also chaotic systems.
What else do we know? We know that the lowest-level systems are fundamentally unchanged from how they were 3.5 billion years ago when cellular life first arose. They may be chaotic but the building-blocks are all very simple. The only real internal changes have been in the organization of the building-blocks. All other changes within cells deal with interactions and mathematically interactions are on a different level.
Most of those lowest-level systems are common to heart cells, skin cells and brain cells. Now, this will include communication mechanisms and those we DO have to consider. Basic housekeeping that is a product only of it being biological can be ignored. Systems specifically activated in neurons and NOT common across all cells also have to be considered, even if housekeeping, as state is persistent in neurons by means of such housekeeping.
Now, the mechanics of these functions aren't what's important. What's important is what they do to the logic of a neuron to make it capable of data processing.
The cell itself is a network of these. In standard computer network terms, you're looking at the equivalent of a multicast-capable routing-capable ad-hoc network of moderate size. This is just for a single neuron, we're not even up to networking these things. Actually, strictly speaking, it's multiple such networks. In biological cells, you've independent chemical and electrical paths. Different latencies and different bandwidths.
Unless there is firm evidence that this is an implementation detail that does not alter the specification, I believe that it is wisest to assume it DOES alter the specification, that signal delays and other signal characteristics are important. Some variables from iteration X of the system are fed into iteration X+1, but others are fed into iteration X+N (where N can't be guaranteed to be a constant). This is what makes it a chaotic system of chaotic systems rather than merely a bigger chaotic system.
Now, the network of cells is basically more multi-path networking where again different types of interconnect have different properties. Further, not only are the nodes in the network effectively mobile and multicast, but the number of nodes is variable.
(We can ignore the number of connections a given neuron has by looking at the superset of functions exhibited by all types of cell in the brain, whether neuron, axion, or wha
People doing useful and interesting research frequently post on Slashdot, so I don't see what your problem is. It doesn't take a genius to mathematically model a brain and that isn't something people have bothered much with doing.
Some things people have tried to do are build models of compartments of the brain (bad idea), simulations of some poorly-specified upper-level functions of the brain (even worse idea) and discrete/binary simulations of individual neurons assuming them to be stateless and/or with a rigid topology (talk about dense).
The first is like trying to build a model of one part of a Mandelbrot set. A complete waste of time, since the maths doesn't work that way. The second is stupid because without a good specification, there's nothing meaningful to simulate. And since neurons are neither discrete, stateless nor in a fixed network (even adult brains have a surprisingly dynamic topology), all you get is a simulation of something that never existed instead of a simulation of the thing you want.
Why do people do these things? Because they're very doable. Neural networks are a doddle to code up, logic chains and decision trees are trivial on a computer, and since more people are interested in medical applications than AI, understanding compartments is far more practical than understanding the brain itself.
In short, people want to be paid far more than they want to discover, especially since discovering the mathematics of the brain won't do you any good as it'll be well outside the capacity of any machine out there (including the 100 million core one) to do anything sensible with such a model. Nobody likes inventing things that can't be used for another 50-100 years.
However, the fact that nobody WANTS a real mathematical model of the brain doesn't change the fact that the brain is an extremely simple device (mathematically-speaking). The unwritten part of the challenge is that they want a mathematical representation they can use and it is that which does not exist and will not exist for at least the next 50 years, simply because of the technology. The maths is a non-issue.
As far as Navier-Stokes is concerned, there are no reasonable assumptions. Particles do not move with a uniform speed, speed follows a bell curve. Well, almost, as a bell curve has infinite tails in both directions but in physics you're bounded. Particles are strictly between 0 and C and cannot take on either value or anything outside of the range.
In practice, since you don't see too many Bose-Einstein Condensates or even hypersonic particles when boiling water for an egg. However, even in a pan of cold water, there'll be water molecules moving fast enough to leave the liquid, and even when the water is boiling, there'll be water molecules that have the kinetic energy of a slug. Not many, but there will be some.
That's your first problem, because the first simplification is to decide what sort of range of speeds particles are likely to move at. The reality is "all of them, at some point or another".
The second problem is this differentiation between compressible and non-compressible fluids. In the same way that speed is non-uniform, density is also non-uniform. That means all fluids will have a mix of the two characteristics.
The third problem, as I've already pointed out, is that the system is chaotic. This means you need an infinitely fine grid and an infinitesimal time interval between iterations. Neither of these is possible. However, chaotic systems don't necessarily improve as you improve resolution, which is why CFD is often far more coarse-grained than you might expect. It has nothing to do with context, or even compute power, it has to do with experimentally finding a resolution where the results are similar (through the property of self-similarity) to what you might get if you could work at infinite resolution.
Self-similarity is NOT the same as identical, though, which is why most competent hardware engineers treat CFD as being a first approximation at
Ideal circles do not exist, that is true. So what? The idea that you need an ideal form is Platonic (it comes from Plato's cave analogy). Does there need to be some ideal, in order for an approximation to exist? (Well, C++ and Smalltalk programmers can skip that question.)
Let's try a different example. Let's go for the Second Law of Thermodynamics. Statistically speaking, it's universally true. There are no exceptions on the macro scale of space/time. If you were to examine a small patch of quantum foam over a few picoseconds, it would be lousy even as an approximation.
Does this mean the Second Law is wrong? No, not really. Does this mean the Second Law is artificial, as it's only an approximation? No, I think you'll find the early universe obeyed it long before there were any observers.
So what does it mean? It really doesn't mean very much at all. It means you're asking the wrong questions and not getting useful answers.
As for the quotes, Albert Einstein was very bad at maths and I think Benoit Mandelbrot (amongst a few thousand other chaos and fractal specialists) would beg to differ on the ineffectiveness of mathematics in biology. It's hardly the mathematics fault that biologists are lousy mathematicians.
My argument is that it is quite immaterial as to whether Pi is universal or not. In any given specific space, there will be -a- constant that denotes the ratio between the circumference and the diameter. The fact that there exists a constant for a given space (whether or not there exists the same constant for all spaces) means that the property of the ratio is fundamental.
Iff* the same constant holds for all spaces, then Pi as we know it is -also- fundamental, but I am unsure this has been proven. My statement that we can split off what is artificial from what is fundamental is unaffected.
*Maths notation: If and only if
The only way it can be proven that mathematics is wholly artificial is to prove that the set of all mathematical "things" that are fundamental is equal to the empty set. ie: there is nothing - not a single property, not a single result - that is true everywhere, including Goedel's Theorum. If even something as simple as Goedel's Theorum is universal, then there exists at least one part of mathematics that is not invented but is wholly natural.
Now, here we run into a problem. If Goedel's Theorum is not a universal result, but an artifice, then it is also false because it would have to be possible to create a counter-example and the theory states no counter-example of this kind can exist.
Surely that seals the argument right there and then. Those who argue mathematics is wholly artificial must be arguing Goedel's Theorum is false. All other cases do not prohibit the theorum from being true. Thus, if there is sound reason for believing the theorum true, there is sound reason for excluding the notion that mathematics is an artifice.
No, the bicycle is equivalent to a number base or a mathematical system. It is an implementation OF an underlying system (in this case, Newton's Laws), but Newton's Laws would still remain exactly the same whether Newton - or indeed bicycles - had ever existed.
The definition is also immaterial, as that too is an implementation detail. The underlying principle would remain unaltered whether the definitions of circumference, diameter or pi had ever been developed.
You are confusing the overlaid system with what it overlays. I'm saying you don't need to. Your argument is that the overlaid system is artificial, an invented product. I'm saying you're entirely correct on that. But what I am also saying is that what the product overlays, what is beneath the terms, the dynamics and the fancy Greek lettering is not artificial but exists whether it is known to exist or not.
The problem with assuming the two layers are the same is that you run into the Anthropomorphic Principle - the universe is the way it is because it produced people capable of seeing it. Let us, for a moment, assume the Many Worlds theory of Quantum Mechanics is correct. Then there are universes OTHER than the one we see and the theory falls down. The same would be true if the model of a multiverse as a foam (where each universe is a bubble in that foam) is correct.
But if you're on this site, you should be familiar with layering anyway. Maths - the fundamental, overarching thing that is shown in all mathematical systems that exist, will exist or ever have existed - is a Layer 1 concept in the OSI model. Concepts like numbers and other fundamental but artificial building blocks are Layer 2, which makes Group Theory a layer 2 switch. Anything and everything that MUST be true because of something in layer 2 is arguably also layer 2, which would include Goedel's Theorum. Anything that is true only in a specific implementation of mathematics is layer 3 or above.
Does using an OSI representation make it easier to see how not all maths is the same?
Slashdot Mirror
If you kept all the violent and really evil, sick, twisted people locked up, who would we have for lawyers? Politicians? Accountants? CEOs?!
The Huge Manatee!
Please, please, think of the scumbags!
(Seriously, I don't have any objection to people who need to be locked up being locked up. I do have an objection to that being the sole purpose of the legal system - people can change and it's fair to give them the means and opportunity even if they stay incarcerated. It also seems reasonable to give those who will be released the motive to try an alternative. Punishment alone never works and resentment is a great way to encourage people to think of other ways of being crappy to others.)
I regard the death penalty as somewhat childish and immature. "If X can't be alive, then... then... Neither Can Yoooooo! So nyah!" The idea that it gives closure to anything seemed to get a kick in the nuts with the Beltway Sniper's execution. If you don't get closure when the other person doesn't cry, then I'm not sure it's "closure" you're looking for. Try looking up "schoolyard bully".
I'm also not keen on the way a lot of these trials are handled, especially the insanity stuff. A person being insane doesn't alter whether or not they did something, it merely alters their culpability. That should be obvious.
Ergo, it follows that insanity should not be a plea in the trial phase but confined strictly to that phase which deals with culpability, the sentencing.
However, I also disagree with this idea that there are two options - total all-out criminal insanity and total all-out sanity. For a start, it doesn't leave you with anywhere to put lawyers or politicians.
I would far prefer to see a system in which sanity is regarded as a sliding scale and where sentencing allows the judge to split the time between punishment, treatment and rehabilitation (as and where appropriate) according to what produces the best outcome overall, rather than according to what gives the weenies in the press box a vicarious thrill.
Obviously, if a person is going to be incarcerated forever, then rehabilitation to the point where the person would be safe outside is not terribly useful. On the other hand, it seems reasonable to assume that having them stew, rebel and resent is both less cost-effective and less mature than encouraging them to make effective use of their abilities.
Just because someone is sealed off from society doesn't mean society can't benefit from their mind - there's probably plenty of intellectuals and artists behind bars.
Ian Brady is probably one of the craziest crazies to be in Broadmoor, but his book on the way serial killers think, feel and act should certainly be at least browsed by psychiatrists and detectives for insights no rational mind could ever have produced. No matter how little value it really is, the chances are really good that it'll do more good than the British Police's DNA database and CCTV camera system.
I'd rather let a hundred cold-blooded killers live in jail and receive at least some respect as a person if it meant that just one of those hundred produced a masterpiece of art or a book that had significance than have all hundred die purely for the viewing pleasure of Weekend Warriors.
In a hundred years time, which makes the difference? Something that might only rarely advance humanity - but when it does, advance it a lot - or something that provides a momentary mental orgasm for a bunch of f'ed-up "witnesses" and some losers outside and that's it?
I don't see why I should pay taxes for someone getting off on watching another die, when I could be paying taxes to give those in prison a chance to do something positive and worthwhile.
Pinhole cameras work great. Admittedly, they don't use cheap plastic lenses in those, but since I dissed the cheap plastic lenses elsewhere, I've got that covered.
Weren't casio watches considered evidence of involvement in terrorism in Afghanistan? And yours works off nuclear stuff, too! Definitely dodgy.
(On the other hand, that is one seriously cool watch. Probably not one I could afford right now, but with sales of anything being so bad this year, you never know.)
Out-nerded you, then. It was temporarily repaired in Logopolis, although the TARDIS only actually changed shape in one story quite a bit later than that and never changed again.
Come to think of it, it was invisible in The Invasion and therefore might have been any shape at all prior to rematerializing as a police box at the very end.
The Sensorites were able to remove the locking mechanism, resulting in a change to the outward appearance. Thus, even then, the chameleon circuit wasn't so stuck that it couldn't adjust the shape accordingly.
For that matter, although The Doctor has claimed the TARDIS can't change shape, it has not retained the same police box exterior, showing that some chameleon capability exists.
My personal suspicion is that the First Doctor forgot how to change the shape and that later Doctors kept the basic design as a cross between a trademark and a banner.
Heh. First, I'm increasingly skeptical of New Scientist. It went downhill after scrapping the Arianne column. Second, computers can't generate accurate information that isn't there, although they CAN synthesize data values that would be valid based on known information.
Using sound as an example, since you know the general shape of a sine wave, it is perfectly valid to non-linearly interpolate a digital recording to produce an analog waveform that would have produced an identical digital recording. What you cannot do is know if this was the actual original waveform.
If you're really clever, you may even be able to spot relationships between different components of the digital recording. For example, you know that any waveform can be split into simpler waveforms (fourier analysis) and you know that some component waves should relate to specific other component waves (many natural sounds will have harmonics, there may be echoes, and so on). This might allow you to identify some reconstructions as more probable than others. There may be other characteristics and meta-information you can use to refine the process further.
Now, I'll grant the possibility of some very, very good reconstructions from a good image. I'll even grant that a sophisticated enough camera may be able to capture meta-data that could be used to produce a brilliant reconstruction.
What I can't possibly accept, though, is a good reconstruction from a lossy capture at low dynamic range and low resolution and an even lossier image compression into JPEG or other lossy image format. Passable, perhaps, but not good.
Now, if someone could build a camera for a phone that could store images in OpenEXR, -and- have the sensor good enough that using an HDR format made sense, -and- have enough memory on the phone that you could store an image of a few megapixels res, THEN I could accept just about any reconstruction claim the inventor cared to make for the images. Mind you, why would you need to reconstruct anything if the camera is doing all that already?!
If all you measure is the impact of a photon of a very specific energy, the event happens or it doesn't. There's not a whole lot of alternative states.
And, no, there aren't going to be a whole lot of analog effects that interfere with the light-sensitive component. You can't have half an electron jump (it's quantized) and you can't have a fraction of a count (it increments by one or it doesn't). Photon counters are remarkably reliable and very accurate. They're also horribly slow (you accumulate state over a long time) which makes them brilliant for astronomy but useless for even still pictures if the scene has any moving component.
Analog effects WILL affect the analog component, but as I clearly differentiate, you're being an idiot if you confuse the two.
Hard radiation can always mangle solid-state circuits, sure, but I wasn't really considering those to be a typical scenario.
Thermal problems really won't affect a photon counter, as that just alters the movement of atoms. If the camera is exposed to enough heat that the electrons migrate, and it's a point-and-shoot (which basically means plastic lens and cardboard or plastic case), the camera is going to have more problems than a spurious value.
(Remember, we are talking point-and-shoot vs camera phone here, not the Hubble Telescope.)
Reference voltages won't affect the count, because you don't use any. Reference voltages will affect the analog circuits, though, but again those are considered separately.
Quantum effects shouldn't be a problem, since you are not measuring anything (you are only counting) and a pixel isn't on the same order of magnitude as the wavelength.
Remind me, is there anything I'm missing on the list?
Now THAT is a valid argument. If the phone camera does everything you want and need for your general use, then it doesn't matter if a PAS camera is better in any technical sense. I'm still not convinced you really get 5MP, even if the box says you do, but if it gives the results you want then who really cares?
Even DACs are digital devices that produce analog signals, but ANY device based on the photoelectronic effect at a fixed frequency is a digital device (it's either on or off). The analog signal generated is dependent on how many photons are collected in a given interval of time.
God, has Slashdot descended to such depths in only a few decades that people haven't looked up the alternatives to CCDs?
According to Susan Foreman, it's also been an Iconic column and a Sedan chair. In one Peter Davison story, it became a grand piano. So there. Nyah.
I am, and I'll accept that you are, but the vast majority of people have a really suspect intellectual capacity, a very shaky grasp on reality (I hear some people even believe politicians, accountants and/or Scientologists) and an attention-span of a 3 year old. On a good day.
When you consider that Oprah Winfrey is considered to be the height of intellectualism on US television and Coast-to-Coast AM has more credible stories than many of the popular news outlets, it's clear that the species has some serious shortcomings.
On the other hand, Neolithic people had perfectly functional and valuable neurosurgeons, had tools with better-than-millimetre precision, were capable of large-scale transatlantic sea voyages and were building some very good echo chambers.
So when you compare modern humanity (by which you realize you're including Palin supporters, rednecks, New York taxi drivers, Frank Bruno, Hulk Hogan and Terry Wogan) with Neolithic and Paleolithic people, I'm really not convinced humans are advanced as you think.
A decent analog watch has superbly-crafted components and therefore deserves the admiration. A decent digital watch needs something marginally better than a 555 timer, but only marginally.
That's not to say I wouldn't gave respect for a digital watch with the accuracy and drift of a scientific high-precision timer. I'd consider it adequately impressive for a digital watch to have a drift of a few nanoseconds a day.
(I'm judging analog differently than digital in the resolution, but really about the same in terms of the effort required to produce a system of a given standard because that's the only measure of craftsmanship you can ever really have.)
What you really want is a James Bond wristwatch with 0.5T magnetic field generator whenever you turn the dial. Saves on having to spend 5 seconds looking at the bloody time when the server room explodes and the users' machines go into meltdown.
Maybe it's the Asperger's in me, but after getting years of verbal grief for being different, I have no problem with neurotypicals begging. Actually, I don't think that is the Asperger's. I think I might just have become more cynical and more anti-social in my old age. (And gerroff the lawn!)
Well, there's that, but also bear in mind that cameras can afford to put a bit more power into the electronics, so that JPEG compression can be of higher quality.
Doubling the number of pixels on the CCD but more than halving the amount of retrievable data stored will give you a net loss of quality. High-res CCDs are relatively cheap and since the phones don't advertise the resolution of the image as stored, it's a great marketing ploy.
It's next to impossible. Phones need to be very small, lightweight and damage-resistant, the electronics need to be exceedingly low-power and the electronics for the camera and the electronics for the radio transceiver can't conflict.
That last requirement means is you use digital devices that produce analogue signals, the resolution on the ADC has to be so crappy that the RFI from the radio doesn't screw up the picture AND the voltage changes when a call is picked up or an alarm goes off or what have you can't throw the ADC.
The low-power means no fancy, power-hungry logic, the software zoom and other floating-point logic won't be terribly high precision, and the image compression algorithm will need to be light on the quality.
The size and damage-resistance impacts what sort of lens you can use, how rigid the structure has to be, how much the user can just seriously screw up the device before the image quality drops. Even for a disposable standalone camera, it's practical to put in some quite acceptable optics.
Even when such devices are of a size comparable to that OF the phone, you've got to remember that the camera is sans radio (or radios, for phones that have bluetooth and/or wifi and/or AM/FM tuners as well as the standard phone radio), sans keyboard, sans quite a bit of space-hungry stuff that phones either need or have as "features".
Talking of sci-fi, phone booths can't become obsolete! They're needed for "The Duelling Machine" and also for The Doctor's TARDIS! I'm pretty sure they'll be needed when the Triffids attack, too, and they could be handy when "A for Andromeda" happens.
Inherent in the idea that Firefox prohibit malicious code breaking into an extension and causing it to reformat /dev/hda1 is the idea that the extension itself cannot reformat /dev/hda1, as it is impossible for Firefox to know what will or will not cause an extension to do something maliciously.
Indeed, if an extension can do anything, it is possible to write an extension that allows an external program to control your computer with the privileges of Firefox AND (this is the important part) it is ALSO possible for a buggy extension to give such control of your computer with those same privileges.
If an extension can do absolutely anything, without any restriction, so can malicious code. The compromise suggestion was to have privileged operations in a distinct process. An extension overall can then still do anything but a given extension component cannot. (You didn't comment on that - not see it or you prefer to find something bad to say?)
Compartmentalizing extensions would, yes, make life much harder for extension writers. It would force them to be disciplined, code properly and not incorporate stupid security flaws. This would eliminate a lot of stupid coders, but I'm having a hard time seeing that as a bad thing.
If that was unacceptable, if you had to have the bad coders, then is sandboxing really so terrible? I honestly can't think of too many things Java application writers can't do, but Java is sandboxed. Forth programmers don't seem to have too many problems either - it's a very popular language for developing very low-level code like BIOSes - and yet also runs entirely in a compartmentalized virtual machine.
Hell, one could argue that any Trusted OS (ie: an OS that runs on top of a very thin security OS that provides all the operations that have security implications) is essentially an OS running in a sandbox. It might be harder to run the latest *ix games or applications under Trusted Solaris than it is under Linux, but unless you can name something you CANNOT run AT ALL, I'm inclined to believe that the limitation you speak of simply does not exist.
So if Firefox sandboxed extensions then it might need to provide some extra functionality via extensions to the existing API. Doesn't sound too horrible and it's certainly not fatal to developers.
So we definitely have multiple ways of improving security without preventing extension writers from doing what they want -- the only thing improving security would impose is HOW extension writers did things. Again, is that a bad thing?
I would rather see the death of bad code than see the death of Firefox because it got a reputation for being worse than IE on security. Particularly if the reputation was not due to Firefox per se but because extension writers were drunk or lobotomized at the time.
There's really no excuse for Firefox to allow at least some of the more common security flaws - or at least allowing those flaws to cause problems.
First, sandboxing of extensions should limit what problems can be caused.
Second, a lot of errors are caused by the overflowing of buffers - a problem that could be limited by the use of stretchy buffers or bounds-checking malloc implementations. Or not allowing direct access to the heap.
Third, Firefox (and indeed all programs) should run on the principle of least privilege. Where some specific subset of program functionality requires significantly greater privilege than the rest, run the subset as a different thread or process at a different level of privilege. By extension (bad pun, I know), extensions could also be run as a different thread or process with even fewer rights. (OS' that don't allow programs to shed rights might be a problem, though.)
Oh, it's practically a given that the non-linearity won't be the same at all scales. But complex behaviour can be produced by very simple non-linear systems - the Mandelbrot Set being the best-known example, so the presence of non-linearity merely creates a problem of practical computability rather than a problem of mathematical computability.
(Remember, to be computable in the mathematical sense, the algorithm has to complete in finite time. Which can mean anywhere from a few picoseconds to an hour after the heat-death of the Universe. To be practical, though, the model must produce results within the time the results are useful. To be commercially practical, it also has to produce results faster than other methods of getting those results.)
So we're looking at nested non-linear systems, no matter what starting point we're using.
Let's start with a bottom-up approach.
In the biological world, each cell has multiple mechanisms running in parallel where each mechanism is non-linear. The cell itself is a non-linear construct of these. There are different types of interconnect and these are also non-linear, so any network of cells is a non-linear construction of non-linear components. The brain has topological constraints, but unless there's grounds for believing those constraints to fundamentally alter the maths, the maths should be independent of implementation details.
This says we're looking at a nesting 3 deep. So we're looking at a chaotic system in which potentially all of the parameters are themselves chaotic systems in which potentially all of the parameters of that are also chaotic systems.
What else do we know? We know that the lowest-level systems are fundamentally unchanged from how they were 3.5 billion years ago when cellular life first arose. They may be chaotic but the building-blocks are all very simple. The only real internal changes have been in the organization of the building-blocks. All other changes within cells deal with interactions and mathematically interactions are on a different level.
Most of those lowest-level systems are common to heart cells, skin cells and brain cells. Now, this will include communication mechanisms and those we DO have to consider. Basic housekeeping that is a product only of it being biological can be ignored. Systems specifically activated in neurons and NOT common across all cells also have to be considered, even if housekeeping, as state is persistent in neurons by means of such housekeeping.
Now, the mechanics of these functions aren't what's important. What's important is what they do to the logic of a neuron to make it capable of data processing.
The cell itself is a network of these. In standard computer network terms, you're looking at the equivalent of a multicast-capable routing-capable ad-hoc network of moderate size. This is just for a single neuron, we're not even up to networking these things. Actually, strictly speaking, it's multiple such networks. In biological cells, you've independent chemical and electrical paths. Different latencies and different bandwidths.
Unless there is firm evidence that this is an implementation detail that does not alter the specification, I believe that it is wisest to assume it DOES alter the specification, that signal delays and other signal characteristics are important. Some variables from iteration X of the system are fed into iteration X+1, but others are fed into iteration X+N (where N can't be guaranteed to be a constant). This is what makes it a chaotic system of chaotic systems rather than merely a bigger chaotic system.
Now, the network of cells is basically more multi-path networking where again different types of interconnect have different properties. Further, not only are the nodes in the network effectively mobile and multicast, but the number of nodes is variable.
(We can ignore the number of connections a given neuron has by looking at the superset of functions exhibited by all types of cell in the brain, whether neuron, axion, or wha
Ok, I would agree with all that. (I can't mod you informative for obvious and non-chaotic reasons.)
People doing useful and interesting research frequently post on Slashdot, so I don't see what your problem is. It doesn't take a genius to mathematically model a brain and that isn't something people have bothered much with doing.
Some things people have tried to do are build models of compartments of the brain (bad idea), simulations of some poorly-specified upper-level functions of the brain (even worse idea) and discrete/binary simulations of individual neurons assuming them to be stateless and/or with a rigid topology (talk about dense).
The first is like trying to build a model of one part of a Mandelbrot set. A complete waste of time, since the maths doesn't work that way. The second is stupid because without a good specification, there's nothing meaningful to simulate. And since neurons are neither discrete, stateless nor in a fixed network (even adult brains have a surprisingly dynamic topology), all you get is a simulation of something that never existed instead of a simulation of the thing you want.
Why do people do these things? Because they're very doable. Neural networks are a doddle to code up, logic chains and decision trees are trivial on a computer, and since more people are interested in medical applications than AI, understanding compartments is far more practical than understanding the brain itself.
In short, people want to be paid far more than they want to discover, especially since discovering the mathematics of the brain won't do you any good as it'll be well outside the capacity of any machine out there (including the 100 million core one) to do anything sensible with such a model. Nobody likes inventing things that can't be used for another 50-100 years.
However, the fact that nobody WANTS a real mathematical model of the brain doesn't change the fact that the brain is an extremely simple device (mathematically-speaking). The unwritten part of the challenge is that they want a mathematical representation they can use and it is that which does not exist and will not exist for at least the next 50 years, simply because of the technology. The maths is a non-issue.
As far as Navier-Stokes is concerned, there are no reasonable assumptions. Particles do not move with a uniform speed, speed follows a bell curve. Well, almost, as a bell curve has infinite tails in both directions but in physics you're bounded. Particles are strictly between 0 and C and cannot take on either value or anything outside of the range.
In practice, since you don't see too many Bose-Einstein Condensates or even hypersonic particles when boiling water for an egg. However, even in a pan of cold water, there'll be water molecules moving fast enough to leave the liquid, and even when the water is boiling, there'll be water molecules that have the kinetic energy of a slug. Not many, but there will be some.
That's your first problem, because the first simplification is to decide what sort of range of speeds particles are likely to move at. The reality is "all of them, at some point or another".
The second problem is this differentiation between compressible and non-compressible fluids. In the same way that speed is non-uniform, density is also non-uniform. That means all fluids will have a mix of the two characteristics.
The third problem, as I've already pointed out, is that the system is chaotic. This means you need an infinitely fine grid and an infinitesimal time interval between iterations. Neither of these is possible. However, chaotic systems don't necessarily improve as you improve resolution, which is why CFD is often far more coarse-grained than you might expect. It has nothing to do with context, or even compute power, it has to do with experimentally finding a resolution where the results are similar (through the property of self-similarity) to what you might get if you could work at infinite resolution.
Self-similarity is NOT the same as identical, though, which is why most competent hardware engineers treat CFD as being a first approximation at
Ideal circles do not exist, that is true. So what? The idea that you need an ideal form is Platonic (it comes from Plato's cave analogy). Does there need to be some ideal, in order for an approximation to exist? (Well, C++ and Smalltalk programmers can skip that question.)
Let's try a different example. Let's go for the Second Law of Thermodynamics. Statistically speaking, it's universally true. There are no exceptions on the macro scale of space/time. If you were to examine a small patch of quantum foam over a few picoseconds, it would be lousy even as an approximation.
Does this mean the Second Law is wrong? No, not really. Does this mean the Second Law is artificial, as it's only an approximation? No, I think you'll find the early universe obeyed it long before there were any observers.
So what does it mean? It really doesn't mean very much at all. It means you're asking the wrong questions and not getting useful answers.
As for the quotes, Albert Einstein was very bad at maths and I think Benoit Mandelbrot (amongst a few thousand other chaos and fractal specialists) would beg to differ on the ineffectiveness of mathematics in biology. It's hardly the mathematics fault that biologists are lousy mathematicians.
My argument is that it is quite immaterial as to whether Pi is universal or not. In any given specific space, there will be -a- constant that denotes the ratio between the circumference and the diameter. The fact that there exists a constant for a given space (whether or not there exists the same constant for all spaces) means that the property of the ratio is fundamental.
Iff* the same constant holds for all spaces, then Pi as we know it is -also- fundamental, but I am unsure this has been proven. My statement that we can split off what is artificial from what is fundamental is unaffected.
*Maths notation: If and only if
The only way it can be proven that mathematics is wholly artificial is to prove that the set of all mathematical "things" that are fundamental is equal to the empty set. ie: there is nothing - not a single property, not a single result - that is true everywhere, including Goedel's Theorum. If even something as simple as Goedel's Theorum is universal, then there exists at least one part of mathematics that is not invented but is wholly natural.
Now, here we run into a problem. If Goedel's Theorum is not a universal result, but an artifice, then it is also false because it would have to be possible to create a counter-example and the theory states no counter-example of this kind can exist.
Surely that seals the argument right there and then. Those who argue mathematics is wholly artificial must be arguing Goedel's Theorum is false. All other cases do not prohibit the theorum from being true. Thus, if there is sound reason for believing the theorum true, there is sound reason for excluding the notion that mathematics is an artifice.
No, the bicycle is equivalent to a number base or a mathematical system. It is an implementation OF an underlying system (in this case, Newton's Laws), but Newton's Laws would still remain exactly the same whether Newton - or indeed bicycles - had ever existed.
The definition is also immaterial, as that too is an implementation detail. The underlying principle would remain unaltered whether the definitions of circumference, diameter or pi had ever been developed.
You are confusing the overlaid system with what it overlays. I'm saying you don't need to. Your argument is that the overlaid system is artificial, an invented product. I'm saying you're entirely correct on that. But what I am also saying is that what the product overlays, what is beneath the terms, the dynamics and the fancy Greek lettering is not artificial but exists whether it is known to exist or not.
The problem with assuming the two layers are the same is that you run into the Anthropomorphic Principle - the universe is the way it is because it produced people capable of seeing it. Let us, for a moment, assume the Many Worlds theory of Quantum Mechanics is correct. Then there are universes OTHER than the one we see and the theory falls down. The same would be true if the model of a multiverse as a foam (where each universe is a bubble in that foam) is correct.
But if you're on this site, you should be familiar with layering anyway. Maths - the fundamental, overarching thing that is shown in all mathematical systems that exist, will exist or ever have existed - is a Layer 1 concept in the OSI model. Concepts like numbers and other fundamental but artificial building blocks are Layer 2, which makes Group Theory a layer 2 switch. Anything and everything that MUST be true because of something in layer 2 is arguably also layer 2, which would include Goedel's Theorum. Anything that is true only in a specific implementation of mathematics is layer 3 or above.
Does using an OSI representation make it easier to see how not all maths is the same?