It's not entirely new. I'm not in the field so I can't say anything about Sargent's merits or standing, but I have met others working on essentially the same thing, although with different approaches.
The basic idea is rather simple. In a conventional solar-cell you have an energy gap (band gap) between two layers. This energy gap is about equivalent in energy to that of sunlight. So when sunlight hits and gets absorbed, the electrons get enough energy to be kicked over the gap, and you have an electrical current.
The problem here is that the gap energy is roughly equivalent to that blue light-UV. (high in energy) So only the blue and UV light in sunlight is being converted to electricity. So you are missing a lot of solar energy in the red and infrared.
The way to fix this is to change the compound, adding another energy level in the middle of the energy-gap, so that the energy required to get from the lower level to the middle level and from the middle to the top correspond roughly to red light. That way, a red light photon can kick the electron from the bottom level to the middle one, and so a second one can then kick it up to the top. (adding a rung to the ladder)
I assume this is what Sargent's 'film' does. Pretty cool.
If your bicept, for example, were to be attached ten times closer to the pivit point of your elbow, your muscle would need to exert ten times the force in order to do the same work.
First off, it's biceps. And the biceps acts in the opposite direction, the fulcrum in that case would be the shoulder, not the elbow. And what you are describing is a fucking arm not an individual muscle. Muscles contract lengthwise. That's all they do. Go get an anatomy textbook. You need one.
How is your reading comprehension? Was this article (or quote thereof) referencing an 'artificial arm'? No. It was talking about an artificial muscle. A fiber which can contract lengthwise, exerting a force, just like a muscle.
So they, in the article are comparing the force and contraction by some strands of this material, and comparing that to the the force and contraction exerted by a similar amount (by weight or thickness, it doesn't reveal) of natural muscle.
Perhaps you should read up on some basic mechanics?
This has nothing to do with leverage. Leverage is when you have a lever, or rather, torque. It is not a general 'force to distance'-converter.
And there's nothing in the article talking about torque. They're talking about longitudal force versus longitudal contraction. No torque involved, and no leverage.
Not to say that programming can't be viewed as an art; It can. But it is also engineering, which imposes more limits on what qualifies as 'good' programming. (In the same way, architecture is more limited than sculpture, because architecture is about creating buildings, which by definition need to have an inside and an outside and a door, at the very least. Whereas a sculpture has no such constraints.)
So what does this matter for open source? Well, in open source development, anybody can (try to) contribute anything. So obviously you need some kind of 'artistic vision' in order to figure out what parts go in and what parts don't. The thing here is that programmers have a far more akin 'vision' on what constitutes 'good art' (or rather a 'good program') than the rest. Does the code run faster? Have less bugs? Is it easier to read, understand and maintain? And such criteria.
With art (music, images, etc) this isn't as easy. Which makes collaboration very difficult. (Imagine ten people creating a collage, each with a slightly different idea of what the completed collage should represent. It wouldn't work.)
So, this means that to get good art, you have to break up the total task into rather large pieces. (Say, 3d-models, interface graphics, intro/cutscene graphics, etc.) Because it simply wouldn't work if 100 people created the 1000 required 3d-models for a game, since the differences in style and form would make the whole so inconsistent in its 'feel' that it'd bring down the impression.
So, you have these rather large, daunting tasks which more-or-less require being done by one or two people ('auteurs' I guess you could say).
And for a game of any size on the scale of today's commercial games, it's a lot of work for so few people, unless they're working full-time. And that of course requires resources and money.
So what I'm saying is basically that the OSS development model doesn't quite work, because the investment money isn't there to have artists working full-time, and you can't really get the thing started by having a large group of part-timers.
Actually, what about conservation of momentum, conservation of energy, etc?
What about it? It is all conserved. I didn't write anything which implies it isn't.
It's just not physically possible for electrons to be sent at such tremendous speeds.
In a conductor, the electrons are not 'moving' in any classical sense. It's an oversimplification. One way to look at it is that when one electron leaves the wire, another jumps over and takes its place, and another takes that ones' place and so on all the way to the other end. (But that's an oversimplification as well, since they are indistinguishable.)
The only plausible scientific explanation is that photons are what is carrying the electricity, not electrons.
Photons cannot carry an electrical current. They have no charge. They do, however, mediate the electromagnetic force, which is the force between charged particles (the electrical field).
But they don't 'carry the electricity'. But you could say that they are what 'pushes' the electricity through the wire.
Electricity is electromagnetic radiation, i.e. light.
No, electricity is the transfer of electric charge, you wrote that yourself. Electrical potential (e.g. voltage) is mediated through photon exchange, and is light.
But just because the potential is felt at the speed of light does not mean that the current does. (which strives to equalize this potential and reach thermodynamic equillibrium)
Quantum tunneling has nothing to do with electrons moving faster than the speed of light.
Really, why not? Are you saying that electrons do move faster than light, but not through tunnelling? Or what?
The electrons themselves move very very slowly compared to the speed of light.
Speaking of 'the electrons themselves' is meaningless, since in a conductor they are delocalized and have no individual identities. (which they don't really have otherwise either)
Speaking of the speed of a current is meaningful, speaking of the speed of the electrons is not. You can measure the current speed and divide that by the charge transferred and arrive at a number for the 'speed of an electron', but it's not a very meaningful number.
Re:GaAs??? GaAs is material of the future...
on
Where's My 10 Ghz PC?
·
· Score: 2, Insightful
superconductors is the way to go for highest speeds/most concentrated processing power, due to extremely small power dissipation and extremely high clock frequencies (60 GHz for logic is relatively easy right now), but the problem is that after someone invests $3B in a modern semiconductor fab they do NOT want to build a $30M top-of the line superconductor fab to compete with it.
I'd think the more likely reasons would have to do, for starters, with consumers not wanting or being able to afford a computer that requires constant cooling with liquid nitrogen (or even worse, liquid helium) to work.
Rather offtopic, but.. That's a meaningless thing to talk about.
The valence electrons (ones carrying current) in a metal are delocalized. What that means is that they don't really have any 'identity'. They don't belong to any particular atom, and you can't locate them all individually. You just have a 'fermi sea' as it's called, a kind of electron cloud.
Now, the analogy to a water pipe isn't bad. But the key difference here is that with a water pipe, you can identify the water going in and coming out (say, add some colorant to the water going in and see when it comes out).
However, there is no way to 'label' the electrons going in and coming out of a section of wire. Not because we don't know any, but because it's theoretically impossible. If it was possible even in theory to distinguish the particles, then the laws of physics wouldn't allow them to be delocalized - and then the current wouldn't flow.
The algorithms chosen *are* important, and in some cases you shouldn't simply reach into the API toolbox and use the third-party solutions. There is no substitute for knowing how to write your own sort routines, specialized linked lists, and binary trees.
However, this leads to people going too far in that direction as well, which is just as problematic. People re-invent the wheel for no better reason than that their own inflated egos make them believe they can do everything better. Or because the available routines differ in some completely insignificant respect from what they'd prefer.
The result: Lots of programs duplicating effort, creating bloat. You have Gnome, KDE, Mozilla and OpenOffice all with their own set of widgets and APIs. It's fragmentation and duplication of effort for no good reason, and so the casual Linux (in this case) user ends up having to store all four of them in memory.
I think that the Not Invented Here syndrome is a far more important factor in creating 'bloat' than people chosing inefficient algorithms.
Well, Java 2D has been part of the Standard Edition since version 1.2, so I'm not sure what you're on about there.
As for the rest.. The strategy is perfectly clear. Sun makes a bunch of java add-ons in the com.sun.* heirarchy and releases them. If they catch on, they eventually get integrated into Standard Edition.
They used some technique of using a diamond point of action, and layering carbon onto it, like a chip fab does. It seemed like they could produce fairly large quantities at minimal cost. I'm not sure if they have been shot by deBeers yet, tho.
Chemical Vapor Deposition. You basically 'zap' the carbon, forming a plasma cloud of charged carbon atoms. Then you let that cloud cool and fall onto the diamond substrate. If you do it 'just right' the carbon atoms in the cloud will settle into the correct positions on the substrate to continue the crystal lattice, and you'll have added a layer of atoms to the crystal. Repeat the process and you get a flawless, single-crystal diamond.
I think the Wierd article (as always) seriously over-hyped the issue though. This is something which is coming no matter what, and DeBeers won't be able to stop it. There are hundreds of researchers all around the world working on CVD diamond research. Simply put, diamond is too useful a compound to be artificially hoarded for jewellery.
Good news for the rest of us, bad news for DeBeers.
How do you get from "works" to "whatever makes me happy"?
Define what was meant by 'works' in a clearer and unambigous manner then.
Historical truths are much more mutable than those of the physical universe. It may be frustrating, but it's true.
Bull. History is part of the physical universe. At a specific place at a specific point in time, an event happened which some humans found noteworthy. That is history.
If it wasn't, how/why could history be so contested?
History cannot be contested. History is a real physical event. You are talking about accounts of history. And these accounts of history can correspond more or less well with what actually happened. That would be what we call the 'truthfullness' of an account.
"Correspondance to Reality" is one of those things that people believe in that they can't prove.
That's another bullshit argument. Nothing can be proved. Nothing you wrote can be proved either. You must assume.
The most fundamental assumption that exists is that there indeed exists an objective physical universe around us. If you deny that, then there is nothing left at all on which you can base any kind of philosophy of anything. You are also being dishonest, because that assumption underlies every single act of human behaviour.
Then there is a thing called language. Language has no meaning in itself. The entire meaning of language is in its correspondence to this objective, physical world.
Put those two together and you have the definition of truth as the correspondence to reality. No, it cannot be 'proven' in any sense, but it is as fundamental an assumption as can be made. Because you cannot even start to make an assumption without first assuming language has meaning and that reality exists. (i.e. that you are saying something to someone)
I think what you are thinking about is that sometime the correspondence between reality and a statement (its 'truthfulness') cannot be ascertained.
To that I say 'So what?'. Human knowledge is never better than an approximation of the 'absolute truth', and I wouldn't claim it was. The ambiguities of language and limitations of the human mind make anything like that impossible anyway.
To continue on the Holocaust example: Destroy all records of it. Kill anyone with knowledge of it. It would still have happened, even if the entire human race stops thinking it did. Unfortunately that's a 'truth' which would then never be known.
But noone claimed absolute knowledge was attainable either.
On the other hand: sacrificing a goat to the spirits is a truth that may help you survive the famine
Is not true. It does not correspond to physical reality - you would be better off eating the goat instead.
No.. The main thing is not about marketing. It's about management decisions vs. engineering decisions.
Say you have Microsoft language A vs. Open Source language B. Now, if B is better from the technical standpoint than A, then B is the better language from the engineers standpoint.
To management, though, that is only one factor in the equation. They have other concerns:
Can we find staff trained in this language?
Will we be able to in five years?
Does the language have support?
Will it have support in five years?
Will the language continue to be developed to meet our needs in the future as well?
These are all valid concerns which need to be adressed for any operating system/language/platform to catch on in the corporate world. Don't be so naive as to think that all management types are PHBs who buy any marketing that comes their way.
In other cultures, knowing why the ride to work drives you crazy is a truth that helps you stay sane.
Truth is any tool that works better.
That's the pragmatist* definition.
And it's really a pretty crappy one, IMHO. Truth, as most people use the word, is a correspondence to reality. Not "Whatever makes you happy".
For instance, we'd all be happier if the Holocaust never had occured. By that definition, it certainly is something which 'Works better' for most people. But it is not the truth in the way I, or most people, use the word. Because the Holocaust did indeed occur.
It's bunk.
(* Philosophical pragmatism.. William James for instance.)
Who determines who the "experts" and "authorities" are? It can't be these same people, that would just beg the question. Or perhaps its the social structures already created that mold and promote expertise.
What most of us call "facts" are the very things which are independent of social structures, cultural contexts and other biases. As such, anyone is in the position to determine who is an expert and who is not. An expert is a person who has a more detailed knowledge of facts than someone else.
Scientific expertise is not some secret society of self-prepetuating illuminati. That's what sets Science apart from its predecessors Alchemy and Magic. It may be a meritocracy to an extent, but it is an open system. You don't need a degree to practice science.
It may be difficult in practice, but there are still plenty of people making contributions without any academic background. And even more academics making contributions within fields in which they have no formal schooling. (The physicist Richard Feynman for instance, helped decypher Maya hieroglyphs)
The Open Source movement is very similar in that respect. Anyone can contribute, but naturally a major contributor will have an easier time getting his patches into the Linux kernel than someone who never contributed anything before.
Wikipedia is different. It is far more difficult for the readers/reviewers to determine who is right and who is wrong, especially in highly specialized areas.
Instead you risk ending up with the very thing you are accusing peer-review of. You create a social structure where the 'facts' are determined by the authority the person holds within the Wikipedia community rather than a consensus on what actually matches reality best.
The worlds largest producer and consumer of films is India. I thought that was well-known? ('Bollywood' doesn't export much though.) Japan and China have pretty big movie industries too, but don't export much either. The USA is undisputably the largest exporter.
But just about all non-third-world countries have a domestic film industry.
The size of it is roughly proportional to the size of the country. E.g. Unsurprizingly, Germany makes more movies than the Netherlands.
Obviously the size of the budget is proportional to the size of the market. So most countries don't produce big-budget blockbuster-type films. It's too expensive, and they can't really compete with hollywood anyway.
So the natural adaption is to produce 'the other stuff' e.g. comedies, dramas, thrillers, and so on. Some of them are very good.
That said, the USA is a very lousy importer of foreign films. While a successful German film will likely get picked up by cinema chains in most of Europe, it won't get any showing in the USA. The exception here is basically if it gets nominated for a 'best foreign film' Oscar, in which case you might just catch it in an arthouse cinema if you're lucky enough to live in a town with one.
If you go to film festivals and so on, you realize a pretty depressing thing. There are very many really good and original movies are being made in small countries which almost noone in the rest of the world gets to see because these movie execs have the idée fixe that people simply won't watch anything that's subtitled. (I think this is why other countries are better than the USA at importing films. Since they're already importing stuff from the USA, subtitling isn't much of an issue.)
Quite true. I just installed XP for the first time myself last weekend, on a new machine. The last Windows I'd used before migrating completely to Linux on my home computer was Win95.
I was surprized by the lack of drivers! Win95/98 included a pretty big set of drivers for the most common hardware. Not always the best drivers, but drivers nonetheless.
Win XP included nothing. Not even for the integrated AC97 sound card the machine had. The AC97 is _incredibly_ common. We're talking about millions of machines here. Same went for the networking. I guess you could point to the built in internet-search, but it's of course not a good answer if you're networking doesn't work. (This was of course fixed by an included CD with drivers.)
So really, Windows has regressed with respect to drivers. Windows doesn't really include any real drivers at all anymore, it seems.
it's difficult to draw any direct causality between its placement and the timing of this article
Well, that's doesn't seem true, considering that the article itself starts off by mentioning the advertisment.
But really.. Why would that mean the article is biased? People around here are always knee-jerking about bias and FUD, but they seldom explain wherein their supposed 'bias' lies.
Isn't Microsoft given opportunity to state their side of the issue in the article?
FACT: IE has several critical security holes. FACT: Some of them are fixed by SP2. FACT: SP2 isn't available for people running older windows versions. FACT: Millions and millions of people are running older windows versions.
It's a very relevant question. And I think most people would agree Microsofts suggestion that people should get a new computer is rather arrogant.
If there's any bias in the article, it's not pro-Firefox as much as it is pro-consumer.
If MS was to do the right thing by their customers, they wouldn't be getting this kind of press.
Oh please. I didn't say "Windows is backwards compatible with every single app".
I said that Microsoft tries hard to keep backwards-compatibility.
But don't take my word for it, I don't work for them. Read Raymond Chen's various blogarticles on the subject. He is one of the poor souls at MS who worked his butt off to try and keep backwards-compatibility.
I'm impressed by what they've done so far, but not the seven years it took them to do it.
Well, this is a bit symptomatic of a lot of OSS projects, they start out with a grand vision and end up planning and then re-planning and throwing out code and never really get off the ground. Some die and stay dead, some get picked up by a group of enthusiasts with a more down-to-earth approach of 'Getting something working now, improvements later.' and the project takes off. (Case study: See Linux vs. GNU Hurd)
I believe this is pretty much what happened with ReactOS (I'm not a ReactOS developer), so I wouldn't hold it against the current crowd too much.
I suspect it will have the same trouble WINE runs into: it's chasing a moving target, and it's way behind.
Ah, the old catch-up argument. It's a valid argument, but it's not as important for API:s as it is for, say, file formats. With the MS Word file format, Microsoft can tweak and alter that all they want, because it's not publicly documented, and they're not that interested in having compatibility with anything other than MS own products. Backwards compatibility isn't important. Heck, they're happy to break it and create incentive for people to buy the new versions. That's a hard act to follow.
With API:s, things are quite different though. Firstly, the '80-20 rule of features' pretty much applies. Most programs don't use the entire API, but a rather small subset.
Secondly, API:s rarely break backwards compatibility. That would break all existing third-party apps and make it difficult for people to migrate. The exact opposite situation to the previous one. So MS bends over backwards to make stuff backwards compatible. Windows 3.0 apps still run on XP.
The APIs are also (relatively) well documented. Sure, there's a lot of undocumented functionality, but most of that is also unused. The implementer has access to the same information as most application-developers.
Another point of difference is that you don't have to be super-fast in implementing new API:s. The day a new Word version hits the street, people will be asking 'Why doesn't this work with OpenOffice?'.
Not as true for APIs. While we all like the latest and greatest, professional developers don't rewrite their programs to use the latest APIs 'just for the heck of it'. There has to be good reason. In fact, you want to avoid using the latest APIs as far as practical and economical, because otherwise, you're going to be shutting-out potential customers running the old OS version. (There are plenty of brand-new apps released today which run on Win98, or Win95 even.)
I think you're selling classical physics short. Newton's laws are hardly just empirical rules of thumb. I think there's hardly anything as beautiful in its simplicity as the principle of least action, and classical mechanics is embodied in it.
Kepler's laws fit your description fairly well, but Newton's laws were and still are rather profound in their scope.
You're absolutely right about this. Newtonian mechanics as a whole is something more than an ad-hoc set of equations. In general, Newton did more than just curve-fitting. He did work from a basic set of postulates when it came to mechanics.
This is why I specifically referred to Newton's Law of Gravity. His gravity law is an empirical rule of thumb. In my opinion.
He couldn't really justify the way he defined the center of gravity, or why the force had to be inverse-square.
But all of his mechanics are not purely empirical, no.
The difference here lies in the fact that those laws are based on physical properties of the universe!
So how do you reconcile this with your previous definition of 'law' being 'the highest designation we [the science and engineering community] can give phenomena' and also mentioning evolution, which is hardly a 'physical property of the universe' by any account.
Moore's law is dependent on humans to develop technology. We could, in theory, stop advancing technology thus stopping Moore's law. We can't, however, prevent gas from being related inversely between pressure and volume.
Moore's law is an approximation of the speed of computer development, made at a specific point in time. Noone, and certainly not Moore himself, would ever claim it would be valid forever.
We can certainly prevent gas from being related inversely between pressure and volume. Try increasing the pressure. The ideal gas law is an approximation, which is actually only valid as the limit when the pressure drops to zero. If you increase the pressure enough, the gas will eventually liquify, something which Boyle's law of gases doesn't account for at all.
Again, a 'law' is an ad-hoc empirical description. All these physical laws have certain ranges of validity.
For the gas law, it's when the pressure approaches zero. Ohm's law only applies to Ohmic materials. Hooke's law only applies to small deformations (and is rather crude despite that), and so on.
Moore's law is only applicable to the late 20th-early 21st century. So?
Slashdot Mirror
It's not entirely new. I'm not in the field so I can't say anything about Sargent's merits or standing, but I have met others working on essentially the same thing, although with different approaches.
The basic idea is rather simple. In a conventional solar-cell you have an energy gap (band gap) between two layers. This energy gap is about equivalent in energy to that of sunlight. So when sunlight hits and gets absorbed, the electrons get enough energy to be kicked over the gap, and you have an electrical current.
The problem here is that the gap energy is roughly equivalent to that blue light-UV. (high in energy) So only the blue and UV light in sunlight is being converted to electricity. So you are missing a lot of solar energy in the red and infrared.
The way to fix this is to change the compound, adding another energy level in the middle of the energy-gap, so that the energy required to get from the lower level to the middle level and from the middle to the top correspond roughly to red light. That way, a red light photon can kick the electron from the bottom level to the middle one, and so a second one can then kick it up to the top. (adding a rung to the ladder)
I assume this is what Sargent's 'film' does. Pretty cool.
Biceps move the lower arm, not the upper.
Yes, it does. And when it does that, where is that force acting from, in mechanical terms? The shoulder.
If your bicept, for example, were to be attached ten times closer to the pivit point of your elbow, your muscle would need to exert ten times the force in order to do the same work.
First off, it's biceps. And the biceps acts in the opposite direction, the fulcrum in that case would be the shoulder, not the elbow. And what you are describing is a fucking arm not an individual muscle. Muscles contract lengthwise. That's all they do. Go get an anatomy textbook. You need one.
How is your reading comprehension? Was this article (or quote thereof) referencing an 'artificial arm'? No. It was talking about an artificial muscle. A fiber which can contract lengthwise, exerting a force, just like a muscle.
So they, in the article are comparing the force and contraction by some strands of this material, and comparing that to the the force and contraction exerted by a similar amount (by weight or thickness, it doesn't reveal) of natural muscle.
Perhaps you should read up on some basic mechanics?
This has nothing to do with leverage. Leverage is when you have a lever, or rather, torque. It is not a general 'force to distance'-converter.
And there's nothing in the article talking about torque. They're talking about longitudal force versus longitudal contraction. No torque involved, and no leverage.
Not to say that programming can't be viewed as an art; It can. But it is also engineering, which imposes more limits on what qualifies as 'good' programming. (In the same way, architecture is more limited than sculpture, because architecture is about creating buildings, which by definition need to have an inside and an outside and a door, at the very least. Whereas a sculpture has no such constraints.)
So what does this matter for open source? Well, in open source development, anybody can (try to) contribute anything. So obviously you need some kind of 'artistic vision' in order to figure out what parts go in and what parts don't. The thing here is that programmers have a far more akin 'vision' on what constitutes 'good art' (or rather a 'good program') than the rest. Does the code run faster? Have less bugs? Is it easier to read, understand and maintain? And such criteria.
With art (music, images, etc) this isn't as easy. Which makes collaboration very difficult.
(Imagine ten people creating a collage, each with a slightly different idea of what the completed collage should represent. It wouldn't work.)
So, this means that to get good art, you have to break up the total task into rather large pieces. (Say, 3d-models, interface graphics, intro/cutscene graphics, etc.) Because it simply wouldn't work if 100 people created the 1000 required 3d-models for a game, since the differences in style and form would make the whole so inconsistent in its 'feel' that it'd bring down the impression.
So, you have these rather large, daunting tasks which more-or-less require being done by one or two people ('auteurs' I guess you could say).
And for a game of any size on the scale of today's commercial games, it's a lot of work for so few people, unless they're working full-time. And that of course requires resources and money.
So what I'm saying is basically that the OSS development model doesn't quite work, because the investment money isn't there to have artists working full-time, and you can't really get the thing started by having a large group of part-timers.
That's my take on it, anyway.
Actually, what about conservation of momentum, conservation of energy, etc?
What about it? It is all conserved. I didn't write anything which implies it isn't.
It's just not physically possible for electrons to be sent at such tremendous speeds.
In a conductor, the electrons are not 'moving' in any classical sense. It's an oversimplification. One way to look at it is that when one electron leaves the wire, another jumps over and takes its place, and another takes that ones' place and so on all the way to the other end.
(But that's an oversimplification as well, since they are indistinguishable.)
The only plausible scientific explanation is that photons are what is carrying the electricity, not electrons.
Photons cannot carry an electrical current. They have no charge. They do, however, mediate the electromagnetic force, which is the force between charged particles (the electrical field).
But they don't 'carry the electricity'. But you could say that they are what 'pushes' the electricity through the wire.
Electricity is electromagnetic radiation, i.e. light.
No, electricity is the transfer of electric charge, you wrote that yourself. Electrical potential (e.g. voltage) is mediated through photon exchange, and is light.
But just because the potential is felt at the speed of light does not mean that the current does.
(which strives to equalize this potential and reach thermodynamic equillibrium)
Quantum tunneling has nothing to do with electrons moving faster than the speed of light.
Really, why not? Are you saying that electrons do move faster than light, but not through tunnelling? Or what?
The electrons themselves move very very slowly compared to the speed of light.
Speaking of 'the electrons themselves' is meaningless, since in a conductor they are delocalized and have no individual identities. (which they don't really have otherwise either)
Speaking of the speed of a current is meaningful, speaking of the speed of the electrons is not. You can measure the current speed and divide that by the charge transferred and arrive at a number for the 'speed of an electron', but it's not a very meaningful number.
superconductors is the way to go for highest speeds/most concentrated processing power, due to extremely small power dissipation and extremely high clock frequencies (60 GHz for logic is relatively easy right now), but the problem is that after someone invests $3B in a modern semiconductor fab they do NOT want to build a $30M top-of the line superconductor fab to compete with it.
I'd think the more likely reasons would have to do, for starters, with consumers not wanting or being able to afford a computer that requires constant cooling with liquid nitrogen (or even worse, liquid helium) to work.
Rather offtopic, but.. That's a meaningless thing to talk about.
The valence electrons (ones carrying current) in a metal are delocalized. What that means is that they don't really have any 'identity'. They don't belong to any particular atom, and you can't locate them all individually. You just have a 'fermi sea' as it's called, a kind of electron cloud.
Now, the analogy to a water pipe isn't bad. But the key difference here is that with a water pipe, you can identify the water going in and coming out (say, add some colorant to the water going in and see when it comes out).
However, there is no way to 'label' the electrons going in and coming out of a section of wire. Not because we don't know any, but because it's theoretically impossible. If it was possible even in theory to distinguish the particles, then the laws of physics wouldn't allow them to be delocalized - and then the current wouldn't flow.
The algorithms chosen *are* important, and in some cases you shouldn't simply reach into the API toolbox and use the third-party solutions. There is no substitute for knowing how to write your own sort routines, specialized linked lists, and binary trees.
However, this leads to people going too far in that direction as well, which is just as problematic. People re-invent the wheel for no better reason than that their own inflated egos make them believe they can do everything better. Or because the available routines differ in some completely insignificant respect from what they'd prefer.
The result: Lots of programs duplicating effort, creating bloat. You have Gnome, KDE, Mozilla and OpenOffice all with their own set of widgets and APIs. It's fragmentation and duplication of effort for no good reason, and so the casual Linux (in this case) user ends up having to store all four of them in memory.
I think that the Not Invented Here syndrome is a far more important factor in creating 'bloat' than people chosing inefficient algorithms.
Well, Java 2D has been part of the Standard Edition since version 1.2, so I'm not sure what you're on about there.
As for the rest.. The strategy is perfectly clear. Sun makes a bunch of java add-ons in the com.sun.* heirarchy and releases them. If they catch on, they eventually get integrated into Standard Edition.
Swing, Sound and ImageIO for instance.
They used some technique of using a diamond point of action, and layering carbon onto it, like a chip fab does. It seemed like they could produce fairly large quantities at minimal cost. I'm not sure if they have been shot by deBeers yet, tho.
Chemical Vapor Deposition. You basically 'zap' the carbon, forming a plasma cloud of charged carbon atoms. Then you let that cloud cool and fall onto the diamond substrate.
If you do it 'just right' the carbon atoms in the cloud will settle into the correct positions on the substrate to continue the crystal lattice, and you'll have added a layer of atoms to the crystal. Repeat the process and you get a flawless, single-crystal diamond.
I think the Wierd article (as always) seriously over-hyped the issue though. This is something which is coming no matter what, and DeBeers won't be able to stop it. There are hundreds of researchers all around the world working on CVD diamond research. Simply put, diamond is too useful a compound to be artificially hoarded for jewellery.
Good news for the rest of us, bad news for DeBeers.
How do you get from "works" to "whatever makes me happy"?
Define what was meant by 'works' in a clearer and unambigous manner then.
Historical truths are much more mutable than those of the physical universe. It may be frustrating, but it's true.
Bull. History is part of the physical universe. At a specific place at a specific point in time, an event happened which some humans found noteworthy. That is history.
If it wasn't, how/why could history be so contested?
History cannot be contested. History is a real physical event. You are talking about accounts of history. And these accounts of history can correspond more or less well with what actually happened. That would be what we call the 'truthfullness' of an account.
"Correspondance to Reality" is one of those things that people believe in that they can't prove.
That's another bullshit argument. Nothing can be proved. Nothing you wrote can be proved either. You must assume.
The most fundamental assumption that exists is that there indeed exists an objective physical universe around us. If you deny that, then there is nothing left at all on which you can base any kind of philosophy of anything. You are also being dishonest, because that assumption underlies every single act of human behaviour.
Then there is a thing called language. Language has no meaning in itself. The entire meaning of language is in its correspondence to this objective, physical world.
Put those two together and you have the definition of truth as the correspondence to reality. No, it cannot be 'proven' in any sense, but it is as fundamental an assumption as can be made. Because you cannot even start to make an assumption without first assuming language has meaning and that reality exists.
(i.e. that you are saying something to someone)
I think what you are thinking about is that sometime the correspondence between reality and a statement (its 'truthfulness') cannot be ascertained.
To that I say 'So what?'. Human knowledge is never better than an approximation of the 'absolute truth', and I wouldn't claim it was. The ambiguities of language and limitations of the human mind make anything like that impossible anyway.
To continue on the Holocaust example: Destroy all records of it. Kill anyone with knowledge of it. It would still have happened, even if the entire human race stops thinking it did. Unfortunately that's a 'truth' which would then never be known.
But noone claimed absolute knowledge was attainable either.
On the other hand:
sacrificing a goat to the spirits is a truth that may help you survive the famine
Is not true. It does not correspond to physical reality - you would be better off eating the goat instead.
Say you have Microsoft language A vs. Open Source language B. Now, if B is better from the technical standpoint than A, then B is the better language from the engineers standpoint.
To management, though, that is only one factor in the equation. They have other concerns:
Can we find staff trained in this language?
Will we be able to in five years?
Does the language have support?
Will it have support in five years?
Will the language continue to be developed to meet our needs in the future as well?
These are all valid concerns which need to be adressed for any operating system/language/platform to catch on in the corporate world. Don't be so naive as to think that all management types are PHBs who buy any marketing that comes their way.
In other cultures, knowing why the ride to work drives you crazy is a truth that helps you stay sane.
Truth is any tool that works better.
That's the pragmatist* definition.
And it's really a pretty crappy one, IMHO. Truth, as most people use the word, is a correspondence to reality. Not "Whatever makes you happy".
For instance, we'd all be happier if the Holocaust never had occured. By that definition, it certainly is something which 'Works better' for most people. But it is not the truth in the way I, or most people, use the word. Because the Holocaust did indeed occur.
It's bunk.
(* Philosophical pragmatism.. William James for instance.)
Because if something can be proven, the issue of belief does not arise.
You need to explain that one further. Just because something can be proven doesn't mean it has been proven (either true or false).
That would be what is called a "guess", "hypothesis" or "conjecture".
Who determines who the "experts" and "authorities" are? It can't be these same people, that would just beg the question. Or perhaps its the social structures already created that mold and promote expertise.
What most of us call "facts" are the very things which are independent of social structures, cultural contexts and other biases. As such, anyone is in the position to determine who is an expert and who is not.
An expert is a person who has a more detailed knowledge of facts than someone else.
Scientific expertise is not some secret society of self-prepetuating illuminati. That's what sets Science apart from its predecessors Alchemy and Magic. It may be a meritocracy to an extent, but it is an open system. You don't need a degree to practice science.
It may be difficult in practice, but there are still plenty of people making contributions without any academic background. And even more academics making contributions within fields in which they have no formal schooling.
(The physicist Richard Feynman for instance, helped decypher Maya hieroglyphs)
The Open Source movement is very similar in that respect. Anyone can contribute, but naturally a major contributor will have an easier time getting his patches into the Linux kernel than someone who never contributed anything before.
Wikipedia is different. It is far more difficult for the readers/reviewers to determine who is right and who is wrong, especially in highly specialized areas.
Instead you risk ending up with the very thing you are accusing peer-review of. You create a social structure where the 'facts' are determined by the authority the person holds within the Wikipedia community rather than a consensus on what actually matches reality best.
Allowing privacy encourages child porn?
No, caching and distributing child porn for others encourages child porn.
So we should ban it b/c it allows privacy?
Privacy of identity (who is distributing what) is not the same thing as privacy of content (what is being distrubuted).
Publicly distributed content does not need to be anonymous to the distributor. But it is on Freenet - you don't know what it is you are distributing.
It's simply a bad solution. The problem which Freenet is supposed to adress is keeping the identities of distributors anonymous.
Their solution - making the content secret from those who distribute it, is simply a bad solution to the problem.
The worlds largest producer and consumer of films is India. I thought that was well-known?
('Bollywood' doesn't export much though.)
Japan and China have pretty big movie industries too, but don't export much either.
The USA is undisputably the largest exporter.
But just about all non-third-world countries have a domestic film industry.
The size of it is roughly proportional to the size of the country. E.g. Unsurprizingly, Germany makes more movies than the Netherlands.
Obviously the size of the budget is proportional to the size of the market. So most countries don't produce big-budget blockbuster-type films. It's too expensive, and they can't really compete with hollywood anyway.
So the natural adaption is to produce 'the other stuff' e.g. comedies, dramas, thrillers, and so on. Some of them are very good.
That said, the USA is a very lousy importer of foreign films. While a successful German film will likely get picked up by cinema chains in most of Europe, it won't get any showing in the USA. The exception here is basically if it gets nominated for a 'best foreign film' Oscar, in which case you might just catch it in an arthouse cinema if you're lucky enough to live in a town with one.
If you go to film festivals and so on, you realize a pretty depressing thing. There are very many really good and original movies are being made in small countries which almost noone in the rest of the world gets to see because these movie execs have the idée fixe that people simply won't watch anything that's subtitled.
(I think this is why other countries are better than the USA at importing films. Since they're already importing stuff from the USA, subtitling isn't much of an issue.)
Quite true. I just installed XP for the first time myself last weekend, on a new machine. The last Windows I'd used before migrating completely to Linux on my home computer was Win95.
I was surprized by the lack of drivers! Win95/98 included a pretty big set of drivers for the most common hardware. Not always the best drivers, but drivers nonetheless.
Win XP included nothing. Not even for the integrated AC97 sound card the machine had. The AC97 is _incredibly_ common. We're talking about millions of machines here. Same went for the networking. I guess you could point to the built in internet-search, but it's of course not a good answer if you're networking doesn't work.
(This was of course fixed by an included CD with drivers.)
So really, Windows has regressed with respect to drivers. Windows doesn't really include any real drivers at all anymore, it seems.
it's difficult to draw any direct causality between its placement and the timing of this article
Well, that's doesn't seem true, considering that the article itself starts off by mentioning the advertisment.
But really.. Why would that mean the article is biased? People around here are always knee-jerking about bias and FUD, but they seldom explain wherein their supposed 'bias' lies.
Isn't Microsoft given opportunity to state their side of the issue in the article?
FACT: IE has several critical security holes.
FACT: Some of them are fixed by SP2.
FACT: SP2 isn't available for people running older windows versions.
FACT: Millions and millions of people are running older windows versions.
It's a very relevant question. And I think most people would agree Microsofts suggestion that people should get a new computer is rather arrogant.
If there's any bias in the article, it's not pro-Firefox as much as it is pro-consumer.
If MS was to do the right thing by their customers, they wouldn't be getting this kind of press.
Oh please. I didn't say "Windows is backwards compatible with every single app".
I said that Microsoft tries hard to keep backwards-compatibility.
But don't take my word for it, I don't work for them. Read Raymond Chen's various blog articles on the subject. He is one of the poor souls at MS who worked his butt off to try and keep backwards-compatibility.
I'm impressed by what they've done so far, but not the seven years it took them to do it.
Well, this is a bit symptomatic of a lot of OSS projects, they start out with a grand vision and end up planning and then re-planning and throwing out code and never really get off the ground. Some die and stay dead, some get picked up by a group of enthusiasts with a more down-to-earth approach of 'Getting something working now, improvements later.' and the project takes off.
(Case study: See Linux vs. GNU Hurd)
I believe this is pretty much what happened with ReactOS (I'm not a ReactOS developer), so I wouldn't hold it against the current crowd too much.
I suspect it will have the same trouble WINE runs into: it's chasing a moving target, and it's way behind.
Ah, the old catch-up argument. It's a valid argument, but it's not as important for API:s as it is for, say, file formats.
With the MS Word file format, Microsoft can tweak and alter that all they want, because it's not publicly documented, and they're not that interested in having compatibility with anything other than MS own products. Backwards compatibility isn't important. Heck, they're happy to break it and create incentive for people to buy the new versions. That's a hard act to follow.
With API:s, things are quite different though. Firstly, the '80-20 rule of features' pretty much applies. Most programs don't use the entire API, but a rather small subset.
Secondly, API:s rarely break backwards compatibility. That would break all existing third-party apps and make it difficult for people to migrate. The exact opposite situation to the previous one. So MS bends over backwards to make stuff backwards compatible. Windows 3.0 apps still run on XP.
The APIs are also (relatively) well documented. Sure, there's a lot of undocumented functionality, but most of that is also unused. The implementer has access to the same information as most application-developers.
Another point of difference is that you don't have to be super-fast in implementing new API:s.
The day a new Word version hits the street, people will be asking 'Why doesn't this work with OpenOffice?'.
Not as true for APIs. While we all like the latest and greatest, professional developers don't rewrite their programs to use the latest APIs 'just for the heck of it'. There has to be good reason. In fact, you want to avoid using the latest APIs as far as practical and economical, because otherwise, you're going to be shutting-out potential customers running the old OS version.
(There are plenty of brand-new apps released today which run on Win98, or Win95 even.)
I think you're selling classical physics short. Newton's laws are hardly just empirical rules of thumb. I think there's hardly anything as beautiful in its simplicity as the principle of least action, and classical mechanics is embodied in it.
Kepler's laws fit your description fairly well, but Newton's laws were and still are rather profound in their scope.
You're absolutely right about this. Newtonian mechanics as a whole is something more than an ad-hoc set of equations. In general, Newton did more than just curve-fitting. He did work from a basic set of postulates when it came to mechanics.
This is why I specifically referred to Newton's Law of Gravity. His gravity law is an empirical rule of thumb. In my opinion.
He couldn't really justify the way he defined the center of gravity, or why the force had to be inverse-square.
But all of his mechanics are not purely empirical, no.
The difference here lies in the fact that those laws are based on physical properties of the universe!
So how do you reconcile this with your previous definition of 'law' being 'the highest designation we [the science and engineering community] can give phenomena' and also mentioning evolution, which is hardly a 'physical property of the universe' by any account.
Moore's law is dependent on humans to develop technology. We could, in theory, stop advancing technology thus stopping Moore's law. We can't, however, prevent gas from being related inversely between pressure and volume.
Moore's law is an approximation of the speed of computer development, made at a specific point in time. Noone, and certainly not Moore himself, would ever claim it would be valid forever.
We can certainly prevent gas from being related inversely between pressure and volume. Try increasing the pressure. The ideal gas law is an approximation, which is actually only valid as the limit when the pressure drops to zero. If you increase the pressure enough, the gas will eventually liquify, something which Boyle's law of gases doesn't account for at all.
Again, a 'law' is an ad-hoc empirical description. All these physical laws have certain ranges of validity.
For the gas law, it's when the pressure approaches zero. Ohm's law only applies to Ohmic materials. Hooke's law only applies to small deformations (and is rather crude despite that), and so on.
Moore's law is only applicable to the late 20th-early 21st century. So?