I admit that I am not doing a very good job at supporting my claim. But I am correct and history proves it even if I can't (or more over I don't feel like it).
If you want a better answer then you should search through the Linux kernel archives and try to find some posts where people with some universal driver scheme try to win the Linux developers over.
Very quickly you will come to understand that for incumbent operating systems like Linux and FreeBSD, a universal driver API is a much bigger hassle than the regular method. And that for the operating systems that the API is least like, the driver will be suboptimal - and suboptimal is not good enough for many pieces of hardware.
Now I am confused. If the API can change all the time, then what good are the drivers? How many API revisions does your kernel have to support before you realise that it isn't worth the hassle?
The video card driver API probably is going to change about every year. And every few years for sound cards.
When a completely new device comes out, do you wait a year for the standards body to give you an API or do you go ahead and just write a real driver?
Think about video cards and how much they change over the years. What if all the new video cards do curved surfaces in the future instead of polygons. You can't predict what they will be doing.
Think about the funky devices that don't fit into general notions. I have a sound card that has 2 dsp's. One is the regular dsp and the second is used in conjunction with a few MB of flash and RAM to emulate a syth in software. How can you be prepared for odd ball hardware like that.
That API would have to be extremely generic and extremely extensible.
If someone could make such a thing, and it actually be useful and not cumbersome, then that person would be a CS god.
Many people had a mathematical understanding of Special Relativity waay before Einstein. You take the rules for Electricity and Magnetism and you immediately see that they don't work right when you add velocities like with Newtonian Mechanics. Newtonian Mechanics has Gallilean relativity. That's right, there was relativity before Einstein.
It is a straight forward exercise to see that Maxwell's Equations (for E&M) have Lorentzian relativity. That's right they have Lorentz symmetry and NOT Einsteinian symmetry. Other people had already figured this out.
It was Einstein that put his foot down and said look here, Maxwell's equations are right and Newton is wrong. He explained the Lorentz transformations as being a very physical thing. This in and of itself wasn't mega-incredibly important, and Einstein never got a Nobel for it.
But he then knew that he had to fix gravity to work with special relativity. And he did that with the new fangled geometry of Riemmann, which was the more generalized form of Poisson's earlier work. Einstien let the metric be dynamic and the rest was just a matter of elbow grease. That is what made him uber. Dirac did not do that.
This is not to say that Dirac wasn't a good physicist or even an equal to Einstein. Dirac was obviously much better at quantum mechanics than Einstein. (Einstein fell behind the times really bad) Dirac did give us the relativistic, quantum theory of the electron and positron.
Dirac probably was an equal to Einstein. But Dirac didn't have a very good personality. He was anti-social to say the least. If you were a student at Florida State and said hi to him, chances were good that he wouldn't respond at all.
You can say that Dirac is underrated by the public, though not by physicists. You can say that Einstien didn't invent but 10% of special relativity. But to say that Einstein stole his work from Dirac is just BS and your link does not justify that statement.
I agree with everything that you have said, factually.
However, there is much room for a similar line of research to be done in the future that is very much like this.
Outerspace gravity experiments can do soooo much more than Earth based ones. Things like detecting gravitational waves at different energy ranges, !detecting the Axion!, and probing gravity at small ranges to detect the compactified dimensions of string theory.
Hopefully this can pave the way for the real stuff.
is only true in the rest frame of the massive particle
(mc) = (E/c) - p
is the full special relativistic formula
also known as
E = gamma mc
or
(mc) = p_t - p_x - p_y - p_z
in the 4-vector notation
or
(mc) = p_0 - p_1 - p_1 - p_1
or
m = (p*G*p)
where we take the inner product with the metric G
in General relativity we just replace the Special relativity "flat" metric with a curved metric (and we take special care wrt nonorthogonal coordinate systems leading to the dual of vectors being covectors and not vectors... blah blah... handwaving)
so E = mc is inherent in GR
Special Relativity is in GR, kinda like Euclidean geometry is in Newtonian Mechanics.
Not really. The main motivation for the general theory was simply that Newtonian gravity (or more specifically, the Newtonian gravitational potential) failed to make predictions which agreed with observation. The most well-known example of this is the precession of perihelion of Mercury. If you're referring to the fact that Newtonian gravity imposes no upper bound on velocities, then you're correct, but this was more an illustration of the fact that Newtonian gravity was largely irreconcilable with special relativity.
Nope. I am afraid that the parent was correct and that you may have misunderstood him.
Einstein's motivation for GR (General Relativity) was that SR (Special Relativity) is inconsistant with NG (Newtonian Gravity). NG does indeed predict faster than light effects. If you wiggle a particle on one side of the galaxy, then a particle on the other side would feel that immediately.
This is a theoretical motivation, and not a physical motivation. Once you have SR, you immediately have to fiddle with gravity. He would have had to do this even if we had no conflicting evidence against NG.
Special Relativity is the special case of General Relativity when there is no matter to bend spacetime.
General Relativity has "constancy of light speed" and "inertial frames" just the same as Special Relativity. Light travels a null-geodesic in both SR and GR. If you had faster than light travel in one theory, then it would be easy to bridge that to the other theory, they go hand in hand. The difference between the two is usually just a matter of changing metrics and covariant derivatives.
It's almost as bad as when you attempt to put a vanilla Hayes modem or tulip NIC in a Windows box.
Sound Cards stopped being SB compatible when games stopped being written for DOS. I doubt any current sound card would accept an SB driver (or Roland or Adlib for that matter), especially seeing as many variations of the SB aren't really compatible in that manner. But I am interested in being proven wrong.
My question is. Why shouldn't devices come with drivers installed on the device themselves in a platform-indepedant language? Let Windows, Linux, FreeBSD, MacOS, or whatever compile the real driver from that abstract driver when first ran. Instead of updating a driver on the OS you could update it in the flash mem of the device and then let it recompile and run the new driver from the device. Then all OS's would have better driver support - even Windows. This wouldn't be to hard to implement as a standard for new hardware so why isn't it done? Legacy hardware could still have the drivers written in this abstract driver language.. you'd just obviously have to keep a legacy driver cache for your OS to compile when it found those devices. You'd also get the benefit that the drivers could always be compiled to get the best use out of your hardware while being a transparent operation to the user.
That has been thought of and even attempted. And each time it has failed a still birth.
Windows and Linux look at devices in a completely different manner. If a device driver were generic enough to work on both platforms then it would be very suboptimal on each platform.
The architectures are also changing. Some (linux) more rapidly then others. There is no way of knowing what features such a generic driver would need in the future. Eventually you will have layers upon layers of abstraction between the generic driver and the kernel.
But more importantly. Once one good OSS driver is written, then porting it is usually a trivial task. All in all, an OSS driver is really a better solution.
When I wrote Molecular Dynamics code I made similar odd-ball macros to optimize all of my vector operations. (By optimize I mean that it is much faster to write, and in my case it was partially unrolling loops)
Sadly this method is way faster than doing it the correct way in C++. (Or at least it was when I did it)
No offense intended. Those jocks probably are smart. Incidentally I was an uber jock in highschool and I know for a fact that Brown and Dartmouth cheat by giving academic scholarships to athelets that do not deserve them. (Dartmouth was the worst btw: they kept sending letters to a B student football player and didn't even accept our fucking valedictorian - man he was pissed when he found out.)
> > LATOR is capable of testing string theory, an exciting but so far merely theoretical development in high energy physics.
> Or rather, it might conceivably be capable of testing some rather speculative models within string theory; there are plenty of other string theory models that LATOR can't test, and no good reason to believe in one over the other. That's one of the problems with string theory: it's too flexible. People can cook up all sorts of artificial string models, but that doesn't mean that any of those models are likely to be true, even if string theory itself is true.
It will test some of the most reasonable/popular models, which is a big step up from having never been tested at all.
I am sorry, but that is not true. There are upper limits on the radii of extra compact dimensions which gravity operates in. These are found by testing gravity at smaller and smaller distances and looking for a deviation from the Newtonian potential - or more exactly the Poisson equation for mass density.
The poster has a valid point. Any test you do of String Theory model is a measure zero set in the entire group of theories.
How come, in America, the Alamo is viewed as being a patriotic thing?
"Hey guys. Those Mexicans don't like me haven mah slaves. In fact if mah slaves run away across the border, Ima gonnah looose alota monay. Let's steal the land and kill the Mexicans if they try an stop us."
Now I honestly don't think America is that bad compared to most countries. At least in America teachers have the freedom to teach you the real Truth, and likely you will meet at least one of those.
If you want a bad country, look at the Japanese. Their history books are bull shit. Their culture is just rotten in that respect.
You have a confounding variable though. American's have a terrible, terrible diet. I would bet that the average American is also fatter than the average person in any of those countries.
When Bush first ran for government in Texas, he ran as an educated person from a top notch school. He talked smart and he lost the election for being an "egg head".
Subsequently Bush ran as an average Southerner of an average intelligence who "knows what's best". He's done pretty well since then.
His persona of being a Southerner of average intelligence, average mannerisms, and average speaking has been a core part of getting him this far. I think it's because those are traits that people of average intelligence don't see as being the most important traits to have.
Instead of intelligence, he has "know how", he "gets things done". Those are the ideals of the common man.
I honestly have no idea how much of this is natural and how much of this is a show. I would guess that his true person is somewhere between the egghead and the arrogant ignoramus.
Bayesian reasoning deals with this problem. Pick up any textbook on modern Bayesian statistics and Bayesian reasoning.
I am familiar with Bayesian statistics as is any student who has taken a class in probability. I'm not sure where you are going with this?
> Einsten's GR is the most elegant theory in physics. You don't seem to be able to distinguish between math and physics.
Actually, that's what I would claim is your problem: you think the differential geometry formulas are the physical theory.
The physical theory is that there is a proportianality between the stress-energy-momentum tensor (which is divergenceless) and a divergenceless tensor of the curvature difference between the dynamic metric and Minkowski space-time (which happens to be the Einstein tensor). That is the physics: it is a statement that energy curves space-time from its empty, Minkowski form.
They aren't. GR is a huge (implicit) set of postulates about the structure of space
As is all physics. What is your point? The only additional postulate in GR is that the background is dynamic and not fixed.
postulates that closely mirror the axiomatic systems underlying differential geometry (not by accident, one might add).
It is Differetial geometry, GR is a geometrical theory.
All the mathematical machinery of differential geometry, which then makes up GR, is just mathematics, with no physical content.
Yes, and?
I'm simply stating that it involves a huge mathematical edifice; if you model space using the methods of differential geometry, you are making a huge set of assumptions about space.
No more than with any other theory.
The fact that you can summarize that as "one equation" tells you nothing about elegance or simplicity of the theory.
I can't prove elegance, but GR is naturally represented in a coordinate free, purely geometrical manner. It is represented by a single, proportianality equation that is simple to understand the meaning of. From the perspective of a mathematician, GR is the most elegant physical theory. Your opinions may be different, but they should at least be justified.
By analogy, GR is only simple in the way a car is simple for its drive; from the driver's point of view, you just sit in it and turn the ignition. Never mind the $20k worth of hardware that is needed to make the magic happen. You seem to forget about what's happening under the covers (which also makes me suspect that you are not actually a mathematician).
You are just making the math is hard argument that you just denied. If you think differential geometry is so easy, then you are now contradicting your self. GR isn't elegant because its easy to do calculations. In fact numerical GR is insanely difficult.
The bipolar transistor was an accidental discovery by Shockley's group trying to implement field effect transistors, and the field effect transistor was patented in 1925, long before Shockley or Bell Labs even got into the game. Furthermore, it was advances in material sciences that made the implementation of the modern FET possible. In the end, Shockley's group accidentally stumbled over the bipolar transistor and they failed developing the FET based on their theoretical work. Of course, that isn't to say that those guys weren't very smart and highly successful, but your history of the transistor as a straight path from theory to implementation is a myth. You can read some of the history here.
Since you are so good at history, answer me this. How good was their original design, and what did Shockley do afterwards that was even more important? We don't use their original design today.
> You have clearly demonstrated that you have no working knowledge of GR.
Another conclusion to which you jumped.
I based it on some of the odd, nonsensical things that you have said.
> In statistics we call this "failure to reject the null hypothesis".
Sorry, but you ("we") are a little behind the times there. Statistical hypothesis testing is still widely used by "working scientists" because it's simple, can be applied cookbook-style, and because it often gives some useful answers. But it is (provably) not a consistent and correct way of updating your degree of belief in a hypothesis or a scientific theory.
(To put this in familiar terms, trying to argue about how scientific theories are validated from the point of view of statistical hypothesis testing is the rough equivalent of arguing about atomic spectra with knowledge of only Newtonian mechanics.)
Scientists use this manner because of the following well known philosophical problem:
I claim that white birds exist.
Person A finds 1000 birds and counts the number of birds that are white. Person A concludes that white birds exist.
Person B finds 1000 things that have the property of whiteness and counts the number of white things that are birds. Person B concludes that white birds do NOT exist.
When you say you are proving a physical theory you are like person B. The untested domain is too large to assert that you have ever proven anything. This is the sad fact of physical science.
I am only classically trained in probability and logic. If some new methods have formed in the last 50 years and nobody bothered to tell me, then please show me the correct answer. Give me some references.
> Einstein's GR is one single equation.
What a bizarre idea. Einstein's GR actually is a highly complex set of assumptions about spacetime (most of them, unfortunately, not explicitly stated). Using layers upon layers of mathematical structures and notation, those assumptions can be summarized in "one single equation", but that's only because of a lot of good will on the part of the reader.
Oh I forgot the one assumption: space-time is a 4 dimensional manifold of the -+++ complex-Euclidean type.
Einsten's GR is the most elegant theory in physics. You don't seem to be able to distinguish between math and physics. Einstein's GR is one proportionality equation. Yes, GR is hard to do, but that is because differential geomoetry is hard. But it is simple enough that I could teach it to a math grad student in less than one hour.
If you want to argue about how complicated GR is then you are making an argument against mathematics (I am a mathematician btw). And if you want to say that math is hard, then I will only respond with a hell-fucking yes math is hard.
> Let's see, what did the physicists do last century? > Semiconductors -> Transistor -> the mother fucking Computer
Trying to lump together solid state physics and general relativity as if successes in one are a validation of the approach of the other is simply dishonest. In fact, the field effect transistor could have been discovered and optimized completely empirically, with essentially no physical insights (I don't remember how it was actually discovered and optimized, but most of solid state physics uses empirical models anyway.)
Straw man. You made a claim about all of physics and I countered that claim.
And I am very familiar with the history of the transitor and you are as wrong as wrong can be. Semiconductor material for the operation of the transistor was predicted by a very bright theoretical physicist. He had a top notch group assigned to him at Bell Labs and it took them years to get an operational transistor. It years for a 2-Nobel winning physicist to discover.
> I claim that Einstein's GR is the most mathematical of any physical theory that has ever existed. If GR isn't mathy then no physical theory is!
My point exactly.
Then I guess you would be in strong disagreement with the great mathematicians such as Newton and Gauss, or even a current F
Slashdot Mirror
Talking about bandwidth here is just like talking about the electric current of one electron.
It is only going to make sense in the context of distribution functions.
Correction: replace Poisson with Gauss
I am so stupid.
I admit that I am not doing a very good job at supporting my claim. But I am correct and history proves it even if I can't (or more over I don't feel like it).
If you want a better answer then you should search through the Linux kernel archives and try to find some posts where people with some universal driver scheme try to win the Linux developers over.
Very quickly you will come to understand that for incumbent operating systems like Linux and FreeBSD, a universal driver API is a much bigger hassle than the regular method. And that for the operating systems that the API is least like, the driver will be suboptimal - and suboptimal is not good enough for many pieces of hardware.
Now I am confused. If the API can change all the time, then what good are the drivers? How many API revisions does your kernel have to support before you realise that it isn't worth the hassle?
The video card driver API probably is going to change about every year. And every few years for sound cards.
When a completely new device comes out, do you wait a year for the standards body to give you an API or do you go ahead and just write a real driver?
It was the way his father raised him. He learned not to speak much.
Think about video cards and how much they change over the years. What if all the new video cards do curved surfaces in the future instead of polygons. You can't predict what they will be doing.
Think about the funky devices that don't fit into general notions. I have a sound card that has 2 dsp's. One is the regular dsp and the second is used in conjunction with a few MB of flash and RAM to emulate a syth in software. How can you be prepared for odd ball hardware like that.
That API would have to be extremely generic and extremely extensible.
If someone could make such a thing, and it actually be useful and not cumbersome, then that person would be a CS god.
... except that he plagiarized Dirac's works...
Many people had a mathematical understanding of Special Relativity waay before Einstein. You take the rules for Electricity and Magnetism and you immediately see that they don't work right when you add velocities like with Newtonian Mechanics. Newtonian Mechanics has Gallilean relativity. That's right, there was relativity before Einstein.
It is a straight forward exercise to see that Maxwell's Equations (for E&M) have Lorentzian relativity. That's right they have Lorentz symmetry and NOT Einsteinian symmetry. Other people had already figured this out.
It was Einstein that put his foot down and said look here, Maxwell's equations are right and Newton is wrong. He explained the Lorentz transformations as being a very physical thing. This in and of itself wasn't mega-incredibly important, and Einstein never got a Nobel for it.
But he then knew that he had to fix gravity to work with special relativity. And he did that with the new fangled geometry of Riemmann, which was the more generalized form of Poisson's earlier work. Einstien let the metric be dynamic and the rest was just a matter of elbow grease. That is what made him uber. Dirac did not do that.
This is not to say that Dirac wasn't a good physicist or even an equal to Einstein. Dirac was obviously much better at quantum mechanics than Einstein. (Einstein fell behind the times really bad) Dirac did give us the relativistic, quantum theory of the electron and positron.
Dirac probably was an equal to Einstein. But Dirac didn't have a very good personality. He was anti-social to say the least. If you were a student at Florida State and said hi to him, chances were good that he wouldn't respond at all.
You can say that Dirac is underrated by the public, though not by physicists. You can say that Einstien didn't invent but 10% of special relativity. But to say that Einstein stole his work from Dirac is just BS and your link does not justify that statement.
I agree with everything that you have said, factually.
However, there is much room for a similar line of research to be done in the future that is very much like this.
Outerspace gravity experiments can do soooo much more than Earth based ones. Things like detecting gravitational waves at different energy ranges, !detecting the Axion!, and probing gravity at small ranges to detect the compactified dimensions of string theory.
Hopefully this can pave the way for the real stuff.
E = mc
... blah blah ... handwaving)
is only true in the rest frame of the massive particle
(mc) = (E/c) - p
is the full special relativistic formula
also known as
E = gamma mc
or
(mc) = p_t - p_x - p_y - p_z
in the 4-vector notation
or
(mc) = p_0 - p_1 - p_1 - p_1
or
m = (p*G*p)
where we take the inner product with the metric G
in General relativity we just replace the Special relativity "flat" metric with a curved metric (and we take special care wrt nonorthogonal coordinate systems leading to the dual of vectors being covectors and not vectors
so E = mc is inherent in GR
Special Relativity is in GR, kinda like Euclidean geometry is in Newtonian Mechanics.
Not really. The main motivation for the general theory was simply that Newtonian gravity (or more specifically, the Newtonian gravitational potential) failed to make predictions which agreed with observation. The most well-known example of this is the precession of perihelion of Mercury. If you're referring to the fact that Newtonian gravity imposes no upper bound on velocities, then you're correct, but this was more an illustration of the fact that Newtonian gravity was largely irreconcilable with special relativity.
Nope. I am afraid that the parent was correct and that you may have misunderstood him.
Einstein's motivation for GR (General Relativity) was that SR (Special Relativity) is inconsistant with NG (Newtonian Gravity). NG does indeed predict faster than light effects. If you wiggle a particle on one side of the galaxy, then a particle on the other side would feel that immediately.
This is a theoretical motivation, and not a physical motivation. Once you have SR, you immediately have to fiddle with gravity. He would have had to do this even if we had no conflicting evidence against NG.
Special Relativity is the special case of General Relativity when there is no matter to bend spacetime.
General Relativity has "constancy of light speed" and "inertial frames" just the same as Special Relativity. Light travels a null-geodesic in both SR and GR. If you had faster than light travel in one theory, then it would be easy to bridge that to the other theory, they go hand in hand. The difference between the two is usually just a matter of changing metrics and covariant derivatives.
+5 Sheesh.
It's almost as bad as when you attempt to put a vanilla Hayes modem or tulip NIC in a Windows box.
Sound Cards stopped being SB compatible when games stopped being written for DOS. I doubt any current sound card would accept an SB driver (or Roland or Adlib for that matter), especially seeing as many variations of the SB aren't really compatible in that manner. But I am interested in being proven wrong.
My question is. Why shouldn't devices come with drivers installed on the device themselves in a platform-indepedant language? Let Windows, Linux, FreeBSD, MacOS, or whatever compile the real driver from that abstract driver when first ran. Instead of updating a driver on the OS you could update it in the flash mem of the device and then let it recompile and run the new driver from the device. Then all OS's would have better driver support - even Windows. This wouldn't be to hard to implement as a standard for new hardware so why isn't it done? Legacy hardware could still have the drivers written in this abstract driver language.. you'd just obviously have to keep a legacy driver cache for your OS to compile when it found those devices. You'd also get the benefit that the drivers could always be compiled to get the best use out of your hardware while being a transparent operation to the user.
That has been thought of and even attempted. And each time it has failed a still birth.
Windows and Linux look at devices in a completely different manner. If a device driver were generic enough to work on both platforms then it would be very suboptimal on each platform.
The architectures are also changing. Some (linux) more rapidly then others. There is no way of knowing what features such a generic driver would need in the future. Eventually you will have layers upon layers of abstraction between the generic driver and the kernel.
But more importantly. Once one good OSS driver is written, then porting it is usually a trivial task. All in all, an OSS driver is really a better solution.
Some weird macros are used:
When I wrote Molecular Dynamics code I made similar odd-ball macros to optimize all of my vector operations. (By optimize I mean that it is much faster to write, and in my case it was partially unrolling loops)
Sadly this method is way faster than doing it the correct way in C++. (Or at least it was when I did it)
I am guessing that the coder is older.
A Harvard degree does not impress me.
You can get through a liberal arts education and still be considered a dumbass in my book.
No offense intended. Those jocks probably are smart. Incidentally I was an uber jock in highschool and I know for a fact that Brown and Dartmouth cheat by giving academic scholarships to athelets that do not deserve them. (Dartmouth was the worst btw: they kept sending letters to a B student football player and didn't even accept our fucking valedictorian - man he was pissed when he found out.)
And how those hairless sticks of meat make me want to retch sulphur.
> > LATOR is capable of testing string theory, an exciting but so far merely theoretical development in high energy physics.
> Or rather, it might conceivably be capable of testing some rather speculative models within string theory; there are plenty of other string theory models that LATOR can't test, and no good reason to believe in one over the other. That's one of the problems with string theory: it's too flexible. People can cook up all sorts of artificial string models, but that doesn't mean that any of those models are likely to be true, even if string theory itself is true.
It will test some of the most reasonable/popular models, which is a big step up from having never been tested at all.
I am sorry, but that is not true. There are upper limits on the radii of extra compact dimensions which gravity operates in. These are found by testing gravity at smaller and smaller distances and looking for a deviation from the Newtonian potential - or more exactly the Poisson equation for mass density.
The poster has a valid point. Any test you do of String Theory model is a measure zero set in the entire group of theories.
How come, in America, the Alamo is viewed as being a patriotic thing?
"Hey guys. Those Mexicans don't like me haven mah slaves. In fact if mah slaves run away across the border, Ima gonnah looose alota monay. Let's steal the land and kill the Mexicans if they try an stop us."
Now I honestly don't think America is that bad compared to most countries. At least in America teachers have the freedom to teach you the real Truth, and likely you will meet at least one of those.
If you want a bad country, look at the Japanese. Their history books are bull shit. Their culture is just rotten in that respect.
More than once I have read a compelling "pro-SCO" post that disappears not long after it appears.
Too bad SCO had nothing compelling in the court room.
OOOOO Burn.
Japan, a country with a very healthy diet, is #1 in infant mortality rate, and #4 in life expectancy.
Therefore I conclude that life expectance and infant mortality rate are completely and 100% dependent upon diet.
Doesn't that seem a bit silly?
It's SCO IP, what did you expect?
And you know what? They live longer than we do.
You have a confounding variable though. American's have a terrible, terrible diet. I would bet that the average American is also fatter than the average person in any of those countries.
When Bush first ran for government in Texas, he ran as an educated person from a top notch school. He talked smart and he lost the election for being an "egg head".
Subsequently Bush ran as an average Southerner of an average intelligence who "knows what's best". He's done pretty well since then.
His persona of being a Southerner of average intelligence, average mannerisms, and average speaking has been a core part of getting him this far. I think it's because those are traits that people of average intelligence don't see as being the most important traits to have.
Instead of intelligence, he has "know how", he "gets things done". Those are the ideals of the common man.
I honestly have no idea how much of this is natural and how much of this is a show. I would guess that his true person is somewhere between the egghead and the arrogant ignoramus.
Bayesian reasoning deals with this problem. Pick up any textbook on modern Bayesian statistics and Bayesian reasoning.
I am familiar with Bayesian statistics as is any student who has taken a class in probability. I'm not sure where you are going with this?
> Einsten's GR is the most elegant theory in physics. You don't seem to be able to distinguish between math and physics.
Actually, that's what I would claim is your problem: you think the differential geometry formulas are the physical theory.
The physical theory is that there is a proportianality between the stress-energy-momentum tensor (which is divergenceless) and a divergenceless tensor of the curvature difference between the dynamic metric and Minkowski space-time (which happens to be the Einstein tensor). That is the physics: it is a statement that energy curves space-time from its empty, Minkowski form.
They aren't. GR is a huge (implicit) set of postulates about the structure of space
As is all physics. What is your point? The only additional postulate in GR is that the background is dynamic and not fixed.
postulates that closely mirror the axiomatic systems underlying differential geometry (not by accident, one might add).
It is Differetial geometry, GR is a geometrical theory.
All the mathematical machinery of differential geometry, which then makes up GR, is just mathematics, with no physical content.
Yes, and?
I'm simply stating that it involves a huge mathematical edifice; if you model space using the methods of differential geometry, you are making a huge set of assumptions about space.
No more than with any other theory.
The fact that you can summarize that as "one equation" tells you nothing about elegance or simplicity of the theory.
I can't prove elegance, but GR is naturally represented in a coordinate free, purely geometrical manner. It is represented by a single, proportianality equation that is simple to understand the meaning of. From the perspective of a mathematician, GR is the most elegant physical theory. Your opinions may be different, but they should at least be justified.
By analogy, GR is only simple in the way a car is simple for its drive; from the driver's point of view, you just sit in it and turn the ignition. Never mind the $20k worth of hardware that is needed to make the magic happen. You seem to forget about what's happening under the covers (which also makes me suspect that you are not actually a mathematician).
You are just making the math is hard argument that you just denied. If you think differential geometry is so easy, then you are now contradicting your self. GR isn't elegant because its easy to do calculations. In fact numerical GR is insanely difficult.
The bipolar transistor was an accidental discovery by Shockley's group trying to implement field effect transistors, and the field effect transistor was patented in 1925, long before Shockley or Bell Labs even got into the game. Furthermore, it was advances in material sciences that made the implementation of the modern FET possible. In the end, Shockley's group accidentally stumbled over the bipolar transistor and they failed developing the FET based on their theoretical work. Of course, that isn't to say that those guys weren't very smart and highly successful, but your history of the transistor as a straight path from theory to implementation is a myth. You can read some of the history here.
Since you are so good at history, answer me this. How good was their original design, and what did Shockley do afterwards that was even more important? We don't use their original design today.
> You have clearly demonstrated that you have no working knowledge of GR.
Another conclusion to which you jumped.
I based it on some of the odd, nonsensical things that you have said.
But if you are soo good
> In statistics we call this "failure to reject the null hypothesis".
Sorry, but you ("we") are a little behind the times there. Statistical hypothesis testing is still widely used by "working scientists" because it's simple, can be applied cookbook-style, and because it often gives some useful answers. But it is (provably) not a consistent and correct way of updating your degree of belief in a hypothesis or a scientific theory.
(To put this in familiar terms, trying to argue about how scientific theories are validated from the point of view of statistical hypothesis testing is the rough equivalent of arguing about atomic spectra with knowledge of only Newtonian mechanics.)
Scientists use this manner because of the following well known philosophical problem:
I claim that white birds exist.
Person A finds 1000 birds and counts the number of birds that are white. Person A concludes that white birds exist.
Person B finds 1000 things that have the property of whiteness and counts the number of white things that are birds. Person B concludes that white birds do NOT exist.
When you say you are proving a physical theory you are like person B. The untested domain is too large to assert that you have ever proven anything. This is the sad fact of physical science.
I am only classically trained in probability and logic. If some new methods have formed in the last 50 years and nobody bothered to tell me, then please show me the correct answer. Give me some references.
> Einstein's GR is one single equation.
What a bizarre idea. Einstein's GR actually is a highly complex set of assumptions about spacetime (most of them, unfortunately, not explicitly stated). Using layers upon layers of mathematical structures and notation, those assumptions can be summarized in "one single equation", but that's only because of a lot of good will on the part of the reader.
Oh I forgot the one assumption: space-time is a 4 dimensional manifold of the -+++ complex-Euclidean type.
Einsten's GR is the most elegant theory in physics. You don't seem to be able to distinguish between math and physics. Einstein's GR is one proportionality equation. Yes, GR is hard to do, but that is because differential geomoetry is hard. But it is simple enough that I could teach it to a math grad student in less than one hour.
If you want to argue about how complicated GR is then you are making an argument against mathematics (I am a mathematician btw). And if you want to say that math is hard, then I will only respond with a hell-fucking yes math is hard.
> Let's see, what did the physicists do last century?
> Semiconductors -> Transistor -> the mother fucking Computer
Trying to lump together solid state physics and general relativity as if successes in one are a validation of the approach of the other is simply dishonest. In fact, the field effect transistor could have been discovered and optimized completely empirically, with essentially no physical insights (I don't remember how it was actually discovered and optimized, but most of solid state physics uses empirical models anyway.)
Straw man. You made a claim about all of physics and I countered that claim.
And I am very familiar with the history of the transitor and you are as wrong as wrong can be. Semiconductor material for the operation of the transistor was predicted by a very bright theoretical physicist. He had a top notch group assigned to him at Bell Labs and it took them years to get an operational transistor. It years for a 2-Nobel winning physicist to discover.
> I claim that Einstein's GR is the most mathematical of any physical theory that has ever existed. If GR isn't mathy then no physical theory is!
My point exactly.
Then I guess you would be in strong disagreement with the great mathematicians such as Newton and Gauss, or even a current F