Nuclear power? I'd say the team led by Italy's Enrico Fermi, or if you look back further, New Zealand's Ernest Rutherford.
Szilard (who was Hungarian and ended up in the U.S. by way of Britain) was the first to propse a nuclear chain reaction. Fermi, who was by then a U.S. citizen, did the necessary science to turn the idea into reality.
So now it comes down to do we go by country of residence, country of citizenship, or country of birth?
A bonus is *not* something that an employer is required or expected to give to an employee. If you wanted more *salary*, you should have negotiated it.
Depends where you work; someplace like investment banking and your bonus can easily be 150% of your base salary. This is the norm-the size of the bonus is directly related to how well the company as a whole did.
A bonus is a special reward -- an employee did something really exceptional and their employer wanted to show their appreciation.
I'm sure all of senior management did something really exceptional this year and none of the drones did anything useful. I'm sure that's the real reason the CEO's bonus went up and the drones are getting the bobble heads.
In my experience, it is extremely rare to find a journal/conference publication that includes enough information in the methods section to allow others to either check or verify the work or use the findings themselves.
Probably depends on the field of research. Working in physics, I've never had that problem.
The only problem is that the developers are nearly all in their 60's and will probably retire soon. And most of this generation (and probably the last one) don't even want to look at anything in COBOL, RPG, CL, or whatever the system's applications are developed with, much less make it a career. Eventually these things will die because nobody will know what to do with them. In 10 years it will be damn near impossible to find people who will work with anything that isn't GUI-based.
Now if the government truly has been witholding monies from really good projects, sure, that's bad. But in my amateur interested following of the space progression, there hasn't been any 'wow' project which has simply been unable to get funding.
You mean like the fast pluto fly-by (the only unvisited planet in the solar system and one we need to visit soon if we want to see it before the entire atmosphere re-freezes) which was never funded due to the ISS induced fund shortage? Actually, the entire unmanned budget and the earth sciences budget in particular has also taken a beating in the last decade as the ISS sucks up a progressively larger chunck of the annual allocation.
All of the top 10 fastest computers in the world are multi-multi-million dollar machines. This is a breakthrough because it represents another milestone in bringing supercomputing accessible.
Unfortunately, not all supercomputers are created equal. For running something like Monte Carlo simulations where each thread of execution (i.e. CPU) proceeds more or less independantly of each other, Beowulf clusters are just the ticket. For other problems (e.g. molecular dynamics simulations with long-range interactions), you pretty much need a shared-memory system otherwise you nuke your network bandwith keeping the system state of every node in sync with every other node.
Real super-computers also have internal bandwidths
that are orders of magnitude faster than your PCI bus (35 Gbyte/sec/CPU for a top-line Cray, even faster for in-cache access). They also include things like vector processors that just fly through repetative mathematical calculations (think loop unrolling, but done in hardware and big chunks of the loop done in a single clock cycle).
In short, yes a Cray is a multi-multi million dollar machine, but for many common scientific computing problems you are getting a level a performance for that money that no cluster has any hope of touching.
What is the advantage of having a hard drive in each node. Can't you boot each node off of a networked image and load the OS and whatever "work" into memory.
Every byte you pull off the hard drive is one less byte you have to suck over the network. Depending on the type of simulation they're running, a file cache might be a useful resource. Also, for some problems, you can use disk space to get around the 32-bit address space limitations of a PC (yes, this is something I worry about in my own [medical imaging] work). Then there's swap space (slow, but still useful when every megabyte counts). In short, a modest hard drive for each node is used for the exact same things it would be in any other networked PC.
In the end, we pulled it off and emerged successful on the project, and we were regarded almost as heroes in house. We are regarded as can-do people that can rise to a challenge, [...]
I'll bet that's what management used to say about the more experienced guys they let go just after you were hired too.
Sokal published intentionally grabled crap to expose the lack of post-modernist rigor. He knew it was incoherent, the publishers didn't and praised its 'logic'.
Also, I might add, in contrast to the Bogdanov's, Sokal owned up the the fakery right away and immediately published a second article (in a different journal) explaining why he did it and why his original article never should have been published in the first place.
Honesty matters. That's what makes Sokal one of the good guys.
The other highly suspcious aspect of this whole affair is that the work was never published online prior to being submitted to a journal.
Some background for non-physicists: thesedays the primary venue for publishing new works is the arXiv [arxiv.org]. Several hundred papers per day are uploaded here in various categories, and it is the de facto standard library of modern research in physics.
This is only true for certain fields of physics. For the areas I've worked in, articles are either seldom (liquid crystal physics) or almost never (diffuse optical tomography) published online first. Quantum gravity, cosmology, and the like are the exception not the rule in their heavy use of the pre-print servers.
PDF isn't a very good format either because Adobe controls the spec. It isn't open.
But it is well documented (and [beer] free!) and Adobe takes pains to make sure it remains backwards compatible. Nor do they have a pattern of arbitrarily changing the specification to break compatibility with competing products as some companies do. Ghostscript, for example, does a fine job reading and writing PDF files, in large part because of the information published by Adobe.
So why isn't there a reward system for cure research? Have people, the government, or better yet governments put up a sum of money to the person, or group of people that create a verifiable cure. Stick it into something guaranteed and let the pot grow. Would drug companies work to cure aids if there was a few trillion dollars at stake?
Oh please, you have to ask this? C'mon, repeat after me, "Private market good, government bad". It must be true, I hear it all the time on Slashdot!
Most patents are filed with the hope that they'll make the inventor money. Either through direct licensing or through (ugh) lawsuits later. So most folks regard the cost of a search and the legal costs of filing as a cost as doing business.
I'm not sure I buy that, especially when you get beyond the individual to the corporate and university realms. I think most patents are actually filed to prevent someone else from getting the patent first. Having to abandon a product because someone else re-discovered your idea and patented it first sucks, so companies patent everything (or try to) in order to protect their R&D investment. Potential licensing revenues, in my experience, have always been a secondary concern.
Raymond Serway's text, "Physics for Scientists and Engineers (with modern physics)" has EVERYTHING that a beginning physics student needs. All that you need to bring to the table is some basic math skills (you can even get by without calculus if necessary). This is (was?) the workhorse textbook for virtually every first-year "real" physics course that I've ever seen.
Halliday and Resnick (Halliday, Resnick, and Walker in the newer editions) is the other standard first-year textbook. Also very good, but it does require some basic knowledge of calculus.
The Dancing Wu Li Masters: Gary Zukav" [quanta-gaia.org] A book about the dynamics of new physics
without mathematics.
The Elegant Universe: Brian Greene" [wwnorton.com] Again, another new physics book with neat pics and no mathmaticas. Specific to Superstrings mostly.
A Brief History of Time: Stephen Hawking" [psyclops.com] A good book about allmost everything between classical physics and the physics of the last few years. I.E. Relativity, quantum mechanics etc.
These are books about physics, not physics books.
You cannot properly teach physics without at least some mathematics. Pop-physics books like "A Brief Histor of TIme" or "QED" present the concepts clearly to non-technical readers, but they do this by leaving out more than they include. To take quantum mechanics as an example, If you can't work out the energy levels of the hyrogen atom by solving the eigenvalue problem, then you don't really understand quantum mechanics. If you don't even know what an eighenvalue is, then you really don't understand quantum mechanics.
-JS
"Mathematics is the language of physics" --unknown
One of the things I *hated* about my high school science classes (and some of my college classes) was that everything we did had been done before.
Having been a physics TA, most students aren't really happy either when, after spending 2 or 3 hours in the lab, they have nothing to show for it because the experiment didn't work.
The point of the labs is (supposed to be) to reinforce concepts you learned earlier in class and to make them a little less abstract. HS and freshman-level labs are rigged to succeed so that the lab itself won't get in the way of what you're supposed to be learning.
Slashdot Mirror
Huh? NTP predates Linux by several years. And it was developed for Un*x, not specifically Linux.
-JS
Radio goes right through dust; light does not. That kills optical right there for deep-space applications.
-JS
Szilard (who was Hungarian and ended up in the U.S. by way of Britain) was the first to propse a nuclear chain reaction. Fermi, who was by then a U.S. citizen, did the necessary science to turn the idea into reality.
So now it comes down to do we go by country of residence, country of citizenship, or country of birth?
-JS
Depends where you work; someplace like investment banking and your bonus can easily be 150% of your base salary. This is the norm-the size of the bonus is directly related to how well the company as a whole did.
I'm sure all of senior management did something really exceptional this year and none of the drones did anything useful. I'm sure that's the real reason the CEO's bonus went up and the drones are getting the bobble heads.
-JS
Libertarian? Nader? And this got mod'ed insightful?!?
-JS
Probably depends on the field of research. Working in physics, I've never had that problem.
-JS
Anything related to counter-terrorism.
-JS
More than possible. I saw Sprited Away in the theater and it is most definitely white in the original print.
-JS
That's not a problem, it's an opportunity.
-JS
If you speak Old English, that is. It's a rather different language from even the English of Chaucer (ca. 1350).
-JS
You mean like the fast pluto fly-by (the only unvisited planet in the solar system and one we need to visit soon if we want to see it before the entire atmosphere re-freezes) which was never funded due to the ISS induced fund shortage? Actually, the entire unmanned budget and the earth sciences budget in particular has also taken a beating in the last decade as the ISS sucks up a progressively larger chunck of the annual allocation.
-JS
Unfortunately, not all supercomputers are created equal. For running something like Monte Carlo simulations where each thread of execution (i.e. CPU) proceeds more or less independantly of each other, Beowulf clusters are just the ticket. For other problems (e.g. molecular dynamics simulations with long-range interactions), you pretty much need a shared-memory system otherwise you nuke your network bandwith keeping the system state of every node in sync with every other node.
Real super-computers also have internal bandwidths that are orders of magnitude faster than your PCI bus (35 Gbyte/sec/CPU for a top-line Cray, even faster for in-cache access). They also include things like vector processors that just fly through repetative mathematical calculations (think loop unrolling, but done in hardware and big chunks of the loop done in a single clock cycle).
In short, yes a Cray is a multi-multi million dollar machine, but for many common scientific computing problems you are getting a level a performance for that money that no cluster has any hope of touching.
-JS
Every byte you pull off the hard drive is one less byte you have to suck over the network. Depending on the type of simulation they're running, a file cache might be a useful resource. Also, for some problems, you can use disk space to get around the 32-bit address space limitations of a PC (yes, this is something I worry about in my own [medical imaging] work). Then there's swap space (slow, but still useful when every megabyte counts). In short, a modest hard drive for each node is used for the exact same things it would be in any other networked PC.
-JS
I'll bet that's what management used to say about the more experienced guys they let go just after you were hired too.
-JS
Also, I might add, in contrast to the Bogdanov's, Sokal owned up the the fakery right away and immediately published a second article (in a different journal) explaining why he did it and why his original article never should have been published in the first place.
Honesty matters. That's what makes Sokal one of the good guys.
-JS
This is only true for certain fields of physics. For the areas I've worked in, articles are either seldom (liquid crystal physics) or almost never (diffuse optical tomography) published online first. Quantum gravity, cosmology, and the like are the exception not the rule in their heavy use of the pre-print servers.
-JS
But it is well documented (and [beer] free!) and Adobe takes pains to make sure it remains backwards compatible. Nor do they have a pattern of arbitrarily changing the specification to break compatibility with competing products as some companies do. Ghostscript, for example, does a fine job reading and writing PDF files, in large part because of the information published by Adobe.
-JS
Usually, the name given is "Generation Y". Not terribly creative, but it seems to have stuck.
-JS
Oh please, you have to ask this? C'mon, repeat after me, "Private market good, government bad". It must be true, I hear it all the time on Slashdot!
-JS
You're thinking of Rutherford, and that experiment did make the top ten list. Thompson discovered the electron (known pre-Thompson as cathode rays).
-JS
I'm not sure I buy that, especially when you get beyond the individual to the corporate and university realms. I think most patents are actually filed to prevent someone else from getting the patent first. Having to abandon a product because someone else re-discovered your idea and patented it first sucks, so companies patent everything (or try to) in order to protect their R&D investment. Potential licensing revenues, in my experience, have always been a secondary concern.
-JS
Addresses on the stack should not be executable in the first place. That's the real fix.
-JS
Halliday and Resnick (Halliday, Resnick, and Walker in the newer editions) is the other standard first-year textbook. Also very good, but it does require some basic knowledge of calculus.
-JS
These are books about physics, not physics books. You cannot properly teach physics without at least some mathematics. Pop-physics books like "A Brief Histor of TIme" or "QED" present the concepts clearly to non-technical readers, but they do this by leaving out more than they include. To take quantum mechanics as an example, If you can't work out the energy levels of the hyrogen atom by solving the eigenvalue problem, then you don't really understand quantum mechanics. If you don't even know what an eighenvalue is, then you really don't understand quantum mechanics.
-JS
"Mathematics is the language of physics" --unknown
Having been a physics TA, most students aren't really happy either when, after spending 2 or 3 hours in the lab, they have nothing to show for it because the experiment didn't work.
The point of the labs is (supposed to be) to reinforce concepts you learned earlier in class and to make them a little less abstract. HS and freshman-level labs are rigged to succeed so that the lab itself won't get in the way of what you're supposed to be learning.
-JS