My perception is that the posts that get modded high up are the ones which are: 1. Reasonably intelligent 2. Among the first, say 20-30 comments posted. According to this hypothesis, other people need not be dumber than the ones which consistently get high mod points, it's just they aren't obsessive-compulsive-/.-reloaders. So to reiterate my point, to get consistently high mod points: 1. Spend a lot of time hitting reload on the/. frontpage to check if there are any new stories. Or in the *NIX tradition, make a script to do it for you and alarm you when a new story has been published. 2. Post a reasonably informed opinion. 3. Do it quickly before your post drowns among all the other comments.
Yes, very true. As I see it, there are currently 3 high end cpu architectures with a future, sparc, ibm:s power and IA-64. Producing (and designing) these chips gets more expensive all the time, and the upcoming extreme ultraviolet lithography is again going to cost even more. So barring any big surprises, these 3 companies are those which I'd bet on to have the funds to play the game till the end of the current silicon technology. Hopefully whatever comes after silicon will have a cheaper entry-to-market so there can be lots of new innovative companies driving innovation and competition forward. And I think it might be very feasible. Look at all the nanotechnology startups, research on molecular transistors, self-assembling systems and of course quantum computing. Very interresting stuff, albeit it will probably be >10 years before anything like that gets out of the research labs.
Regarding supercomputing, you are quite correct in that there is little incentive to develop new vector supercomputers, taking into account increasing costs of design and fabbing the cpu:s. Cray:s upcoming SV2 seems to be an exception, though. Also keep in mind that most supercomputers, like the Cray T3E and the IBM SP, are just commodity cpu:s connected with a bad-ass proprietary interconnect. So essentially, they are just beowulf clusters with a superfast and low-latency interconnect. They are programmed through MPI, the same thing used for beowulf.
While there are few scientifically interesting applications which require essentially no communication between the nodes, like seti@home and distributed.net (if you want to do something scientifically useful with your spare cycles, check out folding@home), most get by with a quite modest amount of communication between the nodes.
What I personally see as interresting in the supercomputing area in the next few years is not the maximum flops of the DOE:s newest lets-simulate-nuclear-explosions-big-ass-fastest-s upercomputer-in-the-world-thingy, but that the increasing affordability of commodity clustering will allow projects with somewhat more down-to-earth budjets to run some really interresting simulations. Also commodity clustering is improving. Things like Infiniband and whatever they are called will probably offer higher bandwidth/lower latency than the current Myrinet / Gigabit ethernet stuff. Also the improved SMP scalability of Linux 2.4 will perhaps make it economically feasible to have, say, 4-cpu nodes in the cluster instead of the current practice of 1-cpu nodes. The downside of this is of course that applications must be able to take advantage of this, either through directly using a combination of MPI and threads/SHM/whatever or through using frameworks like Petsc or POOMA.
As a final note, remember that the tools (i.e. supercomputers) and the way we solve the problems at hand are not developed in a vacuum. The numerical methods that are developed today certainly place a greater emphasis on solving the problem via scaling through clustering than, say, 20 years ago when vector supercomputers where in the same price class as the parallell ones.
Uh oh... Lets do a quick back-of-the-envelope-calculation. Steel, which tanks largely are made of, has a specific heat of about 450 J/(kgK). A tank weights about 60 metric tonnes. To make the crew at least slightly uncomfortable requires say, increasing the temperature by 100K. So to heat the tank 100K we need 2.8GJ energy. If you want all this energy delivered in a reasonably short time period, say for simplicity 1 second, you'd need a laser capable of producing almost 3GW for that 1 second. To put this into perspective, a rather large nuclear power station produces about 1GW of electricity. In conclusion, with todays technology, forget about heating the tank with a laser.
There's a saying that the military is always preparing for the last war. And yes, the logic is flawed. I bet that most (if not all) funding proposals the military is writing these days contains the words "september 11" and/or "terrorism". The military is just riding on public opinion. Not that I blame them. Mentioning "september 11" in the funding proposal is certainly going to increase the chance that your pet project gets funding.
Nice theory, but it will be a *looooong* time before lasers of any practical size will be able to punch through a tank... you know, tanks tend to be quite thick skinned, as opposed to, say, missiles and aircraft.
Well said, I was thinking of the same. Hierarchial management has persisted for thousands of years and bazillions of management books. In fact, most management strategies are rather subtle modifications on the standard hierarchial model. I.e. what is the optimal span of control for some organization, how to communicate between the subtrees (formalized in the matrix strategy) etc. Not that I'm a kernel developer, but I would suggest that Linus would designate say 10 people as his official subsystem maintainers, announce publicly that all patches must go through them, and to be really sure automatically bounce every patch he gets except from his 10 trusted maintainers (should be a really simple procmail script). Actually, this was also sort of suggested on lkml, and Linus own response sort of reflected a line of thought like this IMHO, even if he explicitly stated that it wasn't/shouldn't be a "general at the top"-kind of strategy.
You're partly right. There seems to be about 1.2e9 muslims and 2e9 christians, of which about 1e9 are catholics. So muslims outnumber the catholics, but not all of christianity.
Well I'm terribly sorry if you understood it that way. What I meant was that "real" scientists won't waste time on disproving this. Rather, the burden of proof lies on the inventor of this machine. Until it's proved to work (say, by a published paper detailing the device and the theory behind it), nobody is going to take it seriously. And no, there is no great conspiracy by the government/scientific community/big corporations/whatever. It's just that time is limited. Real scientists have real work to do.
No, nobody will probably bother to prove it's a hoax. You know, most scientists have real work to do. And the world is abound with all kinds of crackpot theories/devices claiming whatever fantastic properties. Rather, it is the job of the inventor to prove that the thing works. To be taken seriously, it means to publish a paper in a primary journal. Of course, any self-respecting editor would probably reject said paper after a rather brief look. And that is the reason why crackpot theories remain crackpot theories, and do not become part of the "established knowledge".
Puh-leeze... That page is, well, utter drivel. Basically a bunch of unsubstantiated claims and of course the usual rambling about conspiracy theories. Until these theories get published in primary journals I prefer not to waste my time on them as they most certainly are just the workings of some daydreaming crackpot. So why are primary journals (i.e. journals like "Physical Review" etc.) so important? Well, for one thing, they usually have very high standards regarding what gets published. And scientists actually read them, in contrast to the crackpot theories which abound on the net. As an example look at cold fusion.
1. Pons & Fleichmann publish their article. I don't remember in which journal it was, but probably the main reason it got so much publicity was that it actually got published in a primary journal as it means that the manuscript passed the peer-review. Most wacko theories don't get this far, as no self-respecting journal will print the kind of drivel they consist of. The fact that P&F got published was probably the result of a rather huge mistake in the review process.
2. Because the article would have been very important had it been true -> lots of publicity
3. Noone was able to reproduce the experiments
4. Closer investigation revealed that the experimental procedure used by P&F was seriously flawed.
5. Claims refuted. End of story. The end result was certainly a rather big status drop for the journal which published the article.
Actually, according to the the README provided with the kernels,/usr/src/linux is _not recommended_, as (depending on distro) you may mess up the header files used by the C library. Another point is that I think/usr/src/linux should be owned by root.src or something like that. So put your tree somewhere else where it doesn't mess up the C library nor your package management system and where you don't need root access./usr/local, your home directory and with some reservation/opt are good choices (I have my kernel related stuff under/opt/kernel).
Uh oh... To correct myself, the He atom is no longer analytically solvable, only the H atom is. Of course depending on whether you are calculating electron orbitals or the nuclear structure. I.e. for electron structure, a proton with only one electron orbiting it, and if you're calculating nuclear structure the deuteron nucleus (1 proton + 1 neutron) are the most complicated things with analytical solutions. The rest is more or less daring approximations, and number crunching...:)
Do you have a problem with the concept of 'width of a single atom' or what are you saying? Or perhaps radius would be a better word? Of course, the radius of an atom is not something which is very precisely defined due to quantum mechanics. IIRC, a common definition of the radius of an atom is the value r for which (formula in LaTex) \int ^{r}_{0}dr\int d\Omega \Psi ^{\dagger }\Psi >0.5 where \Psi is the normalized wavefunction for the atom. Of course, determining \Psi for anything more complicated than a He atom is a many-body problem with no analytical solution (supercomputers here we come!).
Well the 10GHz is the speed of the processor they think they are able to produce with EUV. It has nothing to do with the wavelength used in the etching process. The 10GHz just means that there is a clock thingy inside the processor which says 'tick' 10*10^9 times per second. Although this _does_ create a problem, as you hinted at. As the speed of light in vacuum is constant, it means that the clock signal will only propagate a certain length before the next 'tick'. With a 10GHz clock, the signal will propagate a maximum of about 3cm before the next 'tick' (high school physics, remember; c=lf, where c=speed of wave propagation, l=wavelength, f=frequency). Of course in reality it will propagate even less than 3cm (which is the wavelength you get if c=speed of light in vacuum=3*10^8m/s), as the speed of light inside the chip is somewhat slower than in vacuum. This will mean that the parts of the chip that are further away from the clock will be somewhat out of sync with the parts that are close to the clock. This is something chip designers certainly have to take into account. I'm not sure, maybe it is already an issue today?
Actually, your feeling of safety is damn important, in many cases more important than the actual level of security. For a (long) explanation, see for example Bruce Schneier's Crypto-Gram Newsletter September 30, 2001, and the links therein, especially the ones about airline security.
Re:General Jon Katz
on
The Drone War
·
· Score: 2, Insightful
Exactly what I thought. But hey, when Katz (you know, big ego & clueless) decides to write an yet another "insightful" article on a subject about which he, again, has no clue whatsoever, it is convenient to forget about things that don't really fit into the story (i.e. robots did all the work). And of course foreign people are usually thought of as a lot more expendable than your own, so a reader which is as clueless as the writer might actually oversee the omission of their deaths.
I'm sorry if anybody sees this as trolling, but I seriously think Katz should get a kick in the butt for being a well known (and respected?) journalist and writing drivel like this, totally ignoring the high prize the northern alliance has paid.
I might be wrong of course, but I thought WWI was basically a big "let's slug it out" thing. The classical picture of WWI is "wasting thousands of lives to conquer a few hectares of mud on the frontline, then get pushed back a little later because neither side understood how to exploit an eventual breach in the lines". Remember that blitzkrieg, i.e. the doctrine to rapidly exploit breaches with mobile units was not developed until the 1930s (by general. von. Clausewitz, if my memory serves me right) and successfully used by the Germans in early WWII. Of course, later in WWII essentially all the involved countries had understood the value of strategic mobility and adapted their doctrines for it. To go back to WWI one can of course argue that blitzkrieg was not possible because the tanks at that time were not really good enough. So with the technology at that time, attacking was a rather expensive concept (not a good idea to assault entrenched machine guns with infantry only).
Ah, the obligatory NR bash. Seems that everytime NR is mentioned, someone posts a link to this page...:) Well, I do believe that the criticism is correct in many ways but it misses the point. NR is not meant to be the authorative source on numerical algorithms. What it is is an excellent reference for the "normal" programmer when he/she needs some numerical algorithm. NR algorithms are certainly better than what most people without extensive experience in the field would come up on themselves. Sure, if your primary occupation is coding numerical stuff, you should read the primary journals on numerics and NR will not be necessary for you. And even for the "normal" programmer, downloading something from netlib is probably a better choice than using the NR algorithms directly. Still, NR provides simple to understand explanations of the algorithims, without requiring you to wade through dozens of monographs to understand what you're doing.
That's an interesting thought, but I know of no evidence that it might be true. The passage of time is not a function of the speed of light. Observed passage of time is a function of the relative speed of the observer to the observee, but that's an entirely different thing.
I think you're confusing fundamental stuff with how we choose to measure things. According to general relativity, the speed of light in vacuum is constant. End of discussion. Everything else follows from that. Now, if the general theory of relativity is wrong, it's entirely another matter. But so far general relativity has proven to be correct.
Slashdot Mirror
My perception is that the posts that get modded high up are the ones which are: /. frontpage to check if there are any new stories. Or in the *NIX tradition, make a script to do it for you and alarm you when a new story has been published.
1. Reasonably intelligent
2. Among the first, say 20-30 comments posted.
According to this hypothesis, other people need not be dumber than the ones which consistently get high mod points, it's just they aren't obsessive-compulsive-/.-reloaders. So to reiterate my point, to get consistently high mod points:
1. Spend a lot of time hitting reload on the
2. Post a reasonably informed opinion.
3. Do it quickly before your post drowns among all the other comments.
I think the original poster meant the Motorola 88k...
Yes, very true. As I see it, there are currently 3 high end cpu architectures with a future, sparc, ibm:s power and IA-64. Producing (and designing) these chips gets more expensive all the time, and the upcoming extreme ultraviolet lithography is again going to cost even more. So barring any big surprises, these 3 companies are those which I'd bet on to have the funds to play the game till the end of the current silicon technology. Hopefully whatever comes after silicon will have a cheaper entry-to-market so there can be lots of new innovative companies driving innovation and competition forward. And I think it might be very feasible. Look at all the nanotechnology startups, research on molecular transistors, self-assembling systems and of course quantum computing. Very interresting stuff, albeit it will probably be >10 years before anything like that gets out of the research labs.
s upercomputer-in-the-world-thingy, but that the increasing affordability of commodity clustering will allow projects with somewhat more down-to-earth budjets to run some really interresting simulations. Also commodity clustering is improving. Things like Infiniband and whatever they are called will probably offer higher bandwidth/lower latency than the current Myrinet / Gigabit ethernet stuff. Also the improved SMP scalability of Linux 2.4 will perhaps make it economically feasible to have, say, 4-cpu nodes in the cluster instead of the current practice of 1-cpu nodes. The downside of this is of course that applications must be able to take advantage of this, either through directly using a combination of MPI and threads/SHM/whatever or through using frameworks like Petsc or POOMA.
Regarding supercomputing, you are quite correct in that there is little incentive to develop new vector supercomputers, taking into account increasing costs of design and fabbing the cpu:s. Cray:s upcoming SV2 seems to be an exception, though. Also keep in mind that most supercomputers, like the Cray T3E and the IBM SP, are just commodity cpu:s connected with a bad-ass proprietary interconnect. So essentially, they are just beowulf clusters with a superfast and low-latency interconnect. They are programmed through MPI, the same thing used for beowulf.
While there are few scientifically interesting applications which require essentially no communication between the nodes, like seti@home and distributed.net (if you want to do something scientifically useful with your spare cycles, check out folding@home), most get by with a quite modest amount of communication between the nodes.
What I personally see as interresting in the supercomputing area in the next few years is not the maximum flops of the DOE:s newest lets-simulate-nuclear-explosions-big-ass-fastest-
As a final note, remember that the tools (i.e. supercomputers) and the way we solve the problems at hand are not developed in a vacuum. The numerical methods that are developed today certainly place a greater emphasis on solving the problem via scaling through clustering than, say, 20 years ago when vector supercomputers where in the same price class as the parallell ones.
Uh oh... Lets do a quick back-of-the-envelope-calculation. Steel, which tanks largely are made of, has a specific heat of about 450 J/(kgK). A tank weights about 60 metric tonnes. To make the crew at least slightly uncomfortable requires say, increasing the temperature by 100K. So to heat the tank 100K we need 2.8GJ energy. If you want all this energy delivered in a reasonably short time period, say for simplicity 1 second, you'd need a laser capable of producing almost 3GW for that 1 second. To put this into perspective, a rather large nuclear power station produces about 1GW of electricity. In conclusion, with todays technology, forget about heating the tank with a laser.
Hey, haven't you seen The Terminator? :)
There's a saying that the military is always preparing for the last war. And yes, the logic is flawed. I bet that most (if not all) funding proposals the military is writing these days contains the words "september 11" and/or "terrorism". The military is just riding on public opinion. Not that I blame them. Mentioning "september 11" in the funding proposal is certainly going to increase the chance that your pet project gets funding.
Nice theory, but it will be a *looooong* time before lasers of any practical size will be able to punch through a tank... you know, tanks tend to be quite thick skinned, as opposed to, say, missiles and aircraft.
Protocol 2 is a rather major improvement, imho. Better security, and sftp is very convenient. And well, testing currently has 3.0.2p1 so I'm happy :)
Problem is, antipersonnel mines cost like 1 cent each. For each of these fancy things you buy, I can buy bazillions of landmines. Sad but true...
Well said, I was thinking of the same. Hierarchial management has persisted for thousands of years and bazillions of management books. In fact, most management strategies are rather subtle modifications on the standard hierarchial model. I.e. what is the optimal span of control for some organization, how to communicate between the subtrees (formalized in the matrix strategy) etc. Not that I'm a kernel developer, but I would suggest that Linus would designate say 10 people as his official subsystem maintainers, announce publicly that all patches must go through them, and to be really sure automatically bounce every patch he gets except from his 10 trusted maintainers (should be a really simple procmail script). Actually, this was also sort of suggested on lkml, and Linus own response sort of reflected a line of thought like this IMHO, even if he explicitly stated that it wasn't/shouldn't be a "general at the top"-kind of strategy.
You're partly right. There seems to be about 1.2e9 muslims and 2e9 christians, of which about 1e9 are catholics. So muslims outnumber the catholics, but not all of christianity.
Oh, I'd have guessed it would have looked like, say, this picture... :)
Well I'm terribly sorry if you understood it that way. What I meant was that "real" scientists won't waste time on disproving this. Rather, the burden of proof lies on the inventor of this machine. Until it's proved to work (say, by a published paper detailing the device and the theory behind it), nobody is going to take it seriously. And no, there is no great conspiracy by the government/scientific community/big corporations/whatever. It's just that time is limited. Real scientists have real work to do.
No, nobody will probably bother to prove it's a hoax. You know, most scientists have real work to do. And the world is abound with all kinds of crackpot theories/devices claiming whatever fantastic properties. Rather, it is the job of the inventor to prove that the thing works. To be taken seriously, it means to publish a paper in a primary journal. Of course, any self-respecting editor would probably reject said paper after a rather brief look. And that is the reason why crackpot theories remain crackpot theories, and do not become part of the "established knowledge".
Puh-leeze... That page is, well, utter drivel. Basically a bunch of unsubstantiated claims and of course the usual rambling about conspiracy theories. Until these theories get published in primary journals I prefer not to waste my time on them as they most certainly are just the workings of some daydreaming crackpot. So why are primary journals (i.e. journals like "Physical Review" etc.) so important? Well, for one thing, they usually have very high standards regarding what gets published. And scientists actually read them, in contrast to the crackpot theories which abound on the net. As an example look at cold fusion.
1. Pons & Fleichmann publish their article. I don't remember in which journal it was, but probably the main reason it got so much publicity was that it actually got published in a primary journal as it means that the manuscript passed the peer-review. Most wacko theories don't get this far, as no self-respecting journal will print the kind of drivel they consist of. The fact that P&F got published was probably the result of a rather huge mistake in the review process.
2. Because the article would have been very important had it been true -> lots of publicity
3. Noone was able to reproduce the experiments
4. Closer investigation revealed that the experimental procedure used by P&F was seriously flawed.
5. Claims refuted. End of story. The end result was certainly a rather big status drop for the journal which published the article.
Actually, according to the the README provided with the kernels, /usr/src/linux is _not recommended_, as (depending on distro) you may mess up the header files used by the C library. Another point is that I think /usr/src/linux should be owned by root.src or something like that. So put your tree somewhere else where it doesn't mess up the C library nor your package management system and where you don't need root access. /usr/local, your home directory and with some reservation /opt are good choices (I have my kernel related stuff under /opt/kernel).
Uh oh... To correct myself, the He atom is no longer analytically solvable, only the H atom is. Of course depending on whether you are calculating electron orbitals or the nuclear structure. I.e. for electron structure, a proton with only one electron orbiting it, and if you're calculating nuclear structure the deuteron nucleus (1 proton + 1 neutron) are the most complicated things with analytical solutions. The rest is more or less daring approximations, and number crunching...:)
Do you have a problem with the concept of 'width of a single atom' or what are you saying? Or perhaps radius would be a better word? Of course, the radius of an atom is not something which is very precisely defined due to quantum mechanics. IIRC, a common definition of the radius of an atom is the value r for which (formula in LaTex) \int ^{r}_{0}dr\int d\Omega \Psi ^{\dagger }\Psi >0.5 where \Psi is the normalized wavefunction for the atom. Of course, determining \Psi for anything more complicated than a He atom is a many-body problem with no analytical solution (supercomputers here we come!).
Well the 10GHz is the speed of the processor they think they are able to produce with EUV. It has nothing to do with the wavelength used in the etching process. The 10GHz just means that there is a clock thingy inside the processor which says 'tick' 10*10^9 times per second. Although this _does_ create a problem, as you hinted at. As the speed of light in vacuum is constant, it means that the clock signal will only propagate a certain length before the next 'tick'. With a 10GHz clock, the signal will propagate a maximum of about 3cm before the next 'tick' (high school physics, remember; c=lf, where c=speed of wave propagation, l=wavelength, f=frequency). Of course in reality it will propagate even less than 3cm (which is the wavelength you get if c=speed of light in vacuum=3*10^8m/s), as the speed of light inside the chip is somewhat slower than in vacuum. This will mean that the parts of the chip that are further away from the clock will be somewhat out of sync with the parts that are close to the clock. This is something chip designers certainly have to take into account. I'm not sure, maybe it is already an issue today?
Actually, your feeling of safety is damn important, in many cases more important than the actual level of security. For a (long) explanation, see for example Bruce Schneier's Crypto-Gram Newsletter September 30, 2001, and the links therein, especially the ones about airline security.
Exactly what I thought. But hey, when Katz (you know, big ego & clueless) decides to write an yet another "insightful" article on a subject about which he, again, has no clue whatsoever, it is convenient to forget about things that don't really fit into the story (i.e. robots did all the work). And of course foreign people are usually thought of as a lot more expendable than your own, so a reader which is as clueless as the writer might actually oversee the omission of their deaths.
I'm sorry if anybody sees this as trolling, but I seriously think Katz should get a kick in the butt for being a well known (and respected?) journalist and writing drivel like this, totally ignoring the high prize the northern alliance has paid.
I might be wrong of course, but I thought WWI was basically a big "let's slug it out" thing. The classical picture of WWI is "wasting thousands of lives to conquer a few hectares of mud on the frontline, then get pushed back a little later because neither side understood how to exploit an eventual breach in the lines". Remember that blitzkrieg, i.e. the doctrine to rapidly exploit breaches with mobile units was not developed until the 1930s (by general. von. Clausewitz, if my memory serves me right) and successfully used by the Germans in early WWII. Of course, later in WWII essentially all the involved countries had understood the value of strategic mobility and adapted their doctrines for it. To go back to WWI one can of course argue that blitzkrieg was not possible because the tanks at that time were not really good enough. So with the technology at that time, attacking was a rather expensive concept (not a good idea to assault entrenched machine guns with infantry only).
Ah, the obligatory NR bash. Seems that everytime NR is mentioned, someone posts a link to this page...:) Well, I do believe that the criticism is correct in many ways but it misses the point. NR is not meant to be the authorative source on numerical algorithms. What it is is an excellent reference for the "normal" programmer when he/she needs some numerical algorithm. NR algorithms are certainly better than what most people without extensive experience in the field would come up on themselves. Sure, if your primary occupation is coding numerical stuff, you should read the primary journals on numerics and NR will not be necessary for you. And even for the "normal" programmer, downloading something from netlib is probably a better choice than using the NR algorithms directly. Still, NR provides simple to understand explanations of the algorithims, without requiring you to wade through dozens of monographs to understand what you're doing.
Well, maybe it was too close to "Wolfram Research", the company behind Mathematica.
That's an interesting thought, but I know of no evidence that it might be true. The passage of time is not a function of the speed of light. Observed passage of time is a function of the relative speed of the observer to the observee, but that's an entirely different thing.
I think you're confusing fundamental stuff with how we choose to measure things. According to general relativity, the speed of light in vacuum is constant. End of discussion. Everything else follows from that. Now, if the general theory of relativity is wrong, it's entirely another matter. But so far general relativity has proven to be correct.