Perhaps trolling for a lawsuit is the point of Browns actions? If Linus were to spend a significant part of his time during a couple of years in court instead of making the kernel better, MS and whoever else stands behind adti would benefit a lot more than they would lose. I'm pretty sure that MS considers Brown to be a pretty expendable asset. So what if Brown would ultimately lose the case? Even if Linus were to get millions in compensation, that's pocket money for MS.
In the HPC world, 1000 cpu jobs are not that uncommon. Of course, these apps run on nodes having relatively few cpu:s and communicate using message passing, so various OS imposed scalability limits doesn't really affect them.
Also, the message passing paradigm tends to lead to programs that replicate as much state as practical on each node, so there is usually less time spent waiting on locks compared to typical multithreaded designs.
Well how about this. Where I live, we have a thing called diesel tax on all diesel vehicles (for personal cars it's currently about EUR 500 / year).
About 10 years ago they suddenly doubled this diesel tax, with no prior warning. To add insult to injury, the politicians explained that the tax actually wasn't increased as the amount on the bill was still the same. Instead they just sent you two bills per year instead of one!
As you hinted below, gamers. And immature people who feel that their worth as human beings is measured by their computers. In short, the fanboys who are the target market for all these quadrillion hardware review sites.
I also bought a radeon 9200 for much the same reasons you explained. But actually the significantly faster 8500 is equally well supported by the same drivers as the 9200. You see, the 9200 is a crippled el cheapo version of the 8500. Problem is, you can't find the 8500 anywhere anymore, and it has a fan while my 9200 gets by with just a heatsink.
That's strange, because 4.3 works just fine for me with my radeon 9200. You just have to set the chipid in the config file because 4.3 doesn't recognize the 9200. The 9200 is the same card as the 9000 (except for agp8x instead of agp4x) so it works well as a 9000.
We probably won't know the real number until (like the Soviet Union) after China is liberated from Communism at some future date.
Well, Rummel which you referred to estimates that about 35 million have been excuted in communist China so far. Additionally, the "black book of communism" estimates that IIRC about 60 million people starved to death during Mao:s "great leap forward" (no natural disaster or such, just plain mismanagement of resources in the name of communism).
The "swap=2x RAM" thing is obsolete admin trivia that simply refuses the die. It comes from the days when physical RAM was mapped into swap to simplify the swapping algorithm. If you didn't have at least a 1:1 correspondence between RAM and swap performance would suffer immensly. Starting with Linux 2.4 and up this is simply no longer true, there is no benefit from using excessively large swap partitions. Same goes for Sun OS and the BSDs these days.
Actually, IIRC, the early 2.4 kernels where the only kernels in Linux history that used this sort of swapping algorithm. When the VM was changed in 2.4.10 this behaviour changed back to the old behaviour.
As to whether 2x RAM is a good rule or not, for most uses it's perhaps a bit excessive, but OTOH with harddrives selling for about EUR 0.6/ GB, who cares?
Re:Where many people miss the point...
on
Is Swap Necessary?
·
· Score: 4, Informative
It'd be nice if they had a similar control for filesystem cache.
You're missing the point. That's exactly what the swappiness setting does, indirectly. If you avoid swapping (swappines = 0), the system has less memory left over for filesystem cache. OTOH, if you set swappiness=100, the system has a lot more memory to use for file cache.
The system always tries to use all available memory, and that's a good thing. The question is whether to use extra memory for file cache or for keeping pages in memory.
Of course you can do it in a language like C, but then you lose the ability to express yourself cleanly through code like 'c=a+2*b'.
Well, you could use daxpy from blas I guess..
It seems that after the initial excitement over template metaprogramming some years ago, many people found it too hard to get right, too long compile times etc., so they turned back to trusty old Fortran 95.
But they could tell everyone they were, just have a screen saver that pegs the CPU, tells you that you've rendered X frames, and displays a cool screensaver from the movie!:)
Instead of just looping the CPU, they could covertly run their own version of the seti@home client and crank up some time on the pixar seti@home team..:)
the film studios I'm sure have crazy fiber/multi-gigabit interconnects within their rendering farms.
While the amount of data to move around probably is too much for dialup, gigabit ethernet is certainly fast enough, and dirt cheap as it's integrated on motherboards. If you look at the top500 list, you see that weta digital (the company which did the CG for lord of the rings IIRC) has a couple of clusters on the list, and they have gig ethernet.
Basically, while rendering is cpu-intensive it's not latency sensitive, so there's no point in blowing a huge amount of cash on a high end cluster interconnect.
Russia's had a military force, but the only thing holding it up was the desire of the people to reap the rewards of becoming communists party members. Obviously there was some patriotism, but it was far from a deciding factor in anything but defense.
That's a question of motivation. Of course any soldier, conscript or professional, fights harder when he's defending his homeland than when he's on some empire building mission. Do you think that morale in the EU armies would be high if we would embark on some "follow in Napoleons and Hitlers footsteps and storm the gates of Moscow"-project? This being the third time, perhaps we'd remember to bring winter clothing this time, though.;-)
But the question of motivation regarding attack vs. defense won't make such a huge impact in the feasibility of any plan. Until the early 80:s, the Soviets had hardware that was about equivalent to the west, and they had overwhelming numerical superiority in Europe. I think they could have succeeded (discounting that any such attack would have provoked a nuclear exhange between the US and the USSR).
There military is not much different than it was previously, except that it is lacking in willing manpower.
Yes it is different. They have downsized a lot, and still a large part of their "active" arsenal is nothing but scrap metal. The few things that work are still 1970 level technology.
If Russia were to revert to a military state, it's quite likely that you'd find that they have a rather fearsome military again.
Not with their current economy. Spending 30% of GDP on the military is not a usable long term scenario, as it will crash the economy. Given that the GDP of Russia is currently about equal to Sweden, the EU is more than able to counter any increase in military spending by Russia, if the EU feels there is a need. By the time the economy of Russia is able to provide a military threat to the EU, we can only hope that the people of Russia will choose prosperity (via tight economic integration with the EU and the rest of the world) instead of yet another world domination scheme. Given the horrors the Russian populace has gone through during the past few centuries, they certainly deserve peace and prosperity as much as anyone else, if not more.
No, my point is that the decisions by EU countries to cut back on their militaries means that they will have no resources to fight any war that lands on their doorstep.
Even after these cutbacks, the EU is more than able to fight any neighborhood war. As I explained above, the Russian military is but a shadow of its former self, they won't be any threat until the Russian economy is on par with the EU countries (which at the very least will take decades). Ukraine, or any other small-scale dictatorship, is small and poor, so they are even less of a threat. The mongolian hordes from China will only reach europe after wading through a hailstorm of Russian nukes, so that's not a really realistic scenario either.
The EU doesn't have any imperial ambitions, as opposed to the current US administration, so there is no need for a expeditionary force capable of conquering some banana republic on the other side of the world either.
With no realistic major military threat in sight, it makes sense to spend less on the military and more on say, growing the economy. E.g. if military spending is reduced from 3 % to 2 % of GDP and the 1 % left over is used for growing the economy, it doesn't take long until the 2 % spending matches the former 3 %.
The US has tremendous experience with Guerilla warfare and at least has semi-effective countermeasures.
No. There is no effective military countermeasure against guerilla warfare. What you can do is try to win the support of the populace, thus enabling your police to work more efficiently, and using armored convoys to give some defense against ambushes. OTOH, while burning supply trucks look bad on TV, they don't really pose a major military threat either. Besides, as the EU isn't planning on conquering other countries, there is little chance of having to deal with guerilla warfare in the first place.
That's complete and utter bollocks. Up until the early 1980:s they were perfectly capable of steamrolling europe. The only thing holding them back was NATO nukes as well as the questionable benefit of attacking in the first place, the world domination plans slowly dying out after Stalin died in 1953.
The thing that did them in was that the communist economic system simply sucked. Thanks to the market economy, the west was able to improve their military hardware faster than the Soviets. When warfare became increasingly computerized in the 1980:s they couldn't keep up any longer.
True. But many of them will argue that their current military forces are sufficient to deal with such "primitive" militaries. Too bad they have no experience with Guerilla warfare...
Are you suggesting that blowing zillions of taxpayer money on star wars weaponry will help the EU fight a guerilla war then?
High tech helps very little in a guerilla war, as the Americans are finding out in Iraq at the moment.
My thinking is that it will be a typical Microsoft product. It will be big, bloated, expensive, but easy enough to use that someone without an advanced CS degree may be able to create a supercomputer application after taking some certification courses. Another McEngineer.
Well, most supercomputer applications are made by people without advanced CS degrees. Rather they are scientists, experts in their own field. That, of course, doesn't mean that writing writing numerical code with Fortran using MPI is easy. Far from it. But I don't think MS can do much about making this easier. Writing numerical software is a very different endeavour than banging out forms stuff with VB.
On the contrary HPC is all about custom/very specialised hardware running a very specialised application built for one perpose alone.
Why don't you take a look at the top500 list instead of guessing? Yes there are a small amount of supercomputer only type architectures (vector processors mostly, like the NEC SX and Cray X1), but most are off the shelf RISC, IA-64 or x86 things. For example, 5 of the top 10 computers are either x86 or IA-64, i.e. in theory Windows could run them.
I find it naturally that MS tries its luck in the HPC world, but windows surely does not fit the bill.
Slashdot Mirror
Umm, doesn't webmin already provide this and much more?
Plus, if the central server crashes, you loose a group if workstations rather than one workstation.
Which, in practice, is what happens when you lose your NFS/CIFS server as well. So what's the difference?
Actually, there's another study that supports this claim, even when talking about chicken eggs:
W.N. Thurman, M.E. Fisher, "Chickens, Eggs, and Causality, or
Which Came First?", American Journal of Agricultural Economics,
237-237, 1988.
Perhaps trolling for a lawsuit is the point of Browns actions? If Linus were to spend a significant part of his time during a couple of years in court instead of making the kernel better, MS and whoever else stands behind adti would benefit a lot more than they would lose. I'm pretty sure that MS considers Brown to be a pretty expendable asset. So what if Brown would ultimately lose the case? Even if Linus were to get millions in compensation, that's pocket money for MS.
'Organically grown' plants. No shit?
Actually, I think most organic farms use lots of shit for fertilizer.
In the HPC world, 1000 cpu jobs are not that uncommon. Of course, these apps run on nodes having relatively few cpu:s and communicate using message passing, so various OS imposed scalability limits doesn't really affect them.
Also, the message passing paradigm tends to lead to programs that replicate as much state as practical on each node, so there is usually less time spent waiting on locks compared to typical multithreaded designs.
What really surprises me is that "traditional" tech universities don't hold the top spots.
Perhaps because they rather spend their money on teaching instead of all kinds of frivolous stuff.
It was kind of a kick in the teeth.
Well how about this. Where I live, we have a thing called diesel tax on all diesel vehicles (for personal cars it's currently about EUR 500 / year).
About 10 years ago they suddenly doubled this diesel tax, with no prior warning. To add insult to injury, the politicians explained that the tax actually wasn't increased as the amount on the bill was still the same. Instead they just sent you two bills per year instead of one!
It's probably a better replacement for the fireplace, considering the heat that kind of thing puts out.
Who still does this?
As you hinted below, gamers. And immature people who feel that their worth as human beings is measured by their computers. In short, the fanboys who are the target market for all these quadrillion hardware review sites.
Huh? How on earth could you draw that conclusion from what I wrote?
I said that for RENDER FARMS, gig ethernet is fast enough, and dirt cheap compared to "real" cluster interconnects such as quadrics or myrinet.
Yup. Good idea. Too bad Bill Clinton shut it down.
I also bought a radeon 9200 for much the same reasons you explained. But actually the significantly faster 8500 is equally well supported by the same drivers as the 9200. You see, the 9200 is a crippled el cheapo version of the 8500. Problem is, you can't find the 8500 anywhere anymore, and it has a fan while my 9200 gets by with just a heatsink.
That's strange, because 4.3 works just fine for me with my radeon 9200. You just have to set the chipid in the config file because 4.3 doesn't recognize the 9200. The 9200 is the same card as the 9000 (except for agp8x instead of agp4x) so it works well as a 9000.
We probably won't know the real number until (like the Soviet Union) after China is liberated from Communism at some future date.
Well, Rummel which you referred to estimates that about 35 million have been excuted in communist China so far. Additionally, the "black book of communism" estimates that IIRC about 60 million people starved to death during Mao:s "great leap forward" (no natural disaster or such, just plain mismanagement of resources in the name of communism).
The "swap=2x RAM" thing is obsolete admin trivia that simply refuses the die. It comes from the days when physical RAM was mapped into swap to simplify the swapping algorithm. If you didn't have at least a 1:1 correspondence between RAM and swap performance would suffer immensly. Starting with Linux 2.4 and up this is simply no longer true, there is no benefit from using excessively large swap partitions. Same goes for Sun OS and the BSDs these days.
Actually, IIRC, the early 2.4 kernels where the only kernels in Linux history that used this sort of swapping algorithm. When the VM was changed in 2.4.10 this behaviour changed back to the old behaviour.
As to whether 2x RAM is a good rule or not, for most uses it's perhaps a bit excessive, but OTOH with harddrives selling for about EUR 0.6/ GB, who cares?
It'd be nice if they had a similar control for filesystem cache.
You're missing the point. That's exactly what the swappiness setting does, indirectly. If you avoid swapping (swappines = 0), the system has less memory left over for filesystem cache. OTOH, if you set swappiness=100, the system has a lot more memory to use for file cache.
The system always tries to use all available memory, and that's a good thing. The question is whether to use extra memory for file cache or for keeping pages in memory.
Of course you can do it in a language like C, but then you lose the ability to express yourself cleanly through code like 'c=a+2*b'.
Well, you could use daxpy from blas I guess..
It seems that after the initial excitement over template metaprogramming some years ago, many people found it too hard to get right, too long compile times etc., so they turned back to trusty old Fortran 95.
Could spell trouble for the PG-13 rating..
-"Say, Shrek, did those evil crackers play a trick on us, or are you just happy to see me?"
-" W00t!! Yeah baby, lets get it on!"
But they could tell everyone they were, just have a screen saver that pegs the CPU, tells you that you've rendered X frames, and displays a cool screensaver from the movie!
Instead of just looping the CPU, they could covertly run their own version of the seti@home client and crank up some time on the pixar seti@home team..
the film studios I'm sure have crazy fiber/multi-gigabit interconnects within their rendering farms.
While the amount of data to move around probably is too much for dialup, gigabit ethernet is certainly fast enough, and dirt cheap as it's integrated on motherboards. If you look at the top500 list, you see that weta digital (the company which did the CG for lord of the rings IIRC) has a couple of clusters on the list, and they have gig ethernet.
Basically, while rendering is cpu-intensive it's not latency sensitive, so there's no point in blowing a huge amount of cash on a high end cluster interconnect.
Russia's had a military force, but the only thing holding it up was the desire of the people to reap the rewards of becoming communists party members. Obviously there was some patriotism, but it was far from a deciding factor in anything but defense.
That's a question of motivation. Of course any soldier, conscript or professional, fights harder when he's defending his homeland than when he's on some empire building mission. Do you think that morale in the EU armies would be high if we would embark on some "follow in Napoleons and Hitlers footsteps and storm the gates of Moscow"-project? This being the third time, perhaps we'd remember to bring winter clothing this time, though.
But the question of motivation regarding attack vs. defense won't make such a huge impact in the feasibility of any plan. Until the early 80:s, the Soviets had hardware that was about equivalent to the west, and they had overwhelming numerical superiority in Europe. I think they could have succeeded (discounting that any such attack would have provoked a nuclear exhange between the US and the USSR).
There military is not much different than it was previously, except that it is lacking in willing manpower.
Yes it is different. They have downsized a lot, and still a large part of their "active" arsenal is nothing but scrap metal. The few things that work are still 1970 level technology.
If Russia were to revert to a military state, it's quite likely that you'd find that they have a rather fearsome military again.
Not with their current economy. Spending 30% of GDP on the military is not a usable long term scenario, as it will crash the economy. Given that the GDP of Russia is currently about equal to Sweden, the EU is more than able to counter any increase in military spending by Russia, if the EU feels there is a need. By the time the economy of Russia is able to provide a military threat to the EU, we can only hope that the people of Russia will choose prosperity (via tight economic integration with the EU and the rest of the world) instead of yet another world domination scheme. Given the horrors the Russian populace has gone through during the past few centuries, they certainly deserve peace and prosperity as much as anyone else, if not more.
No, my point is that the decisions by EU countries to cut back on their militaries means that they will have no resources to fight any war that lands on their doorstep.
Even after these cutbacks, the EU is more than able to fight any neighborhood war. As I explained above, the Russian military is but a shadow of its former self, they won't be any threat until the Russian economy is on par with the EU countries (which at the very least will take decades). Ukraine, or any other small-scale dictatorship, is small and poor, so they are even less of a threat. The mongolian hordes from China will only reach europe after wading through a hailstorm of Russian nukes, so that's not a really realistic scenario either.
The EU doesn't have any imperial ambitions, as opposed to the current US administration, so there is no need for a expeditionary force capable of conquering some banana republic on the other side of the world either.
With no realistic major military threat in sight, it makes sense to spend less on the military and more on say, growing the economy. E.g. if military spending is reduced from 3 % to 2 % of GDP and the 1 % left over is used for growing the economy, it doesn't take long until the 2 % spending matches the former 3 %.
The US has tremendous experience with Guerilla warfare and at least has semi-effective countermeasures.
No. There is no effective military countermeasure against guerilla warfare. What you can do is try to win the support of the populace, thus enabling your police to work more efficiently, and using armored convoys to give some defense against ambushes. OTOH, while burning supply trucks look bad on TV, they don't really pose a major military threat either. Besides, as the EU isn't planning on conquering other countries, there is little chance of having to deal with guerilla warfare in the first place.
The Russian military was never much of anything.
That's complete and utter bollocks. Up until the early 1980:s they were perfectly capable of steamrolling europe. The only thing holding them back was NATO nukes as well as the questionable benefit of attacking in the first place, the world domination plans slowly dying out after Stalin died in 1953.
The thing that did them in was that the communist economic system simply sucked. Thanks to the market economy, the west was able to improve their military hardware faster than the Soviets. When warfare became increasingly computerized in the 1980:s they couldn't keep up any longer.
True. But many of them will argue that their current military forces are sufficient to deal with such "primitive" militaries. Too bad they have no experience with Guerilla warfare...
Are you suggesting that blowing zillions of taxpayer money on star wars weaponry will help the EU fight a guerilla war then?
High tech helps very little in a guerilla war, as the Americans are finding out in Iraq at the moment.
My thinking is that it will be a typical Microsoft product. It will be big, bloated, expensive, but easy enough to use that someone without an advanced CS degree may be able to create a supercomputer application after taking some certification courses. Another McEngineer.
Well, most supercomputer applications are made by people without advanced CS degrees. Rather they are scientists, experts in their own field. That, of course, doesn't mean that writing writing numerical code with Fortran using MPI is easy. Far from it. But I don't think MS can do much about making this easier. Writing numerical software is a very different endeavour than banging out forms stuff with VB.
On the contrary HPC is all about custom/very specialised hardware running a very specialised application built for one perpose alone.
Why don't you take a look at the top500 list instead of guessing? Yes there are a small amount of supercomputer only type architectures (vector processors mostly, like the NEC SX and Cray X1), but most are off the shelf RISC, IA-64 or x86 things. For example, 5 of the top 10 computers are either x86 or IA-64, i.e. in theory Windows could run them.
I find it naturally that MS tries its luck in the HPC world, but windows surely does not fit the bill.
See, we agree on something..