This was a long time ago, but I heard that if you run Linux, this isn't as bad. This is because of how Linux uses memory vs Windows: Windows starts allocating memory for programs at the "top" (uncached) end, whereas Linux starts at the "bottom" (cached) portion of memory.
I'm in a similar situation, and the way I look at it is this: it may be uncached RAM, but it's still faster than my hard drive, which is what I would be using if I were running as many programs as I am able to now.
Still, I'm kicking myself for not getting an HX chipset board, way back when. Oh well.
Linux (and *BSD) does offer some advantages over Tru64 and other high-end unices, the most notable of which is control. The big national labs (who are the primary builders of monster clusters such as this one) have the resources to add whatever features they want to the operating systems they use but they are often not allowed to modify the internals of proprietary operating systems. Also, many of these clusters run on Compaq Alpha / Tru64 systems, future support of which is in doubt. With a Free Software solution, the OS vendor will never go out of business (because there is no OS vendor.)
If there's not a lot of data, it doesn't matter much. If there's tons and tons of data, a cluster design is inefficient.
Sometimes a cluster is still the only option, however: when you have tons and tons of data, but high cpu requirements. I'm thinking of things like the 30 teraflop project.
Infiniband was supposed to help with this by blurring the lines between the network and the system bus, but it's going to be a long time before it exceeds the capabilities of current technology.
If you don't have a manual typwriter, bring out that old impact printer that is gathering dust in your closet. A simple Perl script could change the letters you send enough to make them seem less formulaic.
Or, you could spend an hour a week hand-writing letters to various congresspersons, etc. It's not so important that you send letters to 50 people tomorrow and then ignore them for the next year; spacing the letters out over a few months is perfectly ok.
If your handwriting is as terrible as mine, it doesn't matter; simply do this for a few weeks (being careful to write neatly!) and your handwriting will definitely improve. Not only do you get the benefit of having your voice heard, but your handwriting will be nicer as well.
There's another catch, too: most motherboards' PCI busses don't/won't support 3.6 Gbps of throughput.
Honestly, though, if you need that kind of throughput, I can't really see you going with commodity hardware; you're more likely to go with something like this.
Disclaimer: I think that the linked product is really neat, because I work on something similar.
Sometimes, by restricting some freedoms, we ensure greater freedoms. For example, by restricting the freedom of somebody to murder me, the law ensures that I am free to live.
It's true that when I drink Mt. Dew, my code has less bugs. Which is too bad, because for health reasons I can't guzzle the stuff as much as I'd like...
SUN has also had faster machines simply because the BUS was faster
Not to quibble, but while this might have been true a long time ago, it's certainly not true today. In a Sun Fire 6800 you can't write from memory to PCI space at more than 150 MB/sec, which is really terrible for a 64/66 PCI bus. (The PCI to memory speed in that same machine is about 370 MB/sec.)
Supposedly their next PCI controller chip will fix this problem, but that's what they said about the last one...
Breaking and entering is a crime, no matter how you do it.
Wrong. I can break into my own home as much or as little as I want to. I can smash through my windows, I can break my doors down, I can pick my own locks [although the legality of possessing lockpicking tools varies from jurisdiction to jurisdiction], etc. I can even dig a tunnel from my front yard into my basement, using dynamite to blast through the wall!
In many locations in the US, having lockpicks is not a crime (source: MIT lockpick guide). HOWEVER, using lockpicks in association with a crime is an additional offense in itself. The same should be true for software.
I'm with you 100% here... well, 50%, anyway. Committing a crime is committing a crime, and that's that. I'm not so sure that particular methodologies deserve the specific additional punshments associated with them. E.g. if my wife is murdered because of her race, she's still murdered, and locking the murderer up for 20 years or 200 years won't bring her back. But that's another issue entirely: the point is that the tools themselves should not be illegal if they have legal uses.
You make a very valid point. One thing that has always burned me up, from my 1-12 grade years was when some mindless mundane would ask "Is this going to be on the test?" and if the answer was no they'd veg out and not learn what was to come.
Humorous anecdote: I had a similar situation in my 2nd year calculus course at the University of Wisconsin - Madison. It was a lecture about 5 hours prior to an exam and a student asked if a certain type of problem would be on the exam. The professor said no, but it was still important to know, and spent 10 minutes explaining how to do that type of problem. The student replied "thanks, but I just wanted to know if it was on the exam or not." This pissed the professor off so much that he altered the test to include that type of problem.
Let's look at this from a practical standpoint. You say there is nothing to global warming. I say there is. Let's say that we sit back and do absolutely nothing about it. If you are right and I am wrong, then we will have lost nothing. On the other hand, if I am right and you are wrong, we will have lost everything and will endure an eternity of torture and suffering, followed by probable extinction.
Following this logic, everybody should convert to my religion.
I had a professor (at the University of Wisconsin - Madison) who would give us past exams to study for. He claimed that he always included some questions from past exams on his current exams, so studying these past exams was to our benefit. He also claimed that despite this "early warning," the grades on the exams still followed a typical bell curve, with a few doing very well, most doing ok and a few doing extremely poorly.
The semester that I had this class, the two exams were composed entirely(!) (not partially) of questions from previous exams. Nonetheless, the performance curve still matched the curves from previous semesters. The professor claimed that it indicated that no matter what, a healthy percentage of students were too lazy to study, and those few who did study the past exams, as instructed, deserved their higher scores.
Anyway, my point is that it's not cheating or in any way unethical to study the material being tested, as long as it's with the instructor's ok. Since exams don't really measure much of value anyway, I'm not too concerned that having previous exams available makes it easier for people with at least half a brain to score well. (Those people usually score better anyway.)
And that's a problem. Whatever can be done to make it more like engineering and less like pop music can't be that bad.
Any attempt to make it more "formal" and "rigid" will take the fun out of it, and will cause the brilliant people to leave in disgust, only to be replaced by mindless drones who employ such mind-numbing practices
If a set of programming practices gives mindless drones even 20% of the productivity of brilliant programmers, it will have succeeded. And if XP is too "formal" and "rigid" for you, at least its "40 hour work week" (less for non-USAians) would leave you plenty of free time to write your own, more fun software.
Honestly, get real, this is america and you have no voice here (I smell flame, bring it on), unless you pay for the congress person's / president's election, or for the industry parties.)
This touches on something that I've always wondered about. The forces of evil (RIAA, MPAA, etc.) are not the only rich and powerful organizations in the world. Is there no rich person who fights for freedom wherever there's trouble? What happened to the dot-com millionaires?
I hesitate to make blanket statements about my own future political decisions, but if I ever had a couple million dollars, I would be sure to befriend my local congresscritters, make them aware of my views, and contribute to their campaigns in order to help them propagate my views.
Re:Why are we always on the defensive?
on
CPRM Smokescreen
·
· Score: 1
One action might be assassination of key figures, such as Valent, Rosen etc.
If the world does turn into a totalitarian dictatorship, armed insurrection may be justified, but in the meantime I think that this would be a bad idea. It would merely inflame public opinion against us without really accomplishing anything - there would always be somebody willing to step into their shoes to continue the fight against us decent peace-loving people.
I agree somewhat with the poster who said that the war is essentially won and we're just waiting for the dead body to fall. I say "somewhat" because we do want to limit the damage that the RIAA, MPAA, etc. can do before they're irrelevant.
Basically, this kid's whole future will already be known. Where's the excitement in that kind of life?
I think that would be wonderful. I would just love to know what diseases I am more prone to, so that I can deal with them. Alcoholism? Ok, I won't drink. Parkinson's? Ok, in my productive years I can donate to places that do research in that area. Osteoporosis? Ok, I'll drink lots of milk.
There's a difference between knowing one's "whole future" and knowing one's genetic predispositions. The latter is good, the former is impossible.
I'll agree that creating organ factories is not unethical, but increasing a human lifespan drastically might be. What with over 6 billion people in the world, and roughly 30% of them below poverty level (if I remember UN stats right), making people live longer will only make more people starve to death. That warrants some thought, at least.
That is an interesting thought. I would argue that increasing human lifespan is not necessarily unethical, for this reason: the life spans that would probably be extended would be those of the "developed" world, where productivity is higher. Therefore, the average human productivity would increase. The more production exists, the more there is to be distributed to humans, including those who are in the "developing" world.
How to "fairly" distribute this production is another issue entirely, and full of debate, so I won't comment on it. But I believe that it is possible.
The fundamental prerequisite to bringing "third world" countries up to our standard of living is increasing the amount of production that reaches these countries. One possible way to do this is to increase worldwide production.
Slashdot Mirror
This was a long time ago, but I heard that if you run Linux, this isn't as bad. This is because of how Linux uses memory vs Windows: Windows starts allocating memory for programs at the "top" (uncached) end, whereas Linux starts at the "bottom" (cached) portion of memory.
I'm in a similar situation, and the way I look at it is this: it may be uncached RAM, but it's still faster than my hard drive, which is what I would be using if I were running as many programs as I am able to now.
Still, I'm kicking myself for not getting an HX chipset board, way back when. Oh well.
Linux (and *BSD) does offer some advantages over Tru64 and other high-end unices, the most notable of which is control. The big national labs (who are the primary builders of monster clusters such as this one) have the resources to add whatever features they want to the operating systems they use but they are often not allowed to modify the internals of proprietary operating systems. Also, many of these clusters run on Compaq Alpha / Tru64 systems, future support of which is in doubt. With a Free Software solution, the OS vendor will never go out of business (because there is no OS vendor.)
Except for HP, you mean?
Sometimes a cluster is still the only option, however: when you have tons and tons of data, but high cpu requirements. I'm thinking of things like the 30 teraflop project.
Infiniband was supposed to help with this by blurring the lines between the network and the system bus, but it's going to be a long time before it exceeds the capabilities of current technology.
Or, you could spend an hour a week hand-writing letters to various congresspersons, etc. It's not so important that you send letters to 50 people tomorrow and then ignore them for the next year; spacing the letters out over a few months is perfectly ok.
If your handwriting is as terrible as mine, it doesn't matter; simply do this for a few weeks (being careful to write neatly!) and your handwriting will definitely improve. Not only do you get the benefit of having your voice heard, but your handwriting will be nicer as well.
Honestly, though, if you need that kind of throughput, I can't really see you going with commodity hardware; you're more likely to go with something like this.
Disclaimer: I think that the linked product is really neat, because I work on something similar.
Sometimes, by restricting some freedoms, we ensure greater freedoms. For example, by restricting the freedom of somebody to murder me, the law ensures that I am free to live.
It's true that when I drink Mt. Dew, my code has less bugs. Which is too bad, because for health reasons I can't guzzle the stuff as much as I'd like...
Not to quibble, but while this might have been true a long time ago, it's certainly not true today. In a Sun Fire 6800 you can't write from memory to PCI space at more than 150 MB/sec, which is really terrible for a 64/66 PCI bus. (The PCI to memory speed in that same machine is about 370 MB/sec.)
Supposedly their next PCI controller chip will fix this problem, but that's what they said about the last one...
I was just going to ask the same thing. Maybe he meant that the system itself was able to access the disks at 25 MB/sec?
Wrong. I can break into my own home as much or as little as I want to. I can smash through my windows, I can break my doors down, I can pick my own locks [although the legality of possessing lockpicking tools varies from jurisdiction to jurisdiction], etc. I can even dig a tunnel from my front yard into my basement, using dynamite to blast through the wall!
In many locations in the US, having lockpicks is not a crime (source: MIT lockpick guide). HOWEVER, using lockpicks in association with a crime is an additional offense in itself. The same should be true for software.
I'm with you 100% here... well, 50%, anyway. Committing a crime is committing a crime, and that's that. I'm not so sure that particular methodologies deserve the specific additional punshments associated with them. E.g. if my wife is murdered because of her race, she's still murdered, and locking the murderer up for 20 years or 200 years won't bring her back. But that's another issue entirely: the point is that the tools themselves should not be illegal if they have legal uses.
Humorous anecdote: I had a similar situation in my 2nd year calculus course at the University of Wisconsin - Madison. It was a lecture about 5 hours prior to an exam and a student asked if a certain type of problem would be on the exam. The professor said no, but it was still important to know, and spent 10 minutes explaining how to do that type of problem. The student replied "thanks, but I just wanted to know if it was on the exam or not." This pissed the professor off so much that he altered the test to include that type of problem.
Following this logic, everybody should convert to my religion.
Your freedom to swing your fist ends at my nose.
Actually, I think this would *add* value to Microsoft's upcoming OS...
In any case, the extreme programming people had the acronym first.
I disagree. Being in group projects in school taught me an Important Life Lesson: other people are unreliable.
Regarding "cheating" by studying old exams...
I had a professor (at the University of Wisconsin - Madison) who would give us past exams to study for. He claimed that he always included some questions from past exams on his current exams, so studying these past exams was to our benefit. He also claimed that despite this "early warning," the grades on the exams still followed a typical bell curve, with a few doing very well, most doing ok and a few doing extremely poorly.
The semester that I had this class, the two exams were composed entirely(!) (not partially) of questions from previous exams. Nonetheless, the performance curve still matched the curves from previous semesters. The professor claimed that it indicated that no matter what, a healthy percentage of students were too lazy to study, and those few who did study the past exams, as instructed, deserved their higher scores.
Anyway, my point is that it's not cheating or in any way unethical to study the material being tested, as long as it's with the instructor's ok. Since exams don't really measure much of value anyway, I'm not too concerned that having previous exams available makes it easier for people with at least half a brain to score well. (Those people usually score better anyway.)
Patton.
And that's a problem. Whatever can be done to make it more like engineering and less like pop music can't be that bad.
Any attempt to make it more "formal" and "rigid" will take the fun out of it, and will cause the brilliant people to leave in disgust, only to be replaced by mindless drones who employ such mind-numbing practices
If a set of programming practices gives mindless drones even 20% of the productivity of brilliant programmers, it will have succeeded. And if XP is too "formal" and "rigid" for you, at least its "40 hour work week" (less for non-USAians) would leave you plenty of free time to write your own, more fun software.
I don't know, "40HourWorkWeekSoYouCanHaveALifeProgramming" seems extremely catchy to me.
All this is true, but they're doing the right thing in moving to linux for their IA64 products. I'm no IA64 fan, but that's another issue entirely.
This touches on something that I've always wondered about. The forces of evil (RIAA, MPAA, etc.) are not the only rich and powerful organizations in the world. Is there no rich person who fights for freedom wherever there's trouble? What happened to the dot-com millionaires?
I hesitate to make blanket statements about my own future political decisions, but if I ever had a couple million dollars, I would be sure to befriend my local congresscritters, make them aware of my views, and contribute to their campaigns in order to help them propagate my views.
If the world does turn into a totalitarian dictatorship, armed insurrection may be justified, but in the meantime I think that this would be a bad idea. It would merely inflame public opinion against us without really accomplishing anything - there would always be somebody willing to step into their shoes to continue the fight against us decent peace-loving people.
I agree somewhat with the poster who said that the war is essentially won and we're just waiting for the dead body to fall. I say "somewhat" because we do want to limit the damage that the RIAA, MPAA, etc. can do before they're irrelevant.
I think that would be wonderful. I would just love to know what diseases I am more prone to, so that I can deal with them. Alcoholism? Ok, I won't drink. Parkinson's? Ok, in my productive years I can donate to places that do research in that area. Osteoporosis? Ok, I'll drink lots of milk.
There's a difference between knowing one's "whole future" and knowing one's genetic predispositions. The latter is good, the former is impossible.
That is an interesting thought. I would argue that increasing human lifespan is not necessarily unethical, for this reason: the life spans that would probably be extended would be those of the "developed" world, where productivity is higher. Therefore, the average human productivity would increase. The more production exists, the more there is to be distributed to humans, including those who are in the "developing" world.
How to "fairly" distribute this production is another issue entirely, and full of debate, so I won't comment on it. But I believe that it is possible.
The fundamental prerequisite to bringing "third world" countries up to our standard of living is increasing the amount of production that reaches these countries. One possible way to do this is to increase worldwide production.