I still don't understand why the data center needs to get involved. You can just manage your own power, and using a conventional power meter, the data center can just collect the bill for what you use.
You reduce the cost, complexity, and possible points of failure, and you can still perform any necessary mission.
Until the control channel is jammed, of course, or some guy lobs a mortar into C&C. Didn't you ever play those video games where once you killed the boss, all the other enemies just disappeared? Didn't you think, at the time, "man, that's stupid."
Now we're creating exactly that model. Life imitates art, really.
Power capping allows you to use the full 16A that you're paying for.
Alternatively, a power meter will let you pay for the full 16A you use. Networked power meters are certainly less complex than "power capping". Why not just use those?
Oh, come on. Are you seriously claiming that our feelings about our jobs don't affect the way we do them? The OP might claim his performance wasn't affected, but I guarantee he wasn't as enthusiastic as he could have been.
You're right in that we're really talking about perceived treatment. But you can't deny that wages play a huge role in that perception. Underpaying someone is prime evidence that the employer lacks respect for the employee, and when thus underpaid, an employee has every right to feel shafted.
Look: for 30 years now, wages have not kept up with productivity. Our hard work is going toward the enrichment of a few. It's high time that we increase general wages across the board and end the second gilded age.
Always test new RAM before relying on it, even when you're using ECC memory. It doesn't take long, it's free, and saves you a lot of time later. Memtest86 is an excellent free tool for testing memory. Just burn it onto a CD, boot up with it, and let it run for a new minutes.
Bringing up a few counterexamples does not invalidate the OP's thesis. His ideas hardly "fall apart". Of course there will be a few bad apples who are unethical despite being highly-compensated. But in general? He's spot-on.
Could this technique be used for people who are super morbidly obese to kill of the section of their brain than gives them an appetite?
I believe there are drugs that do the same thing chemically and non-destructively. I've met a few people who've used them to great effect. Granted, they have side effects, but wouldn't overwriting part of your brain with NOPs also have some side effects?
Neurosurgery (along with other kinds of brain damage) frightens me like few other phenomena. It's a little bit like saying: "okay, this piece of code in the kernel is crashing. Let's overwrite it with NOPs and see what happens." What if you need that part of your brain? Are you really the same person after the procedure?
I don't know about those distributions, but I backport packages from Fedora to RHEL frequently. It's simple, really: just grab the fedora srpm and run rpmbuild on it. Most of the time, it'll work fine. Occasionally, you might need to adjust the spec file to accommodate some slight differences, but it's not a big deal. You end up with a package that integrates nicely with the package manager, satisfies dependencies in the normal way, and so on.
Also, I'm not sure why the parent is moderated flamebait. It's a legitimate to want to run a stable distribution, but use later versions of particular packages.
Both I and Richard Dawkins would describe ourselves as "tooth fairy agnostics" -- as in, the evidence for God is so conspicuously absent and evidence against such a being is so strong that God's existence is as likely as the tooth fairy. Therefore, it's not worthwhile considering the possibility of the almighty in our daily actions. The distinction between "God doesn't exist" and "God is so unlikely to exist that the possibility isn't worth considering" is academic and pointless, and fighting about it ignores our common cause against the involvement of magical thinking in public life.
PS. Fuck you. At worst, I screwed up - an honest mistake. You were intentionally rude. There are humans behind these letters on the screen, and you need to learn that.
You're right. I shouldn't have let some momentary external frustration seep out onto a public forum. I am sorry.
Mark me as another vote for Nod32. I bought it a month or so ago and I must say well worth it. Beats the pants off all the "free" ones (AVG and the like)
How do you know Nod is actually doing what it claims? It's easy to boost performance by lying about tests. On the other hand, lying is not an option for free scanners.
(why the *hell* did someone at gcc think such a thing should be default anyway?)
Because it makes sense on every modern platform on earth except for strange embedded ones, that's why. This kernel bug is the result of incorrect kernel code, not a GCC bug.
For the sake of argument, let's suppose you're right. (I think it'll be a cold day in hell when the BSDs move away from GCC.) Increasing performance demands will lead to the inclusion of more optimizations in PCC, and these optimizations will lead people like you to make the same complaints about PCC that people make about GCC today.
Really, what you're opposed to isn't GCC, but the notion of an optimizing compiler. Sorry, but history has spoken: the gain of optimization far outweighs the minor cost of forcing people like you to actually learn what's guaranteed by the language and what is not.
NO ONE has time to write all their code in assembly, not even for the kernel
To be fair, the OP wasn't suggesting that programs actually be written in assembly, but rather that programmers learn assembly and know how to debug libraries. That's a sentiment I'll second: of course you don't write programs using machine code operations, but when something breaks, it's quite useful to be able to drop down to assembly in a debugger and see what's actually going on, especially when debugging optimized code.
As for compiler bugs: once in a while, they really do happen. Of course, one's first, second, and even third reaction shouldn't be to blame the compiler, but when all other options are exhausted, the possibility is there.
You're correct, but just to clarify for the mouthbreathers around here, consider the following C code:
void* x = 0; int y = 0; memcpy(&y, &x, sizeof(y));
There's no guarantee that y is still equal to 0. It follows that initializing a structure with memset(&s, 0, sizeof(s)) might not set to NULL (i.e., the null pointer constant called 0) any pointers in s.
In practice, the difference doesn't matter. It's practically always 0. The difference is academic; no modern platform uses a non-zero NULL pointer.
Of course exclusivity deals are terrible for ordinary people. That's not factually in doubt. Instead, I'd like to explore why you'd post a comment like this. As I see it, there are three explanations:
You're a paid shill: not unheard-of. It would explain your posting as an Anonymous Coward, and would explain the completely idiotic thought wrapped up in perfect grammar and spelling.
You're just trolling: in that case, you've succeeded, though you certainly could have done better.
You genuinely believe the schlock you spewed: this is the least likely and most depressing possibility. This belief would indicate that you really do lack even the most basic grasp of your own economic best interests, and possibly some kind of childhood trauma that created in you an unflinching obedience to authority. Do you vote Republican by any chance?
I sincerely hope the correct option is one or two. People who genuinely believe the crap you posted are responsible for most of the human misery in history.
Slashdot Mirror
So it's Japanese? :-)
I still don't understand why the data center needs to get involved. You can just manage your own power, and using a conventional power meter, the data center can just collect the bill for what you use.
Until the control channel is jammed, of course, or some guy lobs a mortar into C&C. Didn't you ever play those video games where once you killed the boss, all the other enemies just disappeared? Didn't you think, at the time, "man, that's stupid."
Now we're creating exactly that model. Life imitates art, really.
Alternatively, a power meter will let you pay for the full 16A you use. Networked power meters are certainly less complex than "power capping". Why not just use those?
Oh, come on. Are you seriously claiming that our feelings about our jobs don't affect the way we do them? The OP might claim his performance wasn't affected, but I guarantee he wasn't as enthusiastic as he could have been.
You're right in that we're really talking about perceived treatment. But you can't deny that wages play a huge role in that perception. Underpaying someone is prime evidence that the employer lacks respect for the employee, and when thus underpaid, an employee has every right to feel shafted.
Look: for 30 years now, wages have not kept up with productivity. Our hard work is going toward the enrichment of a few. It's high time that we increase general wages across the board and end the second gilded age.
Always test new RAM before relying on it, even when you're using ECC memory. It doesn't take long, it's free, and saves you a lot of time later. Memtest86 is an excellent free tool for testing memory. Just burn it onto a CD, boot up with it, and let it run for a new minutes.
Bringing up a few counterexamples does not invalidate the OP's thesis. His ideas hardly "fall apart". Of course there will be a few bad apples who are unethical despite being highly-compensated. But in general? He's spot-on.
I believe there are drugs that do the same thing chemically and non-destructively. I've met a few people who've used them to great effect. Granted, they have side effects, but wouldn't overwriting part of your brain with NOPs also have some side effects?
Neurosurgery (along with other kinds of brain damage) frightens me like few other phenomena. It's a little bit like saying: "okay, this piece of code in the kernel is crashing. Let's overwrite it with NOPs and see what happens." What if you need that part of your brain? Are you really the same person after the procedure?
Laws only apply to little people like us.
You have no idea how right you are. There's an old saying: "The Internet treats censorship as damage and routes around it."
That's not necessarily true either. Consider Diffie-Hellman-style secure key exchange over an insecure channel.
Haystack (full disclosure: I wrote it.)
I don't know about those distributions, but I backport packages from Fedora to RHEL frequently. It's simple, really: just grab the fedora srpm and run rpmbuild on it. Most of the time, it'll work fine. Occasionally, you might need to adjust the spec file to accommodate some slight differences, but it's not a big deal. You end up with a package that integrates nicely with the package manager, satisfies dependencies in the normal way, and so on.
Also, I'm not sure why the parent is moderated flamebait. It's a legitimate to want to run a stable distribution, but use later versions of particular packages.
It's the silly idea that we can affect gender inequality through affirmative action for pronouns.
Both I and Richard Dawkins would describe ourselves as "tooth fairy agnostics" -- as in, the evidence for God is so conspicuously absent and evidence against such a being is so strong that God's existence is as likely as the tooth fairy. Therefore, it's not worthwhile considering the possibility of the almighty in our daily actions. The distinction between "God doesn't exist" and "God is so unlikely to exist that the possibility isn't worth considering" is academic and pointless, and fighting about it ignores our common cause against the involvement of magical thinking in public life.
You're right. I shouldn't have let some momentary external frustration seep out onto a public forum. I am sorry.
How do you know Nod is actually doing what it claims? It's easy to boost performance by lying about tests. On the other hand, lying is not an option for free scanners.
Because it makes sense on every modern platform on earth except for strange embedded ones, that's why. This kernel bug is the result of incorrect kernel code, not a GCC bug.
For the sake of argument, let's suppose you're right. (I think it'll be a cold day in hell when the BSDs move away from GCC.) Increasing performance demands will lead to the inclusion of more optimizations in PCC, and these optimizations will lead people like you to make the same complaints about PCC that people make about GCC today.
Really, what you're opposed to isn't GCC, but the notion of an optimizing compiler. Sorry, but history has spoken: the gain of optimization far outweighs the minor cost of forcing people like you to actually learn what's guaranteed by the language and what is not.
To be fair, the OP wasn't suggesting that programs actually be written in assembly, but rather that programmers learn assembly and know how to debug libraries. That's a sentiment I'll second: of course you don't write programs using machine code operations, but when something breaks, it's quite useful to be able to drop down to assembly in a debugger and see what's actually going on, especially when debugging optimized code.
As for compiler bugs: once in a while, they really do happen. Of course, one's first, second, and even third reaction shouldn't be to blame the compiler, but when all other options are exhausted, the possibility is there.
What makes you say that? if(x) and if(!x) are perfectly good idiomatic C constructs, and their meaning is apparent at first glance.
You're correct, but just to clarify for the mouthbreathers around here, consider the following C code:
There's no guarantee that y is still equal to 0. It follows that initializing a structure with memset(&s, 0, sizeof(s)) might not set to NULL (i.e., the null pointer constant called 0) any pointers in s.
In practice, the difference doesn't matter. It's practically always 0. The difference is academic; no modern platform uses a non-zero NULL pointer.
Of course exclusivity deals are terrible for ordinary people. That's not factually in doubt. Instead, I'd like to explore why you'd post a comment like this. As I see it, there are three explanations:
I sincerely hope the correct option is one or two. People who genuinely believe the crap you posted are responsible for most of the human misery in history.